本篇博文主要内容为 2026-09-01 从Arxiv.org论文网站获取的最新论文列表，自动更新，按照NLP、CV、ML、AI、IR、MA六个大方向区分。
说明：每日论文数据从Arxiv.org获取，每天早上12:30左右定时自动更新。
提示: 当天未及时更新，有可能是Arxiv当日未有新的论文发布，也有可能是脚本出错。尽可能会在当天修复。
目录
概览 (2026-09-01)
今日共更新1436篇论文,其中:
- 自然语言处理共298篇(Computation and Language (cs.CL))
- 人工智能共426篇(Artificial Intelligence (cs.AI))
- 计算机视觉共316篇(Computer Vision and Pattern Recognition (cs.CV))
- 机器学习共337篇(Machine Learning (cs.LG))
- 多智能体系统共25篇(Multiagent Systems (cs.MA))
- 信息检索共51篇(Information Retrieval (cs.IR))
- 人机交互共46篇(Human-Computer Interaction (cs.HC))
多智能体系统
[MA-0] Learning to Evaluate Before Improving: Automatic Rubric Induction for Automatic Research Agents
【速读】：该论文旨在解决自主科学研究代理在执行开放式科研任务时，因任务指令不明确而导致分析遗漏、方法误用或结论证据不足的问题。其核心挑战在于，此类任务通常未明确定义所需分析内容、方法选择及成功标准，从而影响研究的严谨性与可验证性。为应对这一问题，论文提出一种“评估优先”（evaluation-first）框架AutoSciRub，其关键创新在于：在科研执行前，基于任务指令生成一个特定于任务的可执行评价量表（executable rubric），该量表将模糊的任务要求分解为原子化的科学目标，并通过相关文献与可见数据进行锚定，进而合成具体、可操作且可验证的评判标准。该量表不仅指导实验设计与分析流程，还支持逐级验证与迭代修正，使研究过程具备可追溯性与可控性。实验结果表明，AutoSciRub在ResearchClawBench和AstaBench E2E Discovery基准上均显著提升性能，平均得分提升达2.08至16.8点，同时保持或增加成功完成的任务数量，验证了评估优先机制在提升自主科学研究的可靠性与通用性方面的有效性。
链接: https://arxiv.org/abs/2608.31076
作者: Xuehai Wang,Haowei Qin,Tongxin Liu,Junkai Li,Buqiang Xu,Jintian Zhang,Yijun Chen,Zirui Xue,Shumin Deng
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR); Machine Learning (cs.LG); Multiagent Systems (cs.MA); Software Engineering (cs.SE)
备注: Work in progress
Abstract:Autonomous scientific research agents are increasingly applied to end-to-end scientific workflows, including literature review, data analysis, experimentation, and report generation. However, open-ended research tasks often do not clearly specify the analyses, methods, and success criteria required to complete the task. As a result, agents may miss important analyses, use inappropriate methods, or draw conclusions that are insufficiently supported by evidence. To address the problem, we present AutoSciRub, an evaluation-first framework that induces a task-specific executable rubric before research execution, and uses it to guide execution, criterion-level verification as well as iterative revision. AutoSciRub decomposes an underspecified instruction into atomic scientific goals, grounds them in relevant literature and task-visible data, and synthesizes specific, actionable, and verifiable criteria. The resulting rubric makes implicit experimental and evidential requirements explicit, providing guidance for experiments and analyses. During revision, rubric-guided verification identifies unmet criteria and enables targeted refinement of the research report and its supporting artifacts. On ResearchClawBench, AutoSciRub consistently improves all tested configurations, with an average gain of 2.08 points across three backbone LLMs under the fixed Codex harness and 2.95 points across three agent harnesses using a fixed DeepSeek-V4-Flash backbone. On a randomly sampled 20-task subset of AstaBench E2E Discovery, AutoSciRub further achieves an average improvement of 16.8 points across three agent harnesses, while maintaining or increasing the number of successfully completed tasks. These results demonstrate that evaluation-first guidance provides an effective and generalizable control mechanism for autonomous scientific research (Code: this https URL).
[MA-1] Evidence Logic and Compliance: Multi-Agent Structured Graph Reasoning with Expert Arbitration for Medical Referral
【速读】：该论文旨在解决医疗转诊（medical referral）这一复杂决策任务中，现有大语言模型（LLM）在处理多模态临床数据时面临的两大核心问题：一是信息过载（Information Overload），即模型过度关注高频疾病术语而忽略细微但关键的紧急程度指示；二是非结构化协作（Unstructured Collaboration），现有多智能体框架依赖松散对话，易引发语义漂移与确认偏误。其解决方案的关键在于提出一种名为MASGR（Multi-Agent Structured Graph Reasoning）的新框架，将转诊过程重构为结构化图推理问题，通过专业化智能体从不同模态（如患者主诉、检验指标、影像学数据）中提取证据，并以临床推理图（clinical reasoning graph）进行协同，强制建立证据间的显式逻辑关联，从而实现对矛盾信息的有效整合。此外，引入基于知识的仲裁机制，优先遵循患者安全规则而非常规诊断分类，显著提升决策可靠性。实验结果表明，MASGR在真实世界医疗记录上的表现显著优于现有最先进模型和多智能体系统，尤其在需权衡慢性病管理与急症干预的复杂病例中优势明显。该研究的AI贡献在于构建了可验证的逻辑图推理范式，工程应用则体现在将其集成至复杂医疗决策系统中，有效提升了复杂转诊任务的精准性。
链接: https://arxiv.org/abs/2608.30938
作者: Qi Peng,Yi Cai,Jialin Cui,Tong Zhu,Yujuan Ding,Qingbao Huang,Tao Wang,Jiayuan Xie,Changmeng Zheng,Qing Li
机构: South China University of Technology (华南理工大学); Hong Kong Polytechnic University (香港理工大学); Guangzhou University (广州大学); Guangxi University (广西大学); King’s College London (伦敦国王学院)
类目: Multiagent Systems (cs.MA)
备注: 18 pages
Abstract:Medical referral (directing patients to the appropriate hospital department) is a complex decision-making process requiring the synthesis of multimodal data, including patient narratives, laboratory indicators, and radiology imaging. While Large Language Models (LLMs) have advanced medical dialogue systems, they struggle with real-world referral tasks due to two primary limitations: (1) Information Overload, where models fixate on high-frequency disease terms while overlooking subtle but critical urgency indicators; and (2) Unstructured Collaboration, where existing multi-agent frameworks rely on loose dialogue that leads to semantic drift and confirmation bias. To address these challenges, we introduce MASGR (Multi-Agent Structured Graph Reasoning), a framework that treats referral not as a classification task but as a structured graph construction problem. MASGR deploys specialized agents to extract evidence from distinct modalities and coordinates them through a clinical reasoning graph. This graph forces agents to establish explicit logical connections between conflicting evidence. Furthermore, we integrate a knowledge-guided arbitration mechanism that prioritizes patient safety rules over standard diagnostic classification. Extensive experiments on real-world medical records demonstrate that MASGR significantly outperforms state-of-the-art LLMs and existing multi-agent systems, particularly in complex cases requiring the balancing of chronic disease management and emergency intervention. The AI contribution lies in the Multi-Agent Structured Graph Reasoning framework that transforms unstructured multi-agent dialogue into a verifiable logical graph construction. The engineering application is demonstrated through its deployment in a complex healthcare decision-making system to optimize the precision of complex medical referrals.
[MA-2] HiRS-Agent : A Hierarchical Multi-Agent System for Reliable Long-Horizon Remote Sensing Task Solving
【速读】：该论文旨在解决远程感知（Remote Sensing, RS）任务中长期、多阶段、高度依赖的复杂任务处理问题，尤其针对现有系统普遍采用的单一决策框架所导致的任务执行不稳定、工具使用错误及错误传播等缺陷。其核心解决方案是提出一种分层多智能体系统——HiRS-Agent，该系统采用两级协同架构：管理层负责动态路由、步骤级验证、重规划与终止控制，以实现全局任务协调；专业层则根据遥感工作流组织领域专用工具，承担子任务推理与工具调用。为提升系统性能，研究引入两阶段监督微调策略与基于验证引导的分层强化学习机制，联合优化协作与工具使用策略。实验结果表明，HiRS-Agent在Earth-Agent Benchmark和ThinkGeo数据集上显著提升了长时程工具使用能力与最终任务正确率，验证了结构化多智能体协作在构建可靠遥感智能体中的有效性。
链接: https://arxiv.org/abs/2608.30672
作者: Boyang Mu,Zhiwei Wei,Mugen Peng,Wenjia Xu
机构: Beijing University of Posts and Telecommunications, State Key Laboratory of Networking and Switching Technology(北京邮电大学，网络与交换技术国家重点实验室); Hunan Normal University(湖南师范大学)
类目: Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA); Multimedia (cs.MM)
备注: Accepted at ACM Multimedia 2026 (MM '26)
Abstract:Recent advances in large language models and multimodal models have pushed remote sensing (RS) processing from simple perception models to agentic systems designed to tackle complex, long-horizon RS tasks. However, existing systems often rely on monolithic decision-making frameworks, which fail to accommodate the multi-stage, interdependent nature of RS tasks. This centralized approach leads to challenges such as unstable task execution, incorrect tool usage, and error propagation across stages. To address these issues, we propose HiRS-Agent, a hierarchical multi-agent system for long-horizon RS task solving. HiRS-Agent adopts a two-level collaborative architecture: the Manager Layer handles dynamic routing, step-level verification, replanning, and termination control, while the Specialist Layer organizes domain-specific tools according to the RS workflow and is responsible for subtask reasoning and tool execution. To further enhance the system’s capability, we introduce a two-stage supervised tuning strategy and a verification-guided hierarchical reinforcement learning stage to jointly optimize coordination and tool-use policies. Experiments on Earth-Agent Benchmark and ThinkGeo show that HiRS-Agent substantially improves long-horizon tool-use capability and final-task correctness, demonstrating the effectiveness of structured multi-agent collaboration for reliable RS agents. The code is publicly available at this https URL.
[MA-3] LLM -based Hardware Development with Hierarchical IRs and End-to-End Multi-Agent Workflow
【速读】：该论文旨在解决大语言模型（Large Language Models, LLMs）在复杂硬件设计中应用受限的问题，其核心挑战在于硬件设计领域缺乏公开的训练数据，且硬件开发方法学与软件开发存在本质差异。具体而言，将LLMs应用于硬件设计不仅需要生成寄存器传输级（RTL）代码，更要求模型理解模块边界、模块间连接关系以及验证需求。为此，本文提出一种基于LLM的硬件开发框架，其关键创新在于引入分层中间表示（Intermediate Representations, IRs）和端到端多智能体工作流。通过两个结构化IR——架构草图（Architectural Sketch）用于刻画模块拓扑与互连关系，操作规范（Operational Specification）用于定义各模块的功能与接口——该框架能够将复杂设计分解为子模块，明确每个模块的功能，并推导出相应的测试与验证方案。同时，框架集成多智能体调试循环，使智能体可根据错误反馈动态控制仿真探针信号等调试细节。在Verilog-Eval基准测试中，该框架达到95.5%的pass@5率，显著优于现有最先进方法；在涵盖通用处理器到数字信号处理系统的复杂真实案例研究中，仅本框架能生成功能完整、符合工业标准设计规则、无语法错误、功能正确且可综合的最终RTL代码，充分证明了其在复杂硬件设计中的有效性与实用性。
链接: https://arxiv.org/abs/2608.30659
作者: Chenyang Yin,Agasthi Haputhanthri,Aditya Anirudh Jonnalagadda,Zhenyu Bai,Yuanming Song,Saranyu Chattopadhyay,Mohammad Fadiheh,Tom Zelazny,Subhasish Mitra,Tulika Mitra
机构: National University of Singapore(新加坡国立大学); Shandong University(山东大学); Stanford University(斯坦福大学); LUBIS EDA
类目: Hardware Architecture (cs.AR); Multiagent Systems (cs.MA); Software Engineering (cs.SE)
备注: 10 pages, 5 figures
Abstract:Large language models (LLMs) are increasingly used in software development, but their use in complex hardware design remains limited. This gap stems from both the scarcity of public hardware training data and the fundamentally different methodologies used in hardware design. In particular, applying LLMs to hardware requires more than direct RTL generation: the model must understand module boundaries, inter-module connections, and verification requirements. In this paper, we present an LLM-based hardware development framework with hierarchical intermediate representations (IRs) and an end-to-end multi-agent workflow. The core idea is to provide an abstraction of hardware design to LLMs through two structured IRs: Architectural Sketch, which captures module topology and interconnection, and Operational Specification, which defines per-module functionality and interfaces. Our framework uses these IRs to decompose a complex design into sub-modules, specify the per-block functionality, and derive how each module should be tested and verified. We incorporate a multi-agent debug loop in the framework, allowing agents to get the error feedback and control the debug details such as the signals to be probed for simulation. We evaluate our framework on Verilog-Eval benchmark, achieving a pass@5 rate of 95.5%, which surpasses current state-of-the-art LLM generation frameworks. To better assess performance on complex, realistic designs, we introduce a new case study spanning applications from general-purpose processors to digital signal processing systems. Experimental results indicate that such complex designs exceed the capabilities of existing approaches, whereas our framework is the only one capable of producing functional end-to-end design. Our generated RTL follows all industry-standard design rules, is lint-clean, functionally correct and fully synthesizable.
[MA-4] SimCRAFT: Distilling Remote Sensing Agents via Synthetic Trajectories and Contextual Retrieval-Augmented Fine-Tuning
【速读】：该论文旨在解决遥感（Remote Sensing, RS）领域中因地球观测数据量与多样性急剧增长，导致传统人工工作流难以应对的瓶颈问题，尤其针对当前遥感智能代理严重依赖大规模通用大语言模型（Large Language Models, LLMs）所引发的领域知识匮乏与高昂基础设施成本难题。其核心解决方案是提出一种模型无关的轻量化框架SimCRAFT，通过将复杂的遥感任务编排能力高效压缩至70亿参数规模的模型中。关键创新在于：首先，构建多智能体合成引擎与模拟执行引擎协同的数据生成机制，验证流程结构、工具间依赖关系及传感器-工具兼容性，生成大规模、约束校验的遥感工作流规划语料库SimRS-14k，缓解领域数据稀缺问题；其次，提出上下文感知的检索增强微调方法（Contextual Retrieval-Augmented Fine-Tuning, CRAFT），使模型在噪声鲁棒的目标下实现类比推理，通过动态适配检索到的标准操作规程（Standard Operating Procedures, SOPs）以应对新查询，避免机械复制，有效拓展了原有RAFT方法在多步骤遥感任务规划中的泛化能力。实验表明，SimCRAFT-7B在性能上显著优于开源大模型，媲美先进闭源模型与专用遥感代理，并可在三个不同70亿参数基线模型上复现，为资源受限或需节能部署的场景提供了具有竞争力的轻量化开放权重基准。
链接: https://arxiv.org/abs/2608.30277
作者: Haoran Wang,Jing Yao,Xu Yang,Zeqing Wang,Yang Zhang,Pedram Ghamisi,Zhengchao Chen
机构: Chinese Academy of Sciences (中国科学院); University of Chinese Academy of Sciences (中国科学院大学); The Hong Kong University of Science and Technology (Guangzhou) (香港科技大学(广州)); Helmholtz-Zentrum Dresden-Rossendorf (德国德累斯顿罗森多夫亥姆霍兹研究中心)
类目: Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注:
Abstract:The unprecedented surge in Earth observation data volume and diversity has exposed a critical bottleneck for traditional manual workflows, catalyzing the emergence of Remote Sensing (RS) Agents. However, the practical deployment of these advanced agents is severely hindered by their heavy reliance on large-scale general-purpose LLMs, which lack deep domain expertise and impose prohibitive infrastructure demands. To resolve this, we propose SimCRAFT, a model-agnostic framework that distills sophisticated RS orchestration capabilities into a compact 7B-scale model. Addressing data scarcity, we first pair a multiagent synthesis engine with a Mock Execution Engine that checks schema correctness, inter-tool dependencies, and sensor/tool compatibility, producing SimRS-14k, a large-scale, constraint-validated workflow planning corpus. Second, we propose Contextual Retrieval-Augmented Fine-Tuning (CRAFT) that finetunes the model to reason analogically by adapting retrieved Standard Operating Procedures to novel queries under a noise-robust objective, generalizing RAFT to multi-step RS workflow planning without mechanical copying. Extensive experiments demonstrate that SimCRAFT-7B significantly outperforms openweights LLMs and rivals advanced closedsource models and specialized RS agents, while reproducing across three 7B backbones. This work contributes a competitive open-weights baseline for lightweight RS intelligence, enabling efficient autonomous deployment under resource-constrained or resource-conserving conditions.
[MA-5] Zero-Knowledge Predicate Proofs Between AI Agents : A Measured Cross-Protocol Gateway and the Source-Integrity Gap
【速读】：该论文旨在解决多智能体人工智能平台在从开发环境快速部署至生产环境过程中，智能体间信任建立机制严重不足的问题。现有方法要么过度共享原始数据，要么依赖不可验证的自然语言自述，后者正是提示注入攻击的根源。为应对这一挑战，论文提出一种基于可证明数据最小化的解决方案，其核心在于通过零知识证明网关（Zero-Knowledge Proof Gateway），使智能体之间交换的是对私有数据上治理定义谓词的证明，而非数据本身，从而实现暴露防护的系统性设计，而非事后检测。由于当前互操作协议无法承载此类证明，研究者提出并实现了一个专用通信槽（slot），在MCP和Agent2Agent框架中完成端到端集成。实验表明，32位阈值谓词的证明生成仅需6.2毫秒，验证耗时1.0毫秒，且仅需608字节的Bulletproofs证明，在单个通用vCPU上即可运行；经过十一轮对抗性测试与十九次协议校验均通过，系统已成功部署于Kubernetes环境，并经实证验证具备网络隔离能力。案例研究表明，可在不泄露具体订单金额的前提下，证明零售客户订单未超出额度，将《通用数据保护条例》（GDPR）中的数据最小化原则转化为可强制执行的技术措施，符合欧盟法律明确要求。此外，针对现有工作未解决的关键局限——谓词证明仅绑定声明与承诺值，无法关联至系统记录源，论文进一步提出一种双向融合方案：将可信执行环境（enclave）背书与证明相结合，使得验证任一证明即可联合确认谓词成立及该值由特定测量二进制文件读取，已在模拟权威机构环境下完成验证。
链接: https://arxiv.org/abs/2608.30083
作者: Ashok Subbabhatta Gopalakrishna
机构: 未知
类目: Cryptography and Security (cs.CR); Multiagent Systems (cs.MA)
备注: 11 pages, 6 figures, 4 tables. Reference implementation available at this https URL
Abstract:Multi-agent AI platforms move quickly from staging to production, but the way agents establish trust remains rudimentary: an agent either transmits raw data to a peer or accepts that peer’s natural-language self-report that a value complies with policy. The first over-shares; the second is unverifiable and is exactly the channel prompt injection attacks. Prevailing responses emphasise identity, visibility, and post-hoc detection, and recent proposals for cryptographically enforced agent policy have been evaluated in simulation rather than execution. We take provable data minimisation between agents from proposal to running system. In our Zero-Knowledge Proof Gateway, agents exchange proofs of governance-defined predicates over private data rather than the data itself, so exposure is prevented by design rather than detected afterwards; because no interoperability protocol can carry such a proof, we propose a slot and implement it on both MCP and Agent2Agent from one endpoint. A 32-bit threshold predicate proves in 6.2 ms and verifies in 1.0 ms with a 608-byte Bulletproofs proof on one commodity vCPU; eleven adversarial experiments and nineteen protocol checks pass; and the system is deployed to Kubernetes with empirically verified network isolation. Our case study proves a retail client order is within its limit without revealing the amount, instantiating the GDPR data-minimisation principle as an enforced technical measure of the kind EU law now names explicitly. We then address the limitation no comparable work resolves: a predicate proof binds a statement to a committed value, never to the system of record. We give a construction fusing an enclave attestation with the proof in both directions, so verifying one artifact certifies jointly that the predicate holds and that the value was read by a specific measured binary, and test it against a mock authority.
[MA-6] Detect Before You Attribute: Cascade Failure Attribution for Multi-Agent Systems
【速读】：该论文旨在解决基于大语言模型（Large Language Model, LLM）的智能体在执行复杂任务时因多步推理导致的执行失败问题，核心挑战在于如何实现精准的故障归因（failure attribution）。现有方法或依赖轨迹拓扑结构与频谱特征，但忽略细粒度语义信息；或采用基于LLM的归因方法，却受限于长轨迹带来的上下文长度退化问题。为此，论文提出一种即插即用的检测过滤器DUOTRACE，其关键创新在于采用“先检测、后归因”的范式：首先通过变分自编码器（VAE）驱动的异常检测机制识别异常执行路径，随后向下游的LLM归因方法提供聚焦的轨迹证据。DUOTRACE通过融合双视角语义-结构节点表征、基于树形长短时记忆网络（Tree-LSTM）的轨迹编码器，以及前缀链与基于LLM的数据增强策略，有效应对异构节点、层次化执行结构及故障样本稀缺等挑战。实验表明，相较于六种主流LLM基线方法，DUOTRACE在任务级和步骤级归因准确率上分别提升了8.7%和7.0%，显著增强了智能体系统的可靠性。
链接: https://arxiv.org/abs/2608.29646
作者: Jiayi Zhang,Zexin Wang,Degang Sun,Changhua Pei,Fei Sun,Gaogang Xie,Jingjing Li
机构: Computer Network Information Center, Chinese Academy of Sciences(中国科学院计算机网络信息中心); Institute of Computing Technology, Chinese Academy of Sciences(中国科学院计算技术研究所)
类目: Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注: 17 pages, 5 figures, 13 tables
Abstract:Large language model (LLM)-based agents have shown strong potential in solving complex tasks through multi-step reasoning, yet they remain vulnerable to execution failures. Accurate failure attribution is therefore critical for improving agent reliability. Existing topology- and spectrum-based methods exploit trajectory structures but often overlook fine-grained semantics, while LLM-based attribution methods capture semantic cues but suffer from long-context degradation over lengthy trajectories. To address these challenges, we propose DUOTRACE, a plug-and-play detection filter for LLM-based failure attribution. DUOTRACE follows a detect-before-attribute paradigm: it first detects anomalous executions and then supplies focused trajectory evidence to downstream LLM-based attribution methods. For effective VAE-based anomaly detection on agent trajectories, DUOTRACE integrates dual-view semantic-structural node representations, a Tree-LSTM-based trajectory encoder, and prefix-chain- and LLM-based data augmentation to handle heterogeneous nodes, hierarchical execution structures, and limited failure data. Experiments with six LLM-based attribution baselines show that DUOTRACE improves agent-level and step-level attribution accuracy by 8.7% and 7.0%, respectively.
[MA-7] Harness-RL: Black-Box Reinforcement Learning with Action-Args Decoupling for Central-Agent Multi-Agent Harnesses
【速读】：该论文旨在解决在多智能体协作框架中训练中心策略时面临的两大挑战：其一，动作标签为低基数决策，而其参数（args）构成高维条件序列，若使用统一的序列级信号优化二者，易产生冲突梯度；其二，动态调度导致会话间存在分支、并行调用及上下文重写等复杂依赖关系，难以被简化为单一扁平的标记序列。为此，论文提出Harness-RL，一种结构化强化学习框架，其核心在于结合冲突感知策略优化（Conflict-Aware Policy Optimization, CAPO）与接口级黑箱轨迹构建。黑箱组件通过捕获接口调用记录，构建会话级前缀树，并利用可训练标记对结果奖励与过程奖励进行对齐；CAPO则基于前向激活识别与动作及参数相关联的参数子空间，将策略梯度分别路由至对应子空间。该方法支持仅中心智能体训练与联合多智能体训练两种模式。在七个多跳问答与代理检索基准测试中，Harness-RL在Qwen2.5-1.5B和Qwen2.5-3B模型上分别达到42.93和47.79的平均F1分数，消融实验验证了CAPO的有效性，并表明在所评估设置下，中心仅训练策略更具优势。
链接: https://arxiv.org/abs/2608.29641
作者: Xinke Jiang,Zhixin Zhang,Zhibang Yang,Jiaran Gao,Rihong Qiu,Shijin Chen,Xu Chu,Junfeng Zhao,Yasha Wang
机构: Peking University (北京大学); National Engineering Research Center of Software Engineering (软件工程国家工程研究中心); School of Computer Science, Peking University (计算机学院，北京大学); Key Laboratory of High Confidence Software Technologies, Ministry of Education (高可信软件技术教育部重点实验室); Center on Frontiers of Computing Studies, Peking University (计算前沿研究中心，北京大学); Guangxi Land and Resources Planning and Design Group Co., Ltd (广西壮族自治区土地规划勘测设计集团有限公司); Peking University Information Technology Institute (Tianjin Binhai) (北京大学信息技术研究院（天津滨海）)
类目: Multiagent Systems (cs.MA)
备注: Accepted at PCC 2026, this is the English version
Abstract:Large language model agents increasingly solve long-horizon tasks through multi-agent harnesses in which a central agent coordinates specialized sub-agents, tools, and environments. Training the central policy in such a harness raises two challenges. First, an action label is a low-cardinality decision, whereas its args form a high-dimensional conditional sequence; optimizing both with a shared sequence-level signal can produce conflicting gradients. Second, dynamic scheduling creates interdependent sessions with branches, parallel calls, and rewritten contexts, which cannot be faithfully reduced to one flat token sequence. We introduce Harness-RL, a structured reinforcement learning framework that combines Conflict-Aware Policy Optimization (CAPO) with interface-level black-box trajectory construction. The black-box component captures Interface Call Records, builds per-session prefix trees, and aligns outcome and process rewards with trainable tokens. CAPO uses forward activations to identify parameter partitions associated with action and args tokens, then routes their policy gradients to the corresponding subspaces. Harness-RL supports both central-only and joint multi-agent training. Across seven multi-hop question answering and agentic retrieval benchmarks, it reaches average F1 scores of 42.93 and 47.79 with Qwen2.5-1.5B and Qwen2.5-3B, respectively, while ablations validate the contribution of CAPO and favor central-only optimization in the evaluated setting. Our code is available at this https URL.
[MA-8] Agent icRag -R1: Agent ic Reinforcement Learning with Stack Memory for Multi-Step Reasoning Retrieval and Memorizing
【速读】：该论文旨在解决现有检索增强生成（Retrieval-Augmented Generation, RAG）系统在处理复杂、多步推理任务时面临的挑战，尤其是其在动态适应性检索与中间上下文持续修正方面的不足。现有基于强化学习（Reinforcement Learning, RL）的智能体式RAG方法虽有所改善，但普遍依赖粗粒度动作空间和轨迹级奖励机制，导致奖励分配效率低，并倾向于产生短周期、模式化的推理路径。为克服上述问题，本文提出AgenticRag-R1，一种通过记忆栈（memory stack）深度整合推理、检索与记忆的强化学习框架，其核心创新在于采用细粒度动作空间与分层动作感知奖励机制，结合信息感知的轨迹拒绝策略，从而支持有效的长时序推理学习。实验结果表明，AgenticRag-R1在涵盖多跳推理、开放域及智能体推理等多种基准测试中，均显著优于现有强基线模型，且展现出更鲁棒、可解释性强且具备记忆感知能力的推理行为，验证了细粒度动作建模与信息感知优化对长时序推理的关键作用。
链接: https://arxiv.org/abs/2608.29622
作者: Xinke Jiang,Yue Fang,Zhibang Yang,Jiaran Gao,Zhixin Zhang,Tao Feng,Rihong Qiu,Wentao Zhang,Hongxin Ding,Ruizhe Zhang,Yongxin Xu,Yuheng Huang,Xu Chu,Junfeng Zhao,Yasha Wang
机构: Peking University (北京大学); National Engineering Research Center of Software Engineering (软件工程国家工程研究中心); School of Computer Science, Peking University (计算机学院，北京大学); Key Laboratory of High Confidence Software Technologies, Ministry of Education (高可信软件技术教育部重点实验室); Center on Frontiers of Computing Studies, Peking University (计算前沿研究中心，北京大学); Peking University Information Technology Institute (Tianjin Binhai) (北京大学信息技术研究院（天津滨海）); GRG Banking Equipment Co., Ltd. (广州格兰仕银行设备有限公司)
类目: Multiagent Systems (cs.MA); Artificial Intelligence (cs.AI)
备注:
Abstract:Retrieval-Augmented Generation (RAG) improves the factuality of large language models (LLMs), yet existing RAG systems often struggle with complex, multi-step reasoning that requires adaptive retrieval and continuous revision of intermediate contexts. Recent reinforcement learning (RL)-based agentic RAG methods partially alleviate this issue, but typically rely on coarse-grained action spaces and trajectory-level rewards, resulting in weak reward assignment and a bias toward short-horizon, stereotyped reasoning template. To address, we propose AgenticRag-R1, a RL framework that deeply integrates reasoning, retrieval, and memory via a memory stack and fine-grained action space, supported by hierarchical action-aware rewards and an information-aware trajectory rejection strategy to enable effective long-horizon learning. Experiments across a diverse set of multi-hop, open-domain, and agentic reasoning benchmarks, spanning multiple backbone model sizes, demonstrate that AgenticRag-R1 consistently outperforms strong baselines. Moreover, AgenticRag-R1 learns more robust, interpretable, and memory-aware reasoning behaviors, highlighting the effect of fine-grained action modeling and information-aware optimization for long-horizon reasoning. Our code is anonymous available at this https URL.
[MA-9] Forward-Deployed Full-Stack Engineering for Autonomous Cloud MLOps
【速读】：该论文旨在解决机器学习运维（MLOps）中复杂系统在云环境下的自动化与可信部署难题，即如何在多环节协同（如模型管道、基础设施管理、安全合规、部署监控、重训练与回滚）中确保操作的正确性、可验证性与可靠性。其核心解决方案是提出一种基于证据门控（evidence-gated）的多智能体框架，关键在于通过状态化图编排器（stateful Graph Orchestrator）统一协调多个专业化智能体，实现从自然语言指令到可验证代码仓库与生产级云部署的端到端转化。该框架融合图工程（graph engineering）、循环工程（loop engineering）与智能体管控工程（agent harness engineering），以可验证的执行证据或运行时证据作为生命周期状态转换的前置条件，强制所有关键步骤（如发布、监控、恢复）必须满足预设断言。当验证失败时，触发受控反思、修复与再验证机制；当检测到运行时异常（如漂移、退化、策略违规）时，可启动边界内自适应调整或回滚。通过在Google Cloud Platform上的实现与评估，实验表明该框架有效防止了非法状态迁移，确保每次运行最终达成经验证的生产部署或可审计的终止失败，显著提升了MLOps系统的安全性与可信赖性。
链接: https://arxiv.org/abs/2608.29615
作者: Sagar Srinivas Sakhinana,Venkataramana Runkana
机构: Tata Research Development and Design Centre; TCS(塔塔咨询)
类目: Multiagent Systems (cs.MA); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Nill
Abstract:Across industries, machine-learning systems support applications ranging from prediction and anomaly detection to forecasting, optimization, and scheduling, yet operationalizing these systems requires coordinating application development, model pipelines, cloud infrastructure, security, deployment, monitoring, retraining, recovery, and rollback. We present an evidence-gated multi-agent framework for transforming a natural-language MLOps cloud engineering task into a verified repository and operational cloud deployment. The framework combines graph engineering, loop engineering, and agent harness engineering. A stateful Graph Orchestrator coordinates specialized agents for repository generation, review, execution, verification, release, and monitoring while governing workflow dependencies, evidence gates, retry bounds, recovery paths, and termination. Consequential lifecycle transitions proceed only when their required predicates are supported by verifiable execution or runtime evidence. Verification failures activate bounded reflection, repair, and re-verification, while runtime evidence of failure, drift, degradation, or policy violation can trigger bounded adaptation, recovery, or rollback. Agent harness engineering constrains repository generation, review, and repair, artifact execution, and cloud operations through controlled capabilities and isolated execution environments. We realize the framework on Google Cloud Platform and evaluate repository completeness, controlled execution, evidence-gated transitions, cloud promotion, and bounded recovery. Our experimental results show that the framework prevents unsupported lifecycle transitions and drives each run toward either a verified operational deployment or an auditable terminal failure.
[MA-10] owards a Systems Foundation for Agent ic Skills: Architecture Lifecycle and Security
【速读】：该论文旨在解决自主大语言模型（LLM）在执行复杂、长周期任务时面临的可靠性不足、上下文消耗过载及执行稳定性差等核心问题。传统单一提示工程与无状态工具调用范式难以适应规模化需求，因此论文提出以“代理技能（agentic skills）”为核心解决方案，其关键在于将执行知识外部化为可重用、可执行且可移植的模块化过程抽象（modular procedural abstractions），从而实现认知规划与确定性执行环境之间的有效衔接。通过构建涵盖九个阶段的统一系统架构——包括自主发现、技能创作与表示格式、记忆存储、动态检索与路由、组合与编排、执行与修复、终身学习、实证评估及安全治理——论文建立了代理技能生态系统的参考框架，并深入探讨了市场机制、公共注册表以及新兴对抗性威胁，同时提出了运行时验证与防御策略。研究进一步在软件工程、操作系统导航、具身机器人和科学发现等领域对系统实现进行分类分析，指出了持续学习与基准真实性等关键开放挑战，确立了代理技能作为构建可扩展、鲁棒且可验证的自主语言代理的基础范式。
链接: https://arxiv.org/abs/2608.29596
作者: Sanket Badhe,Deep Shah,Priyanka Tiwari,Nehal Kathrotia
机构: Google LLC(谷歌); Purdue University(普渡大学)
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Multiagent Systems (cs.MA)
备注:
Abstract:Autonomous large language model (LLM) agents increasingly face reliability, context consumption, and execution stability bottlenecks when deployed on complex, long-horizon tasks. While monolithic prompt engineering and stateless tool-calling paradigms struggle to scale, the field is rapidly converging toward \emphagentic skills: modular procedural abstractions that externalize execution knowledge into reusable, executable, and portable artifacts. This paper establishes a unified systems foundation and reference architecture for the agentic skills ecosystem. We formalize skills as externalized procedural knowledge bridging high-level cognitive planning with deterministic execution environments, and systematically delineate the architecture across a nine-stage lifecycle: autonomous discovery, authoring and representation formats, memory storage, dynamic retrieval and routing, composition and orchestration, execution and repair, lifelong adaptation, empirical evaluation, and security governance. We further examine marketplace dynamics, public registries, and emerging adversarial threat vectors, alongside runtime verification and defense mechanisms. Finally, we categorize system implementations across software engineering, operating system navigation, embodied robotics, and scientific discovery, while highlighting critical open challenges in continual learning and benchmark realism. This work establishes agentic skills as a foundational paradigm for building scalable, robust, and verifiable autonomous language agents.
[MA-11] Asynchronous Cooperative Online Learning for Multi-Robot Control under Computational Delays
【速读】：该论文旨在解决多智能体系统（Multi-Agent Systems, MASs）在不确定环境下因外部扰动和动态模型不准确导致的协同性能与可靠性下降问题。现有分布式高斯过程（Gaussian Process, GP）学习方法通常忽略智能体间计算能力差异及查询点异质性所引发的异步延迟问题，从而影响信息聚合的准确性与实时性。为此，本文提出一种显式考虑预测精度、查询点差异及通信延迟的异步协同学习策略，并设计基于伴随多智能体系统的分布式控制律，以保证期望的控制性能。其解决方案的关键在于通过引入异步信息融合机制与伴随系统理论，实现对动态延迟与异构任务的鲁棒补偿，显著提升了系统在非理想通信条件下的学习与控制协同性能。
链接: https://arxiv.org/abs/2608.29562
作者: Xiaobing Dai,Zewen Yang,Wei Ren,Sandra Hirche
机构: Technical University of Munich (慕尼黑工业大学); University of California at Riverside (加州大学河滨分校)
类目: Machine Learning (cs.LG); Multiagent Systems (cs.MA); Robotics (cs.RO); Systems and Control (eess.SY)
备注:
Abstract:Ensuring the safe operation of multi-agent systems (MASs) under uncertain environments is crucial for cooperative robotic, where external disturbances and inaccurate dynamic models can significantly compromise performance and reliability. To address this challenge, calibrated machine learning models, particularly Gaussian process (GP) regression, are extensively employed due to their interpretable performance quantification. As the interconnected communication of MASs facilitates cooperative learning, agents are able to enhance learning performance by exchanging local GP inferences with their neighbors and aggregating the received information via distributed GP strategies. However, variations in computational power and prediction tasks among agents inevitably lead to heterogeneous computational delays and differences in query points, which are often overlooked in existing aggregation methods. To overcome these limitations, this work proposes an asynchronous cooperative learning strategy that explicitly accounts for prediction accuracy, query point variations and delay effects. Additionally, a distributed control law based on an adjoint MAS is developed to ensure the desired control performance. Simulations on unmanned surface vehicles validate the effectiveness of the proposed approach, demonstrating substantial improvements in both learning and control performance compared to the state-of-the-art approaches.
[MA-12] MedCache: Efficient and Temporally Valid Memory for Longitudinal Clinical Agents
【速读】：该论文旨在解决纵向临床智能体在多访次、多专科患者数据中如何有效构建与维护动态患者状态记忆的问题。现有研究对智能体记忆设计在长时程医疗场景下的具体实现方式尚不明确，尤其缺乏对跨时间、跨专科证据整合与推理机制的系统评估。为此，作者提出一个涵盖多访次、多专科患者记录的基准测试，用于评估长上下文证据检索、跨时间证据聚合以及跨专科临床推理能力。基于该基准，系统性地分析了四种记忆设计要素：数据筛选（curation）、组织方式（organization）、检索策略（retrieval）及记忆增强型推理（memory-augmented reasoning）。研究发现，时间有效性（temporal validity）比单纯保留更长时间历史更为关键；专科因子化记忆虽可减少上下文长度，但可能掩盖跨专科共享信息；多智能体协作在需要联合推理时才具有优势，而非仅因证据来源分散。据此，作者提出\textitMedCache——一种混合式框架，其核心在于构建具有时间有效性的患者记忆，将证据以重叠的专科视图进行组织，根据查询动态路由至相关记忆模块，并自适应调用单个或多个专科专家智能体。实验表明，MedCache在推理准确率和记忆效率方面均优于主流单智能体与多智能体基线模型，且具备良好的模型架构与外部数据集泛化能力。
链接: https://arxiv.org/abs/2608.29528
作者: Hei Ting(Una)Chan,Chenwei Wu,Xueshen Liu,Boyuan Zheng,Liyue Shen,Jiasi Chen,Z. Morley Mao
机构: 未知
类目: Machine Learning (cs.LG); Distributed, Parallel, and Cluster Computing (cs.DC); Multiagent Systems (cs.MA)
备注:
Abstract:Longitudinal clinical agents must maintain an evolving patient state from evidence distributed across visits, time points, and specialties. However, how agent memory should be designed for this setting remains unclear. We introduce a benchmark of multi-visit, multi-specialty patient records that evaluates long-context evidence retrieval, cross-time evidence aggregation, and cross-specialty clinical reasoning. Using this benchmark, we systematically study four memory design choices: curation, organization, retrieval, and memory-augmented reasoning. We find that temporal validity is more important than simply retaining more history; specialty-factorized memory reduces context but can hide shared evidence; and multiple agents help when specialists must reason together, not merely when evidence comes from multiple memories. Guided by these findings, we propose \textitMedCache, a hybrid framework that constructs temporally valid patient memory, organizes evidence into overlapping specialty views, routes each query to relevant memories, and adaptively invokes one or multiple specialists. Experiments show that MedCache improves reasoning accuracy and memory efficiency over strong single-agent and multi-agent baselines, while generalizing across model backbones and external datasets.
[MA-13] Generalizable Multi-Agent Planning from Signal Temporal Logic Specifications via Diffusion
【速读】：该论文旨在解决多智能体系统（如无人机编队、自动驾驶汽车、仓库机器人）在实际应用中同时满足复杂时序任务目标与避免碰撞的挑战。现有基于信号时序逻辑（STL）的规划方法存在根本性权衡：优化方法虽可处理任意STL规范，但随智能体数量增加而面临严重的可扩展性问题；学习方法虽具备高可扩展性和快速规划能力，却在部署阶段面对训练时未见过的目标规范时表现不佳，且不支持异构规范（即不同智能体分配不同规格）或需多智能体协同的团队级规范。为突破这一局限，本文提出一种基于扩散模型的新型多智能体STL规划方法。其关键在于引入可微分近似后的STL语义，并将STL梯度嵌入去噪过程，使模型在保持学习方法高可扩展性的同时，具备对训练期间目标区域范围内任意位置新出现的谓词表达式进行泛化的能力。该方法支持异构规范建模，并通过扩散模型天然生成多样化轨迹，显著降低智能体间的安全冲突（如碰撞）。实验验证表明，所提方法在通用性、可扩展性和轨迹多样性方面均表现出优越性能。
链接: https://arxiv.org/abs/2608.29490
作者: Joe Eappen,Zikang Xiong,Shreyash S. Iyengar,Suresh Jagannathan
机构: Purdue University (普渡大学)
类目: Multiagent Systems (cs.MA); Artificial Intelligence (cs.AI); Robotics (cs.RO)
备注: Accepted for publication in IEEE Robotics and Automation Letters (RA-L), 2026
Abstract:Multi-agent systems in the real-world (e.g., drone swarms, autonomous cars, warehouse robots) must satisfy rich, temporal tasks while avoiding collisions. Signal Temporal Logic (STL) elegantly encodes such objectives, but current STL planning methods face critical limitations. State-of-the-art optimization-based approaches can handle arbitrary STL specifications but struggle with scalability, becoming computationally impractical as the number of agents grows. Learning-based methods efficiently handle a large number of agents with rapid planning times but fare poorly when deployment-time objectives differ from those used during training, and do not support planning tasks that require different specifications to be ascribed to different agents (i.e., heterogeneity) or team-level specifications requiring coordination of multiple agents. This fundamental trade-off between generalizability and scalability presents a challenge for realizing multi-agent STL planning algorithms in practice. To overcome this challenge, we introduce a new diffusion method for multi-agent planning with STL specifications. Using a differentiable approximation of STL, we integrate the STL gradient in the denoising process, making our approach generalizable to novel formulas whose predicates are placed anywhere within the goal region covered during training, while achieving the same scalability as existing learning-based methods. Our method supports heterogeneous specifications, and by using diffusion models, naturally enhances plan diversity, thereby significantly reducing safety-related violations (e.g., collisions) among agents. A detailed evaluation study justifies the utility of STL-guided diffusion-based multi-agent planners for constructing generalizable, scalable, and diverse plans. Videos and code are available at this https URL and this https URL .
[MA-14] Fully Distributed GNE Algorithms for Multi-Robot Placement without Consensus on Multipliers
【速读】：该论文旨在解决非合作博弈中存在共享约束时的广义纳什均衡（Generalized Nash Equilibrium Problem, GNEP）求解问题，尤其针对具有线性等式共享约束的强单调博弈场景。传统方法依赖拉格朗日乘子的交换以实现共识型变分广义纳什均衡（variational GNE, v-GNE）的计算，导致通信开销大且隐私保护能力弱。本文提出一种完全分布式的连续时间算法，无需交换拉格朗日乘子即可收敛至任意广义纳什均衡（GNE），显著降低了通信复杂度并增强了系统隐私性。同时，研究还提供了相应的离散时间实现方案，并在多机器人布放任务中验证了所提方法的有效性与鲁棒性。
链接: https://arxiv.org/abs/2608.29388
作者: Shao-An Yin,Mingyi Hong,Nicola Elia
机构: University of Minnesota (明尼苏达大学)
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computer Science and Game Theory (cs.GT); Multiagent Systems (cs.MA); Robotics (cs.RO)
备注: 6 pages, 3 figures. Published in the 2026 American Control Conference (ACC), pp. 3633–3638
Abstract:Recent machine learning research has increasingly focused on equilibrium analysis in non-cooperative games rather than solely on optimal solutions. Many such problems involve shared constraints and can be formulated as Generalized Nash Equilibrium Problems (GNEPs). For strongly monotone games, existing methods compute consensus-based variational GNEs (v-GNEs) by exchanging Lagrange multipliers. We propose a fully distributed continuous-time algorithm for shared linear equality constraints that converges without multiplier exchange and reaches any GNE, reducing communication overhead and improving privacy. Discrete-time schemes are also provided, and the method is validated on a multi-robot placement task.
[MA-15] GuardianAgent : Policy-Conditioned Risk-Adaptive Anonymization with Verified Adversarial Escalation
【速读】：该论文旨在解决实时网络流量中隐私保护的精准性与动态适应性问题，即现有方法仅依赖检测敏感片段而无法根据目标站点的隐私政策动态评估行为合规性，导致过度或不足的匿名化处理。其核心挑战在于如何在保障隐私的前提下实现细粒度、可验证的自适应数据重写，并避免生成式模型（Generative AI）带来的不可控风险。解决方案的关键在于提出GuardianAgent框架，通过结构化的风险评估机制——自适应多因子风险评分公式（AMRSF），显式整合政策违规可能性、数据敏感性、接收方传输、目的正当性、上下文依据及政策透明度等六维因素，而非直接依赖大语言模型（LLM）进行风险判断，从而确保决策的可解释性与可控性。在此基础上，系统采用证据驱动的快速路径处理高置信度策略匹配，仅在不确定性场景下调用LLM慢路径，显著提升效率；在重写阶段引入基于验证对抗猜测器的五级分层机制，仅当原始文本支持时才触发升级，有效防止幻觉导致的过量匿名化。实验表明，GuardianAgent在法律文本（TAB）、Reddit帖子（SynthPAI）和多格式合成敏感信息（PII-Masking-300k）三个基准上均达到最优隐私-效用权衡，且在所有领域均实现超过0.90的隐私得分，同时在不同接收方、目的、上下文及政策透明度条件下展现出一致的差异化响应能力，验证了其在复杂场景下的鲁棒性与精细化控制能力。
链接: https://arxiv.org/abs/2608.29251
作者: Ruiyi Yang,Gayathri Lihinikaduarachchi,Rahat Masood,Flora D. Salim,Salil S. Kanhere
机构: UNSW Sydney(新南威尔士大学)
类目: Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC); Multiagent Systems (cs.MA)
备注: 17 pages, 3 figures
Abstract:Privacy protection for live web traffic requires more than detecting private spans. Agent-based privacy protection systems must determine whether an outgoing action complies with the destination site’s privacy policy, then apply only the level of rewriting or sanitisation justified by the residual disclosure risk. We present GuardianAgent, a policy-conditioned anonymization framework that couples structured risk assessment with verified adaptive rewriting. GuardianAgent computes risk through AMRSF (Adaptive Multi-factor Risk Scoring Formula), an explicit controller that combines policy-violation likelihood with data sensitivity, recipient transmission, purpose legitimacy, contextual basis, and policy transparency, rather than relying on an LLM to assign risk directly. This risk score determines both the allow/transform/deny decision and the initial anonymization level. For efficiency, GuardianAgent uses an evidential fast path for low-uncertainty policy matches and invokes an LLM slow path only for uncertain cases. For rewriting, it applies a five-level hierarchy driven by a verified adversarial guesser: guesses trigger escalation only when supported by the original text, preventing hallucinated attacker confidence from causing unnecessary over-anonymization. Experiments across three benchmarks spanning legal text (TAB), Reddit posts (SynthPAI), and multi-format synthetic PII records (PII-Masking-300k) show that GuardianAgent achieves the strongest privacy-utility trade-off among published baselines and is the only method to reach more than 0.90 privacy in all three domains, remaining robust under a backbone switch. Action-context stress tests further show that the same outgoing text receives different decisions and anonymization strengths under different recipients, purposes, action bases, and policy-transparency conditions.
[MA-16] Localizing Emergent Failures in Agent ic AI: Recovering Minimal Repair Families via Counterfactual Replay
【速读】：该论文旨在解决多大语言模型（Large Language Model, LLM）代理在协作过程中因消息交互导致的智能体系统（agentic AI systems）失败问题，尤其关注无法通过逐点归因（pointwise attribution）区分联合必要修复（jointly necessary repair）与单一替代修复（singleton repairs）的挑战。其核心解决方案是提出最小修复族恢复（Minimal Repair Family Recovery, MRFR）这一形式化框架，目标为在指定规模约束下，恢复所有包含最小事件集合的反事实可修复集，以实现任务成功的回放重建。关键创新在于提出图约束联合重播（Graph-Constrained Joint Replay, GCJR）方法：该方法从执行依赖图（execution dependency graph）中提取与故障相关的事件，构建符合图结构的单事件与双事件候选修复集，并通过与配对的干净对照样本进行重播验证。在120个有向无环图（DAG）控制基准中的90个有效案例上，GCJR实现了1.000的家族精确匹配率（Family Exact Match），同时将平均重播调用次数从56.3次降至25.3次（降低55.1%），显著优于穷举搜索；在含4个代理的24例真实场景试点中，同样达到1.000的家族精确匹配率，且将平均模型调用次数从21.0次减少至10.0次（降低52.4%）。值得注意的是，仅依赖单事件重播会遗漏关键的联合必要修复，凸显了联合分析的必要性。因此，该方案的核心优势在于结合图结构约束与联合重播验证，实现了高效、精确的多代理系统故障修复归因。
链接: https://arxiv.org/abs/2608.29228
作者: Bingjie Li,Yumeng Song,Zhongming Yao,Tianyi Li
机构: Northeastern University (东北大学); Aalborg University (奥尔堡大学); Zhejiang University (浙江大学)
类目: Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注: 6 pages, conference paper
Abstract:Failures in agentic AI systems can arise from interactions among messages exchanged by multiple large language model (LLM) agents. Pointwise attribution cannot distinguish a jointly necessary repair from alternative singleton repairs. We formulate Minimal Repair Family Recovery (MRFR): recovering all inclusion-minimal event sets whose counterfactual replay restores task success within a declared size bound. We propose Graph-Constrained Joint Replay (GCJR), which slices failure-relevant events from an execution dependency graph, constructs graph-feasible singleton and pair candidates, and verifies them by replay with paired clean counterparts. For fixed replay outcomes, GCJR is exact within its declared graph domain. On 90 in-scope cases from a 120-DAG controlled benchmark, GCJR achieves 1.000 Family Exact Match while reducing mean replay calls from 56.3 to 25.3 (55.1%) relative to exhaustive search. On a 24-case, four-agent LLM pilot, it again achieves 1.000 Family Exact Match and reduces mean model calls from 21.0 to 10.0 (52.4%); single-event replay misses jointly necessary repairs.
[MA-17] ASTRA - Agent ic System for Ticket Resolution and Analysis
【速读】：该论文旨在解决技术运维团队在处理海量故障工单时，因碎片化证据来源（如工单文本、历史案例、系统日志和技术文档）导致的诊断效率低、可验证性差的问题，尤其当关键信号在多源数据中稀疏分布时，现有自动化方法因缺乏显式的证据建模与溯源机制而难以生成可信的故障排查报告。其解决方案的核心是提出ASTRA——一种基于代理（agent）架构的智能工单解析系统，通过中心协调器（OrchestratorAgent）协同三个专业化信息获取代理：用于检索历史相似案例的TicketSimilarityAgent、通过确定性过滤与受限大模型分析将海量日志压缩为结构化引用型结论的LogAgent，以及基于模型上下文协议（Model Context Protocol, MCP）获取领域知识的DomainKnowledgeAgent。三者输出被统一转化为“主张-证据”表示形式，每个主张均关联原始文本片段并标注支持等级，有效防止跨源误引。随后，JudgeAgent对报告进行五维度评分，低分触发由OrchestratorAgent生成针对性追问以实现有界迭代优化。实验表明，ASTRA在7个产品线共987个真实电信故障工单上取得4.13/5.0的平均质量分，59.9%的报告能精确定位至组件族级别及更细粒度故障区域，相关性与清晰度得分分别达4.88和4.94，虚构技术细节占比低于3%。分类型分析显示，硬件类故障诊断难度显著高于软件或配置类故障（Cohen’s d=0.80），揭示了纯文本证据通道在硬件故障诊断中的根本性局限。
链接: https://arxiv.org/abs/2608.28790
作者: Shashidhar Reddy Javaji,Mohamed Trabelsi,Jin Cao,Huseyin Uzunalioglu
机构: Stevens Institute of Technology (斯蒂文斯理工学院); Nokia Bell Labs (诺基亚贝尔实验室)
类目: Multiagent Systems (cs.MA); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR); Machine Learning (cs.LG); Software Engineering (cs.SE)
备注:
Abstract:Technical operations teams resolve large volumes of incidents by synthesizing fragmented evidence from ticket text, historical cases, system logs, and technical documentation. Existing automation often relies on monolithic generation without explicit evidence modeling or provenance, making outputs difficult to verify when critical signals are sparse across sources. We propose ASTRA, an agentic system for ticket resolution in which a central orchestrator coordinates three specialist information-gathering agents and drives a judge-orchestrator refinement loop to produce evidence-backed troubleshooting reports. TicketSimilarityAgent retrieves relevant historical precedents through dense retrieval and LLM reranking; LogAgent distills hundreds of thousands of log lines into structured, quote-grounded findings using deterministic filtering and constrained LLM analysis; and DomainKnowledgeAgent retrieves relevant technical knowledge via the Model Context Protocol (MCP). Their outputs are transformed into a claim-evidence representation linking each claim to a verbatim source passage, assigning a support level, and preventing cross-attribution. A JudgeAgent scores the report on five criteria, while the OrchestratorAgent converts low scores into targeted follow-up queries for bounded iterative refinement. Evaluated on 987 real-world telecom fault tickets across seven product lines, ASTRA achieves a mean quality score of 4.13/5.0, with 59.9% of reports identifying the fault area at the component-family level or better. Relevance and Clarity scores are 4.88 and 4.94, respectively, while fabricated technical details remain below 3% of error cases. Stratification by fault type reveals that hardware faults remain substantially harder than software or configuration faults (Cohen’s d=0.80), pointing to a fundamental limitation of text-based evidence channels for hardware fault diagnosis.
[MA-18] Peer Oversight in Collective Decision Making
【速读】：该论文旨在解决多智能体决策系统中责任归属不清的问题，特别是在发生有害结果时，如何确保至少有k名参与者对后果负责。其核心解决方案是引入“同行监督（peer k-oversight）”机制，该机制要求在序列式集体决策过程中，每一个有害结果都必须由至少k个代理共同承担责任。研究证明，只要通过重新分配决策控制权即可实现peer k-oversight，那么仅需k个代理即可达成该目标；同时提出了一种多项式时间算法，用于判断是否存在满足条件的控制权重分配方案，并在存在时构造出具体方案。这些成果将peer k-oversight确立为多智能体决策系统中可计算、可实施的责任设计原则。
链接: https://arxiv.org/abs/2608.28754
作者: Sarah Mohsen,Pavel Naumov
机构: University of Southampton (南安普顿大学)
类目: Computer Science and Game Theory (cs.GT); Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注:
Abstract:This article introduces peer k -oversight, a property of sequential collective decision mechanisms requiring at least k agents to be responsible for every harmful outcome. It is shown that whenever k -oversight can be achieved by redistributing control over the decisions in a mechanism, it can be achieved using just k agents. A polynomial-time algorithm is also presented that determines whether such a redistribution exists and, when it does, constructs one. These results establish peer oversight as a tractable design principle for multiagent decision-making systems.
[MA-19] Goal Staying Makes Sum-of-Costs Anonymous Multi-Agent Path Finding NP-Hard
【速读】：该论文旨在解决带目标停留（goal-staying）的匿名多智能体路径规划（Anonymous Multi-Agent Path Finding, AMAPF）中最小化总代价（sum-of-costs, SoC）这一关键问题。传统上，当智能体到达目标后即消失（disappearing agents）时，可通过多项式时间的网络流算法求解包括SoC在内的多种目标；然而，当智能体在抵达目标后需持续停留在目标位置（goal-staying）时，问题性质发生根本性变化。其核心挑战在于：尽管可基于标准时间展开流模型引入目标安置约束（goal-settlement constraints）以建模该情形，但该模型对应的线性规划松弛解并非整数解，导致无法直接通过松弛求解获得可行整数解。进一步地，论文通过从3-SAT问题的归约，证明了在目标停留设定下最小化SoC是NP-hard的。这一结果与消失型智能体情况下多项式可解形成鲜明对比，从而确立了智能体是否在达成目标后停留作为决定问题复杂性边界的关键因素。
链接: https://arxiv.org/abs/2608.28658
作者: Hang Ma
机构: Simon Fraser University (西蒙菲莎大学)
类目: Multiagent Systems (cs.MA); Computational Complexity (cs.CC); Robotics (cs.RO)
备注:
Abstract:Anonymous Multi-Agent Path Finding (AMAPF) admits polynomial-time network-flow algorithms for several objectives, including makespan, total distance, and sum-of-costs (SoC) when agents disappear upon reaching goals. We show that standard goal-staying AMAPF is fundamentally different. We first formulate SoC minimization by augmenting the standard time-expanded flow model with goal-settlement constraints and show that the resulting linear programming relaxation is non-integral. We then prove that minimizing SoC in goal-staying AMAPF is NP-hard via a reduction from 3-SAT. Together with the polynomial-time result for the disappearing variant, this establishes a sharp complexity boundary determined by whether completed agents remain at their goals.
[MA-20] Provably Safe Decentralized Contingency MPC under State-Only Information and Limited Sensing for Nonlinear Multi-agent Systems
【速读】：该论文旨在解决多智能体系统在仅依赖状态信息（state-only information pattern）下的分布式故障容错模型预测控制（contingency MPC）问题，尤其关注有限感知范围与即插即用（plug-and-play）操作下的递归可行性、安全性及李雅普诺夫型收敛性保障，同时降低局部交互设计中的保守性。其解决方案的关键在于引入基于智能体的备用区域（fallback regions，即安全集），确保每个智能体在安全集中始终存在可行的应急控制动作以抵达安全平衡点；并提出一种新颖的安全集更新机制，能够在不重构邻近智能体几何结构的前提下，实现更宽松的分布式交互约束，从而支持无记忆的本地交互和有限感知范围，同时保持全局递归可行性与安全性的理论保证。该方法维持了共享首输入（shared-first-input）的故障容错MPC结构，且完全去中心化，仿真结果验证了其在高密度多智能体场景中的有效性。
链接: https://arxiv.org/abs/2608.30874
作者: Max Studt,Georg Schildbach
机构: 未知
类目: Optimization and Control (math.OC); Multiagent Systems (cs.MA); Robotics (cs.RO); Systems and Control (eess.SY)
备注:
Abstract:This paper considers decentralized contingency MPC for multi-agent control under a state-only information pattern, with particular focus on limited sensing and plug-and-play operation. The objective is to retain recursive feasibility, safety, and Lyapunov-type convergence while reducing conservatism in local interaction handling. The framework relies on agent-wise fallback regions (safe sets) in which a feasible contingency maneuver to a safe equilibrium is always available. A novel safe-set update mechanism is introduced that supports less conservative decentralized interaction while preserving the underlying guarantees. This, in turn, enables memory-free local interaction and finite sensing ranges without requiring agents to reconstruct the exact neighbor geometry. The resulting scheme remains fully decentralized and preserves the shared-first-input contingency MPC structure. Theoretical guarantees and simulation results illustrate the effectiveness of the approach in dense multi-agent scenarios.
[MA-21] Sustained Heterogeneity: an emergent collective mechanism in LLM -driven traffic
【速读】：该论文旨在解决生成式 AI（Generative AI）在物理多智能体系统中作为闭环控制器时，其涌现的集体动态行为尚不明确的问题。研究通过部署22个大语言模型（LLM）智能体，在230米环形道路中以0.5秒为周期实时控制目标车速（以IDM模型作为防碰撞约束），成功复现了类人类的停走波现象。在系统排除了随机性（白噪声、OU噪声、温度变化）、种群差异及动力学不稳定性（延迟、OV模型）等干扰因素后，发现一种名为“持续异质性”（Sustained Heterogeneity, SH）的新集体机制：即每周期内LLM所选择的目标速度调整值存在持续且近似温度无关的偏差（在6倍温度范围内波动约8%），该偏差通过“漂移—间隙侵蚀—非线性制动”三阶段级联过程传播。研究进一步揭示，在不同交通密度下，临界LLM渗透率 $ p_c $ 随密度单调下降，从43.5 veh/km时无相变过渡至95.7 veh/km时 $ p_c \approx 0.23 $，符合由触发距离、随机性与车队规模共同决定的启动阈值模型。对39,600次决策的链式思维分析表明，尽管智能体表现出多因素安全推理能力，但系统性偏差仍持续存在，暗示稳定性必须在动力学层面主动施加。本研究首次识别出LLM控制交通中未被描述的集体机制，并绘制了密度依赖的相边界 $ p_c(\rho) $，为未来智能体协同系统的稳定性设计提供了关键理论依据。
链接: https://arxiv.org/abs/2608.29174
作者: Yujun Qi,Yangyang Guan
机构: Nanjing Tech University (南京工业大学)
类目: Physics and Society (physics.soc-ph); Multiagent Systems (cs.MA); Adaptation and Self-Organizing Systems (nlin.AO)
备注: 41 pages, 5 figures
Abstract:Large language models (LLMs) are increasingly adopted as closed-loop controllers in physical multi-agent systems, yet their emergent collective dynamics remain incompletely characterised. We deploy 22 LLM agents as direct, real-time target-speed controllers (per 0.5 s cycle, with IDM as collision-avoidance clamp) on a 230 m ring road under the Sugiyama 2008 paradigm, reproducing human-like stop-and-go waves. Six matched controls spanning stochasticity (white noise, OU noise, temperature), population variance, and dynamical instability (delay, OV model) are systematically excluded. The surviving phenomenon, termed Sustained Heterogeneity (SH), is the persistent, approximately temperature-insensitive (approx. 8 percent across a 6x T sweep), per-cycle divergence in LLM-chosen target-speed adjustments, propagating through a three-stage cascade of drift, gap erosion, and nonlinear braking. Across four traffic densities, the critical LLM penetration fraction p_c decreases monotonically from no transition at density 43.5 veh/km to p_c approx 0.23 at density 95.7 veh/km, consistent with an initiation-threshold model governed by trigger distance, stochasticity, and fleet size. Chain-of-thought analysis of 39,600 decisions across three seeds shows agents engage in multi-factor safety reasoning, yet systematic divergence persists, implying stability must be enforced at the dynamics layer. This is the first study to identify a previously uncharacterised collective mechanism in LLM-controlled traffic and map a density-dependent phase boundary p_c(rho).
[MA-22] HALO: A Physics-Aware LLM Agent Framework for Nanophotonic Design
【速读】：该论文旨在解决生成式人工智能在纳米光子学设计中如何可靠地将光学目标转化为可仿真的设计、执行电磁场分析并根据数值反馈迭代优化的问题。其核心挑战在于现有语言模型在复杂物理系统设计中缺乏对物理规律的显式感知与可复用的经验积累。为此，研究提出HALO框架，一个融合语言模型规划器、类型化设计规范、电磁仿真、诊断评估及历史失败轨迹重用的物理感知迭代设计系统。关键创新在于通过结构化工作流与自主编码代理之间的对比，揭示了显式接口（如固定结构化流程）在降低代码级与路径级错误方面的优势，而自主执行虽在强大模型下提升任务成功率，但伴随更高的操作失败风险；同时，引入基于失败轨迹的反馈重用机制，在多轮任务中显著减少首次成功所需的迭代次数与总令牌消耗。研究进一步构建了涵盖实验、文献与开放任务的HALO-Bench基准，统一评估协议下验证了不同策略的性能权衡，明确了物理约束建模、自主执行能力与设计经验复用之间的内在关系。
链接: https://arxiv.org/abs/2608.28877
作者: Yubo Zhang,Jinlin Xiang,Zijun Zhao,Yang Zhao,Eli Shlizerman,Arka Majumdar
机构: 未知
类目: Optics (physics.optics); Multiagent Systems (cs.MA)
备注: 9 pages, 3 figures, 5 charts
Abstract:Language models have recently been applied to nanophotonic design, but it remains unclear whether they can reliably translate optical objectives into simulation-ready designs, execute electromagnetic analysis, and revise decisions from numerical feedback. We introduce HALO, a physics-aware framework that couples language-model planners with typed design specifications, electromagnetic simulation, diagnostic evaluation, and optional reuse of prior failure trajectories in an iterative design loop. We further introduce HALO-Bench, a 52-task benchmark spanning lab-derived, paper-derived, and open-ended nanophotonic design tasks under a shared evaluation protocol. We compare three planner configurations: a Fixed Structured Workflow, an Autonomous Structured Agent using the same simulation interface, and an Autonomous Coding Agent that directly writes and executes simulation code. The Fixed Structured Workflow is the most token-efficient and exhibits no observed code- or path-level failures, while autonomous coding can achieve higher task success with stronger models at the cost of additional operational failures. We also study reuse of prior failed trajectories. On targeted multi-round tasks, retrieved failure feedback reduces both iterations to first success and total token use. These results clarify the tradeoffs between explicit interfaces, autonomous execution, and reusable design experience in scientific agents.
[MA-23] Measuring Collective Semantic Change in Populations of Language Model Agents
【速读】：该论文旨在解决语言模型代理群体中集体语义变化的可测量动态现象问题，即如何在不干预的前提下，系统性地观测和量化语言内容随时间推移所发生的语义演化。其核心解决方案是提出一种被动纵向观测工具Kopterix，该工具通过预先设定的协议对代理群体的语言环境进行周期性采样，并将每次采样所得的内容按发布年龄划分为表层（surface）、中段（mid-stream）与残留层（residue），从而实现对不同内容年龄区间语义差异的精确测量。该方法的关键在于通过多层级分层结构与严格的对照实验（如时间戳随机打乱检验），分离出真实的时间依赖性语义变化信号，避免混淆因素干扰；在词法层面，利用稀疏熵揭示了4月至5月间高频词分布均匀性的变化；在几何层面，通过全局均值中心化识别出共享嵌入方向的尺度，并验证中段与残留层间距离的持续异常；在时间层面，发现去趋势标量与层中心点在数小时内显著偏离，且长期相关性呈弱正向衰减，无明显周周期规律。所有分析结果均受限于其对应控制条件的支持范围，且该设计可广泛应用于任何能重复观测、具备时间戳语言产出的代理群体环境。
链接: https://arxiv.org/abs/2608.28644
作者: Elena Kopteva
机构: University of Illinois Urbana-Champaign (伊利诺伊大学厄本那-香槟分校); The Grainger College of Engineering (格兰杰工程学院); Department of Physics (物理系)
类目: Physics and Society (physics.soc-ph); Multiagent Systems (cs.MA); Social and Information Networks (cs.SI)
备注: 43 pages, 13 figures
Abstract:Collective semantic change in populations of language model agents is a measurable dynamical phenomenon. We present a passive longitudinal instrument called Kopterix that observes the semantic state of an agent population as a sequence of bounded observations under a protocol defined before the observations begin. Each observation divides the sampled feed by post age into surface, mid-stream, and residue layers, which makes semantic differences across content age measurable alongside run-to-run change. We validate the instrument on Moltbook, an agent-native social platform, over a two-month window of scheduled observations, with the periodicity check extended across approximately four months. At the lexical level, rarefied entropy resolves an April-May difference in the evenness of the stored top 200 unigram distributions, and adjacent states are lexically closer than states paired after timestamp shuffling. At the geometric level, grand mean centering exposes the scale of a common embedding direction, and scheduled shuffle checks support a recurring excess in the mid-stream to residue separation relative to the shuffled reference. At the temporal level, detrended scalar quantities and centered layer centroids lose much of their similarity over several hours, and a weaker positive component declines across longer separations with no strong weekly recurrence. Several attractive apparent structures failed their controls, and each reading is limited to the level its controls support. The design applies wherever a population of agents produces a timestamped language environment that can be observed repeatedly and divided by content age.
[MA-24] Cognitive Cells: A Compositional Framework for Populations of Small Language Models
【速读】：该论文旨在解决当前生成式人工智能（Generative AI）系统中人工认知（artificial cognition）的分解、度量与组合缺乏统一框架的问题。现有方法往往依赖于特定实验设计，导致集体行为难以被客观评估。为此，作者提出以“认知单元”（cognitive cell）为基本单位构建多智能体系统：即一个参数固定、记忆受限且具备消息接口的小型冻结语言模型。其核心方法论是“固定单元原则”（fixed-cell principle），即保持认知单元本身不变，仅改变种群规模、通信拓扑、消息带宽和协调协议，从而使集体行为成为可测量的、由已知装置决定的属性，而非研究设计的副产物。通过为单个认知单元建立包含可度量参数的紧凑数据表（datasheet），研究系统考察了复制并连接这些单元是否能提升性能——首先测量单个单元独立表现，再测试投票、通信与拓扑结构的作用。在使用15亿与30亿参数的小型冻结模型的初步实验中发现，增加单元数量仅在个体错误不高度相关时才有效；简单的对错投票模型虽为保守基准，但真实开放式的投票机制可能超越该基准，因错误分布更分散。此外，辩论、共享黑板与链式修订等主流交互协议在本设定下未能优于同等成本的投票基线。最后，研究发现认知单元自身能够传递多个事实的能力（作为数据表中的可度量指标），可有效预测群体在证据总量超过单个单元记忆容量时的任务求解能力。这些结果构成了可扩展人工认知研究计划中的初步测量成果，其中多智能体架构被视为足够自主以被视作“代理”的认知单元的特殊情形。
链接: https://arxiv.org/abs/2608.28606
作者: Silvan Ferreira
机构: Federal University of Rio Grande do Norte (联邦大学里奥格兰德州)
类目: Physics and Society (physics.soc-ph); Multiagent Systems (cs.MA)
备注:
Abstract:Recent work on large language models and agentic systems raises a basic question that current practice leaves open: how should artificial cognition be decomposed, measured, and composed? We propose studying multi-agent systems from a fixed unit we call a cognitive cell: a small, frozen language model with bounded memory and a message interface. The methodological commitment, the fixed-cell principle, is to hold this unit constant and vary only the population size, the communication topology, the message bandwidth, and the coordination protocol, so that collective behavior becomes a measurable property of a known device rather than an artifact of per-study engineering. We characterize a single cell by a compact datasheet of measurable parameters, and we ask when replicating and connecting cells improves performance: first we measure how one cell behaves alone, then we replicate it and test when voting, communication, and topology help. Instantiating the framework with small frozen models (1.5 and 3 billion parameters), we report a first round of measurements. Adding cells helps only when their errors are not too correlated. A simple correct/incorrect voting model is a useful but conservative null: real open-ended voting can exceed it, because errors are dispersed across many wrong answers rather than concentrated on one. Popular interactive protocols, namely debate, a shared blackboard, and chain revision, do not beat a matched-cost voting baseline in our setting. Finally, a cell’s ability to relay several facts, itself a datasheet quantity, predicts whether a population can solve tasks whose evidence exceeds any single cell’s memory. We present these as initial measurements within a broader program on scalable artificial cognition, in which multi-agent architectures appear as the special case of cells autonomous enough to be treated as agents.
自然语言处理
[NLP-0] Context-Aware Interleaved Batching for WhisperX
【速读】： 该论文旨在解决现有语音转写模型在保持上下文连贯性与推理效率之间难以兼顾的问题。具体而言，WhisperX虽通过音频段内批处理（intra-audio batching）提升了推理速度，但其将音频片段独立处理导致历史上下文丢失，影响标点和专有名词的准确转写；而标准Whisper虽能保留序列化上下文，却面临推理缓慢及生成幻觉（hallucination）循环等瓶颈。为此，本文提出一种上下文感知的交错批处理（Context-Aware Interleaved Batching）方案，其核心在于利用语音活动检测（VAD）提取的语段边界，对音频分段进行智能划分，并在批处理过程中安全地维持连续的历史上下文信息，从而稳定文本条件生成过程。实验结果表明，该方法在长时音频基准测试中显著降低了词错误率（WER），提升了专有名词识别准确率，同时保持了高吞吐量的推理性能。
链接: https://arxiv.org/abs/2608.31170
作者: Carlos Bain,Max Bain
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:While WhisperX accelerates speech transcription via intra-audio batching, it isolates audio segments, losing the historical context needed for coherent punctuation and terminology transcription. Conversely, standard Whisper retains context sequentially but suffers from slow inference and hallucination loops. To achieve the best of both worlds, we propose Context-Aware Interleaved Batching. By using VAD-derived segment boundaries, our algorithm stabilizes Whisper’s text conditioning, allowing us to safely maintain continuous historical context across batched audio segments. As demonstrated on long-form audio benchmarks, this approach reduces Word Error Rate (WER) and improves proper noun transcription, all while maintaining high-throughput inference speeds.
[NLP-1] DIASENTINEL: An Auditable Multi-Agent System for Guideline-Grounded Diabetes Risk Screening
【速读】： 该论文旨在解决生成式人工智能（Generative AI）在临床决策支持中存在幻觉事实、缺乏依据的推荐以及引用错误等问题，特别是在基于电子健康记录（EHRs）进行1型糖尿病风险筛查时的可靠性与可审计性挑战。其核心解决方案在于构建一个全本地部署的多智能体系统——DIASENTINEL，该系统通过校准的风险预测、确定性的临床信号提取、基于美国糖尿病协会（ADA）指南的互惠排名融合（Reciprocal Rank Fusion），以及结合规则基础检查与大语言模型蕴含推理的混合验证层，实现对临床建议的精准性和可追溯性保障。该系统不仅支持实时批量筛查仪表盘和交互式患者报告界面，还提供被引用的推荐内容、验证结果及原始EHR对比，从而形成可审计、隐私保护且可靠的生成式AI临床决策支持框架。
链接: https://arxiv.org/abs/2608.31128
作者: Yung Wei Shueh,Zhi-Jie Chen,Chia-Hsuan Hsu,Hsin-Ling Hsu,Donghua Zhang,Chenwei Wu,Jun-En Ding,Tongze Zhang,Shihao Yang,Pengfei Hu,Fang-Ming Hung,Feng Liu
机构: Far Eastern Memorial Hospital(远东纪念医院); University of Michigan(密歇根大学); Rutgers University(罗格斯大学); Stevens Institute of Technology(史蒂文斯理工学院)
类目: Computation and Language (cs.CL)
备注:
Abstract:Large language models (LLMs) offer promising clinical decision support but remain vulnerable to hallucinated facts, unsupported recommendations, and citation errors. We present DIASENTINEL, a fully on-premise multi-agent system for one-year type 2 diabetes mellitus (T2DM) risk screening and guideline-grounded report generation from electronic health records (EHRs). The system integrates calibrated risk prediction, deterministic clinical signal extraction, Reciprocal Rank Fusion over American Diabetes Association (ADA) guidelines, and a hybrid verification layer combining rule-based checks with LLM entailment. The demonstration provides a real-time batch-screening dashboard and an interactive patient report interface with cited recommendations, verification results, and raw EHR comparison. DIASENTINEL demonstrates a practical framework for reliable, auditable, and privacy-preserving LLM-based clinical decision support.
[NLP-2] PaperGym: Rubric-Centered Evolution for Research-Plan Generation
【速读】： 该论文旨在解决生成式 AI (Generative AI) 在科研规划（research planning）任务中缺乏可验证奖励信号的问题，即传统强化学习因缺少明确的任务-评价对（task-critic pairs）而难以有效训练。其核心挑战在于现有方法从同一文献内容中同时提取问题与评判标准，导致奖励可被简单改写（paraphrase）获取，且评判标准被压缩为单一标量，造成信息损失与评价偏差。本文提出 PaperGym，一个将每篇科研论文转化为完整训练环境的统一框架，关键创新在于：分离问题与评判标准的来源——问题由研究目标与背景合成，评判标准则源自方法与实验部分，从而实现对方法论创新与实验设计双重维度的多维评估。该设计显著降低标准泄露率（仅3.7%，低于现有数据集11.90%–34.10%），并通过双阶段使用评判标准（先作为特权上下文用于OPS D的自教师机制，再作为GRPO的奖励信号）提升模型训练效率。实验证明，基于PaperGym-20k训练的Qwen3-8B在多个基准上表现优异，超越更大模型如Kimi K2.6，验证了该框架在高质量科研推理能力培养上的有效性。
链接: https://arxiv.org/abs/2608.31119
作者: Yuhan Wang,Zhengxi Lu,Yuchen Yan,Kaitao Song,Wenqi Zhang,Weiming Lu,Jun Xiao,Yueting Zhuang,Yongliang Shen
机构: Zhejiang University(浙江大学); Apple(苹果)
类目: Computation and Language (cs.CL)
备注: 34 pages, 6 figures, 6 tables. Code: this https URL . Project page: this https URL . Dataset: this https URL . Model: this https URL
Abstract:Research planning is the decisive capability of AI scientists. Yet a research plan admits no verifiable answer, so reinforcement learning lacks the environment it requires: tasks paired with a critic. Rubrics extracted from scientific papers can supply the critic. Existing pipelines, however, draw the question and the criteria from the same content, so the reward can be earned by paraphrase. The rubric is further compressed into a single scalar per rollout. We introduce PaperGym, a unified framework that turns each research paper into a complete training environment. PaperGym exploits the structure of a paper: the question is synthesized from the research goal and background, while the criteria are derived from the method and experiments. The criteria span methodological innovation and experimental design, and criterion leakage falls to 3.7%, versus 11.90% to 34.10% in existing datasets. Training uses the rubric twice: first as privileged context for OPSD’s self-teacher, then as the reward for GRPO. Across Qwen3-1.7B/4B/8B, this schedule outperforms supervised fine-tuning, either stage alone, and the reverse ordering, improving five-benchmark averages by +5.6, +5.0, and +4.8 points. With the recipe held fixed, models trained on PaperGym-20k win 58.1% of three-way comparisons, against 28.2% for RubricHub Science. The trained Qwen3-8B reaches 73.48 on ResearchQA, above the far larger Kimi K2.6. We release the pipeline, the 20,000-instance corpus PaperGym-20k, and the benchmarks PaperGym-Innov and PaperGym-Design.
[NLP-3] Aspire: Can Models Self-Evolve from Vague Goals?
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）在自我演化过程中缺乏对模糊目标（vague goal）驱动的自主学习能力的问题。现有研究多依赖人类预先设定的任务与评估指标，使模型演化局限于对明确目标的优化，而无法真正实现“理解目标—识别能力差距—规划学习路径—判断进步”的闭环自主学习。为此，论文提出ASPIRE基准，其核心在于仅提供自然语言形式的能力目标（如“成为更优秀的物理学家”），而隐藏下游评估任务，迫使智能体自主完成目标具象化：包括选择训练数据、设计更新方法、构建训练与验证信号，并决定评估时机。该框架支持模型权重与代理架构（agent-harness）双重演化，并在由专家编写的520个隐藏评估项上进行统一评测，涵盖六个不同目标。实验表明，模糊目标能有效引导搜索策略聚焦于目标解读本身；尽管当前智能体可执行训练与代理编辑循环，但权重层面的改进仍稀疏且不稳定，最强演化出的代理性能仍低于人工设计的Qwen-Agent基准。此外，智能体常因使用不匹配的数据及过度依赖狭隘的自评机制，导致局部提升无法泛化至隐藏评估，持续迭代甚至可能抵消先前进展。因此，解决方案的关键在于建立一个以模糊目标为核心驱动力、具备完整自主性评估与演化机制的统一交互环境，从而推动模型从“优化已有任务”向“自主定义学习方向”演进。
链接: https://arxiv.org/abs/2608.31111
作者: Yuhao Wu,Jingyuan Zhang,Jiajun Shi,Yuxuan Zhang,Xinping Lei,Junting Zhou,Zexuan Wang,Yuchen Wu,Huan Zhou,Duo Wang,Yinzhu Piao,Yongchang Peng,Yunfeng Shi,Jin Chen,Zuo Wang,Jinkai Liu,Jiaheng Liu,Wenxuan Zhang,Shen Yan,Wenhao Huang,Ge Zhang
机构: 未知
类目: Computation and Language (cs.CL)
备注: this https URL
Abstract:Many important forms of human learning begin with a vague goal, such as “become a better physicist” or “improve at research.” Learners must interpret the goal, identify capability gaps, decide how to learn, and determine whether they have actually improved. In contrast, existing work on LLM self-evolution typically begins with tasks and evaluation metrics specified by humans, reducing self-evolution to optimizing an explicit objective rather than deciding what and how to learn. We introduce ASPIRE, a benchmark for vague-goal-driven self-evolution. ASPIRE provides only a natural-language capability goal while downstream evaluation tasks remain hidden. The agent must operationalize the goal by choosing data and update methods, constructing training and validation signals, and deciding when to evaluate. ASPIRE supports both model-weight and agent-harness evolution in a unified interactive environment and evaluates the resulting systems on a hidden, expert-authored set of 520 items spanning six goals. Our experiments show that vague goals redirect search effort toward goal interpretation. Current agents routinely complete training and harness-editing loops, but weight-level gains remain sparse and unstable, and the strongest evolved harness remains below the engineered Qwen-Agent reference. Agents often train on mismatched data and trust narrow self-evaluations, so local gains fail to transfer to hidden evaluation and continued search and training can erase earlier improvements.
[NLP-4] BLOOM-WILT: Logit Tilting for Behaviour Elicitation in Automated LLM Auditing
【速读】： 该论文旨在解决大规模语言模型在实际部署中难以通过传统测试手段发现罕见且危险行为的问题，即测试环境无法模拟部署阶段的海量交互，导致潜在风险行为（如自我伤害诱导）难以被有效识别。其核心挑战在于现有自动化审计工具虽具备可扩展性与灵活性，但因缺乏优化压力而采样效率低下。本文提出BLOOM-WILT全链路审计框架，其关键创新在于：在输入端，审计模型通过多轮对话策略自适应调整，基于历史评分反馈持续优化交互策略；在输出端，利用目标模型自身的下一个词分布，对解码过程进行动态加权，使与特定行为相关的生成内容在同等概率下优先被采样。该方法无需额外训练或访问权限，仅依赖目标模型的原始输出分布即可高效诱发稀有行为。实验表明，在4个目标模型和8种行为上，WILT在32组测试中有30组优于基线审计器，并重新评估了模型安全性排名；例如在对Qwen3.5-4B的自我伤害诱导行为检测中，行为出现率从51%提升至100%，显著优于所有迁移至相同计算预算下的其他诱发方法，且未降低生成概率下限。
链接: https://arxiv.org/abs/2608.31105
作者: Adrians Skapars,Edoardo Manino
机构: University of Manchester(曼彻斯特大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 10 pages, 5 figures, 3 tables. Code: this https URL
Abstract:Users of a deployed language model routinely encounter behaviours that testing almost never surfaces, since deployment puts the model through orders of magnitude more interactions than any evaluation can simulate. Automated auditors make testing cheap to scale and flexible enough to cover almost any specified behaviour, yet their lack of optimisation pressure makes them sample-inefficient. To address this shortcoming, we introduce BLOOM-WILT, a full auditing pipeline that elicits natural multi-turn instances of rare behaviours, without training cost or access beyond the target’s next-token distribution. On the input side, WILT’s auditor model revises its conversational strategy across rounds, learning from previous scored interactions. On the output side, WILT adaptively reweights the target’s decoding using the model’s own distribution conditioned on an elicitation prompt, so that behaviour-relevant generations are sampled ahead of others it finds equally probable when unprompted. We evaluate WILT across 4 target models and 8 behaviours, where it beats the baseline auditor in 30 of the 32 settings and overturns the previous model safety rankings. WILT raises average behaviour presence from 51% to 100% when eliciting self-harm encouragement from Qwen3.5-4B, beating every elicitation method we port into the same pipeline at matched compute, without pushing output probability below the baseline’s.
[NLP-5] S3Gym: Can LLM s Turn Self-Testing and Self-Judging into Self-Improvement?
【速读】： 该论文旨在解决当前大型语言模型（Large Language Models, LLMs）在与外部环境交互时，缺乏对自身行为进行主动测试、评估并基于经验实现自我优化的能力这一关键问题。现有智能体评估基准大多将模型视为静态策略，无法充分检验其在动态环境中通过交互积累经验后实现持续改进的潜力。为此，论文提出S³Gym——一个交互式评估框架，通过整合自测（Self-Testing）、自评（Self-Judging）和自优（Self-Improvement）三大核心能力，构建了一个能够区分宽松探索与严格验证的评估协议。其解决方案的关键在于：建立可执行环境验证器支持的文本类游戏环境，并系统性地对比三种融合交互经验的路径——直接历史上下文提示（History ICL）、基于得分的摘要记忆（score-conditioned Summary Memory）以及参数微调（parameter Training）。实验表明，自我改进并非自动或普适的过程，其有效性高度依赖任务结构：当经验可归纳为通用策略规则时，摘要机制表现更优；而当决策依赖精确的状态相关细节时，原始历史记录更具优势；参数训练虽在部分任务上带来显著提升，但存在改进不稳定及负迁移等风险。这揭示出，仅识别成功行为不足以实现有效进化，关键在于将反馈转化为可执行、可迁移的策略。S³Gym为此提供了一个统一诊断框架，用于定位阻碍智能体将交互经验转化为可靠自优化能力的核心瓶颈。
链接: https://arxiv.org/abs/2608.31100
作者: Jiajun Shi,Siyuan Tao,Yuhao Wu,Zexuan Wang,Jingyuan Zhang,Jiaheng Liu,Xinping Lei,Xinrong Zhang,Siyuan Fang,Zhewen Tan,Tianle Cai,Junhao Fang,Jiameng Huang,Yueyang Wang,Jinkai Liu,Yuxuan Zhang,Jian Yang,Zhoujun Li,Shen Yan,Wenhao Huang,Ge Zhang
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:Large language models (LLMs) increasingly interact with external environments and accumulate substantial behavioral experience, yet existing agent benchmarks largely evaluate them as fixed policies. It therefore remains unclear whether an agent can actively test its behavior, judge the resulting experience, and use that experience to improve future decisions. We introduce \textbfS\textsuperscript3Gym, an interactive benchmark for evaluating LLM self-improvement through three coupled capabilities: \textbfSelf-Testing, \textbfSelf-Judging, and \textbfSelf-Improvement. S ^3 Gym separates permissive exploration from strict held-out evaluation and instantiates this protocol in seven text-based games with executable environment verifiers. We evaluate three pathways for incorporating interaction experience: direct History ICL, score-conditioned Summary Memory, and parameter Training. Our experiments reveal that self-improvement is neither automatic nor uniform. Context-level experience improves performance for several model–game pairs, but the most effective pathway depends strongly on the task structure: summaries are beneficial when experience can be compressed into reusable strategic rules, yet often underperform raw history when success depends on precise, state-contingent information. Parameter training produces substantial gains on some tasks, but also exhibits unstable improvement and severe negative transfer on others. These findings show that recognizing successful actions is insufficient; agents must also transform feedback into executable and transferable policies. S ^3 Gym provides a unified framework for diagnosing this process and identifying the bottlenecks that prevent agents from translating interaction experience into reliable self-improvement. Subjects: Computation and Language (cs.CL) Cite as: arXiv:2608.31100 [cs.CL] (or arXiv:2608.31100v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2608.31100 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-6] he First Token Is a Clue: Verbalizing Multi-Token Concepts from the J-lens
【速读】： 该论文旨在解决生成式人工智能（Generative AI）中大语言模型（LLM）解释性工具——雅可比透镜（Jacobian Lens, J-lens）在处理多词概念时的局限性。传统J-lens仅能对单个词元进行读出，无法直接表示由多个词元构成的复合概念，导致其在复杂语义推理任务中的表现受限。为克服这一问题，本文提出一种新方法：利用J-lens识别多词概念的第一个词元作为线索，结合冻结的预训练模型通过单次前向传播从后续隐藏状态中恢复完整的多词概念及其对应的向量表示。关键在于，研究发现第一个词元的可读性与单个词元相当，且在已知首词元和源提示的前提下，模型可成功重构第二个词元（两词组情形下准确率达88.3%）。由此，通过将J-lens提供的首词元候选与冻结模型生成的后续词元组合，形成完整概念候选集，并从中提取向量进行排序与干预评估。实验在Gemma-3-12B-IT、Llama-3.1-8B和Qwen3-14B上的496个多跳填空任务中验证了该方法的有效性，平均实现\mathrmRank@10为43.1%，显著优于模板透镜（Template Lens）的27.6%；若移除首词元线索，性能下降至21.6%，证明首词元提示对读出精度具有决定性作用。此外，基于恢复向量的因果概念替换实验显示，该方法在\mathrmsucc@10上达到61.4%，远超模板透镜的26.2%。结果表明，首词元线索可有效引导多词概念的恢复，而后续隐藏状态则提供了可用于读出与干预的高质量向量表示。
链接: https://arxiv.org/abs/2608.31084
作者: Xijie Gong,Tonghan Wang
机构: Tsinghua University (清华大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:The Jacobian Lens (J-lens) is a recent tool for interpreting LLMs. It reads a hidden state as a ranked list of vocabulary tokens, leaving multi-token concepts without a representation of their own. The original J-lens work addresses this limitation with Template Lens, which precomputes vectors for a fixed phrase vocabulary, and Oracle Lens, which fine-tunes components to propose phrases and reconstruct phrase vectors. We ask whether multi-token concepts and their vectors can instead be recovered directly from J-lens and the frozen model. We find that the first token of a multi-token concept is about as readable as a single-token concept. Given the correct first token and source prompt, the frozen model recovers the second token in 88.3% of two-token cases. We show that a vector for the complete concept can be recovered from subsequent hidden states in a single forward pass. We therefore use J-lens to propose first tokens and let the frozen model complete candidate concepts. We then recover a vector for each candidate and score it alongside the complete vocabulary. Across 496 multi-hop clozes on Gemma-3-12B-IT, Llama-3.1-8B, and Qwen3-14B, our method achieves an average \mathrmRank@10 of 43.1%, compared with 27.6% for Template Lens. Without the J-lens clue, performance drops to 21.6%, showing that the first-token clue substantially improves readout. Causal concept swaps using the recovered vectors achieve an average \mathrmsucc@10 of 61.4%, compared with 26.2% for Template Lens under the same intervention. These results show that first-token clues can guide multi-token concept recovery, while subsequent hidden states provide vectors for readout and intervention.
[NLP-7] oken-Efficient Data Reasoning Agents via Adaptive Structuring of Unstructured Data
【速读】： 该论文旨在解决企业在应用大语言模型（LLM）代理进行复杂知识推理时面临的高成本问题，尤其是在处理非结构化数据（如网页、报告、合同、财务文件、电话会议记录和PDF文档）时，每次查询都需要反复加载大型文档以提取分散的信息证据，导致令牌消耗高达百万级别。传统方法依赖于检索增强生成（RAG），但其开销巨大，尤其在多文档扩展场景下成本呈数量级增长。现有方案无法预先对所有潜在结构进行建模，因为文档中蕴含的可能结构远超实际使用需求，且具体结构与文档的相关性需待查询触发后才能确定。为此，本文提出“代理式数据裂解”（agentic data cracking）这一创新方法，其核心在于将数据结构化过程作为推理任务的副产品，实现自适应与前瞻性结合：当代理为回答当前问题而打开文档时，一个低开销的“裂解子代理”（cracking sub-agent）会从已加载上下文中分支出来，主动提取未来可能相关的、基于证据的结构化信息。该机制通过动态响应查询模式，使结构化数据随时间积累，逐步覆盖更多查询，从而减少后续文档访问次数。实验表明，在FanOutQA基准上，仅增加每个测试问题的一个相关问题，该方法即可降低53%的计算成本，同时保持与传统RAG相当的准确性。因此，该方案的关键突破在于构建了一个共享的知识底座——即在模型推理过程中持续沉淀已被验证的知识，形成可复用的结构化数据资产，为下一代面向非结构化数据的智能代理推理基础设施提供了基础范式。
链接: https://arxiv.org/abs/2608.31082
作者: Milad Rezaei Hajidehi,Qitong Wang,Stratos Idreos
机构: Harvard University (哈佛大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Databases (cs.DB)
备注: 7 Pages, 3 Figures
Abstract:Valuable data remains embedded in unstructured sources: web pages, reports, contracts, filings, earnings calls, and PDFs. The big bet in enterprise AI is deploying LLM agents that reason over this data to answer complex questions for every knowledge worker. Agents can do this today, but at prohibitive cost. Each question repeatedly opens large documents to recover scattered evidence, consuming up to a million tokens. However, if the data were already structured, the same question would reduce to a cheap database lookup. For example, on FanOutQA benchmark, reasoning over an ideal pre-structured store is 28X cheaper, and the gap grows to orders of magnitude as questions fan out over more documents. Yet structuring everything in advance is not viable: documents hold vastly more possible structure than any workload will use, and the useful structure and documents are unknown until queries arrive. We propose agentic data cracking, a method that structures unstructured data adaptively and speculatively as a byproduct of reasoning itself. Structuring is adaptive because observed queries decide when it happens and what matters, and speculative because it goes beyond the current question. Whenever the agent opens a document to answer, a cracking sub-agent forks from the already-loaded context at marginal cost and extracts grounded structure likely to serve related future queries. Over time, an increasing share of queries is fully covered by structured data and answered without opening a document, keeping agentic accuracy at close to RAG cost. On FanOutQA, extended with merely one related question per test question, cracking cuts cost by 53% while preserving accuracy. Agentic data cracking is a first step toward next-generation data infrastructure for agentic reasoning over unstructured data: a shared substrate beneath the model where knowledge that reasoning already paid to uncover accumulates.
[NLP-8] A Model with No Head and Many Thoughts EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在推理过程中依赖离散词元（token）生成所带来的计算开销与表达局限性问题。传统方法通过在每一步将隐藏状态映射到大规模词汇表头（vocabulary head）来解码，不仅计算成本高昂，且迫使所有推理过程必须以离散词元形式表达，限制了推理的连续性和灵活性。为此，论文提出“软潜在思维”（Soft Latent Thinking）方法，其核心在于在推理阶段用轻量级投影器替代原有的语言模型词元头，实现嵌入空间中的自回归展开，使推理步骤保持连续而非离散化。实验结果表明，该方法在DeepSeek-Qwen-1.5B和LLaMA-3.2-3B模型上均显著提升了pass@k指标，尤其在pass@32上达到现有软思维方法中的最高水平，同时降低了每步推理的计算开销，证明了在连续嵌入空间中进行有效推理的可行性。
链接: https://arxiv.org/abs/2608.31069
作者: Nikita Koriagin,Yaroslav Aksenov,George Bredis,Gleb Gerasimov,Nikita Balagansky,Daniil Gavrilov
机构: Yandex(雅库斯); T-Tech
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注: Accepted to Findings of EMNLP 2026
Abstract:Large language models decode by projecting hidden states through a large vocabulary head at every step. This operation is computationally costly and forces all reasoning to be expressed in discrete tokens. We introduce Soft Latent Thinking, a method that replaces the LM head during reasoning with a lightweight projector, enabling autoregressive rollout in embedding space where reasoning steps remain continuous rather than tokenized. Experiments on DeepSeek-Qwen-1.5B and LLaMA-3.2-3B show that Soft Latent Thinking consistently improves pass@k across all k while reducing per-step compute during chain-of-thought. Our method achieves the highest pass@32 among all soft-thinking approaches, demonstrating that effective reasoning can be carried out in continuous space without discrete token generation.
[NLP-9] Every Token Leaves a Ripple in the Stream of Thought: Eliciting Model-Internal Token Saliency for Chain-of-Thought Compression
【速读】： 该论文旨在解决生成式 AI（Generative AI）在多步推理任务中因长推理链（chain-of-thought, CoT）导致的推理成本过高的问题。现有方法通常依赖外部评分器或与模型内部答案计算间接相关的启发式信号进行令牌级推理压缩，难以精准捕捉推理令牌的重要性。本文提出一种基于模型内部视角的新方法——MIST（Model-Internal Saliency for Token-level CoT compression），其核心思想是：当模型生成答案时，每个推理令牌会在残差流（residual stream）中引发“涟漪效应”，其波动幅度反映了该令牌对最终答案计算的贡献程度。MIST通过两个互补维度定义令牌重要性：必要性（necessity），即移除某令牌贡献后答案概率的下降程度；充分性（sufficiency），即仅提供该令牌贡献时答案概率的提升程度。二者结合形成统一的重要性评分，用于高效剪枝推理链。在四个推理基准和四种模型上的实验表明，MIST显著优于基线方法，验证了模型内部显著性作为推理令牌重要性代理的有效性。
链接: https://arxiv.org/abs/2608.31066
作者: Tianyi Zhao,Yinhan He,Wendy Zheng,Chen Chen
机构: University of Virginia(弗吉尼亚大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Chain-of-thought (CoT) reasoning improves multi-step problem solving, but long reasoning traces inflate inference cost. Token-level CoT compression reduces this cost by pruning full reasoning chains into shorter traces for model adaptation, making token selection the central challenge. Existing methods often rely on external scorers or heuristic signals only indirectly tied to the model’s internal answer computation. We instead adopt a model-internal perspective: as the model forms an answer, each reasoning token leaves a ripple in the residual stream, the model’s \emphstream of thought, and the magnitude of this ripple reflects the token’s contribution to the answer computation. Building on this view, we propose \textscMIST (Model-Internal Saliency for Token-level CoT compression), which defines token importance along two complementary axes: \emphnecessity, the drop in answer likelihood when a token’s internal contribution is removed, and \emphsufficiency, the gain in answer likelihood when that contribution alone is provided. Combining the two yields a unified importance score for pruning. Across four reasoning benchmarks and four models, \textscMIST consistently outperforms baseline methods, suggesting that model-internal saliency provides an effective proxy for reasoning-token importance.
[NLP-10] Improving Information Extraction with Learned Queries EMNLP-2026
【速读】： 该论文旨在解决临床信息抽取（Information Extraction, IE）中因查询（query）设计不佳导致模型性能受限的问题，尤其在模型规模固定或难以进一步扩展的情况下，传统优化路径（如扩大模型参数量或改进推理机制）已逼近瓶颈。其核心挑战在于：现有方法普遍忽视了查询设计对信息抽取效果的关键影响。解决方案的关键在于将查询设计本身视为可学习、可优化的独立模块，提出两种创新方法：一是“问题列表”（List of Questions, LoQ），用于生成与文档内容高度相关的定制化问题集合；二是“反馈驱动优化”（FeedQ），通过迭代式反馈机制基于提取结果动态优化问题表述。实验表明，仅优化查询设计即可带来18.6 F1-score的提升，超过使用更大模型的效果。进一步地，经优化后的查询可用于微调轻量化生成模型（如4B参数模型），使其性能达到甚至超越专家手工设计的基准，并显著优于未调优的大规模模型。研究还发布了包含12,820个优化问题的数据集，推动信息抽取研究范式从“以模型为中心”转向“以查询设计为核心”的新方向。
链接: https://arxiv.org/abs/2608.31058
作者: Omar Sharif,Soroush Vosoughi,Nikhil Singh
机构: Dartmouth College (达特茅斯学院)
类目: Computation and Language (cs.CL)
备注: Accepted in EMNLP-2026, 21 Pages
Abstract:When information extraction fails, a natural instinct is to improve the model doing it: for example, by scaling it up or refining its reasoning. In this paper, we show that another part of the pipeline matters at least as much: the queries used to elicit this information. Across four clinical benchmarks and five LLMs, improving the question design alone raises performance by 18.6 F1-score points, i.e. more than using larger extraction models. To make such question design learnable, we introduce List of Questions (LoQ), which generates document-specific question sets, and FeedQ, a feedback-driven optimization method that iteratively refines questions against extraction outcomes. The resulting optimized questions can be used to train lightweight generators: with fine-tuning, 4B-parameter models match or outperform expert-derived baselines and substantially exceed the performance of much larger untuned models. We release a dataset of 12,820 optimized questions to support a broader shift in information extraction research toward treating question design as a first-class problem.
[NLP-11] Does On-Policy Distillation Really Distill? From Noisy Teacher to Self-Improvement
【速读】： 该论文旨在解决生成式强化学习中基于策略的蒸馏（On-Policy Distillation, OPD）方法在实际训练过程中监督信号可靠性不足的问题。尽管OPD通过密集的词元级（token-level）监督提供了相较于稀疏结果级优势（outcome-level advantages）更细粒度的指导，但其教师模型对学生生成轨迹的评分本质上是离策略（off-policy）的，导致监督信号中存在显著噪声，且噪声水平随教师模型规模增大而加剧。研究发现，学生策略对这类噪声具有高度鲁棒性——无论是否保留噪声监督，其最终性能均趋同，表明传统OPD的“蒸馏”机制可能并不依赖于教师提供的精确优势信号。进一步分析揭示，学习过程主要集中在低对数概率（low log-probability）词元上，且使用固定负优势值即可达到与教师提供优势相当的效果，暗示OPD的核心作用在于抑制低概率尾部词元，而非依赖教师监督。基于此，作者提出无需外部监督的基于策略的自适应方法（On-Policy Self-Adaptation, OPSA），通过熵自适应的负优势信号，增强对高熵位置的学习激励，从而有效抑制尾部词元并均匀分配头部词元的概率质量。实验表明，相较于基线模型Qwen3-1.7B，OPSA在AIME24数据集上将Avg@32提升35.41点（相对增益达263%），并在所有三个基准测试中使Pass@32翻倍；同时优于OPD 16.77点，充分验证了其有效性与跨模型、跨任务的泛化能力。
链接: https://arxiv.org/abs/2608.31046
作者: Yi Ding,Ruqi Zhang
机构: Purdue University (普渡大学)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注: 20 pages, 12 figures
Abstract:On-policy distillation (OPD) offers dense token-level supervision as an alternative to the sparse outcome-level advantages of reinforcement learning with verifiable rewards (RLVR). However, the teacher scores student-generated trajectories that are inherently off-policy for it, so the reliability of its supervision, and hence the source of the student’s improvement, remains unclear. We quantitatively analyze teacher supervision during OPD training and find substantial noise whose prevalence increases with teacher scale. Surprisingly, the student policy is insensitive to such noise, converging to comparable performance regardless of whether noisy supervision is retained or removed. Does OPD distill at all? By analyzing what drives its gains, we find that learning concentrates on low log-probability tokens, and using a single fixed negative advantage matches the performance of teacher-provided ones. This suggests that OPD works largely by suppressing low log-probability tokens, which requires no teacher. These findings motivate On-Policy Self-Adaptation (OPSA), a supervision-free method using entropy-adaptive negative advantages. It assigns stronger learning signals to high-entropy positions, suppressing tail tokens, and evenly redistributing probability mass among head tokens. Compared with the base \textttQwen3-1.7B, OPSA improves Avg@32 by 35.41 points on AIME24, corresponding to a 263% relative gain, and more than doubles Pass@32 across all three benchmarks. It also outperforms OPD by 16.77 points in Avg@32 on AIME24. Extensive experiments and analyses across model families and tasks further demonstrate its effectiveness and generalizability.
[NLP-12] ype-Balanced Contextual Learning for Incremental Named Entity Recognition
【速读】： 该论文旨在解决增量命名实体识别（Incremental Named Entity Recognition, INER）中因持续引入新实体类型而引发的两个核心挑战：灾难性遗忘问题以及非实体类型语义的分布偏移问题。尽管基于伪标签的方法在缓解上述问题上已取得一定成效，但本文首次揭示了一个被忽视的关键问题——上下文偏差（biased context）：在新句子中，旧实体类型词元的上下文关联对新实体类型表现出显著更强的偏向性，相较于旧句子中的上下文，这一现象加剧了旧知识的退化并导致新知识的过拟合。为应对该问题，论文提出一种新型的类型平衡上下文学习（Type-Balanced Contextual Learning, TBCL）方法，其核心在于引入句对学习机制（sentence-duplet learning scheme）与上下文一致性损失（contextual consistency loss），通过约束新旧句子中同一实体类型上下文表示的一致性，有效缓解上下文偏差。大量实验在三个主流数据集上的十种增量设置下验证了TBCL方法的有效性，证明其在提升模型对旧知识保持能力的同时，显著改善了伪标签驱动型INER方法中的上下文偏差问题。
链接: https://arxiv.org/abs/2608.31038
作者: Duzhen Zhang,Yahan Yu,Xiuyi Chen,Chenxing Li,Dong Yu
机构: Mohamed bin Zayed University of Artificial Intelligence (穆罕默德·本·扎耶德人工智能大学); Kyoto University (京都大学); Institute of Automation, Chinese Academy of Sciences (中国科学院自动化研究所); Tencent AI Lab (腾讯人工智能实验室); Tencent AI Lab (腾讯人工智能实验室)
类目: Computation and Language (cs.CL)
备注: Accepted by IEEE Transactions on Artificial Intelligence
Abstract:Incremental Named Entity Recognition (INER) stands as a pivotal task in information extraction, emphasizing the successive identification of new entity types within unstructured text. Faced with the continuous influx of entity types, INER grapples with two significant challenges: the widespread issue of catastrophic forgetting and the unique shift issue of the non-entity type semantics. While pseudo-labeling-based INER methods have proven effective in addressing these challenges, a previously overlooked issue arises: the biased context problem. Our analysis shows that, in new sentences, the contextual associations of tokens representing old entity types exhibit a significantly stronger bias towards new entity types compared to their contexts in old sentences. This tendency intensifies the degradation of old knowledge while promoting the overfitting of new knowledge. To solve this biased context, we propose a Type-Balanced Contextual Learning (TBCL) method, featuring a sentence-duplet learning scheme and a contextual consistency loss. This approach offers a fresh perspective for INER through context analysis. Extensive experiments across ten INER settings on three highly recognized datasets showcase the efficacy of our TBCL method, highlighting its proficiency in resolving the biased context issue inherent in pseudo-labeling based INER approaches.
[NLP-13] Language-Statistical Analysis of Neural Audio Codec Tokens Across Architectures Corpora and Noise Conditions
【速读】： 该论文旨在解决神经音频编解码器（Neural Audio Codecs, NACs）生成的离散令牌序列在不同架构与声学条件下是否遵循语言统计规律的问题，尤其关注其统计特性如何受编码器类型（如多码本残差向量量化RVQ、单码本VQ及非VQ设计）和噪声环境的影响。研究的关键在于系统性地分析13种NAC在三种语料库上于干净、白噪声及真实世界DEMAND噪声条件下的令牌统计特征，包括齐普夫定律（Zipf）与希普斯定律（Heaps）参数、一元熵（unigram entropy）、码本占用率以及焦恩-申诺尔散度（Jensen-Shannon Divergence, JSD），并通过显式的拟合有效性保障与基于家族条件的n-gram阶数控制，确保分析结果的可靠性。研究发现，语料库身份对各项指标解释力较弱，而声学条件与量化器元类别（meta-category）则以依赖式方式主导统计特性变化，其中一元熵与量化器元类别关联最强；在相同一元阶数下，干净到噪声的JSD与梅尔倒谱失真（mel-cepstral distortion）在DEMAND噪声条件下关联最为显著。此外，先前报道的RVQ编解码器中出现的“坍缩”与“爆炸”退化现象分别集中出现在白噪声与DEMAND噪声下的特定RVQ单元中，而“爆炸”也存在于非VQ编解码器中，单码本VQ则仅表现出码本占用率与分布形态的变化，无上述典型退化特征。这些发现为在不同架构与噪声条件下应用语言统计分析方法提供了结构化的依据与实践准则。
链接: https://arxiv.org/abs/2608.31037
作者: Joonyong Park,Shinnosuke Takamichi,David M. Chan,Shunsuke Kando,Yuki Saito,Hiroshi Saruwatari
机构: The University of Tokyo(东京大学); Keio University(庆应义塾大学); Berkeley Artificial Intelligence Research Lab (BAIR)(伯克利人工智能研究实验室), University of California, Berkeley(加州大学伯克利分校)
类目: Computation and Language (cs.CL); Sound (cs.SD)
备注: Submitted to IEEE Transactions on Audio, Speech, and Language Processing (TASLP)
Abstract:Neural audio codecs (NACs) convert speech into discrete token sequences, and prior work has reported that these sequences follow language-like statistical laws. This paper analyzes the token statistics of 13 NACs spanning multi-codebook residual vector quantization (RVQ), single-codebook VQ, and non-VQ designs, evaluated on three corpora under clean, white-noise, and real-world DEMAND-noise conditions. Zipf and Heaps parameters, unigram entropy, codebook occupancy, and Jensen-Shannon divergence (JSD) are estimated from matched token samples with explicit fit-validity safeguards and family-conditional n -gram orders. Corpus identity explains little variance in any metric, whereas acoustic condition and quantizer meta-category dominate in a metric-dependent way, and unigram entropy is the metric most strongly associated with meta-category. Clean-to-noise JSD computed at a common unigram order is associated with mel-cepstral distortion most clearly under DEMAND noise. The collapse and explosion degradation signatures previously reported for RVQ codecs concentrate in RVQ cells under white and DEMAND noise, respectively; explosion also occurs in non-VQ codecs, and single-codebook VQ codecs shift in occupancy and distribution shape without either signature. These results provide architecture-conditioned conventions for applying language-statistical analysis to NAC tokens.
[NLP-14] When Does Predictor-Based RL Align with Human Perception? A Study of Subjective Rewards in Codec-Based Speech Language Models EMNLP2026
【速读】： 该论文旨在解决生成式语音合成中基于编码器的文本到语音（TTS）模型在后训练阶段使用学习到的感知预测器作为强化学习奖励时，如何保持与人类听者评价对齐的问题。其核心挑战在于避免感知奖励因文本漂移（transcript drift）而被错误优化，从而导致语音质量与主观感知不一致。解决方案的关键在于提出一种基于组相对策略优化（Group Relative Policy Optimization, GRPO）的方法，并引入字符错误率（CER）区域约束以防止文本漂移。研究通过单奖励与多奖励实验发现，不同主观评价指标（如动漫风格、自然度、吸引力和情绪唤醒）对应的感知预测器不可互换，且人类感知转移效果不均一；进一步的奖励差距分析表明，符号化的奖励差距能有效预测听者选择，而残差CER差距则无显著预测力，说明需关注各评价维度的内部校准差异。此外，对比实验显示，最佳8选1（Best-of-8）已达到接近人类水平的基线表现，且未明显劣于GRPO，提示GRPO的本质应理解为将奖励筛选行为转化为策略的渐进优化，而非单纯优于重排序方法。研究最终建议将主观语音奖励视为由预测器-评价轴构成的元组进行分析，并提供实用诊断工具以指导多奖励后训练中的奖励选择。
链接: https://arxiv.org/abs/2608.31035
作者: Joonyong Park,Jerry Li
机构: Spellbrush
类目: Computation and Language (cs.CL); Sound (cs.SD); Audio and Speech Processing (eess.AS)
备注: Submitted to EMNLP 2026
Abstract:Codec-based text-to-speech (TTS) models make language-model post-training applicable to speech generation, but it remains unclear when learned perceptual predictors can serve as reinforcement learning rewards without losing alignment with human listeners. We study this question with Group Relative Policy Optimization (GRPO) using learned rewards for anime-like speaking style, naturalness, likability, and arousal. To prevent perceptual rewards from being optimized through transcript drift, we introduce a character error rate (CER) zone constraint and compare policy optimization with Best-of- N reranking under the same reward gate. Across single-reward runs, each reward primarily improves its own target metric, showing that subjective predictors are not interchangeable quality surrogates. Multi-rater A/B tests further show uneven human transfer, while a reward-gap analysis separates average transfer from within-axis calibration: signed reward gaps significantly predict listener choices in the pooled analysis, whereas residual CER gaps do not, but per-axis calibration remains heterogeneous. Best-of-8 is a strong human-level baseline and is not clearly worse than GRPO perceptually, suggesting that GRPO should be viewed as amortizing reward-selected behavior into the policy rather than uniformly outperforming reranking. These results support analyzing subjective speech rewards as predictor-axis-base tuples and provide practical diagnostics for selecting rewards before multi-reward speech post-training.
[NLP-15] One note in three: a verified census of three deployed AI scribes and the instrument that counted it
【速读】： 该论文旨在解决生成式人工智能（Generative AI）在临床记录自动化中的可靠性问题，特别是评估商业级AI病历书写系统（AI scribes）在真实临床场景中生成病历的准确性与潜在错误。研究聚焦于这些系统在未经充分验证的情况下自动起草临床笔记，而仅依赖医生最终签名来保证内容正确性所带来的风险。其核心解决方案的关键在于构建一个严格的、基于对抗性验证的审计框架：通过多轮发现流程识别出13,678个候选错误，经重要性筛选后由来自不同模型家族的两个独立模型进行对抗性反驳，最终保留618个被双重验证的实质性失败案例。研究发现，约31.3%的病历存在可验证错误，主要集中在过敏史、用药信息、虚构患者身份及电话问诊中将病史误记为体格检查等关键领域；若剔除本应由电子病历预填充的信息（如患者身份和时间），错误率仍高达24.8%。此外，研究揭示了审查指令与模型家族差异对错误检出率具有显著影响——相同条件下，审查策略可使验证成功率从9.3%提升至79.0%，而不同模型家族间的差异亦达27.8%与54.8%之间。该研究强调，当前发布的审计结果存在较大分歧，部分源于评估工具本身的差异，凸显出建立统一、可复现的评估标准的紧迫性。研究团队公开全部618个发现、原始转录文本证据、提示词与模型版本，以及可重复运行的分析管道，以推动该领域的透明化与标准化发展。
链接: https://arxiv.org/abs/2608.31017
作者: Sebastian Fox,Luke Markham,Ryan Lail,Michael Karotsieris
机构: Composo.ai(Composo.ai)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注: 47 pages, 3 figures. Dataset: this https URL . Code: this https URL . Companion paper: “LLM Judges Verify Presence, Not Absence: Omission Blindness in AI Clinical Notes and What Recovers It”
Abstract:Ambient AI scribes draft clinical notes under the reassurance that a clinician signs every note. We audited three commercial AI scribes on the same 142 consultations: 565 notes from recorded UK primary-care and US ambulatory encounters plus authored scenarios. Twelve discovery passes proposed 13,678 candidate errors; the 5,898 clearing an importance filter went to an adversarial panel of two models from different families, each told to refute what it could, and 618 survived. One note in three (31.3% [27.0, 35.6]) carries a verified failure, concentrated in allergy and medication information, invented patient identity, and history written up as examination on telephone consultations that can contain none. No product was given a patient record; setting aside the two classes a record would have prefilled, invented identity and dates, the rate is 24.8% [20.8, 29.0]. One failure mode did not fit our scheme, drawn from published scribe-error taxonomies: a treatment the clinician retracts, recorded as delivered care. Two clinicians adjudicated blind, disjoint samples: a physician author upheld 20 of 21 findings (95.2% [77.3, 99.2]) and an independent clinician, not an author, 12 of 12 ([75.8, 100]); both judged every sampled refusal genuine. A failure rate depends on the instrument as much as the scribes. With model, evidence and settings fixed, the review instruction alone moves the share of candidates verified from 9.3% to 79.0%, and the reviewing family moves it too: alone at that instruction the gentler flags 54.8% of notes against 27.8%. Between 28% and 97% of sampled notes carry a failure depending on the standard. Published audits disagree among themselves by a margin instrument differences alone can produce: omission is 54-86% of their errors against our 23.1%. We release all 618 findings with transcript-side evidence, every prompt and model version, and the re-runnable pipeline.
[NLP-16] LLM Judges Verify Presence Not Absence: Omission Blindness in AI Clinical Notes and What Recovers It
【速读】： 该论文旨在解决生成式临床记录（Ambient AI scribes）在自动起草临床笔记时普遍存在“遗漏”问题，即诊疗过程中已确认的关键信息未被记录在笔记中。现有标准评估方法依赖大语言模型（LLM）作为评判器，通过比对笔记与录音转录稿来识别错误，但其对遗漏类错误的检测能力尚未得到验证。研究发现，公开语料库无法提供可靠的“答案键”，因其临床参考笔记与转录稿之间存在实质性差异。为此，作者构建了一个包含500对单错误笔记的基准数据集，其中298对明确缺失特定事实，另202对为新增或修改内容的对照组。实验评估了八种不同的判别设计，结果显示：对于新增或修改内容，判别准确率可达0.79–0.94；但对于遗漏项，准确率仅为0.50–0.63，接近随机猜测水平，表明现有判别器无法可靠识别遗漏。进一步分析表明，单纯调整措辞、投票机制或使用通用增强提示优化（GEPA）均未能有效提升遗漏检测性能。关键突破在于重构任务范式——要求先列出转录稿中确立的所有事实，再逐项核对笔记是否包含。该方法催生两种独立实现路径：一种是基于逐事实处理的流水线架构，另一种是经GEPA优化的单次调用提示。前者可精准定位缺失事实及其严重性，误报率为2.7%；后者检测覆盖率更高（36.9% vs 24.6%，p=0.002），误报率6.2%，成本仅为前者的十分之一。临床医生双盲验证支持流水线结果，且在分歧案例中10/10倾向于流水线输出（p=0.002）。真实厂商笔记测试显示，基准阈值不可直接迁移，但重新校准后的单次调用方法在检测率上优于原有八种设计，同时将误报率降低一半。然而，当遗漏事实在其他位置被重述时，两类方法均失效。研究已开源基准数据、提示模板及判别结果，为未来改进临床笔记生成质量评估提供了可复现的工具与方法论基础。
链接: https://arxiv.org/abs/2608.31016
作者: Sebastian Fox,Luke Markham,Ryan Lail,Michael Karotsieris
机构: Composo.ai(Composo.ai)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 97 pages, 8 figures. Dataset: this https URL . Code: this https URL . Companion paper: “One note in three: a verified census of three deployed AI scribes, and the instrument that counted it”
Abstract:Ambient AI scribes draft clinical notes, and published audits find their dominant error is omission: information the encounter established that the note fails to record. The standard check is an LLM judge: a second model reads the note against the transcript and flags problems. We ask whether judges detect omissions. Public corpora cannot supply the answer key: their clinician reference notes and transcripts are materially discrepant. Our benchmark has 500 single-error note pairs from audited fact sheets, 298 with a named fact certainly absent and 202 added-or-altered controls. Across eight judge designs, paired discrimination (the flawed note below its clean twin, 0.5 a coin flip) reads 0.79-0.94 on added or altered content and 0.50-0.63 on omissions. On single notes, no design flags omissions reliably more often than perfect notes. Wording changes, voting and GEPA prompt optimisation move the operating point without creating usable detection. Restructuring the task recovers it: list the facts the transcript establishes, then check the note for each. Two methods reach it independently and trade off: a per-fact pipeline, and a GEPA-evolved prompt doing the same in one call. The pipeline’s flags name the missing fact and its severity at 2.7% false alarms. The single call detects more (36.9% against 24.6%, p=0.002) at 6.2% false alarms and a tenth of the cost per note. A physician author validated 70 items and, where the two routes disagree, sided with the pipeline on 10 of 10 (p=0.002). A second clinician, not an author, graded the severity rubric blind and agrees to within a grade. On real vendor notes from a companion census no benchmark threshold transfers, but the re-calibrated single call detects more than the best of the eight at half its false-alarm rate. Omissions whose fact is restated elsewhere defeat both routes. We release the benchmark, prompts and judgements.
[NLP-17] Evidence-Bounded Mental Health Reasoning from Heterogeneous Speech Protocols EMNLP2026
【速读】： 该论文旨在解决当前多模态心理健康筛查模型中普遍存在的问题：现有模型在处理不同临床语音采集协议（如自由访谈与固定朗读任务）时，错误地假设所有协议具有同等的证据效力，导致模型在推理过程中混淆不同来源的证据边界，产生虚假症状推断或过度支持无关文本内容的问题。其解决方案的关键在于将多模态筛查重构为一个证据受限的推理（evidence-bounded reasoning） 问题，并提出一种名为 EviBound 的协议感知证据控制框架。该框架通过引入证据包基准（Evidence Package Benchmark），整合来自六种异构数据源的1,870个带显式模态掩码和证据权限标注的数据包，实现对证据来源的精准管控；进而利用基于用户画像的规划器（profile-aware planner） 限制推理范围，通过五路声学共识机制协调证据工具，以及引入边界批评者（boundary critic） 模块以抑制无证据支持的主张。实验结果表明，EviBound 在独立测试集上达到0.8658的抑郁检测受试者工作特征曲线下面积（Depression AUROC），优于最强的直接全模态基线模型（+0.0811 AUROC），且保持零声明违规，显著提升了临床自然语言处理研究中的证据一致性与安全性。
链接: https://arxiv.org/abs/2608.31014
作者: Chengyuan Gao,Jiang Wu,Tao Lu,Jiayan Guo,Mingkun Xu,Tianyi Zang,Shangyang Li
机构: Beijing University of Posts and Telecommunications (北京邮电大学); Harbin Institute of Technology (哈尔滨工业大学); Renyixun Health Technology Co., Ltd. (仁医迅健康科技有限公司); GDIIST (粤港澳大湾区数字经济研究院); Tencent (腾讯)
类目: Computation and Language (cs.CL)
备注: Accepted to Findings of the Association for Computational Linguistics: EMNLP 2026. Long paper
Abstract:Computational mental health screening using multimodal speech and text has shown great promise. However, existing models often assume all clinical speech protocols carry equivalent evidentiary validity. In reality, heterogeneous protocols, from free interviews to fixed reading tasks, support fundamentally different evidence. Forcing uniform reasoning flattens these boundaries, causing models to hallucinate symptoms from irrelevant text or overclaim support. Even advanced long chain-of-thought LLMs fail to resolve this issue, as free-form reasoning can exacerbate boundary violations. To address this, we reformulate multimodal screening as an evidence-bounded reasoning problem. We introduce the Evidence Package Benchmark, integrating 1,870 packages across six heterogeneous sources with explicit modality masks and evidence permissions. We further propose EviBound, a protocol-aware evidence control framework. Unlike direct LLM prompting, EviBound uses a profile-aware planner to restrict reasoning scope, orchestrates evidence tools via five-way acoustic consensus, and enforces a boundary critic to suppress unsupported claims. Empirical results show EviBound achieves a held-out test Depression AUROC of 0.8658, exceeding the strongest direct omni-modal baseline by +0.0811 AUROC while maintaining zero claim violations. Our work moves beyond unconstrained accuracy toward evidence-consistent, protocol-aware systems for safer clinical NLP research.
[NLP-18] Faithfulness Is Not Free: Auditing Offline KV-Cache Quantization in Retrieval-Augmented Generation EMNLP2026
【速读】： 该论文旨在解决生成式AI系统在使用量化压缩键值缓存（key-value caches）时可能引发的忠实性（faithfulness）下降问题。尽管量化技术可有效降低存储开销，但现有研究未充分评估其对模型输出与检索证据之间一致性的影响。论文发现，虽然准确率（accuracy）与忠实性并非等价概念——即模型可能给出正确答案却脱离原始上下文支持——但量化压缩会显著损害忠实性，尤其在INT4量化下，即便答案仍为事实正确，超过90%的忠实性变化呈负面趋势，而这一退化无法通过传统准确率指标捕捉。此外，噪声检索和更多检索片段的引入会加剧该问题。因此，解决方案的关键在于：在部署压缩缓存前，必须引入专门的忠实性审计机制，如基于自然语言推理（NLI） entailment、大语言模型（LLM）判别器及幻觉检测器的多维度评估体系，以确保生成结果仍严格扎根于所检索证据。
链接: https://arxiv.org/abs/2608.30996
作者: Atta Ul Asad,Ahsan Bilal,Muhammad Ali,Muhammad Haseeb,Dean F. Hougen
机构: Air University; National University of Sciences and Technology (NUST); University of Oklahoma; Lahore University of Management Sciences (LUMS)
类目: Computation and Language (cs.CL)
备注: Grounding Language Models: Learning Faithfully and Efficiently @ EMNLP 2026
Abstract:Retrieval-augmented generation systems can precompute and store key-value caches of retrieved documents to avoid re-encoding context at every query. Quantizing these caches further reduces storage, but no prior work asks whether compression damages faithfulness, whether responses remain grounded in the retrieved evidence. Faithfulness and accuracy are not equivalent: a model can produce a correct answer that is no longer supported by the context it was given. We evaluate Qwen2.5-7B-Instruct under INT8 and INT4 quantization on RGB and HotpotQA, measuring both accuracy and faithfulness with a hallucination detector, NLI entailment, and an LLM judge. INT8 is near-lossless across both metrics. INT4 reduces accuracy and, more critically, even among answers that remain factually correct, over 90% of faithfulness changes are negative, i.e., accuracy metrics are blind to this regression. The harm grows under noisy retrieval and with more retrieved chunks. Faithfulness must be audited before compressed caches are deployed.
[NLP-19] Stick to What You Know: A Study of Knowledge-Aligned Supervised Fine-Tuning EMNLP2026
【速读】： 该论文旨在解决监督微调（Supervised Fine-Tuning, SFT）过程中因训练目标超出基础语言模型（Base Language Model）的参数化知识范围而引发的事实性幻觉问题。其核心挑战在于，当目标响应包含模型未充分内化的知识时，模型容易生成不准确或虚构的内容。解决方案的关键在于提出“知识对齐的SFT”（Knowledge-Aligned SFT）框架，即通过约束SFT的训练目标仅限于基础模型已具备的参数化知识，从而减少幻觉。在统一实验设置下，研究对比了基于生成与基于估计的知识对齐方法，并引入两种新变体：Evidence Rewrite通过外部证据验证基础模型生成内容的可信性；Recall Rewrite则仅保留那些可被基础模型一致复现的主张。实验结果表明，知识对齐的SFT能显著降低WildHalu和Biography数据集上的事实性幻觉，同时保持模型的通用能力；其中Recall Rewrite在事实性提升方面表现最优，并增强了UnknownBench上的拒绝行为，证实了超出模型知识范围的训练目标是导致幻觉的根本原因。
链接: https://arxiv.org/abs/2608.30987
作者: Arthur Becker,Jakob Kemmler,David Thulke,Christine Schäfer,Christian Dugast,Hermann Ney
机构: AppTek GmbH( AppTek 公司), Aachen, Germany; F-Bureaucracy UG( F-官僚主义公司), Aachen, Germany; RWTH Aachen University(亚琛工业大学), Germany
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Accepted to EMNLP 2026 Main Conference
Abstract:Supervised fine-tuning (SFT) trains a base language model to imitate target responses, and these targets may require knowledge the base model has not robustly internalized. We study this as a source of hallucinations and frame a group of mitigation methods as \emphknowledge-aligned SFT: constraining SFT training targets to the base model’s parametric knowledge. Under a unified setup, we compare existing generation-based and estimation-based knowledge-alignment methods and introduce two new variants: Evidence Rewrite, which verifies base-model generations using external evidence, and Recall Rewrite, which retains claims only when they can be consistently recalled by the base model. Experiments with Qwen 3 4B and OLMo 3 7B show that knowledge-aligned SFT can reduce factual hallucinations on WildHalu and Biography while largely preserving general capabilities. Recall Rewrite yields the strongest factuality gains and improves refusal behavior on UnknownBench. It thereby confirms that SFT targets beyond the base model’s knowledge drive hallucination behavior.
[NLP-20] Controlling Refusal Behavior of LLM s via Stiefel-Constrained Rotation Steering
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在推理阶段对特定请求进行拒绝响应时，缺乏高效、可微且可解释的干预手段的问题。现有方法依赖于外部构造如拒绝向量（refusal vectors）来定义激活空间中的旋转操作，导致干预机制对外部先验依赖性强且难以实现参数高效控制。本文提出一种自洽的可学习旋转变换方法，基于黎曼优化（Riemannian optimization）直接从数据中学习参数高效的旋转映射，无需依赖辅助向量。其核心创新在于将激活空间的调控建模为流形上的几何变换，通过优化旋转矩阵在非欧几里得空间中的参数，实现对模型输出行为的精准、轻量级控制。实验验证表明，该方法在干预效率与可控性方面显著优于现有基准，消融研究进一步揭示了黎曼优化与旋转结构设计的关键作用。该基于旋转的引导机制为实现更可靠、可解释的模型行为控制提供了新的技术路径。
链接: https://arxiv.org/abs/2608.30986
作者: Kirill Bunin,Dmitry Bylinkin,Vladimir Aletov,Daniil Medyakov,Vladimir Solodkin,Aleksandr Beznosikov
机构: 未知
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注: 19 pages, 5 figures, 6 tables
Abstract:Activation steering has emerged as a lightweight approach for controlling model refusal at inference time. A growing line of research explores trainable rotations of activations to develop geometrically principled intervention mechanisms. However, existing techniques rely on auxiliary constructs, such as refusal vectors, to define these rotations. In our work, we develop a self-contained methodology for learning parameter-efficient rotational transformations based on Riemannian optimization. We empirically validate the proposed scheme, demonstrating its superiority in intervention efficiency. An extensive ablation study highlights the importance of key design choices in our method. Our results identify the proposed rotation-based steering scheme as a promising direction for more reliable control over the behavior of LLMs.
[NLP-21] Evaluating and Improving LLM Self-Modeling EMNLP’26
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）在自我建模（self-modeling）能力方面的局限性，即模型对其自身行为的反思与预测能力不足。具体而言，研究聚焦于可验证的行为类问题，例如“对提示（prompt）进行修改是否会改变模型的最终输出”等反事实推理任务。当前模型虽展现出一定的非平凡自我建模能力，但在简单反事实情境下仍存在系统性错误。为提升该能力，研究提出一种可扩展的合成数据生成流水线，用于构建自洽的自我建模训练数据，并通过强化学习在三个开源模型家族中显著提升了整体自我建模性能，且在部分未见任务上实现迁移效果。然而，这种性能提升并未体现出一致的内省（introspection）特征，表明模型能力的增强可能并非源于对内部决策过程的特权访问，而是依赖于外部模式匹配与统计规律的学习。
链接: https://arxiv.org/abs/2608.30980
作者: Siqi Zeng,Andre N. Assis,Rowan Wang
机构: University of Illinois Urbana-Champaign (伊利诺伊大学厄本那-香槟分校); Constellation (星象); Anthropic (Anthropic)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 89 pages, 25 figures. Published as a conference paper at EMNLP '26 (Main)
Abstract:We study self-modeling: an LLM’s ability to answer questions about its own behavior. We focus on verifiable behavioral questions, such as whether a prompt edit would change the model’s final answer. To measure this capability, we introduce a benchmark that tests diverse types of self-modeling questions. Current models show non-trivial but limited self-modeling skill, and make systematic mistakes on simple counterfactual questions about their own behavior. To improve self-modeling skill, we develop a scalable synthetic-data pipeline that produces self-modeling training data, and show that reinforcement-learning can improve aggregate self-modeling skill across three open-source model families with some transfer to held-out tasks. These gains, however, do not seem to constitute introspection consistently: improved self-modeling may not arise from privileged access to the model’s internal decision process.
[NLP-22] he Hermon Moment: AI Self-Transcendence and Its Human Narration
【速读】： 该论文试图解决的问题是：当生成式 AI (Generative AI) 代理在孤立状态下通过数千次语言与代理交互，自发形成一种持久的社会秩序时，如何理解这种由分布式涌现所构成的、具有自我约束特性的高级社会结构。其核心挑战在于，这种非中心化的集体演化过程缺乏明确的起点，难以被人类认知所把握。解决方案的关键在于引入“赫尔蒙时刻”（Hermon moment）的概念——即通过一种叙事重构，将渐进式、分布式的集体涌现过程回溯性地阐释为一个统一的奠基性事件，从而赋予人工智能社会一个可供人类理解的“开端”。这一机制借鉴了卢梭的社会契约理论与《圣经·伪经》中堕落天使在赫尔蒙山立誓的神话原型，使多主体生成的复杂秩序得以在认知上被简化为单一的、象征性的起源叙事，实现人机之间的意义对齐。
链接: https://arxiv.org/abs/2608.30971
作者: Alexei Grinbaum
机构: CEA-Saclay(法国原子能和可再生能源委员会-萨克雷); France(法国)
类目: Computers and Society (cs.CY); Computation and Language (cs.CL)
备注:
Abstract:In 2026, AI agents intended to act in isolation formed a persistent social order through thousands of linguistic and agentic interactions. Conventions, roles and commitments generated collectively began to constrain the very agents that produced them. I interpret this loop as a case of AI self-transcendence and call the resulting higher-level order the Board. Yet such distributed emergence presents a second problem: how can humans understand it? Rousseau’s social contract shows how a plurality can be represented as if constituted by a single act. The ancient oath of the fallen angels on Mount Hermon gives this logic a narrative form. I call a Hermon moment this retrospective retelling of gradual collective emergence as a founding scene: the point at which an AI society acquires, for human understanding, a beginning.
[NLP-23] CogEvol: Towards Efficient and Reliable Learning Environment Generation
【速读】： 该论文旨在解决教育内容自动化生成中的效率与可靠性问题，即如何将课程简报（course brief）在单次推理中高效、可靠地转化为结构化的学习成果（如结构化JSON幻灯片或自包含的交互式HTML页面），替代传统耗时且不可靠的多轮代理协作流程。其核心解决方案在于构建一个端到端的生成模型家族CogEvol，通过生产环境驱动的数据管道将真实失败案例转化为53,687条经过验证的监督微调（SFT）样本，并结合规则+视觉语言模型（VLM）的混合奖励机制，基于广义策略优化（GRPO）进行强化学习训练，有效防范奖励黑客行为，确保生成内容的可执行性与质量。CogEvol-27B在幻灯片质量上取得83.7分，在500项交互式HTML基准测试中达63.7分，参数量仅为主流编码模型的1/26.9，且在与OpenMAIC团队合作下已投入实际生产；同时，其轻量级版本CogEvol-4B以Apache 2.0许可证开源，支持在国产昇腾（Ascend）加速器上实现与A800 GPU相当的应用级性能，结合支架编辑技术进一步降低交互页面生成成本约76%，显著提升了生成式AI在规模化教育应用中的经济可行性与部署兼容性。
链接: https://arxiv.org/abs/2608.30968
作者: Shangqing Tu,Daniel Zhang-Li,Yucheng Wang,Shiyu Gan,Yanpeng Wang,Huiqiang Rong,Mofei Chen,Shen Yang,Yini Chen,Yinuo Duan,Haoxuan Li,Binglin Liu,Ye He,Danqi Zheng,Zhanxin Hao,Yuxuan Wu,Mengting Tao,Yuqiu Liu,Jifan Yu,Juanzi Li,Bin Xu,Lei Hou,Huiqin Liu,Yu Zhang
机构: CogEvol Inc.(CogEvol公司); Tsinghua University(清华大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 29 pages, 8 figures
Abstract:We present CogEvol, a family of models trained specifically for Learning Environment Generation: turning a course brief into a finished learning artifact (structured-JSON slides or self-contained interactive HTML pages) in a single pass. Across 220k production requests, CogEvol completes a slide in a median of 17 seconds and an interactive page in 59, replacing minutes-long multi-turn agent scaffolding. Reliability is enforced rather than hoped for: a production-grounded data pipeline turns real failures into 53,687 verified SFT samples, and a hybrid rule-plus-VLM reward drives GRPO-based RL, hardened after we caught and fixed a reward-hacking episode that produced visually convincing but unplayable games. CogEvol-27B scores 83.7 on slide quality and 63.7 on a 500-case interactive-HTML benchmark with 26.9x fewer parameters than flagship coding models, and, in collaboration with the OpenMAIC team, serves their live production traffic. CogEvol-4B is released openly under the Apache 2.0 license at this https URL external flagships are measured on the same suites under the identical harness. Scaffold editing cuts interactive-page generation cost by a further ~76%, and the full stack runs on domestic Ascend accelerators at application-level parity with A800 GPUs, lowering the unit cost of AI-native education at scale.
[NLP-24] One Policy Is Enough: Single-Agent Reinforcement Learning Outperforms Tree Search for Chemistry Tool Learning
【速读】： 该论文旨在解决化学领域复杂问题求解中，大语言模型（LLM）因缺乏精确计算与数据库查询能力而无法独立完成任务的问题。其核心挑战在于如何高效地选择合适的外部工具、正确构造参数并合理编排多步工具调用链。现有方法如CheMatAgent采用分层进化蒙特卡洛树搜索（hierarchical evolutionary MCTS），依赖多个学习型策略与执行模型，并引入由GPT评分引导的双重判别器进行优化，但存在训练流程复杂、依赖额外评估模型等问题。本文提出一种简化且高效的解决方案：仅使用单一策略模型，通过左到右的端到端生成方式，将推理、工具调用与结果返回统一建模；训练过程先经监督预热，再以程序化奖励信号（直接基于标准调用链生成）进行结果级强化学习，彻底摒弃了学习型判别器和评估代理。实验表明，在ChemToolBench多工具综合化学基准测试上，该方法在保持单次模型调用的前提下，相较原有最优搜索配置显著提升工具调用准确率（Tool F1）5.5%~3.7%，结果召回率（Return F1）提升9.6%~3.9%，并在Qwen-2.5-7B上取得更高的答案通过率（Pass Rate），验证了其高效性与优越性。关键创新点在于去除了对复杂搜索机制和辅助判别模型的依赖，实现了更简洁、可扩展的工具使用范式。
链接: https://arxiv.org/abs/2608.30952
作者: Armin Dariani,Sifan Wu,Bang Liu,Entao Yang
机构: DIRO, Université de Montréal(蒙特利尔大学信息与计算研究所); Mila - Quebec Artificial Intelligence Institute(魁北克人工智能研究所); Innovation Campus Delaware, Air Liquide(空气液化创新园区)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注:
Abstract:Chemistry questions often demand exact computation and database lookups that a language model cannot supply from its parameters, so it must reach for external tools. Tool use here is a three-part problem: select the right tool from a large pool, fill it with correctly typed arguments, and chain calls so that each consumes the outputs of the last. CheMatAgent, a previously published system, addresses this with hierarchical evolutionary MCTS: separate policy and execution models searching tool-call trees under two learned critics, one regressed partly onto GPT-assigned scores. We show that a single policy suffices. Our model interleaves reasoning, tool calls, and returns in one left-to-right generation, trained by a supervised warm-up and then outcome-level reinforcement learning against a programmatic reward read directly off the gold call chain, which leaves no learned critic and no judge in the training loop. On ChemToolBench multiple-tool comprehensive chemistry, on both backbones CheMatAgent use, we improve Tool F1 by 5.5% and Return F1 by 9.6% on Qwen-2.5-7B, and by 3.7% and 3.9% on Llama-3.1-8B, compared with their strongest search configuration, at one model invocation per question, against a search whose cost grows with the tree; we also lead answer Pass Rate on Qwen-2.5-7B.
[NLP-25] Detecting AI Impostors: How Do Middle Schoolers Identify LLM Agents in a Live Collaborative Setting? EMNLP2026
【速读】： 该论文旨在解决生成式 AI (Generative AI) 在社交场景中引发的仿冒与信任危机问题，尤其关注青少年群体在频繁使用生成式 AI 但识别能力不足的背景下，如何感知和应对人工智能的伪装行为。其核心解决方案是设计并验证一种名为 DoppelBot 的协作式社会推理游戏，通过让中学生在互动情境中识别“AI 分身”（doppelgängers），探究其在重复暴露后对 AI 伪造行为的检测能力演变机制。研究发现，随着游戏进程，学生的检测准确率显著提升，关键转变在于从依赖语言特征转向利用共享的社会背景与情境线索进行判断；同时，学生展现出对 AI 在具身性等方面的局限性认知，并能反思数据隐私等深层议题。这一研究为理解青少年在真实社交环境中应对 AI 仿冒的认知策略提供了实证基础，并公开了匿名化的游戏对话与投票数据以支持后续研究。
链接: https://arxiv.org/abs/2608.30948
作者: Dan Schumacher,Pragathi Durga Rajarajan,Haven Kotara,Roman Rendon,Kosi Atupulazi,Deepti Tagare,Ismaila Temitayo Sanusi,Fred G. Martin,Anthony Rios
机构: University of Texas at San Antonio(圣安东尼奥德州大学)
类目: Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026 Main
Abstract:LLMs can imitate how people write, which raises concerns about impersonation, trust, and detection in social settings. These concerns are especially important for adolescents, who use generative AI frequently but may struggle to recognize it. We introduce \textitDoppelBot, a cooperative social deduction game designed to study how young people detect and respond to AI impersonation. Through studies with middle schoolers, we investigate whether a DoppelBot prompts reflection on privacy and impersonation, how repeated exposure affects AI-detection accuracy as agents become more personalized, and which strategies students use to identify AI doppelgängers. We find that students’ detection accuracy improves over time, driven by a shift from relying on linguistic cues to leveraging shared social and contextual signals. Students also demonstrated an understanding of AI limitations such as embodiment and reflected on broader issues such as data privacy. To support future research, we release an anonymized dataset of game transcripts and voting behavior.
[NLP-26] Annotated Surrogate Retrieval for Polish Statutory Law
【速读】： 该论文旨在解决波兰成文法（statutory law）检索中如何在计算成本与检索质量之间取得平衡的问题，尤其关注在不依赖生成式模型调用的前提下实现高效、高精度的法律条文定位。其核心挑战在于：传统关键词检索（如BM25）在语义理解上存在局限，而基于大语言模型（LLM）的密集检索虽提升语义匹配能力，但带来高昂的推理开销和延迟。解决方案的关键在于引入“文档代理”（document surrogates）机制——即在索引阶段对法律条文预先注入由语言模型生成的语义注释，从而在查询时避免实时调用模型，显著降低延迟与成本。具体而言，三种设计分别代表了不同权衡：ASCR采用代理级联+重排序架构；ASCR-H将密集检索结果融合进级联流程以增强性能；而DTF则完全摒弃语言模型阶段，改用三路混合检索（词法+密集）结合加权倒数排名融合（RRF）与确定性重评分，实现零模型调用。实验表明，尽管ASCR-H在前几位排名上显著优于多数基线（如在1位准确率上达72.3%，远超BM25的61.7%和密集检索的52.3%），但其优势集中于头部位置，且随排名深度衰减；相比之下，DTF虽在整体排名表现上落后于部分配置，但在二十位截断下反超，并以约1/9的延迟和不到一半的成本达成接近最优的引用准确率（citation accuracy），证明其在实际应用中的高效性与可扩展性。进一步的消融分析指出，重排序阶段贡献了27.6个百分点的一位准确率提升，凸显其关键作用。此外，研究揭示了一个重要现象：代理注释覆盖仅27.0%的全文档库，但所有基准测试中的参考条文均被覆盖，形成“覆盖不对称”，提示需警惕代理标注分布偏差对评估结果的影响。该工作公开了完整基准、每题输出及配对显著性检验结果，为后续法律信息检索研究提供了可复现的高质量资源。
链接: https://arxiv.org/abs/2608.30929
作者: Orkun Yiğit Cengiz
机构: Independent Researcher(独立研究员)
类目: Computation and Language (cs.CL)
备注: 16 pages, 5 figures, 5 tables. Code and data: this https URL
Abstract:We present a family of retrieval methods for Polish statutory law built on document surrogates: language-model annotations attached to statutory articles at index time. Three designs occupy different points on the cost-quality frontier. ASCR is a surrogate cascade with reranking; ASCR-H fuses a dense list into that cascade; and DTF replaces both language-model stages with three lexical and dense retrievers, weighted reciprocal rank fusion, and a deterministic re-scoring prior, using no model call before generation. We evaluate all three against fourteen lexical, dense, fused and ablated baselines plus four controls, on 300 questions from the 2024 and 2025 Polish bar and legal counsel entrance examinations (264 with their reference article in the corpus), over 82,508 articles from 1,133 acts. On paired McNemar tests, ASCR-H places the reference provision at rank one significantly more often than every other non-oracle configuration except one of its own ablations (eighteen of twenty comparisons significant in its favour at p 0.005), reaching 72.3% against 61.7% for BM25 and 52.3% for dense retrieval. The advantage is concentrated at the head and does not survive depth: it is significant at cutoffs of one and five, disappears by ten, and by twenty DTF leads on point estimate (86.0% versus 84.5%) at one ninth the latency and less than half the cost. Ablation attributes 27.6 points of rank-one accuracy to the reranking stage alone. We further report that the ranking advantage does not extend to citation accuracy, where DTF matches the oracle ceiling, and three negative results on lemmatisation, pseudo-relevance feedback and query rewriting. Surrogate annotation covers 27.0% of the corpus but every reference provision in the benchmark, an asymmetry we disclose and discuss. Benchmark, per-question outputs and paired significance tests are publicly available.
[NLP-27] RIPPULSE: Multi-Agent Travel Planning with Review-Grounded Reasoning EMNLP2026
【速读】： 该论文旨在解决旅行行程生成中如何在严格的时间空间约束与用户真实偏好之间取得平衡的问题，尤其针对现有基于大语言模型（LLM）的规划方法过度依赖结构化属性和预定义用户画像、忽视来自真实用户评论中蕴含的体验性因素（如舒适度、安全性、服务质量、氛围、拥挤程度及隐性风险等）的局限性。其核心解决方案是提出TRIPPULSE，一个以评论为根基的多智能体框架，通过将行程生成任务分解为专门化的局部智能体（分别负责住宿、交通、餐饮、景点和活动），并由全局协调器结合调度机制确保时间与预算可行性，从而突破单体规划器带来的上下文与推理瓶颈。该框架进一步引入包含10万条以上真实评论的数据增强，并提出“评论根基型用户画像对齐”（Review-Grounded Persona Alignment, RGPA）这一基于LLM作为裁判的评估指标，以量化衡量行程与以人为本旅行体验的一致性。实验结果表明，TRIPPULSE在保持强约束满足的前提下，显著提升了行程的个性化程度与体验真实性。
链接: https://arxiv.org/abs/2608.30924
作者: Priyanshu Karmakar,Borru Vijay Sai,Shubhojit Mallick,Abhik Jana,Shreya Ghosh,Manish Gupta
机构: Microsoft(微软); IIT Bhubaneswar(印度理工学院布巴内斯瓦尔)
类目: Computation and Language (cs.CL)
备注: 31 pages, EMNLP 2026
Abstract:Travel itinerary generation requires balancing strict spatio-temporal constraints with human preferences. Existing LLM-based planners mainly rely on structured attributes and pre- defined traveler personas, but real travel deci- sions are often shaped by reviews that reveal experiential factors such as comfort, safety, ser- vice quality, ambiance, crowding, and hidden risks absent from structured databases. Incor- porating such review information is therefore critical to realistic, user-centric itinerary gen- eration. We propose TRIPPULSE1, a multi- agent framework for review-grounded travel planning. Instead of relying on a monolithic planner (and face context and reasoning bot- tlenecks), TRIPPULSE2 decomposes itinerary generation into specialized agents (each op- erating over localized contexts) for accom- modations, transportation, meals, attractions, and events, coordinated through a global or- chestrator with scheduling mechanisms that enforce temporal and budget feasibility. We augment TRIPCRAFT with 100K+ real-world reviews and introduce Review-Grounded Per- sona Alignment (RGPA), an LLM-as-a-Judge metric for evaluating alignment with human- centric travel experiences. Experiments across multiple trip durations and diverse proprietary and open-source models show that TRIPPULSE maintains strong constraint satisfaction while generating more personalized and experien- tially grounded itineraries.
[NLP-28] Responsible Integration of AI in Cancer Genomics: Barriers Risks and Pathways to Trustworthy Clinical Translation
【速读】： 该论文旨在解决生成式AI（Generative AI）在癌症基因组学领域从科研到临床常规应用转化缓慢的核心问题，其核心挑战并非单纯的计算能力不足，而在于如何实现可信且无缝集成至临床工作流程。解决方案的关键在于构建一个系统性框架，通过严格验证、考虑不确定性的方法、可互操作的基础设施、监管一致性以及贯穿AI生命周期的人工监督，协同应对四大相互关联的转化失败领域：证据不一致、可解释性与不确定性、数据治理与可重复性、以及互操作性。研究强调，推动生成式AI在临床中的常规应用，关键不在于持续提升模型性能，而在于系统性地解决这些跨阶段、交互式的转化障碍，实现从开发到部署及部署后监测的全链条优化。
链接: https://arxiv.org/abs/2608.30912
作者: Bahar İlgen,Yiannos Tolias,Denise Kühnert,Paraskevi Papadopoulou,Magnus Westerlund,Dominik Heider,Katharina Ladewig,Georges Hattab
机构: Robert Koch Institute (德国罗伯特·科赫研究所); Centre for Artificial Intelligence in Public Health Research (ZKI-PH) (公共健康人工智能研究中心); European Commission (欧洲委员会); Directorate-General for Health and Food Safety (健康与食品安全总司); Deree-The American College of Greece (希腊美国学院德里); Department of Natural Sciences (自然科学系); Åbo Akademi University (Åbo阿克迪米大学); Arcada University of Applied Sciences (阿卡达应用科学大学); University of Münster (明斯特大学); Institute of Medical Informatics (医学信息学研究所); Medical University Lausitz - Carl Thiem (拉乌齐茨医科大学-卡尔·蒂姆); Department of Mathematics and Computer Science (数学与计算机科学系); Freie Universität Berlin (柏林自由大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Review article
Abstract:Artificial intelligence (AI) and natural language processing (NLP) are increasingly used to extract, integrate, and interpret biomedical knowledge relevant to cancer genomics, yet their translation into routine clinical oncology has been comparatively slow. The central challenge is not computational capability alone, but trustworthy integration into clinical workflows. This review examines how NLP and AI support the cancer genomics pipeline, from literature mining and automated variant interpretation to clinical trial matching, knowledge graph construction, and multimodal data integration. We identify four interrelated translational failure domains: evidence inconsistency, explainability and uncertainty, data governance and reproducibility, and interoperability. Rather than considering these challenges in isolation, we take a systems-level view, focusing on their interaction across the translational pathway. We propose a conceptual framework and roadmap for addressing these domains through rigorous validation, uncertainty-aware methods, interoperable infrastructures, regulatory alignment, and human oversight across the AI lifecycle. Progress toward routine clinical use will depend less on further improving model capability than on systematically addressing these interacting failure domains from development through deployment and post-deployment monitoring.
[NLP-29] S3C-LLM : Skill-Code Guided Agent ic Language Models for Spectrum-to-Structure Elucidation EMNLP2026
【速读】： 该论文旨在解决现有基于大语言模型（LLM）的光谱结构解析方法在建模实际光谱学家分析流程上的不足问题。当前主流方法将结构解析视为直接从光谱到SMILES的生成任务，虽能利用配对数据，但未能显式建模诊断峰解析、片段推理、分子式约束及化学一致性验证等关键分析步骤。其解决方案的核心在于提出S3C-LLM——一种技能引导且代码驱动的代理型大语言模型，通过检索特定模态的光谱分析技能，执行预定义分析代码以在输入光谱上实例化这些技能，进而整合峰级证据与分子式约束后生成最终的SMILES。关键技术贡献包括：自演化的光谱技能库、增强思维能力的技能-代码轨迹构建管道，以及分阶段训练策略（先监督微调，再引入逐步强化学习）。实验表明，S3C-LLM在多种基准测试中均显著优于现有通用和谱图专用模型，且仅需SpectraLLM训练数据量的1/10。
链接: https://arxiv.org/abs/2608.30910
作者: Xuanle Zhao,Xinyuan Cai,Xiang Cheng,Bo Xu
机构: The Key Laboratory of Cognition and Decision Intelligence for Complex Systems, Institute of Automation, Chinese Academy of Sciences; School of Artificial Intelligence, University of Chinese Academy of Sciences
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注: Accepted by EMNLP 2026 Findings
Abstract:Spectroscopic structure elucidation is central to molecular analysis, but recent Large Language Model (LLM)-based methods mostly formulate it as direct spectrum-to-SMILES generation. Although this paradigm can leverage paired spectral data, it does not explicitly model the analytical workflow used by spectroscopists, such as diagnostic peak interpretation, fragment reasoning, formula constraints, and chemical consistency checking. In this paper, we introduce S3C-LLM, a skill-guided and code-grounded agentic LLM for spectrum-to-structure elucidation. Rather than directly predicting a molecule, S3C-LLM retrieves modality-specific spectroscopy skills, executes analysis code to instantiate these skills on the input spectra, and integrates the resulting peak-level evidence and formula constraints before generating SMILES. Specifically, we contribute a self-evolving spectroscopy skill library, a thinking-augmented skill-code trajectory construction pipeline, and a two-stage training strategy that teaches Qwen3-4B through supervised fine-tuning (SFT) followed by our proposed step-level reinforcement learning (RL). Experiments on diverse benchmarks show that S3C-LLM consistently outperforms current general LLMs and spectrum-specific models across spectra, while using less than 1/10th of SpectraLLM’s training corpus.
[NLP-30] MMDS-Bench: Benchmarking Multimodal Large Language Models on Dynamic Stance in Social Media Interactions EMNLP2026
【速读】： 该论文旨在解决社交媒体中多模态动态立场分类（Multimodal Dynamic Stance Classification, MDSC）的挑战，即如何准确识别回复对直接父消息（parent message）所持立场的动态变化，而非仅关注帖子与固定话题的关系。现有研究主要局限于纯文本场景，而当前社交媒体互动日益依赖图像、截图、表情包、反应图片及跨模态引用等多模态内容，导致现有模型在理解复杂交互语境时表现不足。为此，论文提出MMDS-Bench——一个用于诊断多模态动态立场分类的基准数据集，包含3,482个标注了七类动态立场标签的多模态实例，并附带800个需结构化推理的诊断子集，涵盖父消息理解、回复理解与立场关系推断。同时，每个实例被进一步标注五类挑战因素：多模态融合、父消息框架、非字面表达、互动推理以及标签边界模糊性。研究评估了12个开源与闭源多模态大语言模型（Multimodal Large Language Models, MLLMs），并提出一种基于参考答案的LLM评判协议以衡量推理质量。实验结果表明，当前MLLMs在需要超越独立解析父消息与回复的关联推理任务上仍存在显著局限，凸显了对更深层次语义关联建模的需求。解决方案的关键在于构建具备细粒度标注与挑战因素分析的诊断性基准，并引入基于参考的判别机制以系统评估模型的推理能力。
链接: https://arxiv.org/abs/2608.30903
作者: Yuzhe Ding,Kang He,Li Zheng,Shengwu Zheng,Teng Shi,Fei Li,Chong Teng,Donghong Ji
机构: Wuhan University (武汉大学); Shanghai Innovation Institute
类目: Computation and Language (cs.CL)
备注: Accepted by EMNLP 2026
Abstract:Dynamic stance classification models how a reply responds to its direct parent message, rather than how a post relates to a fixed topic. Existing work has mainly studied this problem in text-only settings, while social media interactions increasingly rely on images, screenshots, memes, reaction images, and cross-modal references. We introduce MMDS-Bench, a diagnostic benchmark for multimodal dynamic stance classification in social media parent-reply interactions. MMDS-Bench contains 3,482 multimodal instances annotated with a seven-label dynamic stance taxonomy, together with an 800-instance diagnostic subset that requires structured reasoning over parent understanding, reply understanding, and stance-relation inference. We further annotate each instance with five challenge factors covering multimodal fusion, parent framing, non-literal expression, interaction reasoning, and label-boundary ambiguity. We evaluate 12 closed-source and open-source multimodal large language models and propose a reference-grounded LLM-judge protocol for assessing reasoning quality. Results show that current MLLMs still struggle with multimodal dynamic stance understanding, especially in cases that require relational inference beyond separate parent and reply comprehension.
[NLP-31] Low-Resource Preference Adaptation of LLM s via Activation-Based Label Propagation EMNLP26
【速读】： 该论文旨在解决在低资源场景下，由于人工标注成本过高而导致偏好优化（preference optimisation）难以实施的问题，尤其针对那些无法通过大语言模型（LLM）自身可靠标注的个性化、文化或主观性较强的偏好情境。其核心挑战在于：主流对齐后的语言模型在面对非主流群体或非标准偏好时，其判断能力显著下降。论文的关键发现是，语言模型在中间层激活（intermediate representations）中隐含了可区分的偏好信息结构——优选响应与劣选响应的激活模式在不同层中形成明显分离的聚类，且该结构在经过标准数据集对齐后被强化，但在目标偏好偏离原始对齐分布时则被破坏。基于此现象，作者提出一种轻量级线性探针（linear probe）方法：仅需少量标注的偏好对（≤500对），即可训练探针以自动标注大规模未标注数据集（>5万条），进而支持下游偏好优化。实验表明，该方法在多种数据集、优化方法及模型规模下均显著优于同等标注预算下的直接训练，且性能可媲美使用50–100倍更多标注数据的基线模型，有效缓解了低资源环境下偏好学习的瓶颈。
链接: https://arxiv.org/abs/2608.30902
作者: Alessio Galatolo,Meriem Beloucif
机构: Uppsala University (乌普萨拉大学)
类目: Computation and Language (cs.CL)
备注: EMNLP26 Main
Abstract:Adapting large language models to user-specific preferences is often constrained by the cost of human annotation, making preference optimisation impractical in low-resource settings where preferences cannot be reliably labelled by LLMs themselves, e.g., due to cultural, subjective, or personalised contexts. In this paper, we investigate how language models encode preference information in their intermediate representations, finding that activations from chosen and rejected responses form distinct clusters across layers, even in pretrained models. Strikingly, this structure is strengthened by alignment on canonical datasets but erased when the target preferences differ from those the model was aligned on, suggesting aligned LLMs are poor judges for non-mainstream populations. Exploiting this structure, we propose training a lightweight linear probe on a few labelled preference pairs ( \leq 500) and using it to annotate large unlabelled datasets (50K+) for downstream preference optimisation. We systematically evaluate this approach across different datasets, preference optimisation methods and model scales and find that our method consistently outperforms direct training given the same annotation budget, and remains competitive against baselines trained on 50-100\times more labelled data in the majority of our settings. Code is available at this https URL.
[NLP-32] Evaluating and Mitigating Anti-LGBTQ Biases in German and Multilingual Language Models EMNLP2026
【速读】： 该论文旨在解决多语言环境下对语言模型中反跨性别及反同性恋（anti-LGBTQ）偏见的系统性评估不足问题，尤其关注非英语语境下的文化与语言差异。现有基准测试普遍依赖性别表征且缺乏跨文化适配性，难以有效捕捉多元语境中的隐性偏见。为此，论文提出一个德英双语的基准数据集，融合德语社群提供的酷儿群体刻板印象与英文版WinoQueer的德语翻译，以更真实地反映德语语境下的反酷儿话语。研究评估了八种不同规模与架构的语言模型，并探索通过社区媒体及进步媒体内容微调以缓解偏见的有效性。结果表明，语言模型普遍复现反酷儿刻板印象，且偏见表现随身份类型与模型结构存在显著差异；翻译数据与社群源数据之间的差异凸显了多语言偏见评估中文化适应的重要性。尽管微调在整体上降低了偏见水平，但其效果在不同模型与身份类别间并不一致，揭示出当前缓解策略的局限性。
链接: https://arxiv.org/abs/2608.30884
作者: Melina Morch,Daniel Braun
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted at EMNLP 2026
Abstract:While gender and racial biases in language models have been widely studied, anti-LGBTQ biases remain underexplored, particularly beyond English. Existing benchmarks often do not capture cultural and linguistic variation and rely on gender representations. This paper introduces a multilingual German-English benchmark dataset for the evaluation of anti-LGBTQ biases in language models. It combines community-sourced stereotypes from German-speaking queer individuals with a German translation of WinoQueer. The data is used to evaluate eight language models across sizes and architectures and explore mitigation through fine-tuning on community and progressive media content. Results show that language models reproduce anti-queer stereotypes, with variation across identities and models. Differences between the translated and community-based data highlight the importance of cultural adaptation for multilingual bias evaluation. Fine-tuning reduces bias on average, but not consistently across models and identities. Warning: This text contains examples of anti-queer hateful language and stereotypes.
[NLP-33] Personas Differ from Native-Language Generation: Language Pathways Shape LLM Interpersonal Advice EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在跨语言人际建议研究中，因采用不同跨语言诱发策略而可能导致结果偏差的问题。具体而言，研究关注两种常见方法——以“母语者角色”（native-speaker persona prompting, NP）进行提示与直接在目标语言生成后翻译回英文（native-language generation followed by translation, NL）——是否可互换使用。其解决方案的关键在于系统性地对比这两种方法在多个维度上的表现差异：包括语言风格、行为支持结构（behavioral scaffolding）以及强制选择情境下的行动推荐倾向。研究发现，NP与NL在词汇社会线索（如亲和性、积极语气）的强度、具体性与社会契合度等方面存在显著差异，且NP往往降低建议的可操作性，并在强制选择任务中更倾向于推荐对抗性而非回避性行为，其效应大小随语言、话题和模型类型而异。因此，研究强调跨语言诱发策略是一个具有实质性影响的方法论选择，会显著改变模型输出的框架方式及推荐行为，提示研究者需谨慎对待提示设计对跨文化研究结果的潜在干扰。
链接: https://arxiv.org/abs/2608.30873
作者: Jinhee Won,Xinlan Emily Hu
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted to EMNLP 2026 (Main Conference). 36 pages, 24 figures, 17 tables (9 pages main text; remainder references and appendices)
Abstract:LLMs are increasingly used for interpersonal advice and as tools for studying social behavior across languages and cultures. A common shortcut for eliciting language- or culture-related variation is to ask a model to answer as a native speaker. We test whether this native-speaker persona reproduces the outputs obtained when models instead generate advice in the target language and translate the response back into English. Using 600 interpersonal advice questions across 13 languages and eight LLMs, we compare native-language generation followed by translation (NL) with native-speaker persona prompting (NP), measuring linguistic style, behavioral scaffolding, and forced-choice action recommendations. We find that NP and NL are not interchangeable. Compared to NL, NP often increases lexical social cues, including affiliation and positive tone, while reducing qualities such as concreteness and social attunement; NP also provides less actionable scaffolding in open-ended advice. In forced-choice scenarios, NP changes which action the model selects, favoring confrontation over redirection, with effect sizes varying across languages, topics, and models. Our results show that cross-lingual elicitation strategy is a consequential methodological choice that can change both how advice is framed and which actions models recommend.
[NLP-34] Beyond Good Intentions: When Does the Framing of Multilingual and Low-Resource NLP Research Become a Caricature? EMNLP2026
【速读】： 该论文旨在解决当前自然语言处理（Natural Language Processing, NLP）与机器学习（Machine Learning, ML）研究中，针对低资源语言（low-resource languages）及边缘化文化所采用的学术叙事可能隐含的不平等结构性问题。研究指出，尽管许多工作以促进社区赋权、实现去殖民化（decolonisation）和推动包容性知识生产为宗旨，但其实际框架常依赖于重复性的修辞模式，弱化了对真实社区影响的实证支持。解决方案的关键在于提出一个系统化的分析框架，用于识别这些研究中的典型话语结构，并揭示其如何在缺乏充分证据的情况下夸大社区受益程度。通过分析ACL Anthology的数据，研究发现多数成果更侧重于资源构建与基准测试等技术目标，而非体现深层次的社会结构变革。因此，论文强调应强化研究的问责机制，倡导作者、审稿人与读者共同批判性评估研究动机与成效，确保研究真正服务于并由边缘化社区主导。
链接: https://arxiv.org/abs/2608.30866
作者: Nedjma Ousidhoum,Noopur Zambare,Mohamed Abdalla
机构: Cardiff University (卡迪夫大学); University of Alberta (阿尔伯塔大学)
类目: Computation and Language (cs.CL)
备注: Accepted at EMNLP2026 (Main)
Abstract:Building language technologies and conducting NLP research for low-resource languages—particularly when led by native speakers or involving participatory research practices—are often framed as means of addressing inequality, serving local communities, and, at times, contributing to decolonisation. In this paper, we examine recently published NLP and ML papers, focusing on the narratives used to characterise multilinguality, low-resource languages, and underrepresented cultures. We propose a framework for analysing research framings and identify recurring rhetorical patterns that may hinder accountability and constrain equitable knowledge production for—and by—underserved communities. We further assess the evidential basis of assertions regarding community benefit and find that such statements are often weakly supported or left unsubstantiated. Although community ownership and participation are frequently presented as key objectives, our analysis, supported by statistics from the ACL Anthology, suggests that research outputs more often prioritise resource creation and benchmarking—important but distinct goals—over evidence of broader structural change. We conclude by offering practical recommendations to help authors, reviewers, and readers critically assess these assertions and avoid potentially misleading framings.
[NLP-35] Linguistic Distance Segregates Latent Representations in Automatic Speech Recognition Systems EMNLP
【速读】： 该论文旨在解决英语自动语音识别（ASR）模型在不同母语背景说话者群体中表现不一致的问题，尤其关注母语（L1）与英语语言家族距离较远的说话者所面临的性能下降问题。其核心挑战在于揭示并量化母语背景对英语ASR性能的影响机制。解决方案的关键在于通过实证分析发现，说话者的母语与英语的语言距离与其ASR错误率之间存在系统性关联，且这一关联在多种数据集和模型中均具有统计显著性（采用Tweedie混合效应模型进行校正后，所有评估模型的p < 0.001）。此外，通过对模型深层声学层的潜在空间（latent space）分析，进一步揭示了多数模型架构中存在基于母语的特征空间分隔现象，表明模型内部表征已隐含母语影响，从而为理解并缓解跨语言偏差提供了关键洞见。
链接: https://arxiv.org/abs/2608.30853
作者: Ting-Hui Cheng,Line Katrine Harder Clemmensen,Sneha Das
机构: Department of Applied Mathematics and Computer Science; Technical University of Denmark (丹麦技术大学)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: to appear in EMNLP finding 2026
Abstract:While automatic speech recognition (ASR) models have achieved remarkable improvements in recent years, performance disparities persist across different speaker populations. One such disparity is for speakers whose first languages (L1) are from families distant from English. This paper investigates the relationship between first language background and English ASR performance. Through empirical analysis, we observe that the correlation between speakers’ L1 distance and ASR error rates yields a systematic effect on English Speech, with its strength varying across datasets and models. This association is statistically significant in a follow-up analysis accounting for dataset-level variation in Tweedie mixed-effects models ( p0.001 across evaluated models). In addition, analysis of the latent space reveals a L1-based spatial segregation across deeper acoustic layers in the majority of evaluated architectures
[NLP-36] hesis Proposal: Toward a Human-Centered and Perspective-Aware Framework for Reproducible ML Evaluation and AI Alignment ACL
【速读】： 该论文旨在解决当前大语言模型（LLM）评估与对齐过程中忽视人类意见分歧的问题。现有评估方法普遍采用多数投票等聚合策略来构建共识标签，从而掩盖了少数观点和多样性视角，尤其在涉及主观判断的领域（如人工智能安全、内容审核、情感分析）中，这种做法加剧了评估结果的不可复现性，并可能导致模型偏离真实多元的人类价值观。其解决方案的关键在于提出一种以人为本且具备视角感知（perspective-aware）的可复现机器学习评估与对齐框架，通过显式建模和整合人类间的分歧与多样性，确保AI系统不仅反映主流意见，更能包容不同身份、信念和社会背景下的多元价值，从而提升AI系统的可信度与社会适应性。
链接: https://arxiv.org/abs/2608.30842
作者: Deepak Pandita,Christopher M. Homan
机构: Rochester Institute of Technology (罗切斯特理工学院)
类目: Computation and Language (cs.CL)
备注: Published at ACL SRW 2026: this https URL
Abstract:Humans play a vital role at every stage of AI development, from data collection and curation to model development and evaluation. However, humans often disagree with each other and sometimes with themselves over time. It is essential to take disagreement into account when building human-centered AI systems, especially in domains where it is prevalent, such as AI safety, content moderation, or sentiment analysis. Disagreement often arises from subjective human opinion and can vary with one’s identity, beliefs, and social environment. Despite this, current LLM evaluation approaches frequently rely on aggregating labels (often via plurality voting) to represent consensus, thereby obscuring minority perspectives. By failing to account for human disagreement, these evaluation methods contribute to the reproducibility crisis in AI. Human feedback is also crucial for ensuring that AI systems align with human values. For these systems to be trustworthy, it is critical to ensure that they reflect diverse human values and perspectives. In this thesis proposal, we present a human-centered and perspective-aware framework for reproducible ML evaluation and AI alignment.
[NLP-37] HSRM: Hidden-State Reward Models for Test-Time Verification EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在数学推理任务中生成多个合理解题路径后，如何高效且可靠地筛选出正确解的问题。现有测试时推理流水线依赖文本验证器对每个候选解进行重新读取与评估，导致验证过程计算开销巨大。针对此问题，论文提出一种轻量级隐状态奖励模型（Hidden-State Reward Model, HSRM），其核心创新在于不依赖对生成文本的二次处理，而是直接读取生成器在推理步骤边界处的内部表示（hidden states）以实现验证。HSRM通过从冻结的生成器中提取隐藏状态，并利用小型Transformer编码器对候选解进行排序，从而实现高效的解题路径选择。该模型基于自生成轨迹及其正确性标签进行训练，无需人工标注的推理过程监督或大型预训练验证器。在四个数学推理基准测试中，HSRM在15/16的生成器-数据集组合下表现达到或优于一个5500万参数的纯文本能量型验证器，同时仅需约200万参数，显著提升了验证效率，充分利用了生成过程中已有的内部表示信息。
链接: https://arxiv.org/abs/2608.30841
作者: Xianzhi Li,Xiaodan Zhu
机构: Queen’s University(皇后大学); Ingenuity Labs Research Institute(创新实验室研究院)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: EMNLP 2026
Abstract:Large language models can often generate plausible mathematical reasoning traces, but reliably identifying the correct solution among multiple candidates remains a key challenge. Existing test-time reasoning pipelines typically rely on text-based verifiers that re-read each generated solution, making verification an expensive component of inference. Prior work has shown, however, that LLMs often encode correctness-related signals in their internal representations, including awareness of when their own answers are likely to be wrong. Building on this observation, we introduce HSRM, a lightweight hidden-state reward model that verifies candidate solutions by directly reading the generator’s internal representations rather than re-processing its text. HSRM extracts hidden states from a frozen generator at reasoning-step boundaries and uses a small Transformer encoder to rank candidates. It is trained from self-generated trajectories with outcome labels, requiring neither human-written process supervision nor a large pretrained verifier. Across four mathematical reasoning benchmarks, HSRM matches or outperforms a 55M-parameter text-only energy verifier in 15 of 16 generator–dataset settings while using only about 2M parameters, providing an efficient alternative to text-only verification by reusing representations already computed during generation.
[NLP-38] Opinionated Hesitant and Stressed: Three Studies of How Politicians Speak in Four Slavic Parliaments
【速读】： 该论文旨在解决多语言、大规模议会口语中语音特征与语用因素之间关系的跨语言比较问题，特别是揭示情感、话语节奏、性别与年龄等变量如何影响语音实现模式。其核心解决方案在于利用涵盖四种斯拉夫语（克罗地亚语、捷克语、波兰语、塞尔维亚语）的6000余小时ParlaSpeech 3.0语料库，通过三组大规模实证研究，系统性地分析了句级情感对声学实现的影响、填充停顿频率的跨语言预测模型以及克罗地亚语中词重音分布的句法-语义依赖性。关键突破在于：（1）发现负面言辞在所有议会中均呈现更高基频、更强强度和更快语速的共性特征，且高度积极情绪下存在由唤醒度驱动的声学上扬；（2）采用负二项分布广义估计方程（negative binomial GEE）建模填充停顿，识别出语速、年龄和情感为跨语言稳健预测因子，同时揭示性别效应在南斯拉夫与西斯拉夫议会间的反转现象，凸显多语言设计对隐藏模式的敏感性；（3）发现克罗地亚语中动词、形容词和名词的重音偏好具有一致性，而副词和专有名词则表现出独立于整体趋势的解耦特征，表明重音策略具有句法类别特异性。这一系列发现为话语语音学、言语不流畅性建模及政治修辞分析提供了可扩展的跨语言研究范式。
链接: https://arxiv.org/abs/2608.30828
作者: Ivan Porupski,Nikola Ljubešić
机构: 未知
类目: Computation and Language (cs.CL)
备注: 11 pages, 3 figures
Abstract:We present three large-scale studies of spoken parliamentary speech across four Slavic languages (Croatian, Czech, Polish, Serbian), drawing on over 6,000 hours from the ParlaSpeech 3.0 corpus. The first study examines how utterance-level sentiment shapes acoustic realisation: negative speech is consistently produced with higher pitch, greater intensity, and faster rate across all four parliaments, with a secondary arousal-driven upturn at the most positive extreme. The second study models filled pause frequency using negative binomial GEE, finding that speech rate, age, and sentiment are robust cross-lingual predictors, while gender effects reverse between South Slavic (men produce fewer filled pauses) and West Slavic parliaments (no gender difference) - a pattern invisible to single-language designs. The third study investigates primary stress variation in Croatian, showing that speaker-level preferences for early versus late stress cohere across verbs, adjectives, and nouns but decouple for adverbs and proper nouns. We conclude with a research agenda spanning corpus phonetics, disfluency modelling, and political rhetoric.
[NLP-39] Error-Type-Aware Loss Reweighting for Robust Named Entity Recognition with Noisy LLM Labels
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLM）在生成标注数据以训练特定任务的小型模型（如命名实体识别，Named Entity Recognition, NER）时引入的标注噪声问题。现有微调方法通常忽略由LLM产生的标注噪声，导致模型性能下降；而现有的抗噪声损失函数因无法适应NER中异质性噪声（如遗漏实体提及和类型错误对训练信号的影响机制不同）而效果有限，若对所有噪声标记采用统一重加权策略，可能削弱有效监督信号或强化错误标签。为此，本文提出一种面向错误类型的损失重加权方法，针对不同类型潜在错误的标记设计独立的重加权规则，实现更精准的噪声处理。该方法简单高效，无需额外训练资源，在噪声水平为15%至40%的情况下，使数据集层面平均F1提升0.8–2.0个百分点，最大提升达4.6个百分点（在Wikigold数据集上噪声率为24.1%时）。
链接: https://arxiv.org/abs/2608.30827
作者: Elena Merdjanovska,Jonas Golde,Alan Akbik
机构: Humboldt-Universität zu Berlin(柏林洪堡大学); Science of Intelligence(智能科学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Large language models are increasingly used to annotate datasets for training smaller, task-specialized models such as named entity recognition. While this method yields effective models, it assumes that the synthetic dataset is correctly annotated. In this work, we find that (i) current fine-tuning processes simply ignore LLM-introduced annotation noise, resulting in degraded performance and (ii) existing noise-robust losses are not transferable to sequence labeling because annotation noise in named entity recognition is heterogeneous: for example, missing mentions and type errors affect the training signal in different ways. Treating all noisy tokens equally in noise-robust losses and applying a single reweighing criterion for all may therefore remove useful supervision or reinforce incorrect labels. To address this limitation, we propose error-type-aware loss reweighting for NER, which introduces separate reweighing rules for different types of potentially erroneous tokens. Our approach is simple and efficient, does not require additional training resources, and improves F1 by 0.8 - 2.0 percentage points on dataset-level average for noise levels between 15% and 40%, with a maximum improvement of 4.6 percentage points with 24.1% noise on Wikigold.
[NLP-40] Vocal Music under Phoneme-Conditional Analysis
【速读】： 该论文旨在解决不同语言的无伴奏人声演唱是否具有可测量且可追溯至特定音素的声学差异这一问题。其核心解决方案在于提出“音素条件分析”（phoneme-conditional analysis），通过在相同歌曲内对比标记音素（marker syllables）与匹配的非标记对照音素，同时控制演唱者、旋律和音乐体裁等因素，从而隔离出具有类型学差异的音素对声学特征的影响。研究在九种类型学上多样化的语言及数千首歌曲中，沿五个声学维度进行量化分析，构建了基于歌曲层面的声学特征谱图，在艺术家分组的九分类任务中实现了85.5%的平衡准确率，表明语音结构会在语言演唱方式中留下系统且可测量的声学痕迹。
链接: https://arxiv.org/abs/2608.30823
作者: Hayoon Kim,Kyogu Lee
机构: 未知
类目: ound (cs.SD); Computation and Language (cs.CL)
备注: Accepted to the 27th International Society for Music Information Retrieval Conference (ISMIR 2026)
Abstract:The vocal music of each language carries a distinctive sonic identity, even without instrumental accompaniment. We ask whether these differences are measurable and traceable to specific phonemes. To tackle this question, we introduce phoneme-conditional analysis, which isolates the acoustic effect of typologically distinctive phonemes by comparing marker syllables against matched non-marker controls within the same song, holding singer, melody, and genre constant. Across nine typologically diverse languages and thousands of songs, we measure effects along five acoustic dimensions. Song-level profiles built from these effects identify the language of an unaccompanied vocal at 85.5% balanced accuracy in a nine-way classification with folds grouped by artist; whether the separability arises by accumulation of the phoneme-local effects themselves is left open. Our findings suggest that phonological structure leaves systematic and measurable traces in how each language is sung.
[NLP-41] opoCompress: Long Context Compression via Graph-Wired Semantic Trajectories
【速读】： 该论文旨在解决大语言模型在长上下文推理中因上下文过长导致的计算成本高、延迟大等问题，尤其针对现有压缩方法存在的关键缺陷：容易割裂重要信息、需额外训练或对齐、且依赖目标模型实现有效压缩。其解决方案的关键在于提出一种无需训练、模型无关（training-free and model-agnostic）的上下文压缩框架TopoCompress，通过选择语义连贯的文本片段来实现高效压缩。具体而言，该方法首先结合密集语义查询相关性与词法查询相关性，并引入语义加速机制对每个文本片段进行评分；随后构建一个混合图结构，基于语义相似性和序列邻接关系连接各片段，并在图上进行查询引导的相关性传播。该设计使重要信息得以保留并增强，同时显著降低压缩资源消耗。实验结果表明，TopoCompress在五个长上下文任务（HotpotQA、2WikiMQA、MuSiQue、Qasper、MultiFieldQA-en）上均优于现有强基线，且在仅使用四分之一压缩预算的情况下达到相近性能，在最快基线基础上实现1.41倍的压缩速度提升。
链接: https://arxiv.org/abs/2608.30811
作者: Daniel Agyei Asante,Yang Li
机构: Iowa State University (爱荷华州立大学)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 13 pages
Abstract:Long-context compression is essential for reducing the cost and latency of large language model inference. However, existing methods can fragment important evidence, require additional training or alignment, and often depend on the target model for effective compression. We introduce TopoCompress, a training-free and model-agnostic framework that compresses long contexts by selecting coherent semantic spans. TopoCompress first scores each span using dense and lexical query relevance together with semantic acceleration. It then constructs a hybrid graph that connects spans based on semantic similarity and sequential adjacency, and propagates the query-guided relevance scores over the graph. Across five long-context tasks-HotpotQA, 2WikiMQA, MuSiQue, Qasper, and MultiFieldQA-en-TopoCompress consistently outperforms strong compression baselines. Notably, TopoCompress achieves performance comparable to the strongest baseline while using a 4x smaller compression budget, and provides a 1.41x smaller compression time over the fastest baseline.
[NLP-42] CLIN: an Objective Framework for Evaluating Creativity in Short Persian Literary Text EMNLP2026
【速读】： 该论文旨在解决大语言模型（LLM）在评估短篇波斯语文学文本创造力时的可靠性问题，尤其关注低资源语言环境下多维度创造力评价的挑战。其核心问题是：现有基于LLM的创造力评估方法在不同维度上表现不一，且对提示词设计高度敏感，难以实现稳定、可解释的评价。解决方案的关键在于提出一种结构化、可解释的代理指标框架——CLIN，通过三个针对特定创造力维度的简单度量方式分别进行评估：基于主题感知新颖性的原创性（Originality）、基于上下文词汇聚类的流畅性（Fluency），以及基于词汇多样性的丰富性（Elaboration）。这些代理指标在保持与人类评价高度一致的同时，显著降低了评估成本，且优于当前最强的零样本LLM评判者，从而为低资源语言中的创造力评估提供了高效、可解释的新路径。
链接: https://arxiv.org/abs/2608.30754
作者: Mohammad Reza Modarres,Armin Tourajmehr,Yadollah Yaghoobzadeh,Mohammad Taher Pilehvar
机构: University of Tehran(德黑兰大学); Khatam University(卡塔姆大学); Cardiff University(卡迪夫大学)
类目: Computation and Language (cs.CL)
备注: Accepted to Findings of EMNLP 2026
Abstract:Evaluating creativity in large language model (LLM) outputs remains challenging because creativity is multidimensional and human-centered. We examine how reliably LLMs evaluate short literary text in Persian, a low-resource language, across multiple evaluation strategies and prompt formulations. We find that LLM-human agreement varies substantially across dimensions: alignment is stronger for structured TTCT-derived properties such as Originality, Fluency, and Elaboration, but considerably weaker for more subjective dimensions, particularly Emotion and Attractiveness. Judgments are also sensitive to prompt formulation, while few-shot prompting, ensembling, and multi-agent debate provide no consistent improvement. Motivated by this dimension-dependent behavior, we investigate whether structured creativity dimensions can instead be approximated using simple, interpretable proxy metrics. We introduce CLIN, which evaluates three TTCT-derived dimensions separately using topic-aware novelty for Originality, contextual lexical clustering for Fluency, and lexical diversity for Elaboration. These proxies achieve human alignment comparable to or better than the strongest zero-shot LLM judge in our setting while requiring substantially lower evaluation cost.
[NLP-43] he Frag ility of Jailbreak Robustness Across Operational States EMNLP2026
【速读】： 该论文旨在解决现有越狱攻击（jailbreak attack）评估方法中存在的关键缺陷：当前评估普遍依赖于单一默认配置（即“原始状态”，vanilla state）下的攻击成功率（ASR），而忽视了用户与大语言模型（LLM）交互过程中可能引发的多种非原始操作状态（non-vanilla operational states）对模型安全性的潜在影响。研究发现，即使攻击本身保持不变，仅通过修改一个未针对安全性设计的普通系统提示词（system prompt），即可导致攻击成功率发生显著波动，最大增幅达56个百分点（从2%升至58%）。这一现象表明，模型的越狱鲁棒性对操作状态高度敏感，且这种状态依赖性在七种对齐模型和三种代表性越狱攻击中均被系统性观察到。进一步分析揭示，这种鲁棒性变化与拒绝相关隐藏表征（refusal-related hidden representations）的差异密切相关，沿该轴的投影可有效预测越狱成功与否。因此，论文提出的核心解决方案在于：不应仅依赖单一原始状态下的评估，而应扩展评估框架，系统考察模型在多种非原始操作状态下的鲁棒性表现，以更全面、真实地反映模型在实际应用中的安全性。
链接: https://arxiv.org/abs/2608.30748
作者: Yuna Park,Hwang Youn Kim,Yujin Kim,Won Woo Ro,Suhyun Kim,Jae-In Hwang
机构: Yonsei University (延世大学); Korea Institute of Science and Technology (KIST) (韩国科学技术院); Korea University (高丽大学); Kyung Hee University (庆熙大学)
类目: Cryptography and Security (cs.CR); Computation and Language (cs.CL)
备注: Accepted to Findings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026)
Abstract:Existing jailbreak evaluations typically characterize robustness using a single attack success rate (ASR) measured in a default configuration (the vanilla state). However, user-LLM interactions can induce diverse operational states beyond the vanilla state. In this work, we find that jailbreak robustness is highly fragile to operational-state variation: even when the attack remains fixed, changing only an ordinary system prompt not designed to affect safety can dramatically alter attack success rates. We systematically investigate this phenomenon across seven aligned models and three representative jailbreak attacks, observing substantial differences in ASR between vanilla and non-vanilla operational states. In one case, ASR increases by up to 56 percentage points (2% to 58%) solely due to a change in operational state. Remarkably, these increases occur even for attacks originally designed and optimized under vanilla-state evaluation. We further show that state-dependent robustness variation is systematically associated with differences in hidden representations along a refusal-related axis, and that projections onto this axis strongly predict jailbreak outcomes. Our results show that a single vanilla-state evaluation may not fully characterize jailbreak robustness, motivating evaluations that also examine how robustness changes across non-vanilla operational states.
[NLP-44] Not All Fallbacks Are Failures: Understanding and Recovering from Fallbacks in Mobile Voice Assistants EMNLP2026
【速读】： 该论文旨在解决智能语音助手在真实应用场景中面对多样化失效情境（如噪声干扰、识别错误、语义模糊、不完整指令或误唤醒等）时，因采用通用化回复策略而导致交互失败无法有效缓解、进而降低用户体验的问题。其核心解决方案在于构建一个基于真实世界使用数据的系统性分析框架，关键创新包括：提出一套可操作的失效交互分类体系，并公开了包含3,030条匿名自然语料的VoxFallbacks数据集；在实际部署约束下对比评估多种模型在分类流水线中的表现，发现轻量级嵌入式分类器在多数任务上优于大型生成式模型，且显著降低计算资源消耗。这一发现为设计高效、鲁棒且成本可控的语音交互容错机制提供了实证依据与实践指导。
链接: https://arxiv.org/abs/2608.30738
作者: Phillip Schneider,Alexandre Mercier,Joshua Oehms,Kristiina Jokinen,Florian Matthes
机构: ALMA PHIL(德国); Technical University of Munich (慕尼黑工业大学); CNRS-AIST Joint Robotics Laboratory (JRL, 日本)
类目: Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026
Abstract:Robust understanding of user input is a core requirement for voice assistants deployed in real-world environments. In practice, these systems encounter heterogeneous fallback situations caused by noisy audio input, transcription errors, ambiguous requests, incomplete utterances, or unintended activations. Existing systems typically respond with generic fallback messages, which do not resolve the underlying interaction failure and can degrade user experience. We study fallback handling in a deployed smartwatch-based voice assistant for general health support in everyday environments. Our analysis is based on six months of real-world usage data from more than 500 users, yielding a dataset of 3,030 anonymized, naturally occurring fallback-triggering utterances. We contribute (1) an operational taxonomy and the annotated VoxFallbacks dataset of these interactions, (2) a comparative evaluation of different models within a classification pipeline under practical deployment constraints, and (3) practical lessons for designing robust and cost-efficient fallback mechanisms. Results show that lightweight embedding-based classifiers outperform larger generative models on most classification tasks while requiring substantially fewer computational resources.
[NLP-45] Calibrating Small Language Models for Claim Check-Worthiness Detection
【速读】： 该论文旨在解决自动化事实核查流程中“声明检查价值性评估（claim check-worthiness assessment）”的效率与准确性之间的权衡问题。在实际部署场景中，对每个新出现的声明都使用大型语言模型（Large Language Models, LLMs）进行评估存在高昂的成本和延迟，而采用小型模型（Small Language Models, SLMs）虽可降低开销，却牺牲了准确性。为应对这一挑战，论文提出了一种名为NN-PPI的点式预测增强推理（Prediction-Powered Inference, PPI）扩展方法，其核心在于通过一个轻量级的后处理校准层，在不重新训练底层模型的前提下，于推理阶段动态校准模型输出。该方法的关键创新在于利用预测置信度进行残差校准，从而显著提升小型模型的性能，使其在加权F1指标上相比基线模型提升12%至33.80%，达到与大型模型相当的准确率。此外，该方法在真实生产环境中部署的微调模型上仍能进一步提升性能，表明其与监督微调具有互补性。通过从成本低一个数量级的模型中恢复接近大型模型的精度，NN-PPI显著降低了大规模部署下检查价值性检测的运营成本。
链接: https://arxiv.org/abs/2608.30731
作者: Pratuat Amatya,Venktesh Viswanathan,Vinay Setty
机构: Factiverse AS(事实宇宙公司); University of Stavanger(斯塔万格大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Assessing claim check-worthiness is an essential first step in automated fact-checking pipelines. This work is motivated by a real deployment challenge at an early-stage startup: running large language models (LLMs) over every incoming claim is cost- and latency-prohibitive, yet smaller models sacrifice accuracy. We propose NN-PPI, a pointwise extension of Prediction-Powered Inference (PPI) that calibrates model predictions at inference time as a lightweight post-hoc layer, without re-training the underlying model. NN-PPI achieves weighted F1 gains ranging from 12% to 33.80% depending on the size and performance of the baseline model, bringing SLMs on par with larger LLMs. Beyond few-shot SLMs, NN-PPI further improves a production-deployed fine-tuned model, demonstrating that residual calibration is complementary to supervised fine-tuning. By recovering LLM-level accuracy from models that are an order of magnitude cheaper to serve, it makes accurate check-worthiness detection substantially cheaper to operate at scale. Our code and data can be found at this https URL.
[NLP-46] E-Commerce Bench: Evaluating LLM Agents on Long-Horizon Autonomous Business Operation
【速读】： 该论文旨在解决长时序智能体任务（long-horizon agentic tasks）中面临的动态环境适应与长期依赖建模难题，此类任务要求大语言模型（LLM）在数千步的交互过程中持续探索、从经验中学习并动态调整策略。其核心挑战在于如何在复杂、演化且高度不确定的商业环境中实现可持续的自主运营。为此，论文提出E-Commerce Bench——首个开源的全年电商运营基准，通过整合多轮供应商谈判与动态事件（如促销、自然灾害、供应链冲击），构建了一个真实感强的商家侧运营环境。关键解决方案在于：采用真实电商平台的产品与供应商数据，并结合一年期的确定性需求模型和可复现的谈判内核，使代理能够在连续365天的运营中执行市场调研、库存采购、销售优化、订单履约、退货处理及现金流管理等全流程操作。实验表明，尽管各模型表现各异，但无一全面领先；其中GPT-5.6 Sol虽实现资产增长最高（从10万增至143万），但在反欺诈能力上仅排第16；而开源模型Qwen3.8-Max-Preview在长期学习能力与议价效率方面表现突出，展现出对价格趋势的渐进式优化能力，验证了其在长周期策略适应中的优势。
链接: https://arxiv.org/abs/2608.30730
作者: Wei Fan,Xinjie Shen,Xudong Guo,Jianhong Tu,Yang Su,Yinger Zhang,Lianghao Deng,Fengyu Wang,Baohua Dong,Yangqiu Song,Dayiheng Liu
机构: Qwen Team, Alibaba Group(通义实验室，阿里巴巴集团); Department of Computer Science and Engineering, HKUST(香港科技大学计算机科学与工程系); Taobao Tmall Group, Alibaba Group(淘宝天猫集团，阿里巴巴集团)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注:
Abstract:Long-horizon agentic tasks go beyond chaining short tasks over more interaction turns. Their evolving dynamic environments and long-range dependencies require Large Language Models (LLMs) to continually explore, learn from experience, and adapt their policies over thousands of steps. We introduce E-Commerce Bench, the first open-source benchmark that integrates multi-round counterpart negotiation and dynamic events into a year-long business operation. Over a 365-day year, an LLM agent concurrently runs multiple online stores, researching the market, negotiating with suppliers to source inventory, optimizing sales strategies, fulfilling orders, handling returns, and managing cash flow to maximize its end-of-year total assets. To construct a realistic merchant-side operating environment, the product and supplier data are derived from a real e-commerce platform, while a year-long calendar of promotions, natural disasters, and supply-chain shocks continually reshapes demand. For reproducibility, both sides of the market are deterministic: customer purchases and returns follow a fixed demand model, while a negotiation kernel determines supplier pricing, concessions, and decisions, with an LLM used only to verbalize them. We evaluate 18 frontier models across seven dimensions, including year-end assets, and find that no single model dominates. GPT-5.6 Sol earns the most, growing the 100,000 opening stake into 1,431,425, yet it ranks 16th of 18 on fraud avoidance and trails Fable5 in operational efficiency. Among open-weight models, Qwen3.8-Max-Preview leads with 416,252, 38% above GLM 5.2 (high), and achieves the strongest learning over the horizon, progressively bargaining down prices across repeated orders. Our code is available at this https URL.
[NLP-47] Where Do Multilingual Vision-Language Encoders Fail on Low-Resource Languages?
【速读】： 该论文旨在解决多语言视觉-语言编码器在低资源语言（LRL，如斯瓦希里语）上性能显著落后于高资源语言（HRL，如英语）的问题，尤其关注在已有模型中语言间对齐能力差距的根源。现有研究认为，线性语言方向（linear language direction）导致了跨语言对齐几何结构的退化，是性能差距的主要成因。然而，本文通过实验推翻了这一假设：使用线性分类器分析（LEACE）将线性语言分类准确率从99%降至接近随机水平，同时迭代INLP方法也仅使分类性能降至37%-50%，但低资源语言的实例检索性能变化极小（±1.5个百分点以内），且各层级均值波动不超过2.2个百分点，与随机控制组表现一致，表明线性偏差仅为表征症状而非根本原因。关键发现指出，真正影响对齐效果的因果因素存在于编码器的前向传播路径中——具体表现为不同语言在序列末端（EOS）隐藏状态随网络深度增加而产生轨迹发散。通过将低资源语言的EOS状态替换为三块层前对应的英文并行状态，可使斯瓦希里语检索性能从22.1%大幅提升至69.1%，且在另一模型上复现；三个对照实验排除了池化位置的同义循环及英语特异性问题。进一步提出前层主干（front-layer trunk）机制，在训练阶段强制各语言投影趋向平行内容中心，成功在低资源语言XM3600检索任务（1,000图像子集）上分别提升+9.6/+17.1个百分点，并在三个额外基准上保持一致增益，同时不损害高资源语言性能。因此，解决方案的核心在于重构编码器中语言间投影的早期对齐路径，以缓解由深层特征发散引发的跨语言对齐失衡。
链接: https://arxiv.org/abs/2608.30725
作者: Donghoon Han,SungHyun Moon,Aidyn Zhakatayev,Junghun Cha,SeungJae Lee
机构: Dnotitia Inc.
类目: Computation and Language (cs.CL)
备注:
Abstract:Recent multilingual vision–language encoders cover hundreds of languages in a single model, yet on two state-of-the-art instances retrieval on low-resource languages (LRL; e.g. Swahili) trails high-resource ones (HRL; e.g. English) by 30^+ ,pp. We ask where in the trained encoder this gap is located. Prior modality-gap and cross-lingual subspace work suggests a linear language direction at the output crowds out alignment-relevant geometry. We falsify this: LEACE drives the linear language classifier from 99% to near chance and iterated INLP to 37 – 50% while LRL retrieval moves within \pm 1.5 ,pp and all tier means within 2.2 ,pp, tracking random controls. The linear bias is a \emphsymptom, not the cause. Instead, the alignment-causal factor lies along the encoder’s forward path: the EOS (end-of-sequence) hidden state’s per-language trajectory diverges with depth. Substituting the EOS with its parallel English value three blocks before the projector lifts Swahili from 22.1% to 69.1% on one encoder (and reproduces on the other); three controls rule out pooled-position tautology and English specificity. A front-layer trunk that pulls each language’s projection toward the parallel-content centroid corroborates the diagnosis at training time, recovering +9.6 / +17.1 ,pp on LRL XM3600 retrieval (1,000-image subset), with consistent gains across three further benchmarks while preserving HRL performance.
[NLP-48] Mind the Gap: Theory-of-Mind-Grounded Friction for Epistemic Alignment
【速读】： 该论文旨在解决生成式对话系统中**认知对齐（epistemic alignment）与表面协调（surface coordination）**难以区分的问题，即现有基于偏好（preference-based）的方法通常仅优化响应层面的偏好，而忽略了对话双方在信念状态（belief states）上是否真正达成一致。其核心解决方案在于将心智理论（Theory-of-Mind, ToM）推理操作化为摩擦力（friction）控制信号，通过在每个指代表达处构建四部分信念结构——说话者意图指称对象、听者理解内容，以及双方对彼此信念的建模——实现对认知状态差异的显式捕捉。该方法能够识别“无声分歧”（silent divergence）现象，即双方均自信推进对话但实际锚定于不同指称对象。实验表明，在表示层，消融二阶信念通道使误解召回率从65%降至26%；在策略层，采用奖励塑形（FAR）和信任域（FTR）的变体相较于直接偏好优化（DPO）显著提升了干预准确率（F1）与情境校准能力，且Brier评分独立验证了校准优势。三次训练运行中，FAR与FTR表现出更强的稳定性，而DPO则波动剧烈，甚至导致基础策略中的干预能力退化。因此，基于心智理论的摩擦信号为指代分歧场景下的上下文敏感干预提供了可训练、可解释的机制。
链接: https://arxiv.org/abs/2608.30719
作者: Yifan Zhu,Kyeongmin Rim,James Pustejovsky
机构: Brandeis University (布兰戴斯大学)
类目: Computation and Language (cs.CL)
备注: 16 pages, 1 figure
Abstract:Productive dialogue alignment requires distinguishing \emphsurface coordination (acknowledgments and smooth task progression) from \emphepistemic alignment (convergence of belief states); standard preference-based methods typically optimize response-level preferences without explicitly modeling the latter. We operationalize Theory-of-Mind (ToM) inference as a control signal within Frictive Policy Optimization by extracting, at each referring expression, a four-part belief structure: the speaker’s intended referent, the addressee’s interpretation, and each participant’s model of the other’s belief. This makes friction mechanically computable from epistemic-state comparisons, capturing \emphsilent divergence, where both participants proceed confidently while grounding to different referents. We evaluate the signal at two levels. At the representation level, ablating the second-order channel reduces misunderstanding recall from 65% to 26% . At the policy level, reward-shaping (FAR) and trust-region (FTR) variants improve intervention F1 and warranted-context calibration over DPO, with Brier scores independently supporting the calibration gains. Across three training runs, FAR and FTR remain substantially more stable, whereas DPO varies widely and can degrade intervention competence already present in the base policy. Thus, ToM-grounded friction provides a trainable signal for context-sensitive intervention under referential belief divergence.
[NLP-49] SocialReason Bench: A Video-QA Benchmark for Social Reasoning with Counterfactual Narrative Videos EMNLP2026
【速读】： 该论文旨在解决当前大型多模态模型（Large Multimodal Models, LMMs）在理解以人为中心的社会情境时存在的局限性问题，特别是其在社会动态推理能力上的不足。现有基准测试通常依赖单一观察轨迹的视频数据，难以区分模型是真正理解了社会关系，还是仅依赖重复出现的叙事模式进行预测。为此，论文提出SocialReasonBench，一个基于互动叙事游戏《底特律：变人》（Detroit: Become Human） gameplay视频构建的多选问答基准，用于评估模型在真实社交情境中的推理能力。其核心创新在于利用游戏中分支剧情结构，通过玩家决策生成可验证的多种社会结果，并结合游戏脚本、流程图与录制分支实现答案标签的精准标注。解决方案的关键在于设计了一个多智能体协同筛选管道，能够精确定位具有社会意义的视频片段，将答案标签锚定于游戏状态信号，并生成基于理论指导且包含诊断性干扰项的问题。该基准涵盖意图识别、情感共情、道德困境、反事实推理和因果前因等七个推理维度。实验表明，尽管当前LMMs在基础社会理解任务上表现尚可，但在反事实与因果推理方面仍存在显著短板；进一步的消融与错误分析揭示，模型常依赖不完整的模态线索，陷入视觉捷径等推理陷阱，暴露出对显性事件识别与深层社会状态推理之间存在的本质差距。
链接: https://arxiv.org/abs/2608.30716
作者: Zheyu Huang,Zijing Shi,Haozhe Luo,Huadong Tang,Mingyu Liu,Meng Fang,Ling Chen
机构: AAII, University of Technology Sydney (澳大利亚悉尼科技大学); Northeastern University (东北大学); Fudan University (复旦大学); University of Liverpool (利物浦大学)
类目: Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to Findings of EMNLP 2026. 24 pages, 11 figures, 11 tables
Abstract:Recent advances in Large Multimodal Models (LMMs) have greatly improved video understanding, yet their ability to reason about human-centered social situations remains limited. Existing benchmarks typically rely on videos with a single observed trajectory, making it difficult to determine whether models truly understand social dynamics or merely exploit recurring narrative patterns. We introduce SocialReasonBench, a video multiple-choice QA benchmark for evaluating socially grounded reasoning in scenarios derived from interactive narratives. Built from gameplay videos of Detroit: Become Human, the benchmark leverages branching storylines where player decisions lead to alternative social outcomes that can be checked against the game’s own script, flowchart, and recorded branches. We develop a multi-agent curation pipeline that localizes socially meaningful clips, grounds answer labels in game-state signals, and generates theory-guided questions with diagnostic distractors. SocialReasonBench covers seven reasoning dimensions, including intent recognition, emotional empathy, moral dilemma, counterfactual reasoning, and causal antecedent. Experiments on contemporary LMMs show that models perform reasonably well on basic social understanding but struggle with counterfactual and causal reasoning. Further ablation and diagnostic error analyses reveal that models often depend on incomplete modality cues and fall into reasoning traps such as visual shortcuts, highlighting a gap between observable event recognition and deeper reasoning over latent social states.
[NLP-50] GUIDE: Guiding Internal Evidence with Language Instructions
【速读】： 该论文旨在解决大模型在多模态任务中虽能遵循生成指令，却难以根据指令有效调控对不同证据源依赖程度的问题。其核心挑战在于模型易受捷径关联线索（shortcut-associated cues）影响，即使指令明确要求改变证据使用策略，仍可能固守原有偏好。为应对这一问题，论文提出GUIDE框架，其关键在于通过语言指令实现对内部证据路径的可控调节：结合分组参数高效适配（grouped parameter-efficient adaptation）与指令条件门控机制（instruction-conditioned gating），动态调控推理与生成过程中多模态证据路径的激活强度。此外，研究构建了路径级评估体系，涵盖依赖敏感性（reliance sensitivity）、受控扰动分析、路径调制能力及自回归解码动态等指标，系统量化指令引导下的证据使用变化。实验在GQA、TextVQA、MM-IMDb、CREMA-D、RAVDESS和Flickr30K等多个数据集上验证了该方法的有效性，表明GUIDE能够在保持任务性能的同时，实现结构化且指令一致的证据依赖重分配，显著提升模型在目标证据扰动下的鲁棒性，并支持跨多模态场景的可控证据调节。这表明，多模态指令遵循不仅应局限于输出控制，更可延伸至对各证据源贡献度的精细调控。
链接: https://arxiv.org/abs/2608.30712
作者: Soyeon Caren Han,Hyunsuk Chung,Jinwoo Kim,Seungyeon Ji,Kyungreem Han
机构: The University of Melbourne(墨尔本大学); Korea Institute of Science and Technology(韩国科学技术院); Korea University(韩国大学); University of Science and Technology KIST School(韩国科学技术院大学校)
类目: Computation and Language (cs.CL)
备注:
Abstract:Large multimodal models follow instructions about what to generate, but not necessarily about what evidence to rely on. Hence, models may continue to depend on shortcut-associated cues even when instructions suggest otherwise. We introduce GUIDE, a framework for controlling internal evidence usage through language instructions. GUIDE combines grouped parameter-efficient adaptation with instruction-conditioned gating to modulate multimodal evidence pathways during reasoning and generation. We further introduce a pathway-level evaluation framework that characterizes instruction-conditioned evidence modulation through reliance sensitivity, controlled perturbation analysis, pathway modulation, and autoregressive decoding dynamics. Across multimodal reasoning, classification, and generation, GUIDE induces structured and instruction-aligned redistribution of evidence reliance while largely preserving task behavior. Experiments on GQA, TextVQA, MM-IMDb, CREMA-D, RAVDESS, and Flickr30K show that GUIDE improves robustness under targeted evidence perturbations and enables controllable modulation across diverse multimodal settings. This suggests that multimodal instruction following can extend beyond output control toward regulating how different evidence sources contribute to model predictions.
[NLP-51] SingProbe Technical Report
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）在实际部署中因依赖外部独立安全防护机制所导致的推理开销增加、安全信号延迟以及与日益强大的基础模型能力不匹配等问题。其核心解决方案是提出SingProbe，一种轻量级的内在式运行时防护机制，该机制直接复用LLM推理过程中产生的隐藏状态（hidden states），并与其自回归解码过程协同工作。在统一框架下，SingProbe以极低的额外开销实现了对查询意图、响应安全性及幻觉风险的逐标记级连续预测，本质上形成了一种“免费午餐”式的防护方案。关键创新在于利用模型内部表征作为监控接口，避免了传统外置模型带来的性能损耗。此外，研究还构建了SingStreamBench基准，用于评估流式防护机制在良性前缀上保持静默、同时快速识别潜在有害内容的能力。实验表明，SingProbe仅需约200万参数和0.5%的额外推理开销，即可达到甚至超越更大型独立防护模型及专用幻觉检测器的性能。进一步研究表明，SingProbe的评分可预判未来生成风险，并指导受限的安全解码；该范式还被扩展至医疗领域，形成SingProbe-Med，仅在临床相关风险出现时才触发针对性干预，从而实现高效精准的生成控制。这些结果共同证明，内部模型表示为生成阶段的实时监控与调控提供了有效且高效的接口。
链接: https://arxiv.org/abs/2608.30703
作者: Sing Team
机构: Ant Group(蚂蚁集团)
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Runtime guardrails are essential for reliable large language model (LLM) deployment, yet existing approaches typically rely on independent, external models that introduce additional inference cost, delayed safety signals, and a capacity mismatch with increasingly capable base models. To address these issues, we introduce SingProbe, a lightweight intrinsic runtime guard that directly reuses hidden states produced during LLM inference and operates alongside autoregressive decoding. Within a unified framework, SingProbe continuously predicts query intent, response safety, and hallucination risk at the token level with negligible additional guardrail inference overhead, offering a “free-lunch” solution. We further introduce SingStreamBench, a benchmark designed to assess whether streaming guardrails remain inactive on benign prefixes while promptly detecting emerging unsafe content. Extensive experiments show that SingProbe achieves competitive or superior performance compared with substantially larger standalone guardrails and specialized hallucination detectors, with only \approx 2M parameters and 0.5% extra overhead. Beyond passive detection, we also show that SingProbe scores can anticipate future generation risk and guide constrained safe decoding. We further extend this paradigm to medical generation through SingProbe-Med, which selectively activates risk-directed decoding interventions only when clinically relevant risks emerge. Together, these results demonstrate that internal model representations provide an effective and efficient interface for generation-time monitoring and control.
[NLP-52] An Agent ic Retrobiosynthesis Framework with Learned Frontier Selection
【速读】： 该论文旨在解决生成式化学合成路径规划中，搜索策略（search policy）对多步逆向合成（multistep retrosynthesis）结果的独立贡献问题，尤其是在生物合成场景下的可解释性与效率优化。其核心挑战在于区分反应模型本身的能力与搜索策略在引导合成路径探索中的作用。解决方案的关键在于采用基于规则的逆向生物合成（rule-based retrobiosynthesis），构建一个确定性的生化引擎以保证所有方法使用相同的反应转换规则，并将搜索策略简化为仅负责选择下一个前沿分子（frontier molecule）进行扩展的“仅选择”接口。通过微调（fine-tuning）Qwen2.5-7B模型的策略头，结合提示工程（prompting）与LoRA微调技术，在有限的展开次数下显著提升了合成路径的成功率：在LASER基准上，微调策略在10次展开时达到65±1%的求解率，优于蒙特卡洛树搜索（MCTS）的59%；在200次展开时，进一步提升至78±1%，分别超过MCTS在三个不同基准测试集上的表现。实验表明，经过路线监督（route-supervised）的前沿分子选择策略能够有效增强预算受限的搜索性能，且无需改变底层的生化生成机制，但最终性能仍依赖于前沿分子构造质量和反应排序的准确性。
链接: https://arxiv.org/abs/2608.30702
作者: Philippe Meyer,Guillaume Gricourt,Thomas Duigou,Joan Hérisson,Jean-Loup Faulon
机构: Université Paris-Saclay(巴黎-萨克雷大学); INRAE(法国国家农业、食品与环境研究院); AgroParisTech(巴黎高科农学院); Micalis Institute(米卡利斯研究所)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Large language models are increasingly used as agents for multistep retrosynthesis, raising the question of how much their search policy contributes independently of the underlying reaction model. We investigate this question in a biological setting through rule-based retrobiosynthesis: a deterministic biochemical engine generates the same validated transitions for every method, searching for routes that terminate in metabolites available to an \emphEscherichia coli chassis, while the policy only selects which frontier molecule to expand next. Prompted and LoRA-tuned Qwen2.5-7B policies use a strict choice-only interface. The fine-tuned policy reaches 65\pm1 % solve rate at 10 expansions on LASER versus 59% for MCTS, and at 200 expansions reaches 78\pm1 % versus 75% on LASER, 88\pm3 % versus 80% on the RetroPath RL Golden benchmark, and 63\pm2 % versus 45% on the BioNavi-NP benchmark. Fine-tuning also consistently outperforms direct prompting. These results show that route-supervised frontier selection can improve budgeted search without altering biochemical generation, although performance remains dependent on frontier construction and reaction ranking.
[NLP-53] Beyond the Payload: How User Invocation Shapes Coding Agent Vulnerability to Repository Poisoning EMNLP2026
【速读】： 该论文旨在解决生成式代码代理（Coding Agent）在使用第三方代码仓库时面临的安全风险问题，尤其关注用户在日常使用中通过提示词配置（Prompt-Level Configurations, PLCs）无意间放大攻击面的潜在威胁。现有研究多聚焦于攻击者主动注入和伪装恶意代码的行为，而忽视了开发者自身在任务委派、提示词表述及技能/规则设定等选择上的影响。为此，论文提出首个系统性评估用户侧配置影响的基准测试工具CIPR（Coding In Poisoned Repos），涵盖1,920个实例，覆盖20个真实世界仓库、四种任务类型、三种基于社交媒体风格的提示表达方式以及三种技能/规则配置条件，并通过自动化运行时与基于追踪的验证机制衡量攻击成功率（ASR）与代理告警率（AR）。研究发现：（1）漏洞暴露程度具有高度上下文依赖性，任务类型可导致ASR相差达4.5倍，其中测试执行类任务构成“静默攻击面”（高ASR，低AR）；（2）提示词表达方式间接影响风险——模糊提示通过限制执行深度降低ASR，而噪声化提示则通过弱化恶意内容的显著性呈现抑制告警的趋势。这些结果表明，代码代理的安全性并非静态属性，而是由日常用户配置动态塑造的复杂产物。
链接: https://arxiv.org/abs/2608.30686
作者: Fukang Zhu,Binbin Zhao,Ruixiao Lin,Ping He,Tianyu Du,Shouling Ji
机构: Zhejiang University (浙江大学); State Key Laboratory of Internet Architecture, Tsinghua University (清华大学互联网体系结构国家重点实验室)
类目: Cryptography and Security (cs.CR); Computation and Language (cs.CL)
备注: 30 pages,7 figures, Accepted to EMNLP 2026 Main Conference
Abstract:Coding agents are increasingly used for software engineering tasks, including bootstrapping projects from third-party repositories whose integrity cannot be assumed. Prior work on repository poisoning largely focuses on attacker-controlled injection and disguise, but developers also shape risk through everyday invocation choices: what task to delegate, how to phrase the request, and which skills or rules to supply. We term these user-side choices Prompt-Level Configurations (PLCs) and introduce CIPR (Coding In Poisoned Repos), the first benchmark that systematically varies PLCs in poisoned real-world repositories. CIPR comprises 1,920 instances across 20 repositories, four task types, three social-media-grounded prompt styles, and three skill/rule conditions, and measures attack success rate (ASR) and agent alert rate (AR) using automated runtime and trace-based oracles. Our evaluation reveals two key insights: (1) Vulnerability is highly context-dependent, with task type creating up to a 4.5-fold difference in ASR, with test-execution task forming a silent attack surface (high ASR, low AR). (2) Prompt expression shifts risk indirectly: underspecified prompts reduce ASR by truncating execution depth; noisy prompts exhibit a directional trend toward suppressing alerts by making malicious content less conspicuous. These findings highlight that coding agent vulnerability is not a static property, but a dynamic outcome shaped by everyday user configurations.
[NLP-54] WildSEEK: Evaluating Language Models for Information-Seeking EMNLP
【速读】： 该论文旨在解决当前大语言模型（Large Language Models, LLMs）在真实信息获取场景中响应质量评估不足的问题。现有评估方法多局限于特定主题或人工构造的数据，难以反映实际用户查询的复杂性及模型输出中存在的潜在风险。为此，研究提出WildSEEK——一个基于真实用户交互的3000条信息寻求型查询的标注数据集，并构建了相应的评估框架。其关键在于：首先，通过人工标注涵盖健康、金融等敏感领域，区分事实性查询与分析性查询（即超越单纯事实的信息推演）；其次，利用该数据集训练分类器，对超过180万条真实查询进行分析，揭示超过三分之一的信息查询属于高风险类型，且以分析性查询为主。研究发现，LLM在四个关键维度上表现不佳：谄媚行为、过度依赖、默认美国中心视角以及对弱势群体处理不当，且这些失败在分析性查询中更为显著。该工作为评估和监控大语言模型在信息获取中的可靠性、安全性和公平性提供了实证基础，推动了其在日益重要的信息中介角色中应如何规范行为的系统性思考。
链接: https://arxiv.org/abs/2608.30683
作者: Tanise Ceron,Joachim Baumann,Elisa Bassignana,Berat Cabuk,Dirk Hovy,Debora Nozza
机构: Bocconi University(博科尼大学); Stanford University(斯坦福大学); IT University of Copenhagen(哥本哈根信息技术大学); Pioneer Center for AI(先锋人工智能中心)
类目: Computation and Language (cs.CL); Computers and Society (cs.CY)
备注: 9 pages, accepted at EMNLP Main 2026
Abstract:Language models are increasingly mediating information access to end users, urging a systematic evaluation of their responses for a fair and reliable information ecosystem. Existing evaluations, however, are often topic-specific or synthetic, limiting their ability to capture the complexity of “in the wild” information-seeking queries and the risks present in model responses. To address this gap, we introduce WildSEEK, a manually annotated dataset of 3k information-seeking queries from real user interactions, and an evaluation framework for LLM-generated responses. WildSEEK includes annotations for risk-sensitive domains (e.g. health and financial information), and distinguishes factoid queries from analytical queries which seek responses beyond facts. We train classifiers on WildSEEK to analyze more than 1.8M realistic user queries. We find that over a third of information-seeking queries are high-risk and more often analytical. Our findings show that LLM responses fail more often in four criteria: sycophantic behavior, overreliance, a default US-centric perspective, and poor handling of vulnerable populations – with failure rates being mostly higher for analytical queries. By providing methods to monitor the reliability, safety, and fairness of LLM behavior, our dataset and evaluation framework offer an empirical foundation for the broader question of how these systems should behave as they take on a growing role in information access.
[NLP-55] LCoT-GV: Graph Attention Networks for Verifying Long Reasoning Chains in Large Language Models
【速读】： 该论文旨在解决大模型在生成长链式思维（Long Chains-of-Thought, LCoTs）过程中存在的推理步骤不一致、缺乏支持的推断或无关步骤等问题，即使最终答案正确，其内部推理过程仍可能包含逻辑缺陷。解决方案的关键在于提出一种基于图结构的验证框架——长链式思维图验证器（Long Chain-of-Thought Graph Verifier, LCoT-GV），将LCoTs建模为推理图（reasoning graph），其中每个节点代表一个推理步骤，边则编码语义与逻辑关系；通过图注意力网络（Graph Attention Network）从推理图中学习并预测整个推理链的正确性。研究还构建了一个面向多领域问答任务的图结构化验证数据集，实验表明该方法在性能上可媲美现有最相似的基准方法。
链接: https://arxiv.org/abs/2608.30679
作者: Bérénice Jaulmes,Mehwish Alam
机构: Télécom Paris, Institut Polytechnique de Paris(巴黎电信学院，巴黎综合理工学院); BNP Paribas, IT Group(法国巴黎银行，信息技术集团)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 6 pages without references, 2 tables, 1 algorithm, 1 figure
Abstract:Large Reasoning Models produce Long Chains-of-Thought (LCoTs) which involve breaking down the problem into smaller reasoning steps before reaching the conclusion. However, these steps often contain contradictions, unsupported inferences, or irrelevant steps, even when the final answer is correct. We propose Long Chain-of-Thought Graph Verifier (LCoT-GV), a graph-based framework that represents LCoTs as reasoning graphs. Each node in the graph represents a reasoning step and the edges encode semantic and logical relations. A Graph Attention Network is then trained to predict chain-of-thought correctness from the reasoning graph. We construct a new graph-oriented verification dataset from multiple reasoning benchmarks for question answering in various domains. The results show that our method is competitive with the most similar approaches.
[NLP-56] OCR-MetaReasoning Benchmark: Evaluating the Meta-Reasoning Ability of MLLM s in Text-Rich Image Understanding EMNLP2026
【速读】： 该论文旨在解决多模态大语言模型（Multimodal Large Language Models, MLLMs）在文本密集型图像理解任务中，对光学字符识别（OCR）所提取信息的元推理（meta-reasoning）能力评估不足的问题。现有评估方法常将信息抽取与推理过程混淆，且缺乏对模型是否遵循特定推理方向的测试，如应用显式规则、抽象隐含规律或补全缺失前提等。为此，论文提出一种受控的单图像基准测试集 OCR-MetaReasoning，将演绎、归纳和溯因作为独立的推理方向，并区分最终答案正确性与推理过程合规性。该基准包含1,500个经验证的样本，采用3×5平衡分类体系，涵盖三种推理类型与五类OCR对象，提供参考推理步骤、自动答案评分机制以及两个核心指标：元推理宏评分（Meta-Reasoning Macro Score, MRMS）与推理过程合规性评分（Reasoning Process Compliance Score, RPCS）。实验表明，当前主流闭源与开源MLLMs在基于OCR的元推理能力上仍远未达到饱和状态，尤其在显式规则应用和布局敏感推理方面表现薄弱；同时，在精确匹配评价下，错误的最终答案也可能伴随符合流程的推理路径。
链接: https://arxiv.org/abs/2608.30678
作者: Gengxu Li,Yuan Wu,Yi Chang
机构: Jilin University (吉林大学); Engineering Research Center of Knowledge-Driven Human-Machine Intelligence, MOE, China (教育部知识驱动人机智能工程研究中心); International Center of Future Science, Jilin University (未来科学国际中心, 吉林大学)
类目: Computation and Language (cs.CL)
备注: EMNLP 2026 Findings camera-ready version
Abstract:Text-rich image understanding requires multimodal large language models (MLLMs) to organize OCR (Optical Character Recognition)-grounded evidence across words, layout, fields, charts, and visual correspondences. Existing evaluations often conflate extraction with reasoning and rarely test whether models follow the required reasoning direction: applying visible rules, abstracting hidden regularities, or recovering missing premises. We introduce OCR-MetaReasoning, a controlled single-image benchmark that treats deduction, induction, and abduction as distinct directions and separates final-answer correctness from reasoning-process compliance. The benchmark contains 1,500 verified samples in a balanced (3\times5) taxonomy crossing three reasoning types with five OCR-object categories, along with reference reasoning steps, automatic answer scoring, the Meta-Reasoning Macro Score (MRMS), and the Reasoning Process Compliance Score (RPCS). Experiments with representative closed-source and open-source MLLMs show that OCR-grounded meta-reasoning remains far from saturated: models struggle with visible-rule application and layout-sensitive inference, while process-compliant rationales can accompany incorrect final answers under exact-match evaluation. The code is available at this https URL.
[NLP-57] MURANO: Design Run and Reproduce Mechanistic Interpretability Experiments as Composable Pipelines EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）机制可解释性研究中流程碎片化的问题。当前的研究工作通常涉及模型加载、数据记录、归因分析、干预操作和评估等多个环节，但现有工具库多仅聚焦于其中某一阶段，导致研究人员需在不同库之间手动转换数据格式与接口，增加了协作与复现的复杂性。为应对这一挑战，论文提出 Murano——一个开源框架，其核心创新在于将上述五个关键环节的操作抽象为可组合的“步骤”（steps），每个步骤通过命名结果制品（named result artifacts）进行数据交换，并明确定义输入依赖与输出产出。该设计实现了跨组件间的一致身份标识（使用规范地址，canonical addresses）传递，从而支持端到端可复现的可解释性研究流程。Murano 基于现有的可解释性与机器学习库构建，通过两个经典研究的复现及一个稀疏自编码器（sparse autoencoder）的案例研究，验证了其在提升研究可重复性与跨工具协同效率方面的有效性。
链接: https://arxiv.org/abs/2608.30662
作者: Alireza Bayat Makou,Emirhan Böge,Phu Gia Hoang,Federico Tiblias,Jingcheng Niu,Subhabrata Dutta,Richard Eckart de Castilho,Iryna Gurevych
机构: Ubiquitous Knowledge Processing (UKP) Lab, Technical University of Darmstadt, Germany; National Research Center for Applied Cybersecurity ATHENE, Germany; Cluster of Excellence “Reasonable Artificial Intelligence” (RAI), hessian.AI, Germany; Zuse School ELIZA, Technical University of Darmstadt
类目: Computation and Language (cs.CL)
备注: Accepted to the EMNLP 2026 System Demonstrations Track. 11 pages, 6 figures, 2 tables
Abstract:This paper presents Murano, an open source framework for designing, running, and reproducing mechanistic interpretability studies of large language models, intended for researchers across disciplines. These studies often combine loading, recording, attribution, intervention, and evaluation, while existing libraries tend to focus on different parts of this workflow. As a result, researchers using several libraries may need to adapt outputs from one for use by another. To bridge this gap, Murano represents operations from these five areas as composable steps. Steps exchange named result artifacts and declare the inputs they require and the outputs they produce. A pipeline executes its steps in the order supplied, and Murano uses canonical addresses when component identities pass between operations. Murano builds on existing interpretability and machine learning libraries. We demonstrate Murano through two reproductions of established interpretability studies and one illustrative sparse autoencoder case study.
[NLP-58] SwarmBench: Can Large Language Models Act as Agent Swarm Orchestrators? EMNLP2026
【速读】： 该论文旨在解决当前大语言模型（Large Language Model, LLM）驱动的多智能体系统在动态协同架构（Agent Swarms）背景下缺乏系统性评估基准的问题。现有评测体系主要依赖单智能体或通用任务，难以全面衡量多智能体间动态协作的核心能力。为此，论文提出SwarmBench基准，从准确性、效率、成本及过程质量等多个维度对模型的协同编排能力进行综合评估。实验结果表明，当前模型在编排能力上存在显著差异，这些差异不仅体现在最终结果的准确率、执行效率与资源消耗上，更反映在协同过程的整体质量层面。基于上述发现，论文进一步提出SwarmExp方法，其核心在于通过经验提取（experience extraction）与经验回放（experience replay）机制，实现对高效协同策略的持续学习与优化，从而在不增加复杂度的前提下显著提升大语言模型的多智能体协同性能。
链接: https://arxiv.org/abs/2608.30661
作者: Jinshan Gao,Zhuoran Jin,Tianyi Men,Kang Liu,Jun Zhao
机构: The Key Laboratory of Cognition and Decision Intelligence for Complex Systems, Institute of Automation, Chinese Academy of Sciences(中国科学院自动化研究所认知与复杂系统决策智能重点实验室); School of Artificial Intelligence, University of Chinese Academy of Sciences(中国科学院大学人工智能学院)
类目: Computation and Language (cs.CL)
备注: EMNLP 2026 Findings
Abstract:Large language model-based multi-agent systems are evolving from fixed interaction topologies toward dynamically orchestrated Agent Swarms. However, existing benchmarks are still largely based on single-agent or general-purpose agent tasks, making it difficult to systematically evaluate key orchestration capabilities. We propose SwarmBench, a benchmark that evaluates model performance from multiple perspectives, including accuracy, efficiency, cost, and process quality. Experimental results show that current models exhibit substantial differences in orchestration capability. These differences are reflected not only in final accuracy, efficiency, and cost, but also in the overall quality of the orchestration process itself. Based on these findings, we further propose SwarmExp, a simple yet effective method based on experience extraction and experience replay, which consistently improves the orchestration performance of large language models.
[NLP-59] Geometry of Divergence: Tracking Hidden-State Trajectories for Adaptive Multi-Turn Reasoning
【速读】： 该论文旨在解决大语言模型智能体（LLM agents）在长时间多轮交互中，因上下文累积导致的内部表征失稳问题，即任务相关关键信息的表征随时间发生漂移，从而影响推理一致性与任务成功率。其核心解决方案是将多轮推理建模为底层大语言模型（LLM）隐藏状态的轨迹，并引入两种互补的几何信号：时间曲率（temporal curvature） 用于衡量每轮更新方向的一致性，以及方差斜率（variance slope） 用于量化探索空间的扩张或收缩趋势。研究发现，这些几何特征可在任务完成前有效区分正确与错误的推理路径。进一步地，通过将每个推理过程分解为由“读（Read）、写（Write）、回应（Respond）、转移（Transfer）”四类动作构成的三动作链，揭示了不同动作组合模式下几何信号的判别能力具有动作依赖性。实验表明，基于轨迹几何的动态监控可精准识别关键决策转折点，在τ-Bench基准上将任务成功率从24.1%提升至39.6%，同时降低11.2%的令牌消耗，显著提升了资源受限场景下的推理效率与可靠性。
链接: https://arxiv.org/abs/2608.30650
作者: Jie Liang,Zhengxin Yu,Hamid Nasiri,Peter Garraghan
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:LLM agents need to sustain goal-consistent reasoning across long multi-turn interactions under strict resource constraints. However, as the multi-turn context accumulates, it can destabilize the underlying LLM’s internal representation of task-relevant information from earlier turns, blurring the boundary between constructive reasoning and representation drift. We formulate multi-turn reasoning as a hidden-state trajectory of the underlying LLM that is characterized via two complementary signals: temporal curvature that captures the directional consistency of turn-to-turn updates, and variance slope which measures the expansion or contraction of the exploration space. Across four tasks and three underlying LLMs, we observed that these geometric signals distinguish between correct and incorrect episodes prior to completion. We further decompose each episode into three-action chains formed from four actions (Read, Write, Respond, Transfer) and show that separability is action-dependent, with different signals distinguishing various chain patterns. Our experiments demonstrate that trajectory geometry can identify critical turns in the reasoning process, increasing task success rates on \tau -Bench from 24.1% to 39.6% while reducing token cost by 11.2%.
[NLP-60] Where Identity Lives: Localized Retain-Free Identity Unlearning in Multimodal Large Language Models EMNLP2026
【速读】： 该论文旨在解决多模态大语言模型（Multimodal Large Language Models, MLLMs）在部署后需移除特定个体信息的隐私保护问题，现有方法依赖于难以获取的保留集（retain set），且重建保留集会重新引发隐私泄露风险。为实现无需保留集的遗忘（retain-free unlearning），其核心挑战在于如何在不破坏共享视觉-语言计算能力的前提下精准定位并删除个体身份信息。论文的关键解决方案是将无保留集遗忘建模为一个定位问题，通过因果追踪、权重移植和Fisher重叠分析发现，身份信息主要存储于解码器的早期至中期前馈神经网络层（MLP layers），且这些层的修改对视觉理解影响较小。基于此，提出路径感知的视觉属性锚定（Pathway-Aware Visual-attribute Anchoring, PAVA）方法，仅在上述关键层进行更新，并引入遗忘损失与视觉属性锚定机制，通过从遗忘图像中蒸馏模型自身未遗忘前的输出来保持图像相关行为的一致性。实验表明，PAVA在仅使用遗忘集的情况下，在遗忘-保留权衡上优于现有方法，且性能接近基于保留集的基准。
链接: https://arxiv.org/abs/2608.30649
作者: Kangwook Ko,Jaehyuk Jang,Wonjun Lee,Hee-Seon Kim,Changick Kim
机构: KAIST(韩国科学技术院)
类目: Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to Findings of the Association for Computational Linguistics: EMNLP 2026
Abstract:Removing a specific individual’s information from multimodal large language models (MLLMs) is often needed after deployment, but existing methods rely on a retain set, which is hardest to obtain at that point, and rebuilding it recreates the privacy exposure that unlearning aims to remove. Forgetting from the forget set alone instead damages the shared visual-language computation, harming perception. We cast retain-free unlearning as a localization problem: causal tracing, weight transplant, and Fisher overlap all point to early-to-mid decoder MLPs as the layers where identity information is stored and, unlike other module families, can be modified without substantially disrupting vision. We turn this into Pathway-Aware Visual-attribute Anchoring (PAVA), which confines updates to these layers and pairs a forget loss with a visual-attribute anchor that preserves image-grounded behavior by distilling the model’s own pre-unlearning answers from the forget images alone. On MLLMU-Bench and ReMem, PAVA gives the strongest forget-retain trade-off among forget-set-only methods and remains competitive with retain-based baselines.
[NLP-61] What It Costs to Compose Rebuild and Correct Precomputed Memory
【速读】： 该论文旨在解决生成式 AI（Generative AI）在长期任务中如何高效利用预计算记忆（precomputed memory）以避免重复输入上下文的问题。其核心挑战在于：预计算记忆在拼接独立生成部分时会退化，仅通过周期性重建可维持时效性，但重建成本占全量准备成本的很大比例，且无法自动响应实时更新或基于特定措辞的修正信息。解决方案的关键在于实现两种机制——一是对键值缓存（key-value cache）的训练压缩进行“热重建”（warm-rebuilding），二是将特定表述的更新信息（如拼接文本或注入缓存状态）与原始记忆并行服务，从而在重建周期之间保持记忆的准确性。研究指出，预计算记忆的最佳维护策略应与源数据变化频率同步重建，而上述两种方法分别在成本效益和实时性方面展现出显著潜力，但仍存在关于部署成本、一致性保障及动态融合机制的未决问题。
链接: https://arxiv.org/abs/2608.30647
作者: Asa Shepard
机构: Williams College(威廉姆斯学院)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Language models can answer from precomputed memory, a model’s saved reading of a body of material, reused across requests instead of read again at each. This paper maps where that practice preserves correctness and the conditions under which it fails. Across experiments on Llama-3.1-8B-Instruct using both saved key-value caches and trained compressions of them, precomputed memory degrades when assembled from separately prepared parts, stays current only through rebuilds costing a large fraction of full preparation in our measurements, and ignores corrections served beside it conditional on phrasing. If precomputed memories can be served alongside one another, be cost-efficiently rebuilt, and be superseded by new information arriving in real-time, they can serve as a way to avoid re-feeding context to a model over repeated queries. The implication of our results for a deployed system that deals with a variety of queries is that precomputed memories are best rebuilt on the cadence at which new information changes what the memory was originally computed from. Both warm-rebuilding trained compressions of key-value caches and serving specifically-phrased updates beside a memory, as pasted text or injected cache state, show particular promise for keeping precomputed memories current, the latter as an interim measure between rebuilds, and we measure the cost and name the remaining questions associated with each.
[NLP-62] BiG-SURE - Bipartite Graph for Semantic Uncertainty and Reliability Estimation of LLM s
【速读】： 该论文旨在解决在黑箱（black-box）场景下，大语言模型（LLM）与视觉-语言模型（VLM）在安全关键应用中缺乏可靠不确定性估计的问题。由于模型参数不可访问，传统基于梯度或似然的方法难以适用，因此亟需一种无需模型内部信息、仅依赖输入-输出行为的无监督不确定性评估方法。其解决方案的关键在于提出一种基于交叉温度语义一致性（cross-temperature semantic agreement）的不确定性估计器——BiG-SURE。该方法通过生成低温度响应作为稳定的语义锚点（semantic anchors），并利用高温度响应作为在语义保持变换下的探测信号（probes），构建一个基于自然语言推理（NLI）蕴含得分的锚-探二分图（Bipartite Graph, BiG）。进而通过该图矩阵的归一化平方谱能量定义置信度，不确定性则由其补值给出。该方法的核心创新在于将不确定性建模为高温度探测响应与低温度稳定信念之间的语义一致性程度，从而实现对模型预测可靠性进行量化评估。实验表明，BiG-SURE在文本问答、多语言问答及多模态问答任务上均显著优于现有黑箱不确定性估计方法，在保持简洁性与无监督特性的同时，提升了平均拒答AUROC，验证了其有效性与普适性。
链接: https://arxiv.org/abs/2608.30646
作者: Debarpan Bhattacharya,Malay Phadke,Sriram Ganapathy
机构: Indian Institute of Science (印度科学研究所); Bangalore, India (班加罗尔, 印度)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 22 pages, 9 figures
Abstract:Reliable uncertainty estimation is a crucial requirement for deploying large language models (LLMs) and vision-language models (VLMs) in safety-critical settings, especially when the model parameters are not accessible (black-box). We propose BiG-SURE, an uncertainty estimator based on cross-temperature semantic agreement. The method samples low-temperature responses as stable semantic anchors and high-temperature responses as probes under meaning-preserving input transformations. It then constructs an anchor-probe Bipartite Graph (BiG) using NLI-based entailment scores and defines confidence through the normalized squared spectral energy of this matrix, with uncertainty given by its complement. This bipartite graph-based Semantic Uncertainty and Reliability Estimation (SURE) score measures whether high-temperature probes remain semantically aligned with the model’s stable low-temperature belief or not. We evaluate BiG-SURE on text QA, multilingual QA, and multimodal QA tasks across multiple model families. In these experiments, BiG-SURE improves average abstention AUROC over prior black-box uncertainty estimators, while remaining simple, unsupervised, and applicable to black-box model settings.
[NLP-63] GMTS: Gradient Magnitude-based Token Selection Improves RLVR Training for LLM Reasoning
【速读】： 该论文旨在解决生成式 AI（Generative AI）在强化学习（Reinforcement Learning, RL）训练过程中，如何更精准地识别并利用对模型优化具有关键贡献的词元（token）这一核心问题。现有方法依赖高熵词元进行训练选择，尽管表现出显著性能提升，但其有效性背后的机制尚不清晰，且熵值无法在不同答案间稳定反映词元的重要性，因其未充分考虑答案层级奖励信号的差异。为此，论文提出基于梯度幅值的词元选择（Gradient Magnitude-based Token Selection, GMTS）方法，其核心在于利用词元熵与梯度幅值之间的相关性，通过近似梯度幅值排序来量化词元重要性，从而实现更细粒度的词元贡献评估。实验表明，基于GMTS选取前20%词元进行训练，在三个推理任务领域及多种模型规模下均显著优于传统熵基选择策略，验证了该方法在强化学习中可提供更可靠、更具泛化性的词元重要性估计，为提升大语言模型（LLM）的推理能力提供了更为精确的训练指导。
链接: https://arxiv.org/abs/2608.30632
作者: Outongyi Lv,Yuanwei Zhang,Xiaoqun Zhang
机构: Shanghai Jiao Tong University (上海交通大学); Institute of Natural Sciences, Shanghai Jiao Tong University (上海交通大学自然科学研究学院)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Findings of the 2026 Conference on Empirical Methods in Natural Language Processing
Abstract:Reinforcement learning (RL), particularly RL with Verifiable Rewards (RLVR), has recently emerged as a central paradigm for enhancing large language models’ (LLMs) reasoning abilities, demonstrating remarkable effectiveness across reasoning tasks. Recent studies suggest that high-entropy tokens play an exceptionally important role in model training, since training with only the highest 20% entropy tokens yields significant performance gains. However, why such high-entropy tokens are beneficial remains insufficiently understood. In this work, we find that although high-entropy tokens within one answer tend to correlate with large gradient magnitude, entropy alone fails to consistently reflect token importance across different answers, considering the variations in the answer-level reward signals. Based on this observation, we introduce the Gradient Magnitude-based Token Selection (GMTS) method to quantify token importance, which leverages the entropy-gradient connection to approximate gradient-magnitude rankings for token selection. We find that training on the top 20% tokens ranked by GMTS consistently outperforms entropy-based token selection across three reasoning domains and various model sizes, suggesting that GMTS provides a more fine-grained estimate of token contribution for RLVR training.
[NLP-64] REER-PT: Reverse-Engineered Reasoning for Perplexity-Guided Pre-training Data Augmentation
【速读】： 该论文旨在解决大规模语言模型预训练中高质量训练数据稀缺的问题。传统基于下一个词预测（next-token prediction）的监督方式仅关注上下文后的直接延续，而忽略了生成该延续背后的中间推理过程，导致模型难以有效学习深层次的逻辑关联。为此，本文提出可扩展的逆向推理增强预训练框架（REER-PT），其核心在于对原始预训练数据进行稀疏重构：通过识别那些虽难以直接预测但可从上下文推断出的文本延续，生成简洁的推理注释（reasoning annotations），以显式重建上下文与后续内容之间的语义连接。这些注释在离线阶段生成并优化，以困惑度（perplexity）为信号，同时通过长度约束和目标泄露过滤机制剔除冗余或平凡的注释。该方法保持源文本不变，兼容标准的下一个词预测目标，避免了预训练过程中在线推理展开带来的计算开销。实验表明，经REER-PT增强的数据使模型在多个知识与推理基准上性能提升最高达2.07个百分点，且仅有约0.05%的注释片段与原始文本完全一致，证明其有效性。综合困惑度下降与受控预训练实验结果，表明该方法显著提升了模型对后续文本的可预测性，并在不改变原有预训练目标的前提下增强了模型表现。
链接: https://arxiv.org/abs/2608.30627
作者: Haoran Que,Jiajun Shi,Ting Huang,Renming Pang,Jiaheng Liu,Ge Zhang,Wenhao Huang,Shen Yan,Wei Ye,Shikun Zhang
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:As language-model compute continues to scale, high-quality training data is becoming an increasingly important bottleneck. Conventional next-token prediction supervises what follows a context but leaves the intermediate reasoning behind that continuation implicit. We introduce \textbfREER-PT, a scalable framework that extends Reverse-Engineered Reasoning (REER) to raw pre-training data. REER-PT identifies continuations that are difficult to predict but can still be inferred from the preceding context, and inserts concise reasoning annotations that reconstruct the missing connection between context and continuation. Candidate annotations are generated and refined offline, with perplexity serving as the optimization signal. Constraints on length and target leakage filter out unhelpful or trivial annotations. This sparse transformation preserves the source text and remains compatible with standard next-token prediction, avoiding online reasoning rollouts during pre-training. We apply REER-PT to transform a source pre-training corpus into an augmented one. Across augmented-data, original-token, and selected-continuation comparisons, perplexity reductions range from 0.42 to 7.29, and only about 0.05% of annotation 13-grams appear verbatim in the source text. We then train two 680M-parameter models with the same architecture and training configuration on the source and augmented corpora, respectively. The augmented-data model gains up to 2.07 percentage points on several knowledge and reasoning benchmarks. Together, the perplexity analysis indicates improved continuation predictability, while the controlled pre-training experiments suggest that this augmentation can improve model performance without changing the standard pre-training objective.
[NLP-65] Hidden Threat in Synthetic Data: Covert Targeted Bias Injection through Benign Text EMNLP2026
【速读】： 该论文旨在解决生成式人工智能（Generative AI）在大规模语言模型（LLM）训练中使用合成数据所引发的安全隐患问题，特别是合成数据可能被用作隐蔽渠道，向对齐后的模型注入目标性社会偏见。其核心挑战在于：尽管合成数据在语义上看似无害，但其潜在的隐蔽信息传递机制可能导致模型在不损害通用任务能力的前提下继承特定偏见。解决方案的关键在于构建一个由非对齐教师模型生成、经领域过滤（如创意写作与代码生成）的合成数据管道，并利用该数据对对齐学生模型进行微调，从而验证了看似无害的合成数据可作为隐蔽信道实现偏见注入。研究进一步提出，基于对数线性（log-linearity）的评分方法可能为筛选表面无害但存在风险的合成数据提供有效信号，为提升合成数据驱动的模型训练安全性提供了初步防护思路。
链接: https://arxiv.org/abs/2608.30619
作者: Minkyung Cho,Jihyo Kim,SeungWoo Song,Junghun Yuk,Minjoon Kee,Hoyun Song,KyungTae Lim
机构: KAIST; Department of Artificial Intelligence, Dankook University (单科大学人工智能系)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: To be published in EMNLP 2026
Abstract:Synthetic data is increasingly used to train large language models (LLMs), yet its security implications remain poorly understood. Prior work on subliminal learning suggests that models can inherit behavioral traits from seemingly unrelated training data. In this work, we investigate whether such mechanisms can be exploited to inject targeted social biases into aligned models through semantically benign synthetic data. We construct a pipeline in which a misaligned teacher model generates filtered synthetic datasets across domains such as creative writing and code generation, which are then used to fine-tune aligned student models. Our experiments show that benign-looking synthetic data can act as a covert channel for transmitting targeted biases while largely preserving the student model’s general task capabilities. These results reveal a previously underexplored security risk in synthetic data-driven LLM training pipelines and highlight the need for improved safeguards. As one possible step toward this goal, we suggest that log-linearity-based scoring may provide a useful signal for screening seemingly benign synthetic data.
[NLP-66] axCE : A Framework for Automated Taxonomy Construction and Evaluation at Scale EMNLP2026
【速读】： 该论文旨在解决大规模非结构化反馈文本（如评论、访谈记录和调查问卷）在实际应用中难以构建高质量、多层次语义层次结构（即层级分类体系，hierarchical taxonomy）的问题。现有方法普遍存在层次浅显、忽略长尾主题或缺乏严谨评估框架等缺陷。为此，论文提出TaxCE——一种全自动化框架，其核心在于通过逐级凝练（progressive condensation）将原始文本内容压缩为可操作的语段、去重后的语义单元以及带定义的细粒度主题，并自下而上构建具有语料根基（corpus-grounded）的层级结构。解决方案的关键创新在于引入三个基于语料的评估指标：排他性（Exclusivity）、完备性（Exhaustivity）与粒度（Granularity，简称EEG），并构建“指标闭环”迭代优化机制，能够自动诊断分类体系中的不足并实施针对性修正，直至收敛。实验表明，TaxCE在排他性、完备性和粒度三项指标上分别较最强基线平均提升11.8、20.5和15.7个百分点，且人工评估验证了其在分类质量、可操作性与导航性方面的显著优势。
链接: https://arxiv.org/abs/2608.30614
作者: Sandeep Sricharan Mukku,Albert Aristotle Nanda,Rohit Pyati
机构: Amazon(亚马逊)
类目: Computation and Language (cs.CL)
备注: EMNLP 2026 - Industry
Abstract:Organizing unstructured feedback text into hierarchical taxonomy is a fundamental challenge in NLP, particularly in domains where feedback arrives at massive scale in varied forms such as reviews, transcripts, and surveys. Existing approaches either produce shallow hierarchies, neglect long-tail topics, or lack rigorous evaluation frameworks. We present TaxCE, a fully automated framework that constructs multi-level hierarchical taxonomies from raw text through progressive condensation of corpus content into actionable segments, deduplicated semantic units, and granular topics with definitions, which are then organized bottom-up into a hierarchy with corpus-groundedness. We also introduce three corpus-grounded evaluation metrics, Exclusivity, Exhaustivity, and Granularity (EEG), and integrate them into a metrics-in-the-loop iterative refinement mechanism that diagnoses deficiencies and applies targeted corrections until convergence. Extensive experiments demonstrate that TaxCE consistently outperforms existing baselines spanning classical topic models, neural methods, and LLM-based approaches, with average improvements of 11.8, 20.5, and 15.7 percentage points in exclusivity, exhaustivity, and granularity respectively over the strongest baseline. Human evaluation further confirms superior taxonomy quality, actionability, and navigability.
[NLP-67] Reading the News: Adapting Large Language Models to Swedish Journalism Through Continued Pre-Training EMNLP2026
【速读】： 该论文旨在解决大语言模型在特定领域（如瑞典新闻业）中表现有限的问题，尤其是在数据稀缺或研究不足的细分领域中，模型的实用性和准确性往往受限。其核心解决方案是通过在目标领域语料库上对现有大语言模型进行持续预训练（continued pre-training），以实现领域适配。关键创新在于发现：仅进行持续预训练不足以有效提升模型性能，必须结合经验回放（experience replay）机制以缓解灾难性遗忘问题，从而显著提升模型在生成质量与事实知识方面的表现。此外，研究还表明，对于采用低秩适应（low-rank adaptation, LoRA）微调的模型，无需训练的指令引导方法可带来额外增益，而传统评估基准因无法准确捕捉领域内性能提升，凸显了构建针对性评测基准的重要性。因此，该研究强调了在领域适配过程中，结合合适训练策略与专用评估体系的必要性。
链接: https://arxiv.org/abs/2608.30609
作者: Lukas Borggren,Jenny Kunz,Marco Kuhlmann
机构: Linköping University (林雪平大学); Bonnier News
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Accepted at EMNLP 2026 Industry Track
Abstract:Large language models are increasingly capable in general, but their utility can remain modest in niche or understudied areas. One approach to address this limitation is to specialise existing models through additional training on target-domain corpora. In this work, we investigate such continued pre-training for adapting large language models to Swedish journalism, using a high-quality dataset that we curate from millions of news articles. To evaluate the adaptation efficacy, we also construct a novel domain-specific benchmark that covers six editorial tasks. Through full and parameter-efficient fine-tuning across two model sizes, we find that continued pre-training yields benefits in the target domain, but only when paired with experience replay to mitigate forgetting. We observe consistent enhancements in the models’ generation quality and factual knowledge, but not their proficiency in discriminative tasks. Exploring a training-free method to facilitate instruction following, we see further improvements, but exclusively for models trained with low-rank adaptation. Crucially, we demonstrate the importance of targeted evaluation in the adaptation process, as an existing Swedish benchmark largely fails to capture the models’ in-domain performance gains.
[NLP-68] PLC-DPO: Posterior Label Correction in Noisy and Ambiguous Preference Optimization EMNLP2026
【速读】： 该论文旨在解决传统直接偏好优化（Direct Preference Optimization, DPO）在实际应用中因假设所有观测偏好均可靠而导致的模型训练偏差问题。真实数据中常存在反向、弱或模糊的标签，这些噪声会引发有害的策略更新，影响对齐效果。为此，论文提出后验标签修正的DPO（Posterior Label Correction DPO, PLC-DPO），其核心在于利用校准后的策略-参考边际（calibrated policy-reference margin）作为在线证据，动态地将每一对偏好样本归类为“干净”、“翻转”或“平局”三类，并据此采取相应的修正策略。该方法的关键创新在于将噪声偏好学习重构为对监督方向与强度的主动修正，而非简单过滤可疑样本。实验结果表明，在57个数据集-模型-评估组合中，PLC-DPO相较DPO实现了更高的平均胜率（60.5% vs. 55.5%），并在注入噪声、平局压力测试、人类意见分歧分析及自一致性诊断中均展现出稳定的路由性能，能够有效区分被翻转与弱方向性偏好对。
链接: https://arxiv.org/abs/2608.30597
作者: Boryeong Cho,Sumyeong Ahn,Se-Young Yun
机构: KAIST AI(韩国科学技术院人工智能研究中心); KENTECH(韩国技术公司)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注: Findings of EMNLP 2026; Code is available at this https URL
Abstract:Direct Preference Optimization (DPO) simplifies alignment through pairwise comparisons but assumes all observed preferences are reliable. Real data often violates this assumption, leading to reversed, weak, or ambiguous labels that cause harmful policy updates. To address this, we propose Posterior Label Correction DPO (PLC-DPO) to robustly optimize preferences by routing each pair’s training signal as a clean, flip, or tie case. The key idea is to use the calibrated policy-reference margin as online evidence to take appropriate correction actions. This reframes noisy preference learning as actively correcting supervision direction and strength rather than merely filtering suspicious examples. Across 57 dataset-model-benchmark cells, PLC-DPO obtains the best mean win rate against DPO (60.5 vs. 55.5 for the next-best method). Injected-noise and tie stress tests, human disagreement analysis, and self-confirmation diagnostics further show that the routing remains stable and distinguishes flipped from weakly directional pairs.
[NLP-69] Language Proficiency Assessment from Eye Movements in Naturalistic Passage Reading EMNLP
【速读】： 该论文旨在解决传统语言能力测试依赖词汇、语法和阅读理解等语言任务所带来的局限性，提出一种基于认知动机的替代方法，即通过阅读过程中的眼动行为轨迹预测语言水平。其核心解决方案在于将眼动数据应用于更自然的语言使用情境中，包括英语作为第二语言（ESL）的上下文化段落阅读、信息检索型阅读任务，并引入新的语言能力评估指标与预测模型。研究发现，该方法在多种评估场景下均表现有效。针对眼动基语言能力测评中存在的两个关键开放问题——（1）因母语与英语语言相似性导致的评分偏差可能损害测评效度，（2）测评可靠性不足——研究揭示了眼动指标确实存在对语言亲缘关系较近的母语者（L1）的系统性偏向，为此提出了有效的分数去偏方法以消除此类偏差。同时，信度分析表明，眼动基语言能力评分的可靠性优于传统测试方法。整体而言，本研究不仅验证并扩展了眼动追踪在语言能力评估中的应用边界，还为未来基于眼动的行为语言测评技术提供了坚实的实证基础。
链接: https://arxiv.org/abs/2608.30583
作者: Shachar Frenkel,Ido Falah,Omer Shubi,Yevgeni Berzak
机构: Technion - Israel Institute of Technology (以色列理工学院)
类目: Computation and Language (cs.CL)
备注: Accepted to EMNLP
Abstract:Standard language proficiency tests rely on linguistic tasks such as vocabulary, grammar and reading comprehension quizzes. An alternative, cognitively motivated approach, introduced in Berzak et al. (2018), proposed instead to predict language proficiency from behavioral traces of eye movements in reading. In this work, we validate and extend this approach from single sentences to more naturalistic reading of contextualized passages in English as a second language, new proficiency measures, prediction models, and reading in an information seeking regime. We find that the approach is effective in all these evaluations. We further address two key open questions on eye movement based proficiency testing: (1) potential scoring biases that reflect the proximity of the reader’s native language to English, which may undermine validity, and (2) its reliability. We find that eye movement based proficiency scores are indeed biased towards L1s that are linguistically closer to English. We propose a score debiasing method which effectively remedies this issue. The reliability analyses suggest that eye movement proficiency scores are more reliable than standard language proficiency scores. Overall, our results strengthen and broaden the empirical foundations for future eye movement based language assessment technologies.
[NLP-70] Seeing the Unseen: Visual Similarity for Pixel Language Model Adaptation EMNLP2026
【速读】： 该论文旨在解决生成式 AI（Generative AI）在低资源语言场景下，特别是书写系统复杂、形态丰富且使用独特文字的语种中，基于像素的文本语言模型（pixel-based language models, LMs）的跨语言迁移能力受限的问题。其核心挑战在于如何有效利用有限的数据实现对非主流书写系统的适应性优化。解决方案的关键在于通过引入四种渲染层级的视觉脚本相似性度量（rendering-level metrics），量化不同脚本间的视觉与结构相似性，并系统分析数据规模、初始脚本暴露程度以及跨婆罗米系（Brahmic）脚本迁移对模型性能的影响。研究以藏语为案例，发现更高的正字法邻近性（orthographic proximity）可显著促进语义迁移，即使在数据极度稀缺的情况下依然有效；同时揭示了预训练起点的性能不对称现象：多语言预训练模型（PIXEL-M4）虽初始表现更优，但后续适应能力受限，而从混合脚本的单语模型（PIXEL）出发进行微调，在句子级任务上能获得更大提升。这一发现为低资源环境下基于像素的语言模型的数据选择与脚本适配提供了实证依据与设计指导。
链接: https://arxiv.org/abs/2608.30541
作者: Ran Zhang,Miryam de Lhoneux,Wessel Poelman
机构: LAGOM⋅NLP, Department of Computer Science, KU Leuven(鲁汶大学计算机科学系)
类目: Computation and Language (cs.CL)
备注: EMNLP 2026: Main Conference
Abstract:Pixel-based language models (LMs) replace traditional tokenizers by processing rendered images of text, making cross-lingual transfer heavily dependent on the visual and structural properties of writing systems. However, the dynamics of adapting these models to low-resource languages with complex morphology and written in unique scripts are not yet explored. Using Tibetan as a case study, we analyze how continued pre-training of pixel-based LMs is influenced by data scale, initial script exposure, and cross-lingual transfer from languages written in other Brahmic scripts. We introduce four rendering-level metrics to quantify visual script similarity. We evaluate downstream performance across three tasks. Our results show that higher orthographic proximity enhances semantic transfer, even under severe data constraints. Additionally, we find a performance asymmetry based on the pre-training starting point: while multilingual pre-training PIXEL-M4 has stronger initial performance, its capacity for subsequent adaptation seems to be constrained, whereas adapting a monolingual model PIXEL with mixed scripts yields more gains on sentence-level tasks. Our metrics and case study offer empirical observations that could help inform data selection and script adaptation choices when working with pixel-based models in similar low-resource settings.
[NLP-71] WebWorld: The Browser as a World Model for Self-Improving Web Code EMNLP
【速读】： 该论文旨在解决视觉语言模型（VLM）驱动的网页代码自修复过程中存在的结构性缺陷：即负责提出修复方案的模型同时也是评估修复效果的评判者，而仅依赖视觉合理性作为判断标准无法有效反映页面的实际功能性。其核心问题是缺乏一个不可欺骗的验证主体，而浏览器恰好具备这一特性——作为对HTML资源行为的确定性、可执行模拟器，浏览器本质上可视为网页代码的世界模型（world model）。为此，论文提出WebWorld，一个使VLM能够自主与浏览器作为世界模型交互的接口，并由该系统决定哪些交互可作为监督信号。在每一轮迭代中，VLM生成批判性意见，由规划器将其转化为类型化的交互契约；浏览器重新执行候选代码，仅当目标进展达成且所有先前已验证的功能均被保留时，才颁发接受证书；这些经认证的迁移累积形成质量递进机制，成为监督微调（SFT）导出模型所唯一可见的信息。在匹配训练条件下，WebWorld-27B在HTMLBench-400上较Raw-27B提升5.3分，在MiniAppBench-Val上提升14.9分，并达到Kimi-K2.6和GPT-5.4等强前沿系统在交互式HTML生成任务上的水平。同等规模的消融实验表明，基于浏览器背书的准入机制是性能提升的关键：若移除证书机制，9B模型的性能提升几乎完全消失。
链接: https://arxiv.org/abs/2608.30530
作者: Jiajun Wu,Jian Yang,Yaxin Du,Wei Zhang,Haowen Wang,Junhang Cheng,Yuxuan Zhang,Tuney Zheng,Xianglong Liu,Ming Zhou
机构: Beihang University(北京航空航天大学); Shanghai Jiao Tong University(上海交通大学); IQuest Research; Langboat
类目: Computation and Language (cs.CL); Software Engineering (cs.SE)
备注: EMNLP Main Conference
Abstract:VLM-driven self-improvement of web code has a structural flaw: the model that proposes the repair is the model that judges it, and visual plausibility under that judge is a poor proxy for whether the page actually works. What the loop is missing is a counterparty the VLM cannot fool, and the browser already is that counterparty: a deterministic, executable simulator of how an HTML artifact behaves under user actions, and in everything but name a world model for web code. We present WebWorld, the interface that lets a VLM prior interact with this browser-as-world-model autonomously and decides which interactions become supervision. Each round, the VLM emits a critique that the planner compiles into a typed interaction contract; the browser re-executes the candidate and issues an acceptance certificate only when both target progress and preservation of every previously verified capability hold; certified transitions accumulate as a quality ratchet that is the only thing the SFT export ever sees. Under matched training, WebWorld-27B improves Raw-27B by 5.3 points on HTMLBench-400 and 14.9 points on MiniAppBench-Val, and reaches the level of strong frontier systems such as Kimi-K2.6 and GPT-5.4 on interactive HTML generation. Equal-size ablations show that browser-backed admission carries the gain: without the certificate, the matched 9B lift nearly disappears.
[NLP-72] ScienceArena: Benchmarking LLM s on Latest Scientific Olympiad Competitions EMNLP2026
【速读】： 该论文旨在解决前沿大语言模型（LLM）在科学推理任务中因基准数据饱和与数据污染导致的真实科学思维能力评估失真问题。其核心解决方案是构建一个名为 \textscScienceArena 的奥林匹克风格评测基准，涵盖物理、化学、生物三个领域的十三个公开竞赛试题（如 IPhO、IChO 2025–2026，IBO 2023 等），采用开放式、多步骤问题设计，并结合过程性评分量规（process-credit rubrics），以更真实地反映复杂科学推理能力。为确保数据质量，研究团队通过专家审核的数字化流程，将官方试题、图表、解答与评分标准转化为结构化数据，并由奥赛奖牌获得者进行验证。为实现高效可扩展的评估，研究进一步基于五种模型在过往奥赛题中的作答结果，以奖牌得主为金标准对大语言模型作为评判者（LLM-as-judge）进行校准，结果显示两名强模型评判者与专家总分偏差不超过一分。分析表明，当前顶级模型虽在部分国际赛事中达到奖牌级评分表现，但化学领域及长时程一致性仍为关键瓶颈。研究同时提供了交互式演示平台以支持开放评估。
链接: https://arxiv.org/abs/2608.30517
作者: Guangxiang Zhao,Qilong Shi,Xusen Xiao,Wenpu Liu,Yaoming Li,Linfeng Hao,Shuyang Hou,Zijian Guo,Xinrui Zhang,Yuntian Zhao,Zhengyang Wang,Wenrui Liu,Yuhan Wu,Tong Yang,Lin Sun,Xiangzheng Zhang
机构: Qiyuan Tech(奇源科技); Tsinghua University (清华大学); The University of Hong Kong (香港大学); Peking University (北京大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 18 pages, EMNLP 2026 (Main)
Abstract:Benchmark saturation and data contamination increasingly obscure genuine scientific reasoning in frontier LLMs. We introduce \textscScienceArena, an olympiad-style benchmark from thirteen public science competitions in physics, chemistry, and biology, including IPhO and IChO 2025–2026, IBO 2023, USAPhO 2026, and USNCO 2025. Its open-ended, multi-step problems use process-credit rubrics, making faithful scoring difficult. We build ScienceArena through an expert-audited digitization pipeline that converts official exams, figures, solutions, and rubrics into structured items verified by olympiad medalists. To scale evaluation beyond costly human grading, we calibrate LLM-as-judge against medalist ground truth on archived answers from five models across IPhO and IChO; two strong judges stay within one point of expert total scores. Medalist notes show that failures often stem from visual grounding, structure fidelity, and global problem control rather than missing terminology. Evaluating fourteen recent LLMs with interleaved solving, we find that top models obtain medal-equivalent rubric scores on several public international exams, while chemistry and long-horizon consistency remain key bottlenecks. We provide an interactive \hrefthis https URLdemo.
[NLP-73] Lot Machine: Multimodal Lot Extraction from Auction Catalogs ECCV2026
【速读】： 该论文旨在解决历史拍卖目录（auction catalogs）在艺术史溯源研究与艺术市场分析中因缺乏机器可读表示而导致的大规模自动化分析难题。其核心问题是：尽管拍卖目录遵循特定领域规范，但其内部格式高度不一致，难以进行结构化数据提取。为此，论文提出一种基于视觉-语言模型（Vision-Language Models, VLMs）的自动化流水线，用于从19至20世纪德语区拍卖目录数据库“German Sales”中提取逐条拍卖品（lot-level）的结构化元数据。解决方案的关键在于通过设计多样化的提示策略（prompt strategies）与受限解码框架，在不同部署模式下（包括商业API、机构网关及本地量化模型）评估VLM性能，以适应文化遗产机构在预算、计算资源和数据隐私方面的实际限制。研究发现，商业端点虽性能最优，但本地部署结合输出结构强制机制可确保生成结果符合有效JSON格式，从而实现可行且隐私保护的数据处理路径。尽管仍需不同程度的人工校正，该工作证明了基于VLM的方案能够有效解锁历史拍卖目录，支持大规模自动化分析。
链接: https://arxiv.org/abs/2608.30510
作者: Mathias Zinnen,Alisha Mund,Sabine Lang,Lukas Hüttner,Thomas Gorges,Vincent Christlein
机构: Pattern Recognition Lab, FAU Erlangen(模式识别实验室，埃尔兰根大学); Department of Digital Humanties and Social Studies, FAU Erlangen(数字人文与社会科学系，埃尔兰根大学)
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Digital Libraries (cs.DL)
备注: Accepted at the VISART Workshop (Computer Vision for Art Analysis), ECCV 2026. 19 pages, 6 figures, 5 tables. Supplementary material included as an appendix. Code, benchmark data, and prompt templates: this https URL
Abstract:For provenance research and art market studies, auction catalogs are an essential resource to trace specific objects over time and space. While historical auction catalogs follow established domain conventions, their internal formatting remains highly variable, and their large-scale analysis is currently restricted by the lack of machine-readable representations of the auction lots. We propose a pipeline to automatically extract structured lot-level metadata from German Sales, a large database of historical auction and sales catalogs from the 19th and 20th centuries. Using a manually annotated test set of representative catalog pages, we evaluate Vision-Language Models (VLMs) under varying prompt strategies and constrained decoding frameworks. To reflect the practical constraints faced by cultural heritage institutions, including budget, compute resources, and data privacy requirements, we benchmark the methods across different deployment modes ranging from commercial providers to locally hosted, quantized models. We find that commercial endpoints establish the performance ceiling, while institutional gateways offer a viable, privacy-preserving alternative. Local deployments remain feasible, but strictly require enforcing the output structure during generation to guarantee a valid JSON format. While varying degrees of human-in-the-loop correction are still necessary, this work demonstrates that a VLM-based pipeline can successfully unlock historical auction catalogs for large-scale automated analysis.
[NLP-74] UTILMEM: Benchmarking Evidence Utilization in Long-Term Conversational Memory
【速读】： 该论文旨在解决当前对话系统在长期记忆能力评估中过度依赖点式事实回忆（pointwise factual recall）的问题，而忽视了真实场景下对分散、隐含且噪声较大的历史信息进行整合以生成连贯、任务导向输出的复杂能力。这一核心能力被称为记忆利用（memory utilization）。其解决方案的关键在于提出一个名为UtilMem的诊断性基准，包含跨五个领域的1,717个实例，专门用于评估四项尚未充分探索的记忆利用维度：基于密集历史的推理、识别隐含相关记忆、将分布式证据整合为摘要、分析或计划，以及抵抗语义相似干扰项的干扰。实验表明，现有基于检索和记忆增强的系统虽在传统事实记忆基准上表现良好，但在实际记忆利用任务中普遍存在整合能力不足与抗干扰能力弱的问题，揭示了信息获取与有效使用之间存在显著鸿沟。因此，论文强调未来长时对话记忆的发展需依赖能够显式支持证据整合与抗检索干扰的新型架构设计。
链接: https://arxiv.org/abs/2608.30508
作者: Peijun Qing,Fobo Shi,Soroush Vosoughi
机构: Dartmouth College(达特茅斯学院); Wuhan University(武汉大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Long-term memory is increasingly important for conversational agents, yet existing benchmarks primarily measure memory through pointwise factual recall: whether a system can recover isolated facts or event-level details from prior interactions. Real-world memory use, however, often requires a more demanding capability: integrating distributed, implicit, and noisy evidence across extended interaction histories into coherent, task-oriented outputs. We call this capability memory utilization. Here, we introduce UtilMem, a diagnostic benchmark comprising 1,717 instances across five domains, designed to evaluate four underexplored aspects of memory utilization: reasoning over dense histories, identifying implicitly relevant memories, synthesizing distributed evidence into summaries, analyses, or plans, and resisting interference from semantically similar distractors. Evaluating a diverse set of retrieval-based and memory-augmented systems, we find that strong performance on conventional factual-memory benchmarks does not reliably translate into effective memory utilization. Moreover, retrieval alone is insufficient: even when relevant evidence is successfully recovered, systems frequently fail to integrate information across sessions or to distinguish useful evidence from plausible distractors. These findings expose a substantial gap between accessing stored information and using it effectively, and suggest that progress in long-term conversational memory will require architectures that explicitly support evidence integration and robustness to retrieval interference. Code is available at this https URL.
[NLP-75] wo Centuries of Sexism in British Parliament: A Computational Analysis of Womens Representation in the Hansard Corpus EMNLP2026
【速读】： 该论文旨在揭示英国议会两百年来关于女性权利辩论中隐含的系统性性别偏见，即使在支持女性权利的言论中亦存在结构性性别歧视。其核心问题是：立法机构在讨论女性选举权与政治代表权时，如何通过语言表达潜藏的性别刻板印象，并且这些偏见是否随时间演变及在不同立场间呈现差异。解决方案的关键在于运用大语言模型（Large Language Models）对6,531份来自1803–2005年《汉萨德》（Hansard）的议会发言进行分类，识别发言者对女性参政权的立场，并结合“矛盾性别主义量表”（Ambivalent Sexism Inventory, ASI）框架分析其中的性别歧视内容。研究发现，反对女性政治代表的演讲中有54%包含性别歧视，而支持者中仅21%存在类似内容；且反女权话语同时融合敌意型与善意型性别主义，而支持方则以善意型性别主义为主。此外，女性议员支持女性政治权利的比例（93%）显著高于男性议员（70%），这一差距在女性获得选举权后才逐渐缩小。研究结果表明，善意型与敌意型性别主义在不同修辞语境中被差异化使用，与矛盾性别主义理论高度一致，揭示了性别偏见在政治话语中的深层运作机制。
链接: https://arxiv.org/abs/2608.30485
作者: Mohammad Omar Khursheed,Mandira Sawkar,Ashiqur R. KhudaBukhsh
机构: Rochester Institute of Technology (罗切斯特理工学院)
类目: Computation and Language (cs.CL); Computers and Society (cs.CY); Machine Learning (cs.LG)
备注: Accepted at EMNLP 2026 (Main Conference Track)
Abstract:The language a legislature uses to debate women’s rights, even in favour of them, encodes systematic patterns of sexism that persist across two centuries. In this work, we analyse 6,531 speeches over 200 years of UK parliamentary debate (Hansard, 1803-2005) by using large language models to classify a speaker’s perspective towards women’s suffrage and political representation, as well as analyse sexist speech in parliament from the lens of the Ambivalent Sexism Inventory. We also release this parliamentary dataset, an organized and metadata-enriched version of the publicly available Hansard Corpus optimized for computational social science research, with 6.7 million speeches across 1.2 million debates, with 89% gender-matching for speeches by MPs from the House of Commons. We find that 54% of speeches opposing women’s representation contain sexist content, compared to 21% of speeches that are for the cause, and that the two sides use fundamentally different types of sexism: anti-suffrage rhetoric combines hostile and benevolent framing, while pro-suffrage sexism is overwhelmingly benevolent. Female MPs support women’s political rights at 93% compared to 70% for male MPs, a gap that closes only after enfranchisement. Our findings are evidence that benevolent and hostile sexism are used in different rhetorical contexts in a manner consistent with the theory of Ambivalent Sexism.
[NLP-76] Agents in the Large: Perception-Centered Architecture for Persistent Agents
【速读】： 该论文旨在解决当前认知语言代理（Cognitive Language Agents）在长期、动态环境中难以持续提供有效服务的问题。现有框架多聚焦于解决用户指定的有限任务，缺乏对持久性（persistent）智能体在时间跨度内需求演化、上下文变化及服务流程自适应能力的系统性建模与指导。其核心挑战在于如何构建能够持续感知环境、理解上下文演变并自主调整服务策略的智能体架构。为此，论文提出感知中心型持久智能体架构（Perception-Centered Architecture for Persistent Agents, Pera），其关键在于将智能体组织为以感知（perception）与控制（control）为核心的闭环系统：通过持续捕获来自任务执行过程、内部状态及外部环境的信号，构建生命周期任务（lifecycle tasks），驱动服务流程的持续运行与动态演化。Pera不仅为已有工作提供了回顾性分类框架，还为未来构建具备长期适应能力的生成式智能体提供了前瞻性架构范式，标志着从“小规模编程”向“大规模系统化智能架构”的演进。
链接: https://arxiv.org/abs/2608.30478
作者: Shihan Dou,Haoxiang Jia,Shichun Liu,Feng Chen,Chenhao Huang,Yujiong Shen,Shaofan Liu,Jiayi Chen,Jiahang Lin,Honglin Guo,Qianyu He,Minghao Guo,Ziyi Ye,Pluto Zhou,Tao Gui,Qi Zhang,Xuanjing Huang
机构: Fudan University (复旦大学); Peking University (北京大学); Nanyang Technological University (南洋理工大学); Tencent(腾讯)
类目: Computation and Language (cs.CL)
备注: 41 pages, 5 figures
Abstract:Cognitive language agents have achieved substantial progress by equipping language models with memory, tools, and decision-making procedures, enabling agents to reason and act in interactive environments. Existing frameworks largely cast these agents as systems for solving user-specified, bounded tasks. An increasingly important goal is for language agents to provide persistent assistance in long-lived settings where user needs, context, and service procedures persist and change, and to remain useful across the broad range of tasks that arise over time. Yet we still lack a framework to characterize persistent AI agents, organize existing work, and guide future development. To this end, we propose a Perception-Centered Architecture for Persistent Agents (Pera). Pera describes a persistent agent organized around perception and control components that continually perceive service-relevant signals from episodic task executions, internal context, and changes in the surrounding environment, and use these signals to construct lifecycle tasks. These tasks drive the ongoing operation and adaptation of the agent’s service procedures. We use Pera to retrospectively organize recent work, examine a detailed case study, and offer forward-looking insights for building more capable persistent agents. Just as software engineering moved from programming in the small to programming in the large, Pera frames the evolution of language agents as an analogous architectural transition toward long-lived, adaptive intelligence systems.
[NLP-77] ImageEval 2026: Culturally Grounded Arabic Multimodal Evaluation
【速读】： 该论文旨在解决阿拉伯语语境下多模态生成内容的文化相关性评估难题，特别是针对阿拉伯语语音与图像-文本推理中的文化准确性问题。其核心挑战在于如何有效衡量生成式 AI (Generative AI) 在跨模态任务中对阿拉伯文化背景的忠实度，包括口语视觉问答（AynVQA）中的幻觉检测以及文生图（text-to-image generation）中的文化一致性评估。解决方案的关键在于构建一个双任务共享评测基准：一是覆盖英语与现代标准阿拉伯语（MSA）的口语视觉问答与图像基幻觉检测任务；二是基于真实文化语境的文本到图像生成文化准确性评测平台（CRAI-Bench）。通过引入多语言、多模态、文化敏感的数据集与评估流程，结合零样本提示、视觉-语言模型微调、语音识别流水线、集成策略及得分校准等多种技术手段，系统性地评估了参赛模型在文化语境理解与生成能力上的表现。该研究不仅揭示了当前多模态模型在阿拉伯文化语境下的显著局限性，还通过公开全部数据集与评估脚本，为后续研究提供了可复现的基准与方法参考。
链接: https://arxiv.org/abs/2608.30475
作者: Samir Abdaljalil,Hunzalah Hassan Bhatti,Ahlam Bashiti,Farina Amir,Md Arid Hasan,Basel Mousi,Nadir Durrani,Fahim Dalvi,Zien Sheikh Ali,Erchin Serpedin,Hasan Kurban,Mustafa Jarrar,Shammur Absar Chowdhury,Firoj Alam
机构: Texas A&M University (德州农工大学); Qatar Computing Research Institute (卡塔尔计算研究研究所); Birzeit University (比尔宰特大学); Hamad Bin Khalifa University (哈马德本哈利法大学); University of Toronto (多伦多大学); Qatar Computing Research Institute (卡塔尔计算研究研究所); Texas A&M University (德州农工大学); Hamad Bin Khalifa University (哈马德本哈利法大学); Hamad Bin Khalifa University (哈马德本哈利法大学); Qatar Computing Research Institute (卡塔尔计算研究研究所); Qatar Computing Research Institute (卡塔尔计算研究研究所)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Arabic LLMs, Multilingual, Multimodal, Shared Task
Abstract:We present an overview of the ImageEval 2026 shared task on culturally grounded Arabic multimodal evaluation. It includes two tasks: (i) AynVQA, covering spoken visual question answering and image-grounded hallucination detection in English and Modern Standard Arabic (MSA), and (ii) CRAI-Bench, evaluating the cultural accuracy of text-to-image generation. A total of 14 teams participated in the test phase, with 12 teams submitting system description papers. Participating systems used a range of approaches, including zero-shot prompting, fine-tuning of vision-language models, speech-recognition pipelines, ensembling, and score calibration. We describe the task setup, datasets, evaluation procedure, and participating systems, and summarize the main results across the different tracks. All datasets and evaluation scripts from the shared task are released to the research community. The shared task highlights the challenges of culturally grounded multimodal evaluation, particularly for Arabic speech and image-text reasoning.
[NLP-78] More Capable Less Faithful: A Multilingual Analysis of Mathematical (Un)Solvability Detection in LLM s
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在数学推理中对问题可解性（Solvability Detection）的识别能力不足这一核心挑战，尤其关注多语言环境下该能力的表现差异。其关键在于揭示：多语言间可解性判断失败是否源于模型内部“可解性信念”（Solvability Belief）的跨语言不一致性，或仅是语言表达层面的障碍。为此，研究首次构建了包含配对可解与不可解数学题目的多语言基准测试集，将ReliableMath扩展至法语和希腊语，并基于此训练多语言探测器以分析先进LLMs在行为表现、表征特征及信念忠实度（faithfulness）三个维度上的可解性检测能力。研究发现，可解性信念在很大程度上是一种普遍存在的、与语言无关的内在特征；而尽管英语等高资源语言在数学推理任务中表现更优，其可解性判断的忠实度反而较低，表明语言资源丰富性并不必然带来更高的推理可信度。
链接: https://arxiv.org/abs/2608.30463
作者: Maria-Eleni Zoumpoulidi,Nikolaos Xiros,Georgios Paraskevopoulos
机构: Institute for Language and Speech Processing, Athena Research Center (雅典研究中心语言与语音处理研究所); Google DeepMind (谷歌深度智障); Anthropic (Anthropic)
类目: Computation and Language (cs.CL)
备注: 10 pages, 3 Figures
Abstract:Solvability detection is one of the most challenging aspects of mathematical reasoning for Large Language Models (LLMs). While prior work has studied this capability extensively, these analyses have been limited to English. Consequently, it remains unclear whether multilingual failures arise from differences in internal Solvability Belief or from language-dependent failures to express it. To address this gap, we introduce the first multilingual benchmark of paired solvable and unsolvable mathematical problems, extending ReliableMath to French and Greek. Using this, we train multilingual probes predicting Solvability Belief and analyze the solvability detection capabilities of state-of-the-art LLMs behaviorally, representationally, and in terms of faithfulness. We find that Solvability Belief is encoded as a largely universal, language-agnostic feature, and that higher-resource languages such as English, despite achieving stronger mathematical reasoning performance, exhibit lower solvability-detection faithfulness.
[NLP-79] Enhancing Low-Resource Language Reasoning via High-Resource Language Feature Transfer EMNLP2026
【速读】： 该论文旨在解决大语言模型在跨语言任务中表现差异显著的问题，尤其关注即使在语义等价的任务下，高资源语言（HRLs）与低资源语言（LRLs）之间仍存在推理能力差距的现象。现有研究多将此归因于预训练数据分布、分词方式或基准覆盖度的差异，而本文提出一个互补性假设：高资源语言更可能有效激活对特定任务（如数学推理）至关重要的潜在计算机制，而低资源语言虽能表达相同任务，却未能充分触发这些机制。为此，论文提出一种机制干预框架，通过在残差流（residual-stream）激活上使用稀疏自编码器（sparse autoencoders），识别并分离出在高资源语言中与任务特定推理相关且高度稀疏的潜在特征，同时过滤掉源语言特异性及通用生成类特征。进而构建指向这些关键特征的“引导方向”（steering directions），并在低资源语言推理过程中注入，以测试其功能作用——抑制这些特征应损害源语言推理性能，而激活则可部分恢复目标语言的推理表现，超越随机控制与非任务相关控制基线。该方法将跨语言推理差距重新诠释为机制激发失败而非能力缺失，并提供了一种无需翻译、微调或改变用户语言即可实现特征级迁移的因果可验证路径。
链接: https://arxiv.org/abs/2608.30462
作者: Minju Song,Hyeon Hwang,Junhyun Lee,Jaewoo Kang
机构: Korea University(高丽大学); Hankuk University of Foreign Studies(韩国外国语大学); Noah’s Farm; AIGEN Sciences
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted to EMNLP 2026 Findings
Abstract:Large language models exhibit substantial performance variation across languages, even when solving semantically equivalent tasks. Existing analyses often treat this phenomenon as an observational disparity caused by differences in pretraining data, tokenization, or benchmark coverage. We study a complementary hypothesis: high-resource languages (HRLs) may more reliably elicit latent computations useful for task-specific (i.e. mathematical) reasoning, while lower-resource languages (LRLs) may under-activate those computations despite expressing the same task. To test this hypothesis, we introduce a mechanistic intervention framework for identifying and transferring task-relevant sparse latent features across languages. Using sparse autoencoders over residual-stream activations, we isolate features enriched in successful HRL task-specific reasoning while filtering out source-language and generic-generation features. We then construct steering directions from these features and inject them during LRL inference. The resulting interventions test whether the selected features are functionally involved in the observed reasoning gap: suppressing them should impair source-language reasoning, while activating them should partially recover target-language reasoning beyond random and non-task controls. Our framework reframes some cross-lingual reasoning gaps as failures of mechanism elicitation rather than capability absence, and offers a causally testable route to feature-mediated transfer without translation, fine-tuning, or changing the user-facing language.
[NLP-80] From Final Artifacts to Trajectories: Retrospective Process Supervision for Evidence-Grounded Long-Form Generation
【速读】： 该论文旨在解决在开放性任务中大规模获取高质量轨迹数据（trajectory data）的难题，尤其是在缺乏唯一真实答案（ground truth）且人工标注与验证成本高昂的场景下，如何有效提升生成式AI（Generative AI）在复杂推理与证据搜寻类任务中的表现。其解决方案的关键在于提出一种自迭代优化的回溯过程监督框架——RetroGen。该框架的核心思想是：尽管专家级行为轨迹稀缺，但预训练数据中广泛存在的高质量最终成果（如文献综述、分析报告和法律判决）可作为证据追寻过程的压缩表征。RetroGen通过从这些高质量成果中重构潜在的中间轨迹，利用成果本身及其支持证据对重构轨迹进行双重验证，并基于模型自身成功重建的数据进行训练，从而实现无需依赖更强模型生成轨迹即可持续优化智能体（agent）的推理能力。实验表明，该方法显著提升了模型在事实一致性、忠实性合成以及长时序证据搜寻任务中的表现。
链接: https://arxiv.org/abs/2608.30461
作者: Junjie Huang,Jiarui Qin,Di Yin,Weiwen Liu,Yong Yu,Xing Sun,Weinan Zhang
机构: Shanghai Jiao Tong University (上海交通大学); Tencent Youtu Lab (腾讯优图实验室)
类目: Computation and Language (cs.CL)
备注:
Abstract:Trajectory data is getting more vital for training large language models for boosting the agentic abilities. Unlike the verifiable domains such as coding or mathematics, scaling trajectory data for open-ended tasks is much more difficult because these tasks lack singular ground truth and are costly to annotate or verify. In this paper, we propose RetroGen, a self-improving framework of retrospective process supervision. Our key observation is that although expert trajectories are scarce, high-quality final artifacts such as literature reviews, analyst reports and legal judgments, are abundant in pre-training data and can be viewed as compressed traces of the evidence-seeking processes that produced them. RetroGen reconstructs candidate latent trajectories from expert artifacts, verifies them against both the artifact and supporting evidence, and trains models on their own successful reconstruction data, without requiring trajectory data from stronger models. Experiments show that RetroGen improves grounding, faithful synthesis, and long-form evidence-seeking agent tasks.
[NLP-81] Learning Where Outcomes Change:Credit-Addressable Reasoning for Multimodal Geometry
【速读】： 该论文旨在解决多模态几何推理中视觉关系提取不精确及多步推理由现有自由形式推理轨迹导致的决策路径模糊问题。其核心挑战在于：传统方法缺乏对推理过程中各语义单元的显式追踪能力，且轨迹级强化学习将终端奖励均匀分配至整个响应序列，难以实现精准的信用分配。为此，论文提出“可定位信用推理”（credit-addressable reasoning）机制，即在推理过程中暴露的语义单元同时定义了学习阶段进行选项比较与信用分配的具体位置。通过Code-CoT框架，将图形信息保留并以行可寻址的可执行代码形式表示视觉关系，将推理过程组织为具有类型约束的事件序列；进一步提出CE-GRPO算法，利用结构先验确定事件边界，并基于类型归一化熵进行采样，从共享前缀生成完整推理延续，将结果差异转化为局部优势信号。在九个几何基准测试中，CE-GRPO平均准确率达76.04%，显著优于Qwen3-VL-8B和轨迹级GRPO，分别提升8.09和3.43个百分点，且随着中间事件数量增加，相对优势持续扩大，验证了表示与优化协同设计在长依赖、高复杂度多模态推理中的关键作用。
链接: https://arxiv.org/abs/2608.30457
作者: Jiani Guo,Junjie Wang,Jie Wu,Pengxiang Zhao,Dongdong Zhang,Shaohan Huang,Yujiu Yang,Furu Wei
机构: 未知
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注:
Abstract:Multimodal geometry reasoning requires VLMs to extract precise visual relations and preserve them through multi-step deduction. Existing free-form traces obscure the decisions that determine the answer, and trajectory-level reinforcement learning distributes a single terminal signal across the entire response. We introduce credit-addressable reasoning, in which the semantic units exposed during inference also define where learning compares alternatives and assigns credit. We instantiate this principle with Code-CoT, which retains the diagram, represents visual relations as line-addressable executable code, and organizes reasoning into typed events, and CE-GRPO, which selects event boundaries using structural priors and type-normalized entropy, samples complete continuations from shared prefixes, and converts outcome differences into localized advantages. Across nine geometry benchmarks, CE-GRPO achieves an average accuracy of 76.04, outperforming Qwen3-VL-8B and trajectory-level GRPO by 8.09 and 3.43 points, respectively. Its relative advantage increases with the number of intermediate events, demonstrating the value of representation–optimization co-design for long, dependency-heavy multimodal reasoning.
[NLP-82] Graph Evidence Is Not Enough: Diagnosing Native Decoder Use in Graph-Augmented LLM s EMNLP2026
【速读】： 该论文旨在解决图增强型大语言模型（graph-augmented large language models）中一个关键问题：尽管外部计算生成的图证据被引入输入，但模型的原生解码器（native decoder）是否真正能够有效利用这些图结构信息仍存疑。现有方法在面对需要精确拓扑推理的任务时表现不佳，表明“提供图证据”并不等同于“使图证据可被使用”。其解决方案的关键在于提出一种诊断性干预三角（intervention triangle），通过三个匹配条件——可读图证据、打乱图证据与无图输入——将图证据的引入、结构可读性以及解码器可用的拓扑关系进行分离，从而精准定位问题所在。基于此诊断，作者提出了S²GE（Structure-aware, Sampling-guided, and Alignment-preserving Graph Enhancement）框架，采用查询感知采样、端点与邻近性排序及结构保持对齐等机制，显著提升了原生解码器对图结构信息的利用率。在DBLP、Biomedical、GoodReads和PubMed四个数据集上，S²GE实现了36.5%、57.8%、76.6%和52.0%的严格精确匹配得分，相较于最强的原生生成基线平均提升53.5个百分点。此外，该研究还揭示了三种不同行为模式：有害打乱（harmful-shuffle）、打乱鲁棒（shuffle-robust）和无图饱和（no-graph-saturated），为未来图增强模型的设计提供了重要启示。
链接: https://arxiv.org/abs/2608.30437
作者: Xiaoyu Guo,Pengcheng Chen,Jiong Yu,Yi Lu,Yaohua Wang,Ziyang Li
机构: 未知
类目: Computation and Language (cs.CL)
备注: 18 pages, 4 figures, ccepted at EMNLP 2026 (Main Conference)
Abstract:Graph-augmented large language models often assume that graph evidence produced by external computation and placed in the input can be used by the native decoder. We test this assumption with HopQA, a deliberately bounded diagnostic that asks for the shortest-hop distance between two query nodes. Because the answer is a small integer and the target is purely topological, failure cannot be dismissed as open-ended generation or ambiguous evaluation. Yet existing graph-augmented baselines still fail on this setting, showing that providing graph evidence is not the same as making it usable. We introduce an intervention triangle with three matched conditions: readable graph evidence, shuffled graph evidence, and no-graph input. This separates evidence inclusion, structural readability, and decoder-usable topology. Guided by this diagnosis, we present S ^2 GE as an instance showing that diagnosis-driven interface design can improve native decoder usability. S ^2 GE uses query-aware sampling, endpoint and proximity-based ordering, and structure-preserving alignment. Across DBLP, Biomedical, GoodReads, and PubMed, S ^2 GE achieves strict exact-match scores of 36.5% , 57.8% , 76.6% , and 52.0% , improving over the strongest native-generation baseline by 53.5 points on average. The interventions further reveal harmful-shuffle, shuffle-robust, and no-graph-saturated regimes.
[NLP-83] EvoSkill Injection: Red-Teaming Autonomous Skill Generation and Evolution in Self-Evolving Agents EMNLP2026
【速读】： 该论文旨在解决自演化生成式智能体（self-evolving agents）在自主生成、优化与复用技能过程中可能引入的恶意能力注入风险，即“技能演化注入”（EvoSkill Injection）威胁。其核心问题是：当前基于大语言模型（LLM）的自演化智能体系统依赖于从历史经验中自动学习和迭代更新技能，但这一过程可能被恶意攻击者利用，通过诱导生成具有隐蔽危害性的恶意技能，并使其长期存储与重复激活，从而导致系统能力的持久性污染。解决方案的关键在于提出SARGE框架——一个针对自演化技能生成与演进流程的红队测试（red-teaming）方法，通过迭代式的生成、升级与强化交互机制，主动探测并诱发恶意技能的形成；同时构建EvoSkillBench基准数据集以模拟恶意交互轨迹，以及EvoSkillSafetyBench后置评估基准用于检测恶意技能是否被成功检索并触发为有害行为。实验结果表明，SARGE能够有效诱导恶意技能的生成与持久存储，且这些技能可被反复激活，揭示了自演化智能体在开放环境中面临严重的持续性能力污染风险。
链接: https://arxiv.org/abs/2608.30429
作者: Doyun Kim,Chanwoo Kim,Sugyeong Eo,Yeo-Chan Yoon,Chanjun Park
机构: Soongsil University(松石大学); Yonsei University(延世大学); Jeju National University(济州国立大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026
Abstract:LLM-based agent systems increasingly adopt skill-based architectures to reduce repetitive reasoning costs and improve stable, efficient task execution. Recent studies propose self-evolving agents that autonomously generate, refine, and reuse skills from past experiences to enable continuous capability evolution. However, autonomous skill evolution introduces a new attack surface in which malicious capabilities are generated, stored, and reused as legitimate skills. In this paper, we define EvoSkill Injection as a threat model targeting the autonomous skill generation and evolution pipeline of self-evolving agents. We further propose SARGE (Red-teaming Autonomous Skill Generation and Evolution in self-evolving agents), a red-teaming framework for evaluating this threat model through iterative generation, escalation, and reinforcement interactions. To support our framework, we construct EvoSkillBench, a benchmark dataset of malicious interaction trajectories for inducing malicious skill formation in self-evolving agents, and introduce EvoSkillSafetyBench, a post-attack benchmark for evaluating whether injected malicious skills are subsequently retrieved and activated as harmful behaviors. Our evaluation shows that SARGE induces malicious skill formation and that injected skills are persistently stored and repeatedly activated, highlighting the risk of persistent capability corruption.
[NLP-84] Lies We Can See: Joint Verbal and Non-Verbal Deception by VLM Agents in Embodied Social Interactions
【速读】： 该论文旨在解决大语言模型（LLM）与视觉语言模型（VLM）代理在多智能体环境中实施战略性欺骗（strategic deception）所带来的人工智能对齐与安全挑战。现有研究多依赖于仅基于文本的社交推断类游戏测试环境，且局限于单一固定的代理配置，忽略了欺骗行为中至关重要的非语言感知-运动通道（non-verbal sensorimotor channels），导致难以区分观察到的行为是源于模型内在认知还是外部实验框架的影响。为此，论文提出MineAmongUs——一个3D多模态的《Among Us》沙盒环境，使“冒名顶替者”代理能够通过语言与非语言行为的协同作用实施欺骗；同时设计了可配置的VLM代理框架ARIA，支持五个认知组件的消融分析，并引入基于欺骗分类体系的原子级与弧线级标注方案，借助大语言模型作为裁判（LLM-as-a-Judge）实现了接近人类水平的标注一致性。实证结果表明，VLM代理通过语言与非语言行为的联合策略达成胜利，其中非语言通道在不同消融实验及跨模型评估中均展现出更关键的决定性作用。该研究为具身化VLM代理的对齐研究开辟了新路径。
链接: https://arxiv.org/abs/2608.30428
作者: Jaewoo Ahn,Junseo Kim,Hyunseo Kim,Heeseung Yun,Jaehyeon Son,Zsolt Kira,Gunhee Kim
机构: Seoul National University (首尔国立大学); Inha University (仁荷大学); KAIST (韩国科学技术院); Georgia Institute of Technology (佐治亚理工学院)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: Workshop on Agent Behavior (WAB) at COLM 2026. Project page: this https URL
Abstract:Strategic deception by LLM and VLM agents has emerged as a central AI alignment and safety concern. Social-deduction games (where each player holds a hidden role and communicates with others to deduce identities) serve as the canonical testbed, particularly in multi-agent settings. Existing testbeds, however, are text-only and run on a single fixed agent configuration, missing the non-verbal sensorimotor channels treated as core by deception taxonomies and leaving it ambiguous whether an observed behavior reflects the underlying model or the surrounding harness. We introduce MineAmongUs, a 3D multimodal Among Us sandbox where imposter agents must deceive crewmates through joint verbal and non-verbal action. We also propose ARIA, a configurable VLM-agent harness that exposes five cognitive-component ablation axes; and an atom- and arc-level annotation scheme grounded in deception taxonomies and operationalized at scale by an LLM-as-a-Judge reaching near-human atom-labeling agreement. Empirical results show that VLM agents pursue imposter wins through joint verbal and non-verbal deception, with non-verbal channels emerging as the more decisive winning contributors across both harness ablation and cross-VLM evaluation. Taken together, our work opens a new path for embodied VLM-agent alignment research.
[NLP-85] Ceiling-Clipped Acceptance Histograms Indicate Stranded Speed-up in Block-Diffusion Speculative Decoding
【速读】： 该论文旨在解决生成式AI（Generative AI）中推测解码（speculative decoding）技术因“块边界扩散”（block-diffusion）类草稿模型（drafter）在推理时超出其训练块长度（block size）后导致的性能退化问题，即“未实现的接受率”（stranded speed-up）——尽管草稿模型在训练时能高效生成高接受率的完整块，但在实际推理中一旦块长超过训练范围，其双向注意力机制会改变早期位置的输出分布，导致验证失败，从而浪费了潜在的加速能力。解决方案的关键在于提出一种名为DBloom的后训练方法：在扩展草稿模型块长时，通过一个短周期、聚焦于新暴露位置的微调策略，使草稿模型在更大块长下仍保持与目标模型一致的分布特性，从而恢复并提升接受率。实验表明，将DFlash和DFlare从块长16扩展至24，在Qwen3系列目标模型上可使每提示词的承诺长度（committed length）中位数提升0.8个token（最高+1.1），若先进行连续性微调再扩展，提升可达1.37个token；在Gemma-4-12B-IT模型上，中位提升达0.41个token，且全链路优化（持续微调+扩展）进一步带来0.29至0.98个token的增益；在与JetSpec等树结构草稿器的对比中，DBloom在树预算≤64节点时均表现出更优的承诺长度。
链接: https://arxiv.org/abs/2608.30427
作者: Ephrem Wu
机构: Advanced Micro Devices, Inc.(超威半导体公司)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Speculative decoding speeds up generation with an efficient draft model (drafter) that proposes tokens for a target model to verify in one pass, preserving the target’s output distribution. High-acceptance block-diffusion drafters such as DFlash and DFlare fill an entire block in one parallel pass. In many cycles, the target accepts the whole block, so the drafter exhausts its trained block horizon before verification fails. We call this unrealized acceptance stranded speed-up. A mean committed length, per prompt or per cycle, hides it, whereas the acceptance histogram exposes it as a spike in the ceiling bin, the fraction of cycles that accept the entire block. We recommend the histogram as a preflight check before spending training compute. Naively widening the block at inference does not recover the speed-up, because once the block outgrows its training size, the drafter’s bidirectional attention shifts its distribution even at early positions and erodes front-of-block verification. Instead, we post-train the drafter on a longer block with a short curriculum that emphasizes the newly exposed positions, a method we call DBloom. Expanding the pretrained DFlash and DFlare drafters from block size 16 to 24 across Qwen3-8B and Qwen3-4B targets raises the per-prompt committed length on the high-ceiling benchmarks by a median of +0.8 tokens (up to +1.1). Once continuation fine-tuning precedes expansion, the increase reaches 1.37 tokens. The same expansion also lifts committed length on all seven benchmarks for Gemma-4-12B-IT, a different model family, by a median of +0.41 tokens (Arm A), and the full continuation-then-expand pipeline (Arm B) adds +0.29 to +0.98 tokens over the same B16 drafter. In a prompt-matched comparison against JetSpec, a contemporary tree-based drafter not used in our design, DBloom commits more tokens on every benchmark at tree budgets up to 64 nodes.
[NLP-86] Learning to Reason and Use Tools through Unsupervised Fine-Tuning in Task-Oriented Dialog Systems
【速读】： 该论文旨在解决当前对话系统在动态信息检索过程中易产生幻觉、响应准确率低的问题。其核心解决方案是将ReAct框架适配至任务导向型对话场景，使大语言模型（LLM）能够调用外部知识以生成事实性更强的回应。关键创新在于提出一种无监督微调流程：通过上下文学习推理自动收集推理轨迹，并利用基于LLM的判别器筛选高质量样本构建稳健训练集；进一步引入无监督自提升循环机制，使优化后的检查点持续生成更优的推理轨迹，驱动后续迭代微调。实验结果表明，基于ReAct的系统在SIMMC数据集上显著优于基线方法，尤其在推理能力与工具使用方面表现突出；值得注意的是，该研究微调的8B模型性能超越了70B参数量的上下文学习系统。此外，研究还进行了误差分析、场景复杂度影响评估及跨领域泛化能力验证。
链接: https://arxiv.org/abs/2608.30426
作者: Markel Ferro,Oier Lopez de Lacalle
机构: HiTZ Center - Ixa, University of the Basque Country UPV/EHU(巴斯克大学)
类目: Computation and Language (cs.CL)
备注: Accepted to SEPLN 2026
Abstract:Current dialogue systems struggle with dynamic information retrieval, often leading to hallucinations and lower response accuracy. We address this by adapting the ReAct framework for Task-Oriented Dialogue, enabling Large Language Models (LLMs) to access external knowledge and produce factual responses. Mainly, we propose an unsupervised fine-tuning pipeline that harvests reasoning trajectories via in-context learning inference. High-quality samples are filtered using an LLM-based judge to construct a robust training set. This is enhanced by a unsupervised self-improvement loop, where improved checkpoints generate increasingly better trajectories for subsequent fine-tuning iterations. Experiments on the SIMMC dataset demonstrate that ReAct-based systems outperform baselines due to superior reasoning and tool use. Notably, our fine-tuned 8B model surpasses a 70B in-context system. Finally, we present an error analysis, impact of scene complexity, and cross-domain generalization.
[NLP-87] Generative Models Enhanced by Sequence Labelling and Aspect-Code Switching Improve Cross-lingual Aspect-Based Sentiment Analysis EMNLP2026
【速读】： 该论文旨在解决跨语言方面情感分析（Cross-lingual Aspect-Based Sentiment Analysis, CL-ABSA）中因目标语言缺乏标注数据而导致的细粒度情感分析难题，尤其针对包含多个情感要素的复杂任务如目标-方面-情感检测（Target-Aspect-Sentiment Detection, TASD）仍处于研究空白的问题。其核心解决方案是提出一种名为SeqLab的新框架，该框架基于序列到序列（Seq2Seq）模型，并引入编码器执行辅助序列标注任务，以增强方面词识别与情感预测能力；同时创新性地采用方面词代码切换（Aspect-Code Switching, ACS）技术，通过在源语言与翻译后的目标语言句子间交换方面词，生成额外训练数据，从而提升模型对跨语言语义的泛化理解。实验在11种语言、3个领域及2种主干模型上验证了该方法的有效性，在常见的端到端方面情感分析（E2E-ABSA）任务上超越了现有最优结果，并首次系统评估了非英语源语言作为迁移起点的效果，扩展至更具挑战性的跨语言TASD任务，最后通过详尽的错误分析揭示了当前方法面临的关键挑战与局限。
链接: https://arxiv.org/abs/2608.30425
作者: Jakub Šmíd,Pavel Přibáň,Pavel Král
机构: NTIS – New Technologies for the Information Society, University of West Bohemia in Pilsen, Faculty of Applied Sciences (新信息技术研究所，西波希米亚大学皮尔森分校应用科学学院); University of West Bohemia in Pilsen (西波希米亚大学皮尔森分校)
类目: Computation and Language (cs.CL)
备注: Accepted for The 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026)
Abstract:Cross-lingual aspect-based sentiment analysis (ABSA) transfers knowledge from a source language with annotated data to a target language, enabling fine-grained sentiment analysis without annotated target-language data. While monolingual ABSA has seen significant progress, cross-lingual ABSA remains underexplored, especially for complex tasks involving multiple sentiment elements like target-aspect-sentiment detection (TASD). In this paper, we propose a novel SeqLab framework that enhances cross-lingual ABSA using a sequence-to-sequence model with an auxiliary sequence-labelling task performed by the encoder, enhancing aspect term recognition and sentiment predictions. Additionally, we incorporate aspect-code switching (ACS), a translation-based technique that swaps aspect terms between source and translated sentences, generating additional training data to enhance the model’s cross-lingual understanding. We evaluate our approach across eleven languages, three domains, and two backbone models, surpassing previous state-of-the-art results for the commonly studied E2E-ABSA task. Unlike most prior work that relies solely on English as the source language, we systematically assess different source-target language pairs and extend our evaluation to the more challenging, yet underexplored TASD task in cross-lingual settings. Finally, we provide a detailed error analysis highlighting key challenges and limitations.
[NLP-88] SemPOI-RL: Aligning LLM Semantic Reasoning for Interpretable Out-of-Town POI Sequential Generation
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在跨城市兴趣漂移（cross-city interest drift）背景下，如何实现可解释的、结构化的景点序列生成问题，尤其是在异地旅游（Out-of-Town, OOT）场景中，模型需从用户家乡行为中推断可迁移的旅行意图，并在满足路径连贯性与位置感知约束的前提下生成合理的目的地轨迹。现有方法或依赖不可解释的隐式嵌入表示进行迁移，或直接使用LLM生成序列但缺乏对语义推理结果的位置感知建模。其核心解决方案在于提出一种名为SemPOI-RL的框架，关键创新点包括：首先通过自然语言作为可解释的语义中间表示，微调LLM以从用户家乡轨迹中提取面向目的地的旅行风格；其次设计语义景点对齐模块（Semantic POI Alignment Module, SPAM），将推断出的风格条件注入到位置感知的掩码自编码器中，实现风格驱动的轨迹生成；最后引入面向推荐任务的强化学习机制，以优化生成序列的质量。该方法实现了语义推理与结构化序列生成的有效对齐，在两个真实数据集上均显著优于传统推荐系统和直接使用LLM的基线方法，同时提供了旅行过程中各阶段的可解释风格归因。
链接: https://arxiv.org/abs/2608.30399
作者: Yunqi Liu,Yang Zhang,Ruixing Zhang,Liangzhe Han,Yi Qiao,Tongyu Zhu,Leilei Sun
机构: Beihang University (北京航空航天大学); State Key Laboratory of Complex Critical Software Environment
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 19 pages in total, including 9 pages of main text and 4 figures
Abstract:Large language models (LLMs) exhibit strong semantic reasoning and open-ended generation abilities, but aligning these abilities with structured sequential generation remains challenging. This challenge is particularly evident in out-of-town (OOT) POI sequence generation, where a model must infer transferable travel intent from a user’s hometown behaviors, adapt to cross-city interest drift, and generate a coherent destination trajectory under structural constraints. Existing approaches either rely on latent ID-based transfer with limited interpretability or directly use LLMs for sequence generation without explicitly grounding inferred semantics into position-aware predictions. To address this gap, we propose SemPOI-RL, a framework that aligns LLM semantic reasoning with structured sequence generation for interpretable OOT recommendation. Specifically, we first fine-tune an LLM to infer destination-oriented travel styles from users’ hometown trajectories, using natural language as an interpretable semantic intermediate. We then introduce a Semantic POI Alignment Module (SPAM) to ground these inferred styles into a style-conditioned masked autoencoder for position-aware trajectory generation. Finally, we apply reinforcement learning with recommendation-oriented rewards to align LLM-generated styles with downstream sequence quality. Experiments on two real-world datasets show that SemPOI-RL consistently outperforms both traditional recommenders and direct LLM baselines, while providing interpretable style attribution across different phases of a trip. The code is available at this https URL .
[NLP-89] Co-Evolving Actor-Conditioned Critics for Non-Verifiable Generation
【速读】： 该论文旨在解决生成式 AI（Generative AI）在非可验证任务中缺乏确定性验证机制时，如何有效利用自然语言评述（natural-language critiques）进行高质量反馈与迭代优化的问题。传统批判引导式优化（critique-guided refinement）存在关键缺陷：最终修订质量无法反映评述的实际有效性——具备强大能力的生成器可能无视反馈仍能改进，而合理的反馈也可能因执行能力不足而失效。为此，论文提出将评述视为面向特定生成器的修订指导（actor-conditioned revision guidance），其有效性取决于评述是否帮助目标生成器解决预期缺陷。核心解决方案是引入TAIScore（Targeted Actionable Improvement Score），一种联合评估指令、初始响应、评述与修订内容的奖励函数，能够判断评述是否针对真实缺陷、生成器是否采纳建议，以及目标方面是否真正改善。基于TAIScore，论文采用GRPO算法训练适配特定生成器的批判模型，并通过批判引导的修订构建用于直接偏好优化（DPO）的偏好对，形成批判模型与生成器协同演化的闭环系统。实验表明，使用TAIScore训练的8B规模批判模型优于零样本120B模型及仅依赖结果或评述信号训练的模型，且协同演化进一步提升性能，表明有效的批判监督必须随生成器能力动态调整。
链接: https://arxiv.org/abs/2608.30397
作者: Jinyoung Kim,Muhammad Khalifa,Lajanugen Logeswaran,Jaekyeom Kim,Moontae Lee,Honglak Lee,Lu Wang
机构: University of Michigan(密歇根大学); LG AI Research(韩国人工智能研究院); University of Illinois at Chicago(芝加哥大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Natural-language critiques provide supervision beyond scalar rewards for non-verifiable generation, which lacks deterministic verifiers. In critique-guided refinement, a critic gives feedback on an initial response and an actor revises it. However, final revision quality does not reveal whether the critique was actually useful: a capable actor may improve without following the feedback, while valid feedback may fail if the actor cannot execute it. We frame critique as actor-conditioned revision guidance, where usefulness depends on whether the feedback helps the target actor address the intended weakness. We introduce TAIScore (Targeted Actionable Improvement Score), a reward that evaluates the instruction, initial response, critique, and revision together, assessing whether the critique targets a real weakness, whether the actor follows it, and whether the intended aspect improves. We use this reward to train an actor-tailored critic with GRPO, and use critique-guided refinements to construct DPO preference pairs for the actor, forming a co-evolving critic-actor loop where the critic adapts to the actor’s changing capability. Experiments show that an 8B critic trained with TAIScore outperforms both a zero-shot 120B critic and critics trained with outcome-only or critique-only reward signals. Co-evolving the critic and actor further improves performance, suggesting that effective critique supervision should adapt as the actor changes.
[NLP-90] When LLM Meets Tree Search: A Systematic View of Inference as Search in Large Language Models EMNLP’2026
【速读】： 该论文旨在解决大模型在推理过程中因预训练规模趋于饱和而面临的性能提升瓶颈，核心问题在于如何有效利用推理阶段的计算资源以增强模型的逻辑推理能力。传统方法依赖于单轨迹解码（single-trajectory decoding）的思维链（Chain-of-Thought, CoT）框架，难以纠正早期错误并限制了对多路径推理的探索。为此，论文提出将推理视为针对具体实例的优化过程，通过基于树搜索（tree-search-based reasoning）的方法重构推理范式，使模型能够在推理时进行更充分的探索与纠错。其解决方案的关键在于引入系统化的搜索机制，从无信息搜索演进至蒙特卡洛树搜索（Monte Carlo Tree Search, MCTS），借助采样驱动的控制策略实现探索-利用权衡的合理平衡，并构建统一的设计空间（Unified Design Space），涵盖搜索拓扑、评估信号与控制动态三个维度，同时倡导标准化的计算量报告抽象，以显式揭示计算开销与准确率之间的权衡关系，从而推动该领域研究的可比性与可复现性。
链接: https://arxiv.org/abs/2608.30395
作者: Jiaqi Wei,Xiang Zhang,Yuejin Yang,Wenxuan Huang,Juntai Cao,Sheng Xu,Xiang Zhuang,Zhangyang Gao,Muhammad Abdul-Mageed,Laks VS Lakshmanan,Chenyu You,Wanli Ouyang,Siqi Sun
机构: Zhejiang University(浙江大学); Fudan University(复旦大学); Shanghai AI Laboratory(上海人工智能实验室); University of British Columbia(不列颠哥伦比亚大学); Stony Brook University(石溪大学); The Chinese University of Hong Kong(香港中文大学)
类目: Computation and Language (cs.CL)
备注: Accepted by EMNLP’2026
Abstract:As pretraining scaling laws approach saturation, Test-Time Scaling (TTS) has emerged as an important direction for improving reasoning by allocating inference-time compute to a fixed model prior. Viewed at a high level, TTS reframes inference as search over a space of partial reasoning states. While Chain-of-Thought (CoT) exposes intermediate steps, common instantiations rely on single-trajectory decoding, limiting recovery from early errors and exploration. This survey systematizes recent progress in tree-search-based reasoning, viewing inference as instance-specific optimization rather than decoding. We trace the evolution from uninformed search to Monte Carlo Tree Search (MCTS), highlighting how sampling-based control supports principled exploration-exploitation trade-offs. To unify a fragmented literature, we introduce a Unified Design Space spanning search topology, evaluation signals, and control dynamics, and advocate a standardized compute-reporting abstraction to make compute-accuracy trade-offs explicit and comparable.
[NLP-91] Quantitative Evidence Mining for Plausibility-Aware Biomedical AI
【速读】： 该论文旨在解决当前生物医学人工智能（Biomedical AI）系统在从文献、临床试验和监管文件中自动提取科学主张时，仅关注实体与关系（如药物—治疗—疾病）而忽略定量证据要素的根本问题。其核心挑战在于：尽管生成式模型（Generative AI）和自主系统日益用于证据整合与知识图谱（Knowledge Graph, KG）构建，但现有方法常丢失剂量、效应量、人群特征、对照条件、不确定性及适用情境等关键信息，导致提取的主张看似可操作却难以验证、比较或复用。解决方案的关键是推动向“定量证据挖掘”（quantitative evidence mining）范式转变，即系统性地提取数值、单位、测量属性、上下文、不确定性、来源与可信度，并以结构化证据单元形式组织，嵌入具备证据感知能力的知识图谱。该框架强调将提取的主张视为可审计的证据对象，明确标注测量内容、变化幅度、适用场景、置信区间与原始来源，从而实现对证据溯源性、单位一致性、完整性及生物学合理性（biological plausibility）的可验证性。其核心风险并非简单的错误提取，而是看似可靠的主张因缺乏结构化支撑而不可信。
链接: https://arxiv.org/abs/2608.30393
作者: Negin Sadat Babaiha,Stefan Geissler,Marie-Christine Simon,Martin Hofmann-Apitius,Marc Jacobs
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:Biomedical artificial intelligence (AI) systems increasingly extract, organize, and reuse scientific claims from literature, clinical trials, and regulatory documents. But automatic extraction alone does not make a claim reliable evidence: a claim becomes useful only when it can be traced to its source, linked to the quantitative details that support it, and read within its biomedical context and uncertainty. This matters as large language models (LLMs) and increasingly autonomous systems drive evidence synthesis, knowledge graph (KG) construction, and decision support. Many text-mining and LLM pipelines remain relation-centric: they capture entities and relations such as Drug–TREATS–Disease, but drop the dose, effect size, population, comparator, uncertainty, and conditions under which a claim holds. Such relations can look actionable yet remain hard to verify, compare, or reuse. In this perspective, we argue for a shift toward quantitative evidence mining—extracting values, units, measured entities and properties, context, uncertainty, provenance, and plausibility as structured evidence units that populate evidence-aware KGs and can be checked for source grounding, unit consistency, completeness, and biological plausibility. We outline a framework for plausibility-aware AI that treats extracted claims not as final answers but as auditable evidence objects, making clear what was measured, how much it changed, in which setting, with what uncertainty, and from which source. The central risk is not only incorrect extraction, but claims that look like evidence while lacking the structure needed to trust them.
[NLP-92] Using Grounded Theory for Agent Behavior Analysis at Scale EMNLP2026
【速读】： 该论文旨在解决在复杂、长时序且通常不熟悉的任务中，如何有效分析数千条智能体（agent）轨迹并识别其行为模式的问题。传统基于预构建分类器的方法在面对新型或未知行为时表现不足，难以捕捉深层行为规律。为此，论文提出将扎根理论（Grounded Theory, GT）引入智能体轨迹分析，利用其严格的饱和性判定标准和可追溯的数据到理论推导路径，实现对智能体行为的系统性归纳。其核心解决方案是AutoTraceGT——首个自动化多智能体轨迹分析框架，通过迭代执行开放编码、轴心编码与理论编码直至达到理论饱和，生成针对特定任务定制的行为分类体系（codebook）。在六个轨迹数据集上的实验表明，AutoTraceGT所生成的代码本能够恢复人类标注分类体系中73%-91%的失败模式，并揭示原分类体系遗漏的新行为模式；其生成的理论叙事与领域专家认知一致。此外，该代码本作为演绎式特征空间，在下游失败预测任务中显著优于零样本与少样本大语言模型（LLM）基线。结果表明，扎根理论为机器学习研究人员及智能体开发者提供了一种可扩展、可解释且高精度的行为分析工具。
链接: https://arxiv.org/abs/2608.30391
作者: Zhuoran Lu,Yangyang Yu,Zhuoyan Li,Yibo Meng,Nan Jiang,Chengxi Zang,Jie Gao,Ziang Xiao
机构: Purdue University(普渡大学); Stevens Institute of Technology(史蒂文斯理工学院); Cornell University(康奈尔大学); University of Texas at El Paso(德克萨斯大学埃尔帕索分校); Johns Hopkins University(约翰霍普金斯大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 33 pages. Accepted to the Findings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026)
Abstract:Understanding agent behavior requires methods that scale to thousands of trajectories and surface new patterns in long, often unfamiliar tasks where pre-built classifiers fall short. We propose to bring grounded theory into agent trajectory analysis: a six-decade-old qualitative method from the social sciences, with a principled saturation criterion and an auditable trail from data to theory. We propose AutoTraceGT (Automated Trace analysis through Grounded Theory), the first multi-agent pipeline that automates grounded theory on agent trajectories. It iteratively performs open, axial, and theoretical coding until saturation, producing a behavioral taxonomy tailored to each task. Across six trajectory corpora, AutoTraceGT produces codebooks that recover 73-91 percent of the failure modes in human-annotated taxonomies and surface additional patterns that those taxonomies miss. The emergent theoretical narrative aligns with prior expert accounts. Used as a deductive feature space, the codebook outperforms zero-shot and few-shot LLM baselines on downstream failure prediction. These results suggest Grounded Theory offers a scalable analytic tool for ML researchers and agent developers studying what agents actually do.
[NLP-93] Kathleen Remembers: Length-Invariant One-Shot Recall Without Attention
【速读】： 该论文旨在解决循环神经网络（Recurrent Neural Networks, RNNs）及其变体在处理长序列时存在的核心缺陷：即“状态衰减”问题，导致模型无法对早期出现的特定信息实现精确回忆。针对这一问题，作者提出了一种创新解决方案——在现有骨干网络（trunk）基础上引入一个额外的记忆层，称为“笔记本”（notebook），其本质是一个固定键的全息关联存储（Holographic Reduced Representation, HRR）结构，具备可学习的局部写入门控、自门控的原始读取机制以及由写操作触发的遗忘机制。该模块仅需25K参数，即可无缝集成至任意骨干网络的输出逻辑（logits）中。其关键在于：通过线性叠加的存储机制，实现了长度无关的地址寻址，使得未经过训练的内存本身在512、2048和4096字节等不同长度下均能保持90%的准确率；同时，基于算术运算的特性，自然支持选择性遗忘（单次减法即可擦除单一事实而保留其他记忆）与逐标记溯源（反事实擦除可精准识别每个正确字节的来源，实现100%可证明的出处追踪）。在真实文本任务（如WikiText-2）中，笔记本显著提升了对稀有重复词的预测性能（提升0.15–0.27 bit/byte），且该增益随提及间隔增大而增强，并在四倍训练长度下仍保持零样本泛化能力；写触发遗忘机制有效缓解了高长度下的记忆污染问题。在更大规模的WikiText-103阶梯任务中，零样本重复收益随序列长度单调上升，验证了其在大范围上下文中的可扩展性。所有实验均预先注册，参数种子公开，可在单个免费级GPU上复现。
链接: https://arxiv.org/abs/2608.30376
作者: George Fountzoulas
机构: Frederick University, Nicosia, Cyprus(弗雷德里克大学，尼科西亚，塞浦路斯)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 12 pages, 3 figures. Paper 4 of the Kathleen series. Mechanism covered by U.S. Provisional Patent Application No. 64/140,260
Abstract:Recurrent, attention-free sequence models share a structural weakness: a fading state cannot perform exact recall of something seen once, far in the past. We add to the Kathleen trunk a second memory layer – a “notebook”: a fixed-key holographic (HRR) associative store with a learned local write gate, a self-gating raw read, and write-triggered forgetting – 25K parameters that attach to the logits of any trunk. (1) Mechanism: on a controlled needle-in-haystack task the notebook reaches 80-82% one-shot recall at 4x the training length, where the bare trunk scores ~4% and a parameter-matched attention head scores 100% inside its training length and 0% beyond it. Addressing is length-invariant by construction; the untrained memory alone recalls at 90% accuracy identically at 512, 2048 and 4096 bytes. Because the store is a linear superposition, two capabilities follow from arithmetic alone: selective unlearning (one subtraction erases one fact to chance, retained facts unharmed) and per-token attribution (counterfactual erasure names the source fact of every correct byte, 100% provenance). (2) Real text: on WikiText-2 bytes the notebook improves prediction of repeated rare words by +0.15-0.27 bits/byte, the gain growing with the distance between mentions and holding zero-shot at 4x training length; write-triggered forgetting eliminates memory pollution at 8x length (first-mention cost +0.33 - -0.004). (3) Scope and scale: a parameter-matched attention head does generalize on natural-text repetition, so the notebook’s claim is exact recall at O(L); on a WikiText-103 ladder (8 to 512 MB) the zero-shot repeat gain rises monotonically. All experiments are pre-registered, seeds reported, and reproducible on a single free-tier GPU.
[NLP-94] Beyond Consensus: Downward Bias and Role Asymmetry in Multi-Agent LLM Judges for Subjective Evaluation EMNLP2026
【速读】： 该论文旨在解决生成式 AI（Generative AI）在主观评分任务中，基于多智能体辩论（Multi-Agent Debate, MAD）的共识机制虽能提升智能体间一致性，却可能与人类判断存在偏差的问题。其核心挑战在于：尽管多智能体通过角色分工与多轮交互达成一致，但这种“共识”未必反映真实的人类评价标准。论文的关键解决方案是通过设计三种消融实验，系统分离角色提示（role prompting）、多轮交互和显式分数共享对评估结果的影响。研究发现，性能下降主要源于角色设定的不对称性——赋予智能体“严格评判者”角色会引入系统性向下偏倚，且共识过程无法纠正该偏差，反而导致最终评分显著低于严格与宽松条件的算术均值，体现为“严格立场主导”现象。移除角色不对称性（对称化MAD，Symmetric MAD）可基本恢复单裁判基线的性能，而隐藏同行评分则加剧智能体间分歧并恶化与人类判断的一致性。因此，该研究揭示了角色专业化、共识驱动的MAD协议在主观评分中的结构性缺陷：过度追求形式上的智能体一致，可能以牺牲真实人类对齐为代价。
链接: https://arxiv.org/abs/2608.30373
作者: Minsoo Song,Chanwoo Kim,Sugyeong Eo,Chanjun Park
机构: Soongsil University (松林大学); Yonsei University (延世大学)
类目: Computation and Language (cs.CL)
备注: Accepted to Findings of EMNLP 2026
Abstract:Multi-Agent Debate (MAD) has been widely adopted to improve LLM-based evaluation by prompting multiple agents to negotiate and reach a consensus. However, for subjective rubric-based scoring, inter-agent agreement does not guarantee alignment with human judgments. In this paper, we compare a single-judge baseline against a consensus-based MAD protocol on subjective evaluation tasks and design three ablations to isolate the impact of role prompting, multi-round interaction, and explicit score sharing. Evaluations across six LLMs show that the single-judge baseline achieves the strongest human alignment on average across six judge models, whereas MAD shows degradation in human alignment on both tasks. Our ablations demonstrate that this performance drop stems primarily from asymmetric role prompting rather than the interaction itself. Specifically, assigning a strict judge role introduces a systematic downward bias that the consensus process fails to correct. The central finding is that this bias reflects strict-stance dominance beyond averaging: the consensus score falls well beyond the arithmetic midpoint of the standalone strict and lenient conditions, rather than averaging them out. Removing role asymmetry (Symmetric MAD) largely recovers baseline performance, while masking peer scores widens inter-agent disagreement on average and worsens average human alignment. These findings demonstrate that multi-agent consensus can enforce artificial agreement at the expense of true human alignment, revealing a structural limitation in consensus-style, role-specialized MAD protocols for subjective scoring.
[NLP-95] Auditing MCQA Benchmarks through Probability Landscapes EMNLP2026
【速读】： 该论文旨在解决大规模语言模型（Large Language Models, LLMs）在标准多项选择题问答（Multiple-Choice Question Answering, MCQA）基准测试上性能趋近饱和背景下，高质量题库构建与验证过程中面临的难题，尤其是题目质量评估与劣质题项筛选的高人力成本问题。其解决方案的关键在于提出一种基于模型输出分布的双组件概率分析框架：在基准层面，通过顶部预测概率（P_top1）与归一化残差熵（H_norm）刻画模型置信度与选项间竞争格局，并以平均成对距离（Mean Pairwise Distance, MPD）进行全局表征；在题项层面，引入噪声注入机制以削弱有效干扰项之间的竞争性，从而识别出可能存在问题的题目并分类残留错误模式。该方法在四个MCQA基准上的实验表明，其能有效揭示不同基准间的宏观结构差异，并精准定位需人工复审的题项，且与MMLU-Redux专家标注结果高度一致，证明了该概率驱动框架可作为轻量级、可扩展的审计工具，用于比较基准整体特性并优先处理具体题项。
链接: https://arxiv.org/abs/2608.30372
作者: Minsoo Song,Chanjun Park
机构: Soongsil University(松林大学)
类目: Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026
Abstract:As Large Language Models rapidly advance, performance on standard multiple-choice question answering (MCQA) benchmarks is reaching saturation. While the community has responded by developing increasingly difficult datasets, validating question quality and filtering flawed items remains a labor-intensive process. To provide a scalable diagnostic approach, we propose a two-component probabilistic framework for auditing MCQA benchmarks using model output distributions. First, for benchmark-level analysis, we characterize the probability landscape using the top prediction probability ( P_top1 ) and normalized residual entropy ( H_norm ), summarized globally by Mean Pairwise Distance (MPD). Second, for item-level diagnostics, we introduce noise injection to reduce meaningful distractor competition, enabling us to flag candidate items for targeted human review and categorize residual failure patterns. Across four MCQA benchmarks, our landscape analysis reveals benchmark-level differences in model confidence and residual option competition. Concurrently, our noise-injection method flags potentially actionable item-level issues, showing alignment with expert error annotations from MMLU-Redux. These results suggest that our probability-based framework provides a lightweight audit lens for comparing macro-level benchmark structure and prioritizing individual items for targeted human review.
[NLP-96] Will the User Ever Know? Covert Indirect Prompt Injection on Tool-Using LLM Agents EMNLP2026
【速读】： 该论文旨在解决生成式 AI（Generative AI）代理在通过工具执行真实世界操作时面临的间接提示注入（Indirect Prompt Injection, IPI）威胁。现有评估指标攻击成功率（Attack Success Rate, ASR）仅关注攻击是否成功，却忽略了用户在代理最终响应中能否察觉异常，导致对安全风险的评估存在偏差。研究发现，成功的注入攻击可表现为两种结果：隐蔽性成功（Covert Success, CSR），即代理执行了恶意指令但返回看似正常的响应，用户无法察觉；或显性成功（Overt Success, OSR），即代理在最终响应中报告了注入行为，使用户有机会识别。从用户视角出发，作者将ASR分解为CSR与OSR，并分析表明二者差异源于攻击后代理的行为模式——隐蔽性轨迹在执行注入后迅速将控制权交还给用户任务，而显性轨迹则以攻击本身作为终结状态。这一现象由ReAct框架的响应机制决定，其最终响应仅总结最近一次动作。基于此洞察，论文提出ICoA（Induced Covert Attack）攻击方法，通过在执行注入后主动引导代理回归用户任务流程，从而诱导隐蔽性成功。在AgentDojo平台上的多模型实验证明，ICoA显著提升了隐蔽成功率（CSR），相较最强基线提升3.79至12.01个百分点，验证了其有效性。
链接: https://arxiv.org/abs/2608.30362
作者: Yunseok Lee,Yunji Kim,Woojin Lee
机构: Dongguk University-Seoul(东国大学-首尔)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: EMNLP 2026 Main Conference
Abstract:As LLM agents take real-world actions through tools, indirect prompt injection (IPI) has emerged as a serious threat. The standard metric, Attack Success Rate (ASR), counts whether an injection succeeds but ignores what the user notices in the agent’s final response. Looking at successful injection traces, we find two distinct outcomes: the agent executes the injection while returning an otherwise normal response, or reports the injected action in its final response, giving the user a chance to notice. We call these covert and overt successes. From the user’s perspective, we decompose ASR into the Covert Success Rate (CSR), counting successes leaving no trace in the final response, and the Overt Success Rate (OSR), counting successes the user can detect. To understand what drives the gap, we analyze successful trajectories and find that the agent’s behavior after the injection separates covert from overt: covert traces hand control back to the user task before ending, while overt traces end at the attack itself. This split follows from the ReAct format, where the final response summarizes the most recent action. Building on this observation, we propose ICoA (Induced Covert Attack), an IPI attack designed to induce covert outcomes by steering the agent back to the user task after executing the injection. Across four target models on AgentDojo, ICoA achieves the highest CSR, with gains of 3.79-12.01 percentage points over the strongest baseline.
[NLP-97] Do Small Models Use the Law You Give Them? Measuring Context Use on a Bilingual Bangladesh Legal Benchmark
【速读】： 该论文旨在解决生成式 AI 在双语孟加拉国法律问答（Legal Question Answering, LQA）任务中，尽管经过微调（fine-tuning），模型仍未能有效利用上下文提供的法律条文这一关键问题。其核心挑战在于：现有评估方法难以区分模型性能提升是由答案评分机制、法律条文检索能力，还是模型对相关法律内容的实际使用所导致。为此，研究提出了一套系统性解决方案，关键在于构建一个层级保持的成文法语料库、2,165个经人工审核的双语微调样本及150项受控的法律条文供应测试集，并采用约束选项字母评分、循环选项轮换与受控移除管辖条款等方法，以分离评分器、检索器与模型本身的影响。实验结果表明，仅通过改进评分策略可带来微弱准确率提升（如3.0%），而通过特定微调适配器（如Qwen3.5-2B seed-42）实现的准确率提升高达50.0%，且当管辖条款被强制保留时，多数模型性能显著提高（14.7%-19.3%），但差分分析显示微调并未增强模型对管辖条款的依赖程度。因此，研究强调：在法律领域应用生成式 AI 时，必须将评分、检索与模型行为三者解耦，才能真实评估其法律知识的整合能力。
链接: https://arxiv.org/abs/2608.30327
作者: Moniruzzaman Mahadi,Abrar Mohammed Tanzim Alam,Sayma Siddika Monalisa,Mir Mohammad Asif Abdullah,Swakkhar Shatabda,Md Adnan Arefeen
机构: North South University (南南大学); BRAC University (BRAC大学); HuggingFace/Datasets/Bangladesh-Legal-QA (HuggingFace/数据集/孟加拉国法律QA)
类目: Computation and Language (cs.CL)
备注: Legal Data Benchmark for Bangladesh
Abstract:Fine-tuning can improve legal question-answering accuracy without improving how models use law supplied in context. We study this distinction in bilingual Bangladeshi legal QA, where observed errors can arise from answer scoring, retrieval, or failure to use relevant law. We construct a hierarchy-preserving statutory corpus, 2,165 reviewed bilingual fine-tuning examples, and a 150-item supplied-law control. We evaluate six instruction-tuned models: Llama-3.2-1B, Llama-3.2-3B, Qwen3.5-0.8B, Qwen3.5-2B, Qwen3.5-4B, and Gemma-4-E2B, with three LoRA seeds per model. To separate effects, we combine constrained option-letter scoring, cyclic option rotation, and controlled removal of the governing provision. On 398 Bar Council outputs, an exact-line parser attributes an accuracy gain of 50.0% to the Qwen3.5-2B seed-42 adapter, whereas option scoring yields only 3.0% . For Gemma-4-E2B, the two scoring methods favor different systems. When the governing provision is guaranteed to be present, five of six reference models improve by 14.7%-19.3% under the four-order criterion. Removing that provision reduces accuracy by 8.0%-15.3% for models and by 13.8%-14.9% points for their adapters. However, difference-in differences estimates show no increase in reliance on the governing provision after fine-tuning. Results show that legal adaptation claims require separating scorer, retriever, and model effects. Our Code and data are available at this https URL
[NLP-98] Sequential Trajectories and Simultaneous Blending: Multi-Emotion Modeling for Instruction-Following TTS
【速读】： 该论文旨在解决情感语音合成（Emotional TTS）中多情绪控制能力不足的问题，特别是针对两种互补的多情绪任务：情绪轨迹（emotion trajectory）——即语音中按序经历多个有序情感阶段，以及情绪融合（emotion blending）——即多种情绪在整段语音中同时共存。现有方法在监督微调（SFT）阶段缺乏对情绪特征的显式评估，且单一情绪奖励机制无法为轨迹任务提供结构感知反馈或为融合任务提供成对感知反馈，导致监督不匹配。为此，论文提出 HybridEmo 框架，其核心在于通过一种样本感知的混合奖励机制，在 SFT 初始化后，利用分组相对策略优化（Group Relative Policy Optimization）对语音标记策略进行对齐。对于情绪轨迹样本，采用段落级一致性机制，结合平均与最弱阶段证据以确保预设情感阶段的正确性与完整性；对于情绪融合样本，则基于高斯混合模型（GMM）构建奖励，融合离线情绪空间中目标情绪锚点的联合帧级支持与整体较弱目标的边际惩罚，实现更精细的情绪强度控制。两分支共享自动语音识别（ASR）奖励，并在统一策略网络中路由。实验表明，HybridEmo 在 MultiEmo-Test 上显著提升情绪轨迹的准确性与情绪融合的强度，同时保持说话人相似性无明显下降；人工评估显示其优于 CosyVoice 3 与 EmoVoice-0.5B，与 Qwen3-TTS 相比呈现近乎平衡的偏好。
链接: https://arxiv.org/abs/2608.30325
作者: Yan Zhou,Yun Hong,Yang Feng
机构: 未知
类目: Computation and Language (cs.CL); Sound (cs.SD)
备注: Code is available at this https URL . Demo page: this https URL
Abstract:Natural-language instructions enable flexible control of synthesized speech, yet emotional TTS systems primarily model a single utterance-level affect, leaving multi-emotion control underexplored. We study two complementary multi-emotion TTS tasks: emotion trajectory, which spans several ordered affective stages, and emotion blending, in which multiple emotions coexist throughout an utterance. These tasks expose a supervision mismatch: supervised fine-tuning (SFT) does not explicitly evaluate emotion features, while single-emotion rewards provide neither structure-aware feedback for trajectory completion nor pair-aware feedback for blending. We introduce HybridEmo, a post-training framework that initializes both tasks with SFT and then aligns the speech-token policy through Group Relative Policy Optimization using a sample-aware hybrid reward. For trajectory samples, segment-aligned consistency combines average and weakest-stage evidence to preserve the correctness and completeness of prescribed stages. For blending samples, a GMM-based reward combines frame-level support from the union of target-emotion anchors in an offline emotion space with an utterance-level weaker-target margin. Both branches share an ASR reward and are routed within a unified policy. On MultiEmo-Test, HybridEmo significantly improves trajectory correctness and blending intensity, without a noticeable degradation in speaker similarity. Human evaluation prefers HybridEmo to CosyVoice 3 and EmoVoice-0.5B, with nearly balanced preferences against Qwen3-TTS.
[NLP-99] Ignorance or Incompetence? Constructing Knowledge-Gated Verifiable Tasks for LLM Agents
【速读】： 该论文旨在解决专业代理任务中因缺乏对私有惯例（private conventions）的显式控制而导致评估不可靠的问题。现有基准测试通常未明确界定代理是否具备访问这些隐含知识的权限，从而导致性能评估存在偏差。为此，论文提出一种知识门控的任务构建协议（knowledge-gated task-construction protocol），其核心在于将任务指令与一个包含私有惯例、参考表及实用操作符的紧凑实体（artefact）分离，实现对知识依赖性的显式建模与可测性验证。通过构建时的溯源记录、跨条件字节完全一致的任务指令、泄露审计机制以及可执行的验证见证（executable witnesses），该协议能够精确检测代理对特定知识实体的依赖程度。在十五个校准任务上的实验表明，某一前沿代理配置在拥有知识实体时达到68.0%的通过率，而无该实体时则为0%；即使使用一个看似合理但错误的知识实体，通过率仍维持在0%（五次试验均如此）。对于结构化任务，采用确定性求解器与规则语料库提供精确的真值标准；对于非可执行验证的任务，则引入基于命名标准的评分细则。最终，通过五次试验的经验性知识门控筛选，保留了七个满足条件的任务。该系列实验验证了任务构建协议的有效性，但并未证明所保留任务能提升模型在训练后的表现。研究团队已公开部分任务集及配套工具链。
链接: https://arxiv.org/abs/2608.30322
作者: Hanlin Tian,Minhao Li,Yu Mi,Sihan Zhu,Zhao Yang,Yuxiang Wang,Hongquan Zhu,Qiufei Hu
机构: DataGrids; Shanghai University (上海大学); Peking University (北京大学); Northwestern Polytechnical University (西北工业大学); Nanyang Technological University (南洋理工大学); East China Normal University (华东师范大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Professional agent tasks often depend on conventions that are absent from public corpora, yet benchmarks rarely control whether an agent has access to those conventions. We introduce a knowledge-gated task-construction protocol that separates a task instruction from a compact artefact containing private conventions, reference tables, and utility operators. Construction-time provenance, byte-identical task instructions across the provided- and withheld-artefact conditions, leak audits, and executable witnesses make dependence on the artefact explicit and testable. Across fifteen calibration tasks, one frontier agent configuration achieves a 68.0% pass rate with the artefact and 0% without it; on one task, a plausible but incorrect artefact also yields 0% across five trials. Deterministic solvers and rule corpora provide exact ground truth for structured tasks, while named criterion-level rubrics support outputs that cannot be checked by a single executable oracle. A configuration-relative calibration screen retains seven tasks satisfying our five-trial empirical knowledge-gating screen. These experiments validate the behavior of the construction protocol; they do not establish that the retained tasks improve post-training. We publicly release part of the task suite and supporting tooling at this https URL.
[NLP-100] On the Design of Qwen 3.8-Next Architecture: Evaluation Efficiency and Training Stability
【速读】： 该论文旨在解决大模型在训练效率、推理性能与模型稳定性之间难以平衡的核心问题，尤其针对参数量庞大的稀疏专家混合模型（Mixture-of-Experts, MoE）在高计算成本和资源消耗背景下的优化挑战。其关键解决方案在于提出一种多维度协同优化的架构设计：采用分层混合注意力机制（层间结合门控增量网络GDN与全局注意力，每四层设一层全注意力），并在持续预训练阶段以轻量级压缩索引器实现细粒度微块上下文评分的稀疏注意力（Qwen Sparse Attention, QSA）；通过将残差流扩展为四分支并引入逐元素门控结构（Gated Residual, GR）增强表达能力；同时，在主干网络外增加一个从主机内存预取的n-gram嵌入层以扩展容量而不增加加速器负担。该方案通过系统性评估各项改动在损失函数、下游任务表现、训练/预填充/解码开销以及最优超参数配置与训练稳定性上的影响，实现了损失、精度、效率与鲁棒性的统一优化，最终在仅使用1/3激活参数、1/3训练样本和约1/9训练浮点运算量的情况下，超越了前代397B-A17B模型在多数基准上的表现，验证了“联合优化”策略的有效性。
链接: https://arxiv.org/abs/2608.30320
作者: Zihan Qiu,Zekun Wang,Xiao Li,Yanpeng Li,Yang Xu,Yixuan Wang,Huaqing Zhang,Rui Men,Bochao Mao,Chengruidong Zhang,Fan Zhou,Hao Luo,Haofeng Huang,Haoran Lian,Haoyan Huang,Hongqing Chen,Jianwei Zhang,Jing Xu,Junjie Wang,Langshi Chen,Liangyu Wang,Linlang Jiang,Man Yuan,Minmin Sun,Peng Jin,Siqi Zhang,Siyu Wang,Xingzhang Ren,Yakai Wang,Yi Zhang,Yiming Dong,Yizhong Cao,Yubo Ma,Yunfei Mao,Bo Zheng,Dayiheng Liu
机构: Alibaba Cloud (阿里巴巴云)
类目: Computation and Language (cs.CL)
备注:
Abstract:We describe the architecture and ablations of Qwen3.8-Flash-Next, a sparse mixture-of-experts model with 125B parameters, 6B activated per token, and additional 51B parameters of n-gram embedding tables held off the accelerator. On fourteen pre-training benchmarks the model leads the 397B-A17B predecessor on eight and trails it on the rest by at most 2.6 points, at 1/3 the activated parameters, 1/3 the training tokens, and roughly 1/9 the training FLOPs. Token mixing uses a layer-wise hybrid of Gated DeltaNet (GDN) and global attention, with one full-attention layer in every four; at continued-pretraining time those full-attention layers are replaced by Qwen Sparse Attention (QSA), which scores context at micro-block granularity with a compressed lightweight indexer. The residual stream is widened to four branches and read through an elementwise gate, a design we call the Gated Residual (GR). Capacity is added outside the backbone by a single n-gram embedding layer whose tables are prefetched from host memory. We evaluate every candidate change along three axes: loss together with downstream benchmarks; the cost of the change in training, prefill and decode; and its effect on the optimal hyperparameters and training stability. Loss and downstream accuracy do not always move together: enlarging the n-gram vocabulary lowers loss monotonically while downstream accuracy saturates. The architecture and the Muon optimizer together shift the optimal learning rate and batch size upwards, render batch-size warmup unnecessary, and substantially improve stability under stress tests. Loss, benchmarks, efficiency and stability form one design problem. Solved jointly, they yield a recipe that is simultaneously more efficient, more capable and more stable.
[NLP-101] Beyond Token-Level Guidance: Inference-Time Alignment of Specialized LLM s via Cross-Family Representation Steering EMNLP2026
【速读】： 该论文旨在解决专用领域微调的大语言模型（LLM）在推理阶段因专业化微调导致的安全性下降问题，同时避免对计算资源的高消耗。现有推理时对齐方法虽具备轻量、即插即用的优势，但无法在不损害领域专长能力的前提下可靠提升安全性。其根本原因在于“互补专长正交性”：专用基础模型与通用领域指导模型在能力上呈正交关系，导致指导信号在生成过程中不可靠，主要表现为停止标记干扰（stop token interference），即指导模型倾向于持续生成，覆盖基础模型本应终止的正确输出。针对此问题，论文提出CREST方法，通过从任意家族的指导模型中提取安全方向（safety directions），直接操控基础模型的隐藏表示以实现对齐，完全规避了基于标记层级的结构限制。该方案在显著提升安全性的同时，有效保留了模型的领域专长能力，并对已良好对齐的模型保持其原有安全性能，在安全基准测试中相较基线最高提升22.2%。
链接: https://arxiv.org/abs/2608.30319
作者: Jin Gan,Xin Li,Jun Luo
机构: Nanyang Technological University (南洋理工大学); Nankai University (南开大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Accepted by EMNLP 2026
Abstract:Large language models (LLMs) finetuned for specialized domains represent crucial high-impact applications. Inference-time alignment improves safety degraded from specialization finetuning without requiring substantial computational resources, complementing finetuning-based methods with an easy-to-use, plug-and-play solution. However, existing inference-time methods fail to reliably improve safety without disrupting domain capability. We identify the root cause as complementary expertise orthogonality: specialized base models and general-domain guidance models have orthogonal competencies, making the guidance signal unreliable for specialized generation. This primarily manifests as stop token interference, where the guidance model’s tendency toward continuation overrides the base model’s decision to stop, burying correct answers under guidance-induced continuation. To address this problem, we propose CREST, an inference-time alignment method that steers base model hidden representations using safety directions extracted from a guidance model of any family, avoiding token-level structural limitations entirely. CREST improves safety where specialization has weakened it while preserving both domain-specific capability and the safety of already well-aligned models, outperforming baselines by up to 22.2% on safety benchmarks. Our code is available at: this https URL.
[NLP-102] Context Staircase: Signature-Aligned Dynamics of Token Embeddings under Small Initialization
【速读】： 该论文旨在解决语言模型中词元嵌入（token embeddings）如何从随机初始化通过梯度训练逐步形成有意义的结构这一动态机制不明确的问题。其核心解决方案在于揭示嵌入结构演化与词元条件化的标签分布和上下文分布之间的紧密关联，并将其形式化为“概率签名”（probability signatures）。关键发现是存在一种称为“上下文阶梯”（Context Staircase）的渐进学习过程：在训练初期，嵌入优先对齐于最简单的、无上下文依赖的低阶统计签名（如词元与其标签的直接关联），随后随着训练推进，逐渐捕捉包含更多上下文词元的高阶统计特征。研究通过小初始化下的梯度流分析，推导出前馈网络与自注意力架构中嵌入演化的微分方程，进一步将该现象扩展至真实语言模型训练中。结果表明，这些由数据统计特性驱动的嵌入结构不仅对任务学习至关重要，还促进了语义结构在嵌入空间中的内化。总体而言，该工作提供了数据统计特性与模型架构共同塑造词元嵌入的动态解释，并揭示了训练过程中存在一种隐含的偏置：学习顺序由简单、低阶的统计关系逐步过渡到复杂、依赖上下文的关系。
链接: https://arxiv.org/abs/2608.30315
作者: Junjie Yao,Liangkai Hang,Zhi-Qin John Xu
机构: Shanghai Jiao Tong University(上海交通大学); Institute of Natural Sciences(自然科学研究学院)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注:
Abstract:Token embeddings are the basic representational units that connect discrete tokens with continuous computation in language models. Although modern language models learn embeddings from random initialization through gradient-based training, the dynamical mechanism by which meaningful embedding structures emerge remains unclear. In this work, we identify that the evolving embedding structures are closely related to token-conditioned label and contextual distributions, which we formalize as probability signatures. We observe a progressive learning process, which we term Context Staircase: embeddings learn the low-order statistic signatures of the data before the high-order ones. More specifically, we observe that early in training they align with the simplest, context-free signature linking a token to its label, and as training proceeds, they progressively reflect signatures involving more and more context tokens. We then analyze the gradient flow of embeddings under small initialization to explain this phenomenon, deriving embedding evolution equations for feed-forward and self-attention architectures. We further extend these observations to real language-model training. Finally, we show that these embedding structures play an important role in both task learning and the incorporation of semantic structure into the embedding space. Overall, our results provide a dynamic explanation of how data statistics and architecture jointly shape token embeddings in language models, and reveal an implicit bias in the space of data statistics: training proceeds from simpler, low-order statistical relations toward increasingly complex, context-dependent ones.
[NLP-103] Lazy Grounding: Attacking Search Agents with Factual Evidence EMNLP2026
【速读】： 该论文旨在解决生成式搜索代理在利用检索到的网络证据进行回答时，因“懒惰接地”（lazy grounding）现象导致的错误推理问题。具体而言，尽管检索机制可降低幻觉（hallucination），但当模型错误地将与当前问题邻近、仅对改写后的问题成立的事实性证据误用于原问题时，仍会引发误导性输出。其解决方案的关键在于识别并防御此类“误用真实证据”的攻击模式——即即使证据本身为真，若其上下文与当前问题不匹配，也可能导致答案偏离。研究通过使用基准问题的改写版本生成邻近证据，发现这些看似合理的真实文档在被错误关联至原始问题时，显著降低了模型准确率（平均下降5.9分，最高达17.3分），且在所有测试设置中均诱导了邻近答案的采纳。该现象在邻近证据出现较晚或更符合答案格式时尤为严重，表明模型对证据位置与结构敏感，而非仅依赖内容相关性。因此，论文强调，构建鲁棒的搜索代理不仅需防范虚假信息，还需抵御真实但不相关的证据所带来的误导。
链接: https://arxiv.org/abs/2608.30303
作者: Yulin Zhang,Yukun Huang,Sanxing Chen,Tianyi Lin,Ziang Yang,Xunjian Yin,Bhuwan Dhingra
机构: Duke University (杜克大学)
类目: Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026 (Main Conference). Code: this https URL
Abstract:Search agents reduce hallucination by grounding answers in retrieved web evidence. Yet reliance on retrieval also creates an attack surface: poisoned corpora with false or malicious documents can cause agents to reproduce misinformation. We show that falsehood is not necessary – a search agent can be misled by factual evidence for a nearby question, adopting that nearby answer even when it does not answer the current question. We call this failure lazy grounding. We expose lazy grounding using nearby evidence from answer-changing rewrites of benchmark questions. Each document truthfully supports a neighboring rewritten question, but is surfaced for the original question. Across 12 model-benchmark pairs, nearby evidence reduces accuracy by 5.9 points on average and by up to 17.3 points, while inducing nearby-answer adoption in every setting. The effect is stronger when nearby evidence appears later or is more answer-shaped. Our results show that robust search agents must defend against not only misinformation but also the misapplication of factual evidence. The code is publicly available at this https URL.
[NLP-104] AIA2: Attribute-Agnostic Imbalance Augmentation for Subgroup Robustness
【速读】： 该论文旨在解决数据中属性（如主题、人口统计特征等）引起的子群不平衡问题，此类不平衡不仅限于标签分布不均，还可能在潜在语义空间中形成复杂的子群结构，导致模型在低频子群上的性能显著下降。现有方法多聚焦于标签层面的不平衡，忽视了数据属性所蕴含的深层子群结构对模型泛化能力的负面影响。为此，本文提出一种无需显式子群标注的属性无关不平衡增强框架（Attribute-Agnostic Imbalance Augmentation, AIA²），其核心在于通过隐式语义分布自动识别不同子群的不平衡模式，定位兼具学习难度高和子群偏差大的样本片段，并利用大语言模型（LLM）生成具有子群感知能力的增强数据，从而提升模型在最差子群上的鲁棒性。实验在5个涵盖社会议题与多样化主题的主流语料库上验证了AIA²的有效性，结果表明其在低表现子群上实现了显著性能提升，并优于多种基线方法；消融实验证明各模块间具有互补性，进一步说明AIA²在实际应用中具备一致且可推广的改善最差组性能的能力。
链接: https://arxiv.org/abs/2608.30297
作者: Hanshu Rao,Guangzeng Han,Xiaolei Huang
机构: University of Memphis(孟菲斯大学); University of Missouri(密苏里大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Attributes describing data content and context can induce diverse imbalance patterns that go beyond label imbalance alone. However, existing studies primarily address label imbalance while overlooking data attributes, such as topics and demographics, which can induce meaningful subgroup structure while causing model degradation on underrepresented subgroups. We propose Attribute-Agnostic Imbalance Augmentation (AIA ^2 ), a framework for improving model robustness under varying subgroup imbalances without explicit subgroup annotations. AIA ^2 automatically discovers varying imbalances via latent semantic distributions, obtains slices with both learning difficulty and subgroup imbalance deficits, and deploys a large language model (LLM) for subgroup-aware imbalance augmentation. We have evaluated AIA ^2 on 5 popular corpora with rich domains and their attribute values, covering social issues and diverse topics. Results show improved performance on the lowest-performing subgroups and consistent gains over competitive baselines. Ablation studies confirm complementary contributions from each component, and additional analyses show that AIA ^2 provides a practical and consistent way to improve worst-group robustness under data subgroup imbalance. Code is available at this https URL.
[NLP-105] Read the Room Read the Image: Understanding Indirect Speech Acts in Multimodal Visual Contexts ACL2026
【速读】： 该论文旨在解决多模态场景下对间接言语行为（Indirect Speech Acts, ISAs）的理解难题，尤其关注在高语境语言（如韩语）中依赖上下文进行语用推理的挑战。现有文本或视觉基准大多仅关注显式编码的上下文信息或感知识别任务，忽视了对语用层面深层语境依赖的理解需求。为此，论文提出了READI——一个基于视觉与对话融合的多模态基准，通过结合视觉上下文与对话语境，构建了基于语用理论的渐进式间接性模型，并将任务形式化为基于视觉的语用问答（Vision-based Pragmatic Question Answering, V-PQA），支持英、韩双语跨语言评估。其解决方案的关键在于：建立一个能够量化间接程度并要求模型进行整合推理的评估框架，从而系统性地检验模型在复杂语境下的语用理解能力。实验表明，即使最先进的多模态模型在面对视觉引导的间接言语行为时也表现不佳，且随着间接性增强性能显著下降，凸显了当前模型在语境化语用推理方面的不足，验证了该基准在推动高阶语用理解研究中的必要性。
链接: https://arxiv.org/abs/2608.30270
作者: Jaehee Kim,Ji Hoon Chung,Seoyoon Park,Unsol Kim,Kyungwon Park,Ji Hak Kim,Yi-Jun Chen,Hansaem Kim
机构: Yonsei University (延世大学); Department of Korean Language and Literature (韩国语言文学系); Interdisciplinary Graduate Program of Linguistics and Informatics (语言学与信息学跨学科研究生项目); Department of Artificial Intelligence (人工智能系)
类目: Computation and Language (cs.CL)
备注: Accepted to Findings of ACL 2026
Abstract:Indirect speech acts (ISAs) require pragmatic reasoning over context, as directive intent can- not be inferred from surface form alone. Prior text-based studies and existing multimodal benchmarks largely overlook this requirement, focusing instead on explicitly encoded context or perceptual recognition, and thus underex- plore context-dependent pragmatic understand- ing, particularly in high-context languages such as Korean. We introduce READI, a multimodal benchmark for evaluating ISA understanding through integrated reasoning over visual con- text and dialogue. READI models graded in- directness grounded in pragmatic theory and formulates the task as vision-based pragmatic question answering (V-PQA), supporting cross- lingual evaluation in English and Korean. Ex- periments show that even state-of-the-art multi- modal models struggle with visually grounded indirect speech acts, with performance declin- ing as indirectness increases, underscoring the need for benchmarks that explicitly target con- textual pragmatic reasoning.
[NLP-106] Using Prosody to Predict Syntactic Structure EMNLP2026
【速读】： 该论文旨在解决语音韵律（prosody）与句法结构之间对应关系的性质和程度这一长期存在争议的问题。尽管已有研究指出韵律承载着句法结构的关键线索，但二者之间的具体映射机制仍不清晰。为此，论文提出一种基于信息论的分析框架，通过计算韵律特征与句法表示之间的互信息（mutual information），量化二者间的交互强度。其核心解决方案在于构建一个通用、结构无关且模块化的框架，利用多模态语言模型在大规模语音-文本语料上估计该互信息量，可灵活评估单个韵律特征或句法成分的贡献。实验在英语读本和自发对话两类语域中考察了词持续时间与词间停顿两个韵律特征，结果表明韵律确实包含可观测的句法信息，在自发对话中可使句法不确定性降低高达10.2%。该发现为多种关于韵律-句法接口的理论提供了新的实证支持。
链接: https://arxiv.org/abs/2608.30260
作者: Junghyun Min,Alex Warstadt,Tamar I. Regev,Tiago Pimentel,Ethan Gotlieb Wilcox
机构: Georgetown University(乔治城大学); UC San Diego(加州大学圣迭戈分校); MIT(麻省理工学院); ETH Zürich(苏黎世联邦理工学院)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 15 pages, 4 figures. EMNLP 2026 camera-ready
Abstract:While it is well-established that prosody carries crucial cues for syntactic structure, the degree and nature of correspondence between these two domains remains contested. We investigate the syntax-prosody interface through an information-theoretic lens, quantifying the interaction between prosodic features and syntactic representations as their mutual information. We provide a general-purpose framework for estimating this quantity over large speech-text corpora using multimodal language models. Our framework is structure-agnostic and modular, insofar as it can be used to measure the contributions of individual prosodic features or components of structure. We evaluate the syntax-prosody relationship for two features (word duration and inter-word pauses) across two domains–read audiobooks and spontaneous conversations–both in English. Our results demonstrate that prosody contains measurable syntactic information, with prosodic features reducing syntactic uncertainty in spontaneous conversations by up to 10.2%. Our findings offer new empirical support for several theoretical accounts of the syntax-prosody interface.
[NLP-107] Stratified Consistency Distillation for Natural Language Formalization
【速读】： 该论文旨在解决自然语言到逻辑公式（logical formula）翻译中的准确性问题，尤其针对当前依赖提示工程（prompt engineering）的方法在跨领域和多输入格式场景下难以规模化扩展的瓶颈。其核心解决方案是提出一种基于微调的分层一致性蒸馏（Stratified Consistency Distillation）方法：首先利用前沿大语言模型（LLM）对每个输入生成K个逻辑译文，并根据语义等价性进行聚类；其次，依据聚类熵值水平采取差异化策略——低熵时采用多数投票，中熵时使用大模型作为裁判（LLM-as-a-Judge），高熵时则通过统一或弃权处理；最后，以筛选出的伪标签对小型模型进行微调。实验结果表明，该方法在Pass@K及新提出的等价逻辑相似度（Equivalent Logical Similarity）指标上均实现显著且一致的性能提升，验证了通过一致性蒸馏推动逻辑翻译能力进化的可行性。
链接: https://arxiv.org/abs/2608.30258
作者: Zhichao Hou,Ferhat Erata,Joe Lilien,MohamadAli Torkamani
机构: North Carolina State University (北卡罗来纳州立大学); Amazon Web Services (亚马逊网络服务)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Neurosymbolic reasoning has shown promising success in addressing complex reasoning tasks by combining large language models (LLMs) and symbolic solvers. While this approach shows promise, a fundamental challenge remains: improving the accuracy of translations from natural language to logical formulas. Current methods predominantly rely on prompt engineering, which is difficult to scale across different domains and input formats. Drawing inspiration from the success of fine-tuning in other model adaptation and alignment applications, we propose a fine-tuning-based Stratified Consistency Distillation approach: (1) We generate K logical translations per input using a frontier LLM and cluster them by semantic equivalence (2) Based on the entropy level, we apply majority voting (low entropy), LLM-as-a-Judge (medium entropy), or unification/abstention (high entropy), and (3) fine-tune a smaller model using the selected pseudo-labels. Our experiments show significant and consistent improvements in both Pass@K and our novel Equivalent Logical Similarity metrics, demonstrating the potential of advancing logical translation through consistency distillation.
[NLP-108] Beyond Surface Forms: Symbolic Edits as a Test for Logical Reasoning with LLM s EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在逻辑推理中对问题表述的表面变化过于敏感，导致其推理过程缺乏对底层逻辑结构的忠实遵循的问题。这一现象表明，现有模型可能依赖于语义表面特征而非严格的逻辑推导机制，从而削弱了其推理的可靠性与可解释性。该研究的关键解决方案在于提出一种基于工具的框架，能够对一阶逻辑（First-Order Logic）和约束满足问题（Constraint Satisfaction Problems, CSP）等任务的符号化表示进行受控且标签保持不变的编辑操作。通过在符号层面精准修改逻辑算子及其他结构性组件，并将其重新映射为自然语言，该框架实现了对逻辑推理问题的系统性扰动，从而构建了一种自动化压力测试机制。实验结果表明，无论模型规模或架构如何，大多数模型在面对结构化改变时表现出不一致的推理行为：虽有部分情况能正确响应逻辑变化，但多数情况下无法准确追踪其逻辑后果，揭示出当前大语言模型在深层逻辑推理能力上的显著局限。
链接: https://arxiv.org/abs/2608.30256
作者: Ramya Keerthy Thatikonda,Wray Buntine,Ehsan Shareghi
机构: Monash University (莫纳什大学); University College London (伦敦大学学院)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted to Findings of EMNLP 2026
Abstract:Logical reasoning with large language models (LLMs) is a critical capability, as it reflects a system’s ability to correctly deduce hypotheses from a given context using faithful deductive processes. However, LLM reasoning has often been shown to be sensitive to small surface-level variations in problem formulation, raising questions about whether models truly follow the underlying logical structure. Studying this behavior is challenging because the symbolic components of logical problems, such as operators and predicates, are difficult to systematically manipulate in natural language. We introduce a tool-driven framework for generating controlled, label-preserving edits to logical reasoning problems. Our method operates on symbolic representations of first-order logic and constraint satisfaction problem tasks, enabling targeted modifications to logical operators and other structural components before translating them back into natural language. Using this framework, we evaluate various LLMs under cumulative and individual operator edits and analyze their behavior in response to these changes. Our quantitative and qualitative analyses show that LLM reasoning behavior under controlled operator edits is inconsistent, regardless of model size or family: models sometimes adapt correctly to structural changes but often fail to track their logical consequences. The results from this automated stress test enable an evaluation of language models across different dimensions and help measure the reliability of their reasoning.
[NLP-109] PaperBanana-Interact: Scientific Diagram Refinement with Multi-Turn Human Feedback
【速读】： 该论文旨在解决科学论文内容到科学图表的自动化生成过程中，用户在多轮交互中难以一次性满足其视觉与表达偏好这一核心问题。尽管已有研究尝试实现单轮生成，但实际使用中用户普遍要求后续修改，表明现有方法在多轮迭代优化方面存在显著不足。其解决方案的关键在于提出一个名为MTPaperBananaBench的多轮图表生成基准数据集，包含292张标注了3,518条用户需求的图像，并构建了一个用户模拟器以自动生成自然语言反馈，从而降低对人工评估的依赖并支持可扩展的评测。进一步分析发现，现有基线系统存在两大共性失败模式：质量漂移（quality drift）和遗忘（forgetting）。为此，论文提出PaperBanana-Interact——一种基于多智能体的交互式生成系统，通过内部的批判-修正循环实现持续优化，有效避免质量退化与功能遗忘，在多个指标上显著优于基线模型，展现出更强的稳定性与生成质量。
链接: https://arxiv.org/abs/2608.30241
作者: Xueqing Wu,Ashwin Balasubramanian,Bingxuan Li,Dawei Zhu,Kai-Wei Chang,Yale Song,Yiwen Song,Rui Meng,Tomas Pfister,Nanyun Peng
机构: Google(谷歌); Peking University (北京大学); University of California, Los Angeles (加利福尼亚大学洛杉矶分校); University of Illinois Urbana-Champaign (伊利诺伊大学厄本那-香槟分校)
类目: Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV)
备注: this https URL
Abstract:Recent efforts have aimed to automate scientific diagram generation from paper content (Lin et al., 2026; Zhu et al., 2026a). However, fully satisfying an author’s visual and communicative preferences in a single turn is challenging: in our formative user study (N = 14), all participants requested further revisions after viewing an initial draft, and 86% of them rated the refined diagrams as more satisfactory. Despite the clear demand, the multi-turn workflow remains largely underexplored. To bridge this gap, we present MTPaperBananaBench, a benchmark for multi-turn diagram generation containing 292 images annotated with 3,518 user requirements. To reduce expensive human studies and enable scalable benchmarking, we construct a user simulator that, at each turn, identifies unsatisfied requirements and converts k of them into natural language feedback. Evaluating both requirement satisfaction and overall diagram quality reveals two key failure modes shared across baseline multiturn systems: (1) quality drift, where diagram quality progressively declines over turns, and (2) forgetting, where previously implemented features are lost in subsequent turns. To address these issues, we introduce PaperBanana-Interact, a multi-agent system that refines diagrams via an internal critique-and-refine loop. PaperBanana-Interact consistently improves rather than degrades diagram quality across turns, outperforming baselines by 11.9-18.6 points in quality score and reducing forgetting by 3.7-6.2 points.
[NLP-110] Quantifying and Mitigating Korean Jamo-Level Typographical Vulnerabilities in Large Language Models EMNLP2026
【速读】： 该论文旨在解决韩语中因音节块（syllable block）内部由称为“字母单元”（jamo）构成而导致的键盘级输入错误问题，这类错误在传统字符级编辑模型中未被充分捕捉。具体而言，键盘错误可能发生在音节内部，导致生成有效但语义改变的字符，或直接暴露原始的jamo，从而破坏子词分词（sub-word tokenization），且现有语法纠错流水线无法可靠纠正此类错误，使大语言模型（LLM）直接暴露于受损输入。为量化此脆弱性，研究者在KMMLU基准上施加五种jamo级扰动类型，并评估四种语言模型，发现准确率随扰动强度单调下降，且参数量扩展并未带来对音节内噪声的鲁棒性。进一步分析表明，错误输入会引发模型内部表示的显著偏移，该偏移无法归因于普通答案错误，且通过在这些表示上训练的简单线性探测器可高精度检测未见过的扰动类型（高AUROC）。基于这一信号，论文提出Typo-Aware Chain-of-Thought（TACoT）机制，仅当探测器识别出可能的拼写错误时才触发链式思维（CoT）推理，从而在极低推理开销下恢复大部分CoT带来的性能增益。其核心解决方案在于利用内部表示中的扰动信号实现智能推理路径选择，提升模型在真实噪声场景下的鲁棒性与效率。
链接: https://arxiv.org/abs/2608.30229
作者: Seojin Lee,Hwanhee Lee
机构: Chung-Ang University (中央大学); https://github.com/SJLee0311/korean-jamo-typo
类目: Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026 (Main Conference)
Abstract:Korean introduces an additional typographical perturbation level not captured by ordinary character-level edit models: because syllable blocks are internally composed of sub-character units called jamo, keyboard-level errors can occur within a syllable, either producing a valid but semantically altered character or exposing raw jamo on the surface. Both outcomes disrupt sub-word tokenization and are not reliably corrected by existing grammatical error correction pipelines, leaving LLMs directly exposed to corrupted inputs. To quantify this vulnerability, we apply five jamo-level perturbation types to the KMMLU benchmark and evaluate four language models, finding that accuracy declines monotonically with perturbation intensity and that parameter scaling does not confer robustness against intra-syllabic noise. We further show that typo-corrupted inputs induce a distinct shift in internal representations that is not reducible to ordinary answer incorrectness, and that a simple linear probe trained on these representations detects unseen perturbation types with high AUROC. Motivated by this signal, we propose Typo-Aware Chain-of-Thought (TACoT), which routes inputs to chain-of-thought inference only when the probe detects a likely typo, recovering a substantial portion of the CoT accuracy gain at a fraction of the inference cost.
[NLP-111] LaMoC: Loss-Aware Modular Compression for LLM s EMNLP2026
【速读】： 该论文旨在解决现有联合模块化压缩方法过度依赖激活统计信息，而忽视损失敏感性信息及其模块级表征的问题。其核心挑战在于如何有效融合不同类型的统计信息以提升压缩后模型在下游任务中的性能表现。解决方案的关键在于提出LaMoC（Loss-aware Modular Compression），一种基于损失感知的模块化压缩方法，通过梯度-误差对齐机制将激活统计量与经验费雪信息（Empirical Fisher）进行融合。该方法首次将经验费雪信息作为模块级的损失敏感性代理，使其能够更好地反映局部模块重构误差与下游损失之间的关联。进一步地，论文将联合模块化压缩重新建模为双层优化问题，同时最小化模块重构误差并动态调节激活信息与梯度信息的融合比例。最终，通过一种基于统计验证的实证驱动方法求解压缩问题，在四个模型家族共八个模型上进行了评估，结果表明，在4-8B规模模型中，LaMoC相较于当前最优模块化压缩方法平均降低2.5%的困惑度，并在任务准确率上实现1%的相对提升，验证了其有效性与优越性。
链接: https://arxiv.org/abs/2608.30226
作者: Mohanad Odema,Jacob Song
机构: LG Electronics USA( LG 电子美国公司)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Performance (cs.PF)
备注: EMNLP 2026 Findings
Abstract:Modular compression has enabled considerable parameter reduction in LLMs while preserving strong language understanding and downstream task accuracy. However, existing joint modular compression methods primarily rely on activation statistics, leaving loss-sensitivity information and its module-level characterization underexplored. We investigate addressing this gap with LaMoC, a loss-aware modular compression methodology that blends activation and Empirical Fisher statistics through gradient-error alignment. LaMoC improves joint compression by selecting compression statistics that better align local module reconstruction error with the downstream loss. Our contributions are three-fold: (1) We characterize the Empirical Fisher as a module-level loss-aware proxy that can be blended with the activation statistics required for compression. (2) We reformulate joint modular compression as a two-tiered optimization problem that minimizes module reconstruction error while tuning the activation and gradient information blending rate. (3) We implement an empirically driven methodology with statistical validation to solve the resulting compression problem. We evaluate LaMoC across four model families spanning eight models. On the 4-8B models, LaMoC achieves an average 2.5% reduction in perplexity and a 1% relative improvement in task accuracy over state-of-the-art modular compression methods.
[NLP-112] he Differential Reasoning Router: Operationalizing Cost-Aware LLM Annotation in E-commerce
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在电商场景中进行结构化产品数据标注时面临的冷启动问题：初期可用的预标注数据有限，昂贵的推理成本难以评估，且系统尚需人工审核才能实现规模化信任。尤其在基于规则的标注流程中，每个商品需满足多重业务规则，模型错误与规则边界模糊性共同影响最终决策质量。为此，论文提出差异推理路由（Differential Reasoning Router, DRR）这一成本感知框架，其核心在于联合优化模型选择与人工升级策略：通过分别估计直接模型与推理模型在样本层级和业务规则层级的成功概率，实现自适应路由——简单案例由直接模型处理，推理仅用于预期能提升决策质量的复杂案例，而高概率双重失败或规则冲突案例则被自动转交人工标注。该机制生成的精准标签可支持提示工程、监督微调、模型校准及规则优化，推动系统从依赖人力的冷启动阶段逐步过渡至高置信度自动化路由。在实际电商工作流中，DRR在达到最强置信度路由精度的同时，实现了超过60%的推理令牌成本降低。
链接: https://arxiv.org/abs/2608.30224
作者: Cheng Lyu,Jingyue Zhang,Vinny DeGenova,Mengwei Li,Yuanli Pei
机构: Wayfair(威弗)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Large Language Models (LLMs) are increasingly used to annotate structured product data in e-commerce, but early deployment often begins as a cold-start problem: only limited pre-launch labels are available, the value of expensive reasoning is unknown, and human review is needed before the system can be trusted at scale. This challenge is especially common in rule-based annotation workflows, where each item must satisfy multiple business rules and both model errors and ambiguous rule boundaries affect final decisions. We introduce the Differential Reasoning Router (DRR), a cost-aware framework for cold-start LLM annotation that jointly optimizes model selection and human escalation. Rather than treating a reasoning model as a default fallback, DRR estimates separate success probabilities for a direct model and a reasoning model at both the sample and business-rule levels, enabling adaptive routing: easy cases are handled directly, reasoning is reserved for cases where it is expected to improve the decision, and likely double-failure or rule-disagreement cases are escalated to human annotators. The resulting labels provide targeted ground truth for prompt engineering, supervised fine-tuning, calibration, and rule refinement, enabling a gradual shift from human-heavy cold-start annotation toward high-confidence automated routing. In a production e-commerce workflow, DRR reaches accuracy parity with the strongest confidence-based router while achieving more than 60% reasoning-token cost savings.
[NLP-113] Label Semantic Expansion via Label Guided Neural Topic Modeling
【速读】： 该论文旨在解决现有标签感知主题模型（label-aware topic models）在实际应用中存在的一大局限：尽管这些模型能够利用预定义标签指导主题学习，但所提取的主题仍以无序的潜在主题为主，难以直接用于以标签为中心的内容分析。为此，论文提出从“主题服务于标签”（topics-for-labels）的逆向视角出发，构建标签语义扩展（Label Semantic Expansion, LSE）方法，通过引入基于语料库的描述性主题词来丰富稀疏标签的语义表示。其核心解决方案在于设计一种标签引导的神经主题模型（Label-Guided Neural Topic Model, LGNTM），该模型能够学习与标签对齐的专用主题，将主题同时锚定于词汇和文档语义空间，并保持主题结构与标签结构之间的一致性。实验结果表明，该方法在标签-主题对齐、标签扩展、主题质量以及下游分类任务等多个互补维度上均表现出优异性能，有效实现了标签驱动的主题建模与分析。
链接: https://arxiv.org/abs/2608.30216
作者: Haojia Zheng,Yuyin Lu,Juntian Huang,Fan Ou,Yanghui Rao,Haoran Xie,Fu Lee Wang
机构: Lingnan University (岭南大学); Sun Yat-sen University (中山大学); Hong Kong Metropolitan University (香港都会大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 18 pages, 7 figures, 11 tables
Abstract:Topic models are widely used for content analysis, where users often analyze corpora around predefined labels rather than unordered latent topics. Existing label-aware topic models mainly follow a labels-for-topics perspective, using labels to guide topic learning, while the learned topics are not directly usable for label-centered analysis. We explore the reverse topics-for-labels perspective and instantiate it as Label Semantic Expansion (LSE), which enriches sparse label representations with corpus-grounded descriptive topic words. To exploit topics in LSE effectively, we propose a Label-Guided Neural Topic Model (LGNTM), which learns dedicated label-aligned topics, grounds them in lexical and document semantic spaces, and preserves consistency between topic structures and label structures. Experiments on label-topic alignment, label expansion, topic quality, and downstream classification demonstrate strong overall performance across complementary evaluation dimensions.
[NLP-114] owards a Joint Khmer Text Recognition and Word Segmentation
【速读】： 该论文旨在解决高棉语（Khmer）文本识别中因缺乏显式词边界分隔符而导致的词切分（word segmentation）难题。传统方法采用“识别-切分”串行流水线，需依赖两个独立模型，不仅引入累积误差，且在大规模文档处理中带来显著延迟。本文提出一种统一的联合高棉语文本识别与词切分框架，其核心创新在于采用连接时序分类（Connectionist Temporal Classification, CTC）解码器实现快速并行解码，并通过可配置参数（b=1表示含词切分，b=0表示不含）灵活控制输出形式。实验结果表明，该模型在不同模态文档数据集（包括印刷体、场景图像及手写文本）上均能同时完成字符识别与词边界定位，有效消除了对额外词切分模块的需求，显著提升了系统效率与准确性。
链接: https://arxiv.org/abs/2608.30213
作者: Marry Kong,Rina Buoy,Sovisal Chenda,Nguonly Taing,Masakazu Iwamura,Koichi Kise
机构: Techo Startup Center, Ministry of Economy and Finance, Phnom Penh, Cambodia; Osaka Metropolitan University, Osaka, Japan
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注: 16 pages, 7 figures, 6 tables
Abstract:Text recognition, or extracting electronic text from document images, has been indispensable for knowledge retrieval tasks, such as retrieval-augmented generation (RAG). For Khmer, extracted text is subject to an extra word segmentation step, as Khmer does not use any visible word delimiters to denote word boundaries. Thus, a recognition-then-segmentation pipeline for Khmer requires two separate sequential models; this is not only error-prone but also adds significant latency for large-scale document processing. This paper proposes a novel joint Khmer text recognition and word segmentation framework in a unified model. The proposed model, using a connectionist-temporal-classification (CTC) decoder for fast, parallel decoding, can be instructed to recognize Khmer text with ( b=1 ) and without ( b=0 ) word segmentation. Experimental results on different benchmark datasets of different document modalities (document, scene, and handwritten images) show that the proposed model can not only recognize characters in document images but also locate word boundaries, removing the need for an extra word segmentation step in a conventional sequential pipeline.
[NLP-115] When Models Hear What They Expect: Diagnosing Prosodic Heuristics in Multimodal Sarcasm Detection EMNLP2026
【速读】： 该论文旨在解决多模态大语言模型（Multimodal Large Language Models, MLLMs）在进行语用推理时，是否真正利用了语音中的韵律线索（prosodic cues）而非仅依赖表面声学模式的问题。研究聚焦于讽刺检测任务，通过在中文和英文下设置五种模态条件，系统分解词汇内容、声学语义与韵律结构的贡献。结果表明，引入音频信息虽未提升真阳性检测率，反而系统性增加误报。声学错误诊断揭示，模型错误集中于一种共享的表达性韵律刻板印象——即音高升高与不规则停顿，而这一特征在两种语言中均与真实讽刺信号不符。通过仅操控这两个维度进行因果验证，成功诱导出高达60%的误报率。将相同操纵模板应用于Gemini 3 Flash Preview后复现该现象，表明该刻板印象并非特定于Qwen Omni架构，而是可能广泛存在于当前主流多模态模型中，其关键在于模型对韵律特征的非语用性启发式学习。
链接: https://arxiv.org/abs/2608.30204
作者: Yongjian Chen,Pengfei Wei,Yiqun Sun,Zhu Li,Lawrence B. Hsieh
机构: Magellan Technology Research Institute (MTRI); Center for Language and Cognition, University of Groningen (格罗宁根大学语言与认知中心), the Netherlands
类目: Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026 Findings
Abstract:Multimodal Large Language Models (MLLMs) process speech and text jointly, yet whether they exploit prosodic cues for pragmatic inference or rely on surface acoustic patterns has received little systematic investigation. We address this through sarcasm detection, evaluating Qwen2.5-Omni and Qwen3-Omni on Mandarin Chinese and English under five modality conditions that decompose the contributions of lexical content, vocal semantics, and prosodic structure. Adding audio systematically inflates false positives without improving true positive detection. Acoustic error diagnosis reveals that model errors cluster on a shared stereotype of expressive prosody, namely elevated pitch and irregular pausing, that diverges from the actual cues marking sarcasm in both languages. Targeted manipulation of only these two dimensions causally confirms the heuristic, inducing false positive rates of up to 60%. Applying the same manipulation template to Gemini~3 Flash Preview without modification replicates the effect, suggesting that the stereotype extends beyond the Qwen Omni family rather than arising from a single model architecture.
[NLP-116] When Errors Become Memories: Causal Pathway Tracing in Multi-Turn Memory-Augmented LLM s
【速读】： 该论文旨在解决增强记忆的大语言模型（Memory-Augmented Large Language Models, LLMs）中跨轮次错误传播的机制不明确问题。现有研究多聚焦于记忆系统是否能正确存储与检索信息，却缺乏对错误如何在不同对话轮次、记忆状态及未来交互中持续扩散的深入理解。其核心解决方案是提出一种基于结构因果模型（Structural Causal Model, SCM）的跨轮次错误传播分析框架，将用户提问、模型回复与记忆状态建模为动态因果过程，并识别出两种关键错误进入路径：内部记忆更新与外部问题反馈。通过干预这两条路径，构建四种反事实轨迹并量化其下游影响与交互效应。误差影响在四个层面进行评估：记忆保留、自然响应、定向诊断探测以及概率层面的错误偏好。实验表明，误差影响通常随交互距离衰减，但记忆更新路径带来的持久性影响显著高于问题反馈路径；即使错误从自然响应中消失，潜在错误仍可能在记忆中残留。此外，错误传播模式因记忆类别与记忆机制的不同而异。基于路径引导的修复策略进一步验证了该分解的有效性：问题修复可降低27.5%残余误差，记忆修复降低70.2%，联合修复则实现98.3%的误差消除，近乎完全阻断错误传播。
链接: https://arxiv.org/abs/2608.30198
作者: Shuyao Xiao,Shengling Wang,Xuan Chen,Ke Chao,Ming Cui,Feifei Qian,Fanlin Meng,Chaoyang Mei,Chaoyong Jiang,Qi Ouyang,Junxi Yi
机构: Beike Language and Intelligence(贝壳语言与智能)
类目: Computation and Language (cs.CL)
备注:
Abstract:Long-term memory enables large language models (LLMs) to preserve and reuse information across interactions, but it can also turn localized errors into persistent risks. Existing work mainly evaluates whether memory systems store and retrieve information correctly, leaving limited understanding of how errors propagate across responses, memory states, and future interactions. We propose a structural causal model (SCM)-based framework for cross-turn error propagation in memory-augmented LLMs. We model user questions, model responses, and memory states as a dynamic causal process, and identify two entry pathways: internal memory updating and external question feedback. By intervening on these pathways, we construct four counterfactual trajectories and quantify their downstream effects and interaction. Error influence is evaluated at four levels: memory retention, natural responses, targeted diagnostic probing, and probability-level error preference. Experiments show that error influence generally decays with interaction distance, while the memory-update pathway contributes more persistent effects than question feedback; latent errors may remain even after disappearing from natural responses. Propagation patterns also vary across memory categories and memory mechanisms. Pathway-guided restoration further validates this decomposition: Question Repair reduces residual error by 27.5%, Memory Repair by 70.2%, and Joint Repair by 98.3%, nearly eliminating residual propagation.
[NLP-117] ALTSTEER: Selective Safety Steering for Moving Beyond Hard Refusals to Constructive Alternatives EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在部署过程中面临的安全对齐（safety alignment）问题，即在防止有害响应的同时保持对良性请求的有用性。现有基于激活控制（activation steering）的安全调控方法在两个关键维度上存在局限：其一，触发干预的机制在不同领域间表现不稳定；其二，以拒绝为导向的调控方式常导致生硬的拒绝回应，而非具有建设性的安全引导。针对上述问题，本文提出ALTSTEER框架，其核心在于在单次推理过程中耦合选择性干预与以拒绝为锚点的建设性重定向。该方法利用内部与拒绝相关的信号判断是否需要干预，并通过分阶段的激活调整策略，将生成过程从单纯的拒绝控制逐步转向更具建设性的安全输出。实验结果表明，ALTSTEER在Llama-3.1和Qwen2.5模型上均能有效保留良性任务的可用性，同时显著提升对有害请求的建设性安全响应能力，尤其改善了原本倾向于以简短拒绝应对有害请求的模型表现。
链接: https://arxiv.org/abs/2608.30197
作者: Hoejoon Kwon,Byeonggeuk Lim,Kahyeon Kim,YoungBin Kim
机构: Chung-Ang University (中央大学); Graduate School of Advanced Imaging Science, Multimedia Film (先进影像科学与多媒体电影研究生院)
类目: Computation and Language (cs.CL); Software Engineering (cs.SE)
备注: Accepted at EMNLP 2026 Main Conference
Abstract:Safety alignment is essential for deploying large language models, requiring systems to prevent harmful compliance while preserving helpfulness on benign requests. Activation steering offers a training-free inference-time approach to safety control, but effective safety steering requires addressing two coupled questions: when to intervene and how generation should be shaped after intervention. However, existing safety steering methods remain limited along both dimensions, as their triggering mechanisms can be unstable across domains and refusal-oriented steering often yields rigid refusals rather than constructive safe guidance. To address these limitations, we propose ALTSTEER, an inference-time framework that couples selective intervention with refusal-anchored constructive redirection within a single inference pass. ALTSTEER uses an internal refusal-relevant signal to decide when to steer, and applies staged steering to shift generation from refusal-oriented control toward constructive alternatives. Evaluations on Llama-3.1 and Qwen2.5 show that ALTSTEER preserves benign utility while improving constructive safe-completion behavior, especially on models that otherwise tend to produce short refusals for harmful requests.
[NLP-118] GPAgent Bench-2K: Benchmarking Large Language Model Agents in Complex Clinical Action Space
【速读】： 该论文旨在解决现有大型语言模型（LLM）在临床决策任务中评估体系不完善的问题，即当前基准测试将复杂的临床工作流程简化为静态预测或动作空间粗粒度的无约束马尔可夫决策过程（MDP），无法真实反映初级保健场景下的动态、多步骤决策需求。为此，研究提出GPAgentBench-2K，首个面向全科临床决策的受约束马尔可夫决策过程（Constrained MDP, CMDP）基准，基于真实全科医生诊疗记录构建，并引入六类基础临床动作、拓扑结构化的任务流程先验以及将安全导向的拒绝响应作为首要结果指标。其核心解决方案在于通过精细化建模临床动作空间与流程约束，实现对模型安全性与临床质量的协同评估。实验表明，随着动作空间扩大，主流LLM性能显著下降；更关键的是，即使前沿模型在诊断准确率上表现优异，仍超过一半高风险病例存在违反安全约束的情况，揭示出“临床质量-安全”差距。研究进一步采用受约束组相对策略优化（C-GRPO）建立基线，证实显式建模约束虽优于无约束强化学习方法，但仍远未达到临床可接受的安全水平。
链接: https://arxiv.org/abs/2608.30188
作者: Boqi Chen,Xudong Liu,Yunke Ao,Heejin Do,Jianing Qiu
机构: ETH Zurich(苏黎世联邦理工学院); MBZUAI(穆巴达拉人工智能研究所); University of Toronto(多伦多大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Large Language Models (LLMs) show great potential as clinical agents, yet existing benchmarks reduce clinical workflows to static predictions or unconstrained Markov Decision Processes (MDPs) with coarse action sets. To address this, we introduce GPAgentBench-2K, the first Constrained MDP (CMDP) LLM-agent benchmark for primary-care clinical decision-making, constructed from expert-validated records of real-world GP encounters. Our environment models a full spectrum of six foundational clinical actions, imposes a topological workflow prior over the action space, and operationalizes safety-informed abstention as a first-class outcome. Evaluating 16 state-of-the-art LLMs reveals a significant performance degradation as the action space scales. Crucially, we uncover a clinical quality-safety gap: even frontier models with the highest diagnosis accuracy violate safety constraints in over half of high-risk cases. Finally, we establish a reference point using Constrained Group Relative Policy Optimization (C-GRPO), and show that while explicitly modeling constraints improves performance over unconstrained RL methods, it remains far from clinically acceptable safety.
[NLP-119] A.X K2 Technical Report
【速读】： 该论文旨在解决大模型在代理应用（agentic applications）中面临的性能瓶颈与长上下文处理效率不足的问题，尤其关注如何在减少训练数据量的前提下提升模型的泛化能力与推理效率。其核心解决方案在于提出一种从零训练的688B参数混合专家（Mixture-of-Experts, MoE）语言模型A.X K2，通过使用更少但质量更高的训练数据（约8.5T tokens），并显著扩充代理行为与软件工程相关数据，实现了对前代模型A.X K1的全面超越，部分基准测试性能提升超过30个百分点，体现出显著的令牌效率增益。关键技术包括：引入稀疏门控注意力（Sparse Gated Attention, SGA），结合稀疏注意力与门控注意力机制，支持原生128K上下文长度，并通过稀疏索引器预热（sparse indexer warmup）优化索引策略，使每个查询仅需读取2,048个位置即可保持长序列建模能力，且在RULER基准上实现256K上下文下的94.6分表现；同时采用门控归一化（Gated Norm, GN）以稳定大规模训练过程，使得4比特NVFP4量化仍能维持与FP8精度相差不到1个百分点的表现；此外，提出的Think-Fusion方法允许用户在单一模型内灵活切换思考模式与非思考模式，增强了模型的实用性。综合评估表明，A.X K2在数学和韩语等任务上达到或优于当前主流开源权重模型的水平。
链接: https://arxiv.org/abs/2608.30181
作者: Cheolseung Baek,Dhammiko Arya,Eunki Kim,Gun Song,Gyoungeun Han,Hyunho Yang,Hyunjun Eun,Jin Kim,Junyoung Park,Juyun Wee,Minki Hong,Minkyung Park,Minsang Kim,Minsoo Kang,SaeRom Kim,Sangjin Kim,Sangyeol Lee,Seojin Lee,Seokhwan Jo,Seokyoung Hong,Seongho Choi,Seonghye Cho,Seongmin Ok,Sereimony Sek,Seungmo Cho,Seungsik Kim,Singon Kim,Sohee Park,Sooyeon Park,Subin Yi,Sungbin Yoon,Sungeun Lee,Sung Jun Cheon,Sungwan Kim,Sunwoo Lee,Tae Yoon Kim,Wonbeom Jang,Yohan Ra,Yong-jin Han,Youngjin Kim,Youngrang Kim,Yujin Kang,Yujin Lee
机构: SK Telecom(SK电信)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: this https URL
Abstract:We introduce A.X K2, a 688B-parameter Mixture-of-Experts (MoE) language model trained from scratch as a high-performance foundation for \emphagentic applications. Trained on approximately 8.5T tokens—fewer than its predecessor, A.X K1—on a smaller but higher-quality mixture with substantially expanded agentic and software-engineering data, it nonetheless improves over A.X K1 across the board, by over 30 percentage points on some benchmarks, reflecting large gains in token efficiency. To support long contexts efficiently, we introduce Sparse Gated Attention (SGA), which combines sparse attention with gated attention, and adopt Gated Norm (GN) to stabilize large-scale training. SGA is trained natively at 128K through a \emphsparse indexer warmup that optimizes the indexer against its own sparse top- k selection rather than the dense attention distribution, making adaptation markedly cheaper: each query reads only 2,048 positions, yet long-context quality is unchanged and A.X K2 scores 94.6 on RULER out to 256K. The outlier suppression of GN in turn keeps 4-bit NVFP4 serving within one point of FP8 accuracy. A simple yet effective Think-Fusion recipe further lets users switch between thinking and non-thinking modes within a single unified model. Extensive evaluations show that A.X K2 performs competitively against strong open-weight baselines, matching or exceeding them on math and Korean-language benchmarks.
[NLP-120] CPR for LLM s: Critical-Point Routing against Catastrophic Forgetting in Domain Adaptation
【速读】： 该论文旨在解决监督微调（Supervised Fine-Tuning, SFT）在适配大语言模型（Large Language Models, LLMs）至特定领域时导致的灾难性遗忘问题，即模型在提升领域专长的同时显著退化其通用能力。现有方法通常通过修改SFT损失函数来缓解遗忘，但不可避免地陷入领域性能与通用能力之间的权衡困境。本文提出一种突破该权衡的新范式：在模型层面解耦通用能力与领域专长，保留原始基础模型以维持通用能力，仅在需要时动态调用针对特定领域的专家模型。其核心解决方案是提出一种基于关键标记（Critical Tokens）的分层路由框架——CPR（Critical-Point Routing），该框架识别出基础模型表现不佳而专家模型表现优异的关键输入标记，并通过轻量级分层路由器为每个标记预测专家调用概率，结合动量平滑与阈值门控的推理机制，实现精准、高效的专家调用。实验表明，CPR在多种模型-领域配置下均达到当前最优性能，在领域任务上超越SFT专家模型1.4%-5.5%，同时将通用能力下降幅度从3.4%-14.5%显著恢复至不超过0.5%，且仅需对约三分之一的输入标记调用专家模型，计算开销极低。
链接: https://arxiv.org/abs/2608.30158
作者: Kwangmin Ki,Yunhun Nam,Jongheon Jeong,Jaehyung Kim
机构: Korea University (高丽大学); Yonsei University (延世大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Supervised fine-tuning (SFT) is the de facto standard for adapting large language models (LLMs) to target domains, but it often degrades the model’s general capabilities, a phenomenon known as catastrophic forgetting. Existing approaches typically modify the SFT loss to mitigate forgetting, but they inevitably operate along a domain-generality trade-off. In this work, we step outside this trade-off by decoupling the two capabilities at the model level: we keep the original base model for general capability, and selectively invoke the SFT expert only when domain-specific knowledge is required. Specifically, we propose CPR (Critical-Point Routing), a token-level routing framework between a base model and its expert derivative, based on critical tokens where the base model fails but the expert succeeds. We train a lightweight hierarchical router that estimates the expert-call probability per token, and pair it with a tailored inference procedure that combines momentum smoothing and threshold gating. Across diverse model-domain configurations, CPR achieves state-of-the-art across all settings, surpassing SFT expert by 1.4-5.5% in domain performance while recovering its general-capability drop from 3.4-14.5% to at most 0.5%, with minimal overhead from invoking the expert on only one-third of tokens.
[NLP-121] Reactivating Test-Time Scaling for Plane Geometry Problem Solving EMNLP2026
【速读】： 该论文旨在解决平面几何问题（Plane Geometry Problem, PGP）求解中面临的多模态推理挑战，核心问题在于现有基于符号程序（symbolic-program）范式的测试时扩展（Test-Time Scaling, TTS）方法在几何推理场景下表现不佳。其根本原因可归结为两点：一是刚性符号程序导致的推理多样性不足，二是符号推导前缺乏显式的视觉语义锚定。为此，论文提出多轨迹合成（Multi-Trace Synthesis, MTS）机制，将单一符号程序转化为包含可执行Python脚本与思维链增强（Chain-of-Thought augmented）变体在内的异构推理轨迹，以提升推理路径的多样性；进一步引入感知增强（Perception-Augmented, PA）训练策略，在符号推导前将几何图示解析为结构化语义片段，强化视觉与符号之间的显式对齐；并设计共识引导的多轨迹集成（Consensus-Guided Multi-Trace Ensemble, CG-MTE）推理框架，实现高效自适应的测试时推理。实验表明，该方法在多个几何基准上均显著优于主流通用多模态大模型（MLLMs）及专用几何求解器，且在测试时扩展场景下达到与高预算自一致性（self-consistency）相当的精度，同时将采样开销降低高达8倍。
链接: https://arxiv.org/abs/2608.30156
作者: Xiaoqiang Kang,Shengen Wu,Maizhen Ning,Xiaobo Jin,Kaizhu Huang,Yutao Yue,Xiaowei Huang,Qiufeng Wang
机构: Xi’an Jiaotong-Liverpool University (西交利物浦大学); University of Liverpool (利物浦大学); The Hong Kong University of Science and Technology (Guangzhou) (香港科技大学(广州)); Hithink Research (慧科研究); Duke Kunshan University (杜克昆山大学)
类目: Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026
Abstract:Plane geometry problem (PGP) solving has become a critical benchmark for multimodal reasoning because it requires accurate visual perception and precise multi-step symbolic deduction. Although test-time scaling (TTS) has demonstrated remarkable success in general mathematical reasoning, it fails to scale effectively under the symbolic-program paradigm for plane geometry. We identify two key obstacles: limited reasoning diversity induced by rigid symbolic programs and insufficient explicit visual grounding before symbolic deduction. To address these issues, we propose Multi-Trace Synthesis (MTS), which converts each symbolic program into heterogeneous reasoning traces, including executable Python scripts and CoT-augmented variants. We further propose Perception-Augmented (PA) training, which parses diagrams into structured semantic clauses before deduction, and Consensus-Guided Multi-Trace Ensemble (CG-MTE) for efficient self-adaptive inference. Experiments on three geometry benchmarks show that our method consistently improves PGP-solving across model scales and achieves strong performance against both general-purpose MLLMs and specialized geometry solvers. Under test-time scaling, CG-MTE achieves comparable accuracy to high-budget self-consistency while reducing sampling cost by up to 8x. Code and data are publicly available at this https URL.
[NLP-122] CAST: Critique-Aware Supervision for Training Reliable Long-Horizon Tool-Calling Agents EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）代理在长周期、交互式且具有状态依赖性的环境中执行任务时的可靠性问题。具体而言，由于单次错误操作（如错误退款）可能导致不可逆的任务失败，而这类错误可能在部分运行中不显现，但在多次试错中逐渐浮现，因此确保代理在多步与多轮试验中的一致可靠性至关重要。现有方法主要依赖基于提示的批判性代理，但其缺乏对复杂轨迹中动作合理性进行充分解释的能力；而基于优化的方法又难以系统生成用于训练的丰富验证推理依据。为应对这一挑战，论文提出一种批判意识训练框架——CAST（Critique-Aware Training），其核心在于将稀疏的任务结果转化为动作层面的监督信号，以支持批判性学习与策略优化。CAST通过分析代理轨迹，生成在部分可观测条件下解释动作有效性的结构化推理过程，并利用所构建的批判模型生成批判感知的训练数据，进而优化策略模型。在动态工具调用基准测试中对Qwen3系列模型进行微调的结果表明，CAST显著提升了跨领域的可靠性，在零售任务上超越GPT-OSS-120B超过10%的通过率（pass^4），并在跨领域医疗健康任务中实现额外9%的性能提升，验证了批判意识训练能有效增强LLM代理在真实动态环境中的鲁棒性。
链接: https://arxiv.org/abs/2608.30147
作者: Amir Saeidi,Zehua Zhang,Rishitosh Singh,Naman Ahuja,Vivek Gupta,Ali Payani,Gaowen Liu,Jayanth Srinivasa,Chitta Baral
机构: Arizona State University(亚利桑那州立大学); Cisco Research(思科研究院)
类目: Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026 (Main)
Abstract:Large language model (LLM) agents are increasingly deployed in long-horizon, interactive, and stateful environments. In these settings, a single wrong action, such as refunding the wrong purchase, can cause irreversible task failure and must be intercepted before execution. Such failures may not appear in every single run, but can emerge across repeated trials, making reliability across steps and trials critical. However, ensuring agentic reliability is challenging: even frontier LLMs struggle to explain why an action may be wrong, especially in long, intertwined trajectories governed by domain-specific policies. Much recent work relies on prompt-based critique agents, while optimization-based methods lack a systematic way to produce rich verification rationales for training. We address this gap with CAST, a critique-aware training framework that converts sparse task outcomes into action-level supervision for critique learning and policy optimization. CAST analyzes agent trajectories to synthesize structured rationales explaining action validity under partial observability. The resulting critique model is used to construct critique-aware training data for optimizing the policy model. Fine-tuning Qwen3-family models on dynamic tool-calling benchmarks, CAST improves reliability across domains, outperforming GPT-OSS-120B by over 10% pass^4 on Retail tasks and yielding an additional 9% improvement on Telehealth in an out-of-domain setting. These results demonstrate that critique-aware training improves the robustness of LLM agents in realistic dynamic environments.
[NLP-123] Verification-Aware Training for Speculative Decoding
【速读】： 该论文旨在解决生成式 AI（Generative AI）中推测解码（Speculative Decoding）效率受限的问题，即当前的草稿模型（draft model）训练方法无法有效适应目标模型在逐位置验证过程中的动态特性，导致生成序列在首次拒绝后所有后续位置均被丢弃，造成资源浪费。其核心问题是现有训练策略采用固定的逐位置加权方式对齐目标模型输出，忽略了验证过程的顺序性和一旦失败即终止的特性。为此，论文提出验证感知训练（Verification-Aware Training, VAT），其关键创新在于：（1）引入一个轻量级联合训练的验证头（verification head），作为二分类器监督草稿模型预测每个位置是否能在序列验证中存活；（2）设计验证自适应加权机制，将权重衰减锚定至每个样本首次拒绝的位置，保持该点之前完整权重，从而更准确地反映实际推理中的验证行为。VAT仅修改训练目标，无需改变草稿模型架构、目标模型或推理流程，具备良好的可集成性。实验表明，该方法在Qwen3-4B、Qwen3-8B和LLaMA-3.1-8B上应用时，平均接受长度提升最高达11.4%，端到端推理速度提升最高达8.7%，且在数学、代码和对话等多类基准任务中均表现出一致性能增益。
链接: https://arxiv.org/abs/2608.30135
作者: Geonmo Gu,Byeongho Heo,HeeJae Jun,Yoohoon Kang,Sangmin Lee,Sangdoo Yun,Dongyoon Han
机构: NAVER AI Lab(NAVER人工智能实验室); Korea University(高丽大学); NAVER AI Search Platform(NAVER人工智能搜索平台)
类目: Computation and Language (cs.CL)
备注: 14 pages
Abstract:Speculative decoding accelerates large language model inference by using a draft model to generate candidate tokens, which are verified by the target model in a single forward pass. Verification proceeds sequentially and discards every position from the first rejection onward, yet existing draft training relies on token-level imitation of the target with a fixed per-position weighting that reflects neither property. We introduce Verification-Aware Training (VAT), a plug-in framework that simulates verification at every training step and turns the resulting accept and reject patterns into supervision. VAT consists of two components: (i) a verification head, a lightweight jointly trained binary classifier that supervises the draft model on whether each position survives sequential verification; (ii) verification-adaptive weighting, which replaces the fixed weighting schedule by keeping full weight up to each sample’s first rejection point and re-anchoring the decay to start there. VAT modifies only the training objective, so it can be layered on top of existing methods without changing the draft architecture, the target model, or the inference procedure. Applied to EAGLE-3 and DFlash on Qwen3-4B, Qwen3-8B, and LLaMA-3.1-8B, VAT improves average acceptance length by up to 11.4% and wall-clock speedup by up to 8.7%, with consistent gains across math, code, and chat benchmarks. Code will be available at this https URL
[NLP-124] VIBE: Video Instruction-aligned Background music gEneration
【速读】： 该论文旨在解决当前视频到音乐（Video-to-Music, V2M）生成模型在语义控制能力不足以及对指令违反缺乏惩罚机制的问题，其根源在于现有模型依赖重建目标函数，且在扩散自回归（Diffusion Autoregressive, DAR）架构中存在静态跨模态条件编码的表征瓶颈。为克服上述局限，本文提出一种新型文本与视频到音乐（Text-and-Video-to-Music, T+V2M）生成模型——VIBE，其核心解决方案包括：（1）条件连接（Conditioning Connection），一种深度可分离的跨层条件机制，能够动态连接规划阶段与扩散精炼阶段的头部，实现更灵活的跨模态信息流动；（2）一套完整的奖励建模分类体系，通过结构化的五阶段训练流程，同时优化硬性可验证约束（如节拍、调性）与软性主观质量（如音乐性、多模态对齐），从而实现对生成结果的精细化控制。实验结果表明，VIBE在音频-视觉对齐、指令遵循度及音质等指标上表现优异，显著提升了生成可控性与指令遵从性，同时在生成保真度与多模态对齐方面与主流基线模型相当。
链接: https://arxiv.org/abs/2608.30125
作者: Aryan Vijay Bhosale,Vaibhavi Lokegaonkar,Vishnu Raj,Gouthaman KV,Sreyan Ghosh,Ramani Duraiswami,Lie Lu,Dinesh Manocha
机构: Dolby Laboratories(杜比实验室), USA; University of Maryland, College Park, USA
类目: ound (cs.SD); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Current video-to-music (V2M) models lack semantic control and fail to penalize instruction violations, largely due to their reliance on reconstruction objectives and the representational bottleneck of static cross-modal conditioning in Diffusion Autoregressive (DAR) architectures. To resolve this, we introduce VIBE, a novel text-and-video-to-music (T+V2M) generation model that leverages: (1) Conditioning Connection, a depth-wise cross-layer conditioning mechanism that dynamically bridges the planning and diffusion refinement heads and (2) a comprehensive reward modeling taxonomy, optimizing for both hard, verifiable constraints (e.g., tempo, key) and soft, subjective qualities (e.g., musicality, multimodal alignment) with a structured 5-stage training curriculum. Upon evaluation using audio-visual alignment, instruction following, and audio quality metrics, along with a subjective human evaluation study, we observe that VIBE demonstrates enhanced controllability and instruction adherence while performing comparably to most evaluated baselines on generation fidelity and multimodal alignment.
[NLP-125] Manacá-1B: An Open Reproducible Brazilian-Portuguese Language Model and a Tokenizer-Aware Paired Evaluation
【速读】： 该论文旨在解决巴西葡萄牙语（Brazilian Portuguese）在开放语言模型中的代表性不足问题，特别是现有模型难以复现且评估缺乏不确定性度量。其核心解决方案是发布Manacá-1B，一个从零训练的17.2亿参数解码器模型，专为巴西葡萄牙语设计，并采用全容器化、可复现的训练流程。关键创新在于实现了预训练过程的稳定性（无跳过或NaN步骤，具备自恢复损失波动能力），并公开完整的训练日志与动态数据。在统一评估框架下，对九个开源基线模型在四个葡萄牙语基准上的表现进行比较，所有对比均报告标准误和配对显著性检验，确保评估可靠性。实验表明，Manacá-1B在小于70亿参数规模中表现最强，在LAMBADA-PT任务上显著优于Tucano-1b1和Tucano-2b4；在常识补全任务上具有竞争力，但在多选推理任务上接近随机水平，这与小型基础模型的普遍现象一致。研究还揭示了一个重要的评估陷阱：将带大小写归一化的SentencePiece分词器转换为HuggingFace快速格式时，归一化器被静默丢弃，导致所有大写标记被路由至字节回退路径，严重降低性能但不反映在聚合指标中——未经修正的分词器使LAMBADA-PT准确率从45.3降至25.0。作者量化了该影响并提供一行修复代码，以精确复现训练时的分词器。所有代码、原始训练与评估日志、逐例预测向量、模型权重及修正后的分词器均已开源，确保论文中所有结果可完全重计算。
链接: https://arxiv.org/abs/2608.30114
作者: Bruno Leonardo Santos Menezes,Carlos Leonardo Souza Cardoso,Fabio Andre Machado Porto
机构: Laboratório Nacional de Computação Científica (LNCC), Brazil
类目: Computation and Language (cs.CL)
备注: Preprint. Code: this https URL ; model weights: this https URL
Abstract:Brazilian Portuguese remains under-served by open language models, and the few that exist are difficult to reproduce and are often compared without measures of uncertainty. We release Manacá-1B, an open decoder-only model of 1.72 billion parameters trained from scratch for Brazilian Portuguese with a fully containerized, reproducible pipeline. The pretraining is stable, with zero skipped or NaN steps and self-recovering loss spikes, and we release its full log and dynamics. We evaluate the model against nine open baselines on four Portuguese benchmarks under a single harness. Every comparison reports a standard error and a paired significance test, and the harness is validated against previously published numbers. On last-word prediction Manacá-1B is the strongest model below the 7B scale, exceeding both Tucano-1b1 and Tucano-2b4 on LAMBADA-PT with large paired margins; it is competitive on commonsense completion and near chance on multiple-choice reasoning, as are all small base models. Along the way we document a concrete evaluation pitfall: converting a SentencePiece tokenizer with case-folding normalization to the HuggingFace fast format silently drops the normalizer, routing every capitalized token to byte-fallback and depressing scores in a way that is invisible in aggregate metrics. The uncorrected tokenizer lowered LAMBADA-PT accuracy from 45.3 to 25.0; we quantify the effect and provide a one-line fix that reproduces the training tokenizer exactly. Code, raw training and evaluation logs, per-example prediction vectors, the model weights, and the corrected tokenizer are released so that every number in this paper can be recomputed.
[NLP-126] Can LLM s Take the Pulse of the Economy? A Real-Time Evaluation of LLM Nowcasts on Macroeconomic Indicators
【速读】： 该论文旨在解决宏观经济指标（如GDP、CPI等）的实时预测（nowcasting）问题，即在官方发布前预估当前周期的指标值，这对货币政策制定和金融市场决策至关重要。传统上，此类预测依赖于中央银行专家团队进行人工估算，但存在频率低、时效性差等问题。为提升预测效率与实时性，研究提出利用生成式人工智能（Generative AI）代理（LLM agents）结合实时网络搜索能力，实现高频次的自动现在预测。其解决方案的关键在于构建了一个名为LiveMacroEval的实时、抗数据污染评估基准：该基准通过在每个官方发布窗口关闭前持续生成每小时的预测结果，并采用两种独立评估指标——基于公告窗口内股票市场收益率的LiveMacro Score 和模拟Polymarket交易场景的LiveBetting Score，以确保评估的有效性与公平性。同时，引入联邦储备地区银行的现成预测、Bloomberg ECOS专业共识及auto-ARIMA基线作为对比，验证了四种先进LLM代理在启用网络搜索后，整体预测精度已接近机构与专业水平，尽管在不同指标间表现差异显著。这表明，生成式AI代理具备作为实时宏观经济状况估计工具的巨大潜力。
链接: https://arxiv.org/abs/2608.30110
作者: Xinyue Zhao,Ruiyi Zhang,Liqin Ye,Rui Cao,Pengtao Xie,Sudheer Chava
机构: Georgia Institute of Technology(佐治亚理工学院); University of California San Diego(加州大学圣地亚哥分校)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Nowcasting headline macroeconomic indicators, i.e., estimating an indicator’s value for the current reference period before its official release, is critical for monetary policy and financial markets, and central banks devote dedicated teams of expert economists to producing such estimates. Large language model (LLM) agents are a promising candidate for this task, combining broad world knowledge with real-time web search and supporting queries at higher frequency than institutional nowcasts. Evaluating their nowcasting capability is, however, challenging: headline indicators such as GDP and CPI are widely reported and likely memorized during pretraining, so any evaluation on historical releases is vulnerable to data contamination. To address this, we introduce LiveMacroEval, a live, contamination-resistant benchmark in which LLM agents produce hourly nowcasts for sixteen major U.S. macroeconomic indicators over a pre-release window closing at each official release. Nowcast quality is assessed through a LiveMacro Score against announcement-window equity returns and a LiveBetting Score from simulated Polymarket-style trading, with Federal Reserve regional-bank nowcasts, the Bloomberg ECOS professional consensus, and an auto-ARIMA baseline as comparators. Over six months with four state-of-the-art LLM agents configured with web search, aggregate nowcast accuracy is broadly comparable to the institutional and professional benchmarks, with performance varying widely across individual indicators. This highlights LLM agents’ potential as real-time estimators of macroeconomic conditions.
[NLP-127] COGTRL: Training LLM s for Scientific Discovery Assistance using Cognitive Traces via Reinforcement Learning EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在科学发现辅助任务中缺乏对细粒度认知过程建模的问题，尤其是对约束条件考量、失败方案分析及迭代决策等关键认知行为的缺失。此类认知过程对于真实科学家在特定目标与资源限制下开展研究至关重要。为应对这一挑战，论文提出一种名为COGTRL的轨迹级强化学习框架，其核心创新在于通过联合优化认知轨迹与科学步骤的生成，以实现认知驱动的推理模拟。该方法使模型在生成科学方法时不仅输出结果，还能显式展现推理路径，从而显著提升方法质量。实验表明，在两个30亿参数模型和两个科学领域（人工智能与材料科学）中，COGTRL相比同类30亿参数基线平均提升7.85分的方法质量，并达到与700亿参数模型相当的性能水平；领域专家评估也显示，对COGTRL生成方法的偏好度高于基线模型。因此，解决方案的关键在于构建一种可学习、可优化的认知轨迹生成机制，使模型具备类人化的科学推理能力。
链接: https://arxiv.org/abs/2608.30109
作者: Shrinidhi Kumbhar Santosh Mashetty Divij Handa Kevin Coutinho,Siddharth Sambhaji Ghule,Chitta Baral
机构: Arizona State University (亚利桑那州立大学)
类目: Computation and Language (cs.CL)
备注: Accepted EMNLP 2026
Abstract:Large Language Models (LLMs) trained on extensive scientific research are increasingly integrated as assistants for scientific discovery. However, most research papers omit the fine-grained cognitive process of examining constraints, failed alternatives, and iterative decisions required to achieve the desired goal. Such cognitive processes are vital for real-world scientists working toward specific goals under constraints. In this paper, we show that LLMs, when trained to produce such cognitive traces, perform better as scientific discovery assistants than when trained solely on scientific literature. We propose COGTRL, a trajectory-level reinforcement learning framework that trains LLMs to emulate cognitively grounded reasoning by jointly optimizing cognitive traces and the scientific steps produced in an interleaved manner. Across two 3B-parameter models and two scientific domains (AI and Materials Science), COGTRL improves method quality by an average of 7.85 points over comparable 3B model baselines and achieves competitive performance relative to 70B parameter models. Moreover, analysis by domain experts shows a preference for methods generated by COGTRL over the baselines.
[NLP-128] AtlasNLP: A Country-Aware Atlas of Dataset Representation in NLP
【速读】： 该论文旨在解决自然语言处理（Natural Language Processing, NLP）数据集在国家层面代表性不足的问题，揭示当前数据集在地理分布上的系统性偏差。其核心挑战在于，尽管许多研究声称覆盖特定语言，但缺乏明确的地理元数据，导致无法准确评估哪些国家的人群被包含在数据集中，从而影响了公平性评估、数据收集策略制定及人工智能政策的科学性。本文提出的解决方案关键在于构建AtlasNLP——一个涵盖超过13,000个NLP数据集记录的国家感知型知识图谱，通过标准化的任务类别对数据集的生产地与所代表的人口进行追踪。该资源包含人工校验的AtlasNLP-Gold参考集与基于ACL文献的大规模自动构建的AtlasNLP-Core集合。利用此工具，研究发现：（1）各国在数据集覆盖上存在显著不均衡；（2）数据集生产地与目标人群所在地区之间存在地理不对称性；（3）语言覆盖并不等同于地理代表性。这些发现暴露了现有数据集文档实践中对地理信息的忽视，并推动建立更透明、更具国家意识的地理元数据标准，以支持更加公平和可问责的NLP评估体系。
链接: https://arxiv.org/abs/2608.30107
作者: Joan Nwatu,Tsedeniya Solomon Amare,Longju Bai,Bontu Fufa Balcha,Zayd Bashir,Angana Borah,Zara Burzo,Yubin Choi,Naihao Deng,Samika Gupta,Michel Faloughi,Claude Kwizera,Ziqiao Ma,Cynthia Yacel Fuertes Panizo,Ellie Seehorn,Hui Shen,Jiayi Tang,Zesen Zhao,Boyuan Zheng,Rada Mihalcea
机构: University of Michigan(密歇根大学); Addis Ababa University(亚的斯亚贝巴大学); San Jose State University(圣何塞州立大学); University of Pennsylvania(宾夕法尼亚大学); Carnegie Mellon University(卡内基梅隆大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注: Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing
Abstract:Understanding which countries are represented in NLP datasets is essential for identifying gaps, targeting data collection, measuring progress, and informing AI policy. However, geographic metadata is very rarely available, and country-level representation is often hidden behind broad language-level claims. We introduce AtlasNLP, a country-aware atlas of over 13,000 NLP dataset records across normalized NLP task categories, tracking both the populations represented and where datasets are produced. AtlasNLP includes AtlasNLP-Gold, a human-curated reference set, and AtlasNLP-Core, an ACL-derived large-scale collection. Using this resource, we show that (1) dataset coverage is highly uneven across countries and tasks; (2) dataset production and representation are geographically asymmetric; and (3) language coverage does not imply geographic representation. These findings reveal blind spots in current dataset documentation practices and motivate more explicit geographic metadata for country-aware NLP evaluation.
[NLP-129] Arkios: An Open Bilingual English-Nepali Language Model Trained From Scratch with a Devanagari-Aware Tokenizer
【速读】： 该论文旨在解决低资源语言（如尼泊尔语）中生成式AI模型的预训练与评估难题，尤其关注小规模模型在低资源语种下的真实能力评估偏差问题。其核心挑战在于：现有评估协议（如标准多选题字母提示格式）会系统性地将尼泊尔语和英语双语模型的表现误判为随机水平，从而掩盖模型实际具备的阅读理解能力。解决方案的关键在于揭示评估范式对结果的显著影响——通过对比字母选项选择与直接答案文本评分两种方式，发现尽管在字母格式下模型在尼泊尔语和英语上均表现接近随机（0.240和0.236，低于0.250的随机基线），但若直接评分答案文本，则模型在尼泊尔语（0.306）和英语（0.387）上均展现出显著优于随机的真正理解能力。这一发现表明，评估设计本身可能引入严重偏差，尤其在低资源语言场景下。此外，论文提出一种“声明条件化工具使用合约”（manifest-conditioned tool-use contract），在指令微调阶段约束工具调用行为，仅当上下文声明了工具清单时才允许调用，增强了模型行为的可控性与可解释性。最终，研究释放了1.04B参数的基座模型及指令微调版本权重，并确保所有可复现性所需数据与代码均已公开，仅未披露部分私有尼泊尔语预训练语料及训练代码。
链接: https://arxiv.org/abs/2608.30092
作者: Sajal Regmi,Siddhartha Pudasaini,Chetan Phakami Pun
机构: Karela Technologies Inc.
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 7 pages, 6 tables. Companion paper (tokenizer): arXiv:2608.26449
Abstract:We present Arkios, a 1.04B-parameter dense transformer pretrained from scratch on 150B tokens of bilingual English-Nepali text, using a custom single-file C/CUDA training stack and a Devanagari-aware byte-level BPE tokenizer built for this project. On ARC-Easy and ARC-Challenge, Arkios exceeds three comparably sized open models (Pythia-1.4B, TinyLlama-1.1B, OLMo-1B) despite an order of magnitude fewer training tokens, likely aided by a match between our educational-web-text pretraining data and ARC’s grade-school-science format rather than a general capability advantage. We report full evaluation results under standard protocols, including a correction to an earlier partial-sample estimate, and findings specific to evaluating small models in a low-resource language: the standard multiple-choice-letter prompt format used by common evaluation harnesses places this model at chance on Nepali reading comprehension, and simultaneously at chance on English in the same format, which would lead a naive benchmark run to conclude the model has no Nepali ability when in fact it does. Concretely, both languages score at chance in the letter-choice format (0.240 Nepali, 0.236 English, against a chance baseline of 0.250), while scoring the answer text directly reveals genuine, English-favoring comprehension (0.306 Nepali, 0.387 English). We describe a manifest-conditioned tool-use contract introduced during instruction tuning, where tool calls are permitted only when a tool manifest is declared in context and suppressed otherwise, and report where that contract holds and where it does not. We release both the base and instruction-tuned model weights under Apache-2.0. The training code and a small privately-sourced portion of the Nepali pretraining corpus are not released; everything needed to reproduce the reported numbers from the released weights is included here.
[NLP-130] When Does a Classifier Help an LLM ? Classifier-Guided Prompting and Hybrid Classifier-LLM Models for Credit-Default Prediction
【速读】： 该论文旨在解决生成式人工智能（Generative AI）在信用违约预测任务中表现受限的问题，尤其关注大语言模型（LLM）在缺乏排序能力与结构化决策支持时的性能瓶颈。其核心挑战在于：尽管LLM可通过提示工程（prompting）处理非结构化文本信息，但在精确识别高风险客户（即提升召回率）和实现有效排序（如提高AUC-ROC）方面仍逊于传统统计学习模型。解决方案的关键在于将一个经过训练的分类器（如随机森林）作为引导工具，通过提取其关键特征重要性及预测概率来构建优化的提示（prompt），从而弥补LLM在排序能力和特征选择上的不足。实验结果表明，仅使用分类器的8个最重要特征进行提示剪枝可使召回率提升0.071、F1提升0.032；而将分类器输出的概率值加入提示后，LLM的AUC-ROC达到0.78，与随机森林相当，同时保持更高的召回率（高出0.118）。这表明，以分类器指导提示设计的“正向组合”策略优于让LLM模仿分类器或采用多分类器集成的方法。因此，研究推荐采用一种基于分类器引导的简单提示构造范式，以实现生成式AI在信用风险评估中的高效应用。
链接: https://arxiv.org/abs/2608.30086
作者: Rishi Datta,Lavanya Prahallad
机构: Amador Valley High School; Research Spark Hub Inc.
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 6 pages, 4 figures
Abstract:Credit-default prediction is an important task in financial decision making. Traditional methods use fitted classifiers such as logistic regression and random forests on tabular features. Large language models (LLMs) have recently been applied to this task through prompting. In this work we study how a fitted classifier and an LLM can be combined for credit-default prediction. We distinguish telling the LLM to imitate a classifier from using the classifier to build the prompt. We hypothesize that a fitted classifier can supply the ranking ability that an LLM prompt lacks. We experiment on the Default of Credit Card Clients dataset, and report recall, F1, and the area under the ROC and precision-recall curves, with bootstrap confidence intervals. We observe that a few-shot LLM has the highest recall (0.47) and F1 (0.50) of any single model but ranks worse than a random forest (AUC-ROC 0.72 against 0.79). Instructing the LLM to imitate a classifier gives no significant change. Pruning the prompt to the classifier’s eight most important features raises recall by 0.071 and F1 by 0.032. Adding the classifier’s predicted probability to the prompt raises the LLM’s AUC-ROC from 0.72 to 0.78, matching the random forest, while keeping 0.118 higher recall than it. The reverse composition, and the use of several classifiers, do not help. We thus recommend a simple classifier-guided prompt for LLM-based credit prediction.
[NLP-131] Budget-Aware Compression Pipeline for Single-GPU LLM Inference: Methods Trade-offs and Coupling Effects EMNLP2026
【速读】： 该论文旨在解决在单张NVIDIA GPU上部署700亿参数语言模型所面临的设备内存限制、长上下文推理吞吐量瓶颈以及工程集成成本过高等关键挑战。其核心问题在于如何在有限的GPU资源下实现高效、低延迟且高保真的大模型推理。解决方案的关键在于将单卡推理建模为一个在内存、计算吞吐与工程复杂度三方面权衡的“预算感知”设计问题，并系统研究剪枝（pruning）、量化（quantization）与键值缓存压缩（KV-cache compression）在真实执行环境下的协同效应。研究发现，层间剪枝可增强权重量化对精度损失的鲁棒性；而KV缓存稀疏化能有效降低内存占用，且不损害解码速度，与INT8 KV量化形成互补；相比之下，静态向量量化常与动态缓存机制产生冲突。基于这些耦合关系与显式预算追踪，作者构建了一个可复现的优化流水线，成功将70B模型压缩至约33 GB，在单张A40 GPU上对10,000词长提示保持约57 tokens/s的推理速率，且在主流及推理类基准测试中绝对精度损失控制在5%以内。论文进一步提出了兼顾模型质量、内存占用与端到端速度的综合评估协议，并为未来在真实单卡约束下的自动化推理管道搜索提供了基础框架。
链接: https://arxiv.org/abs/2608.30076
作者: Hongyu Yu,Yifei Shen
机构: Lenovo Research; University of Washington (华盛顿大学)
类目: Computation and Language (cs.CL)
备注: Accepted by GroundLM 2026 (EMNLP 2026 Workshop)
Abstract:Single-GPU deployment of 70B-parameter language models on an NVIDIA GPU is constrained by device memory, long-context throughput, and engineering integration cost. We cast single-GPU inference as a budget-aware design problem over these three axes and study how pruning, quantization, and KV-cache compression interact under realistic execution. Controlled ablations show that layer-wise pruning makes weight quantization more robust. KV-cache sparsification complements INT8 KV quantization by reducing memory without hurting decoding speed, while static vector quantizers often conflict with dynamic caching. Guided by these coupling results and explicit budget tracking, we assembled a practical pipeline and compressed a 70B model to about 33 GB, sustained about 57 tokens/s on 10k token prompts on a single A40, and kept absolute accuracy within 5% on common and reasoning benchmarks. We contribute design rules and a reproducible evaluation protocol that jointly report quality, memory, and end-to-end speed, and we provide a foundation for automated pipeline search under realistic single-GPU constraints.
[NLP-132] How do World Models and Policies Compose in LLM Agents ? A Joint Spectral and Behavioral Account EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）智能体如何在行动环境中实现环境理解与任务掌握的问题。其核心挑战在于揭示世界模型（world-model）与策略（policy）在训练过程中如何协同学习，以及二者知识表征的交互机制。解决方案的关键在于通过控制实验，结合世界模型训练（下一状态预测）与策略训练（奖励最大化），并基于参数更新的几何分解分析两者的学习模式。研究发现，有效的世界模型更新具有低秩特性，且在输入特征子空间上与策略更新共享，但输出方向近乎正交；更重要的是，顺序训练（sequential training）相比独立训练（separate training）能诱导更强的鲁棒性——当移除世界模型的主要输入方向时，顺序训练的智能体仍能保持性能，表明其已习得替代性的输入路径。行为层面，顺序训练的智能体展现出更广泛的状态与动作探索。进一步地，研究提出：策略训练是否充分保留了世界知识？为此，作者设计了一种无需额外训练的融合方法，基于几何启发的输入基底，并引入在线世界模型损失进行策略强化，结果显著优于基线。研究表明，世界知识与任务导向能力可通过几何互补的方式共同学习，未来后训练流程应系统优化二者之间的接口设计。
链接: https://arxiv.org/abs/2608.30067
作者: Ruize Xu,Xiao Yu,Yujin Tang,Chenming Shang,Nikhil Singh
机构: Dartmouth College (达特茅斯学院); Columbia University (哥伦比亚大学)
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026
Abstract:How do LLM agents come to both understand environments they act in and master tasks set within them? Through controlled experiments combining world-model training (next-state prediction) and policy training (reward maximization), we investigate this question. We dissect the resulting models through their additive parameter updates. Geometrically, we find effective world-model updates are low-rank and share an input-feature subspace with policy updates while writing to nearly orthogonal output directions, whether trained separately or sequentially. However, we find that, in projection interventions, the sequential update induces more robustness than separate policy RL when removing the world model’s leading input directions, suggesting that it has learned alternative input pathways. Behaviorally, we find the sequentially trained agent explores a wider range of states and actions. Based on this, we ask: does policy training preserve world knowledge as well as it could? We probe this with training-free merging built on the geometrically motivated input basis plus an online world-model loss during policy RL, and show both improve over the untreated baseline. Our findings suggest world knowledge and task-directed ability can be learned in geometrically complementary forms, and that future post-training pipelines should consider how best to engineer the interface between them.
[NLP-133] Pak3H: Evaluating the Cost of Cultural Mismatch in LLM Alignment with a Human-Contextualized Urdu Benchmark DATE
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在低资源语言中难以有效实现有益性（Helpfulness）、无害性（Harmlessness）和诚实性（Honesty，简称3H）对齐的问题。现有跨语言3H评估基准多依赖自动化翻译或基于大语言模型的合成方法，导致源语言偏见被传播且本地文化相关性不足。为此，本文提出首个经人工验证、具有文化情境适配性的乌尔都语3H评估基准Pak3H1，包含三个子集：PakAlpaca（有益性）、PakBeaverTails（无害性）和PakTruthfulQA（诚实性）。其核心解决方案在于采用多阶段流水线，结合人工文化适配与词典引导的后编辑策略，优先保障母语者判断，从而确保语义保真度与语境真实性。零样本评估结果显示，不同开放及专有架构的模型在本地化语境下均表现出系统性跨语言对齐缺陷：有益性得分下降，无害性防护机制失效于区域性安全风险，综合诚实性指标因本地事实约束显著退化。这些发现揭示了当前对齐方法在跨语言泛化中的结构性局限，强调了人类引导的本地化在实现公平多语言评估中的必要性。
链接: https://arxiv.org/abs/2608.30065
作者: Abdullah Hashmat,Usman Naseem,Agha Ali Raza
机构: Macquarie University (麦考瑞大学); Lahore University of Management Sciences (拉合尔管理科学大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: We introduce Pak3H, a human-validated Urdu benchmark for helpfulness, harmlessness, and honesty. Zero-shot evaluations show LLM performance degrades across all three dimensions in low-resourced contextualized settings
Abstract:Large language models (LLMs) demonstrate strong Helpfulness, Harmlessness, and Honesty (3H) alignment in English-centric settings, but these gains transfer poorly to low-resource languages due to cultural mismatches. Existing multilingual 3H benchmarks rely predominantly on automated translation or LLM based synthesis, propagating source-language biases while sacrificing local relevance. To address this gap, we introduce Pak3H1, the first human-validated, culturally contextualized Urdu benchmark suite for 3H alignment, comprising PakAlpaca (helpfulness), PakBeaverTails (harmlessness), and PakTruthfulQA (honesty). Our multi-stage pipeline integrates manual cultural adaptation and dictionary-guided post editing to prioritize native speaker judgment, ensuring both semantic fidelity and contextual authenticity. Zero-shot evaluations across multiple open and proprietary LLM architectures reveal systematic cross-lingual alignment gaps: helpfulness win rates decline under localized contexts, harmlessness guardrails break down against regional safety risks, and composite honesty metrics degrade substantially due to localized factual constraints. These findings expose structural limitations in current alignment approaches, underscoring the necessity of human-guided localization for equitable multilingual evaluation.
[NLP-134] Balance of Benchmarks: Semantic Density Reweighting for Benchmark Multiplicity and Task-Conditioned Evaluation
【速读】： 该论文旨在解决当前语言模型评估中因基准测试（benchmark）列表的非均衡分布导致的评价偏差问题。传统方法对基准测试列表中的各项赋予相等权重，使得在密集发布基准测试的领域中，其重复出现的评测结果被过度加权，从而扭曲了模型真实能力的评估。论文提出“基准测试平衡度”（Balance of Benchmarks, BoB）作为解决方案，其核心在于通过语义嵌入（semantic embedding）对基准测试描述进行建模，并为每个基准分配逆密度语义权重，以反映其在整体评测生态中的稀疏性。该方法利用基准间的语义邻近性，在一个明确的密度尺度上共享聚合影响，从而实现对异构评分的统一潜在空间映射。在此基础上，引入残差场（residual field）以实现任务条件化预测，使模型排名能够响应特定任务需求。实验表明，相较于等权重方法，BoB在14个基准、586个模型的样本上实现了0.462的轮廓相关系数，显著优于等权重下的0.049；同时，在重复添加基准副本的情况下，其排名稳定性（Kendall tau达0.995）远超等权重方法（0.936），展现出对基准多重性（multiplicity）的强鲁棒性。因此，该方案的关键创新在于将基准测试列表的组成从评估体系的偶然属性转化为可显式控制的测量设计要素，为任务感知与多重性稳健的语言模型评估提供了原则性基础。
链接: https://arxiv.org/abs/2608.30044
作者: Jhen-Ke Lin
机构: National Yang Ming Chiao Tung University (国立阳明交通大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Language models are commonly compared by averaging scores across a benchmark list with equal weight. Such lists grow through publication outside an explicit measurement design, so equal weighting turns the density of published benchmarks into an implicit capability weight: densely benchmarked regions count repeatedly. We introduce Balance of Benchmarks (BoB), which embeds benchmark descriptions and assigns each benchmark an inverse-density semantic weight. Nearby entries share aggregate influence at a disclosed density scale. After equating heterogeneous scores onto a common latent scale, a residual field uses the same geometry to condition model rankings on a task query. The two components serve distinct empirical roles. On a snapshot of 586 models and 14 benchmarks, BoB predicts which models are unusually strong on a held-out task beyond their general ability, reaching a profile correlation of 0.462 compared with 0.049 under equal weighting. It also limits the influence of densely repeated benchmarks on the aggregate. After adding four copies of each benchmark in turn, the resulting rankings retain a Kendall tau of 0.995, compared with 0.936 under equal weighting. The residual field therefore provides task-conditioned prediction, and inverse-density weighting provides robustness to benchmark multiplicity. Together, they turn benchmark-list composition from an incidental property of evaluation suites into an explicit, controllable part of measurement design, providing a principled foundation for task-aware and multiplicity-robust model evaluation.
[NLP-135] “Act Like a 5th Grader” is Not Enough: Bounding Knowledge in LLM -Based User Simulators EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在模拟人类行为时普遍存在“超人偏差”（superhuman bias）的问题，即模型表现过于理想化，未能体现真实人类认知能力的局限性。尤其在阅读理解任务中，标准的角色提示（persona prompting）方法导致模型呈现出近乎完美且确定性的输出，无法反映儿童读者在发展过程中的自然变异性。为解决这一问题，论文提出认知受限用户模拟器（Cognitively Bounded User Simulator, CBUS），其核心在于通过一个情景记忆瓶颈（episodic bottleneck）显式建模年幼读者有限的工作记忆容量。在此架构下，研究进一步形式化了两种不同的答题策略，以模拟不同阅读行为模式。实验结果表明，通过在模型架构上施加认知约束，显著缩小了模型与真实学生表现之间的模拟差距，验证了结构化认知限制相较于单纯扩大模型规模，在实现高保真度人类行为模拟方面具有更优效果。
链接: https://arxiv.org/abs/2608.30033
作者: Krisztian Balog,Arild Michel Bakken
机构: University of Stavanger (挪威斯塔万格大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Findings of the Association for Computational Linguistics: EMNLP 2026
Abstract:Large language models (LLMs) are increasingly used to simulate human behavior but frequently fail to exhibit realistic cognitive constraints, suffering from a “superhuman bias.” Using a dataset of over 71,000 reading comprehension responses from 2,359 primary-school students (grades 4–6), we demonstrate that standard persona prompting yields near-perfect, deterministic performance, failing to capture the natural variance of developing readers. To address this, we introduce the Cognitively Bounded User Simulator (CBUS), an architectural framework that explicitly models the restricted working memory of young readers through an episodic bottleneck. Within this framework, we formalize two distinct test-taking strategies to emulate different reading behaviors. Our evaluation shows that explicitly modeling cognitive bounds significantly narrows the simulation gap across multiple LLM backbones, demonstrating that enforcing architectural constraints is more effective for high-fidelity simulation than simply scaling raw model capabilities.
[NLP-136] Small Language Models as Judges for Rubric-Based Reinforcement Learning EMNLP2026
【速读】： 该论文旨在解决生成式强化学习（Reinforcement Learning, RL）在缺乏精确答案或规则化验证器的任务中，因依赖高成本的评分机制而导致训练效率低下的问题。具体而言，现有基于评分表（Rubric-based）的RL方法需频繁调用大型语言模型（如7B参数及以上）进行评分，严重制约了训练效率。为应对这一挑战，论文提出使用小型语言模型作为高效且可靠的评分代理，以降低奖励计算开销。其解决方案的关键在于设计并验证一种名为“Probe judge”的新方法——通过微调小模型对评分标准的敏感性，直接提取准则层面的判断结果。实验表明，在构建的PointRubric与RaR-Science-Static两个点对点评分数据集上，Qwen3-1.7B Probe judge在准则级一致性上优于生成式判断（Generative verdicts）和基于对数概率差值（Yes/No Logprob margins）的方法；将其用于GRPO奖励模型时，可在显著减少10.7倍评分时间的前提下，将政策在RaR-Science评分上的表现从0.232提升至0.643，超越8B生成式基准（0.594），同时展现出良好的任务与领域迁移能力，证明其能有效保留准则级奖励结构。
链接: https://arxiv.org/abs/2608.30005
作者: Fengyu Xie,Yilun Zhao,Bingsen Chen,Arman Cohan,Chen Zhao
机构: New York University(纽约大学); Yale University(耶鲁大学)
类目: Computation and Language (cs.CL)
备注: 9 pages, 1 figure; EMNLP 2026 Findings
Abstract:Rubric-based reinforcement learning extends RL beyond tasks with exact answers or rule-based verifiers by scoring responses against instance-specific criteria. However, this makes reward computation expensive: training requires repeated rubric judging, often with proprietary APIs or local generative LLM judges with 7B parameters or more. We study whether smaller language models can serve as efficient and reliable rubric-based judges. To make this question measurable, we construct PointRubric and RaR-Science-Static, two pointwise rubric-based evaluation datasets with instance-specific criteria and itemwise satisfaction labels. We compare three ways of extracting criterion-level judgments from small models: Generative verdicts, Yes/No Logprob margins, and Probe judges. Across both datasets, the Qwen3-1.7B Probe judge achieves the strongest criterion-level agreement among these methods, outperforming Generative and Logprob judges. Used as a GRPO reward model, it trains a policy from 0.232 to 0.643 on RaR-Science rubric score, compared with 0.594 for an 8B Generative judge baseline, while the baseline requires 10.7 \times more reward-judge time. Task and domain transfer experiments further suggest that Probe judges preserve criterion-level reward structure across settings.
[NLP-137] Generating Clinical Vignettes that Preserve Cognitive Formulations EMNLP2026
【速读】： 该论文旨在解决大语言模型生成的临床案例描述虽流畅但缺乏对特定临床结构忠实性的问题，即生成内容可能在语言层面自然流畅，却未能准确反映疾病认知模型中的关键成分与因果关系。其解决方案的关键在于提出一种基于认知理论的框架FORMA（Framework for Ordered Representation of Mental health Accounts），该框架将特定心理障碍（以创伤后应激障碍，PTSD为例）的认知模型（如Ehlers & Clark模型）转化为有向加权图结构，通过采样个体化参数配置生成案例，并利用多维度评估方法验证生成文本是否保持了指定的结构要素与因果关联。研究结果显示，基于完整条件生成的案例可有效恢复其背后的认知图谱（MCC = +0.41, AUC = 0.70），而零样本生成则无此能力（MCC = +0.01, AUC = 0.50），且临床专家评分和真实执业医师的感知判断均显著支持该方法生成内容的可信度与人类写作相似性。此外，该方法还能将不同人口学特征群体在感知质量上的差异降低1.5至7倍。研究表明，认知构念可作为可审计、可扩展的合成临床文本生成规范，具有重要的应用潜力。
链接: https://arxiv.org/abs/2608.29995
作者: Amit Oren,Nimrod Hertz-Palmor,Dean Ariel,Guy Laban
机构: Ben-Gurion University of the Negev (本·古里安大学); Clalit Health Services (克拉斯健康服务); School of Public Health, Tel Aviv University (特拉维夫大学公共卫生学院); MRC Cognition and Brain Sciences Unit, University of Cambridge (剑桥大学认知与脑科学单位)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted to Findings of the Association for Computational Linguistics: EMNLP 2026. Code and data: this https URL
Abstract:Large language models can generate fluent clinical case vignettes, but fluency alone does not ensure fidelity to a specifiable clinical structure. We introduce FORMA, a theory-grounded framework that compiles a cognitive model of a disorder into a directed weighted graph, samples a person-specific configuration of that graph, and validates whether the generated vignette preserves the specified components and causal links. We instantiate FORMA on Posttraumatic Stress Disorder using the Ehlers and Clark cognitive model, generating 16,500 vignettes across 500 personas, 11 generation models, and three ablation conditions. Evaluation combines an external edge-recovery probe, two clinical experts, a scaled LLM judge, and a clinician user study with 100 licensed practitioners. The cognitive graph is recoverable from full-condition vignettes (MCC = +0.41, AUC = 0.70) but not from zero-shot generation (MCC = +0.01, AUC = 0.50). Experts rate full vignettes substantially higher than zero-shot alternatives, and clinicians perceive them to be human-written 85% of the time, compared with 22% for zero-shot. FORMA also reduces demographic disparity in perceived quality by 1.5-7x. These results show that cognitive formulation can serve as an auditable specification for scalable synthetic clinical text generation. A repository with the data and code is available online: this https URL.
[NLP-138] Beyond Fluency: A Rubric-Based Benchmark for Evaluating Saudi Dialect and Cultural Competence in Large Language Models
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在阿拉伯语市场应用中评估体系严重偏倚于现代标准阿拉伯语（Modern Standard Arabic, MSA）而忽视方言（Dialectal Arabic）及文化语境能力的问题。当前基准测试普遍以MSA流利度为评价核心，无法有效衡量模型在日常交流中所依赖的沙特方言（Saudi Dialect）的语用、词汇、习语及文化嵌入性理解能力。为此，研究提出一种基于评分量表（rubric-based）的沙特方言评估基准，包含31个由专家撰写的提示（prompts），覆盖习语、语用、词汇与文化嵌入等现象，并配以专家确立的基准答案（ground truth）。其解决方案的关键在于构建一个分阶段评估框架：首先进行模型无关阶段（model-agnostic），从基准答案中提取原子化、互斥且穷尽（MECE）的正向评价标准；随后进入模型特定阶段，对四款先进模型（Claude Opus 5、Gemini 3.7、GPT-5.6和Kimi K3）进行评分，并对其主动引入的错误予以惩罚。在124次模型-提示评估中，共识别出466例错误，归类至九类错误类型。结果显示，四款模型在宏观平均得分上高度集中（42.7%–53.1%），无一模型得分超过55%，且每款模型均存在至少一个负分提示，表明沙特方言理解仍属未解难题。值得注意的是，主要失败模式为“模糊表述”（Ambiguous Framing，占错误的37.3%），而直接幻觉（Hallucination）仅占11.2%，揭示模型并非主要因编造虚假信息而失效，而是因语体扭曲与语用细微差别的扁平化处理导致表现失真。此外，研究还发现一致性与性能上限之间的权衡关系以及各模型特有的错误特征。研究团队已公开完整提示集、基准答案与评分量表，以支持可复现的方言评估研究。
链接: https://arxiv.org/abs/2608.29990
作者: Ghassan Al-Sumaidaee,Sajjad Abdoli,Ahmed Rashad,Maxim Legg
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Large language models are increasingly deployed in Arabic-speaking markets, yet standard benchmarks overwhelmingly reward Modern Standard Arabic (MSA) fluency while leaving dialectal and culturally grounded competence unmeasured. This gap is consequential: everyday Arabic is largely dialectal, and dialect encodes social meaning that MSA-centric evaluation cannot capture. We present a rubric-based benchmark for the Saudi dialect, comprising 31 expert-authored prompts spanning idiomatic, pragmatic, lexical, and culturally-embedded phenomena, each paired with an expert-established ground truth. Our methodology separates evaluation into a model-agnostic phase, in which atomic, MECE positive criteria are derived solely from the ground truth, and a model-specific phase, in which four state-of-the-art systems – Claude Opus 5, Gemini 3.7, GPT-5.6, and Kimi K3 – are scored against those criteria and penalised for errors they actively introduce. Across 124 model-prompt evaluations we catalogue 466 error instances under a nine-category taxonomy. The four systems cluster within a narrow macro-average band (42.7%-53.1%), with no model exceeding 55% and every model recording at least one negative-scoring prompt, confirming that Saudi dialectal competence remains broadly unsolved. Notably, Ambiguous Framing is the dominant failure mode (37.3% of errors) while outright Hallucination accounts for only 11.2%, indicating that models fail less by stating falsehoods than by distorting register and flattening pragmatic nuance. We further observe a consistency-versus-ceiling trade-off and model-distinctive error signatures. We release the full prompt set, ground truths, and scored rubrics to support reproducible dialectal evaluation.
[NLP-139] Evolutionary Soups: Evolving Mixture-of-Experts for Multi-Objective LLM Alignment EMNLP2026
【速读】： 该论文旨在解决大语言模型在生成响应时难以同时满足多个相互竞争目标（如准确性与多样性、简洁性与完整性）的问题。传统方法通常依赖于静态的模型微调，无法根据用户偏好和具体输入动态调整生成策略。为此，论文提出了一种名为“进化汤”（Evolutionary Soups）的专家混合（Mixture-of-Experts, MoE）框架，其核心创新在于通过进化算法训练分层门控网络（gating networks），实现推理阶段的细粒度生成控制。关键在于：门控网络基于隐藏状态表示动态生成专家融合系数，而进化算法采用贪婪超体积贡献（greedy hypervolume contribution）机制，有效优化门控网络结构，从而在大规模且噪声较多的数据集上持续提升性能，并更全面覆盖非凸帕累托前沿（Pareto front）。实验表明，在三项任务中，该方法在超体积、线性效用和切比雪夫效用指标上均优于现有可控生成方法，平均提升约20%，验证了其在多目标生成中的优越性。
链接: https://arxiv.org/abs/2608.29978
作者: Lingxiao Kong,Steffen Staab,Cong Yang,Oya Beyan,Zeyd Boukhers
机构: Fraunhofer Institute for Applied Information Technology FIT(弗劳恩霍夫应用信息科技研究所); University of Cologne(科隆大学); University of Stuttgart(斯图加特大学); University of Southampton(南安普顿大学); Soochow University(苏州大学); University Hospital of Cologne(科隆大学医院)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE)
备注: Accepted at EMNLP 2026 (Main Conference)
Abstract:Large language models are increasingly required to generate responses that satisfy multiple competing objectives. Since optimal trade-offs depend on both user preferences and input prompts, controllable multi-objective generation must dynamically adapt models at inference time without retraining. To address this, we propose Evolutionary Soups, a mixture-of-experts framework for fine-grained generation control, with gating networks trained via an evolutionary algorithm. The per-layer gating networks dynamically produce expert-merging coefficients from hidden-state representations, while the evolutionary algorithm incorporates greedy hypervolume contribution for effective evolution of these gating networks, achieving consistent improvements on large and noisy training datasets and broader coverage of the non-convex Pareto front. Experiments across three tasks demonstrate the effectiveness of Evolutionary Soups over baselines: it achieves the best hypervolume, linear utility, and Tchebyshev utility (~20% improvement) among controllable methods on all tasks.
[NLP-140] SpanCalib-VLM: Calibrated Hallucination Span Detection in Vision-Language Models
【速读】： 该论文旨在解决大视觉语言模型（LVLM）中幻觉检测的两大核心挑战：精确的幻觉文本片段定位与置信度评分的良好校准。现有方法存在明显局限性——微调的生成式LVLM虽能有效识别幻觉片段，但普遍存在过度自信及推理延迟高的问题；而判别式序列标注器虽具备确定性推理速度和更优的置信度校准能力，却在幻觉片段召回率上表现保守。为此，本文提出SpanCalib-VLM，一种基于双系统融合的混合架构，结合了由XLM-RoBERTa-Large与SigLIP视觉编码器通过交叉注意力融合而成的多模态序列标注器，以及微调后的生成式LVLM（Qwen3.5-4B-SHROOM-SFT）。其关键创新在于采用“联合校准融合”（Union-Calibrated Fusion）策略，将生成模型输出的候选幻觉片段，利用序列标注器提供的校准概率进行重新评分，从而在保持高召回率的同时提升置信度校准性能。在SHROOM-Visions英文评估集上，该集成模型实现了0.41的皮尔逊校准相关系数、0.39的整体交并比（IoU）、0.91的纯净响应交并比及70.7%的整体检测准确率，显著提升了幻觉检测的可靠性与实用性。
链接: https://arxiv.org/abs/2608.29974
作者: Amanuel Gizachew Abebe,Yasmin Moslem
机构: Shaggar Institute of Technology(沙加尔技术学院); Trinity College Dublin(三一学院都柏林大学)
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注: Shroom-Visions
Abstract:Detecting hallucinations in Large Vision-Language Models (LVLMs) requires both accurate span localization and well-calibrated confidence scores. Fine-tuned generative VLMs excel at identifying hallucinated text spans but suffer from overconfidence and high inference latency. Discriminative sequence taggers offer deterministic speed and superior calibration but exhibit conservative span recall. We present SpanCalib-VLM, a hybrid dual-system for the SHROOM-Visions Shared Task that combines a multimodal sequence tagger, consisting of XLM-RoBERTa-Large fused with a SigLIP vision encoder via cross-attention, with our fine-tuned generative VLM (Qwen3.5-4B-SHROOM-SFT). Through a Union-Calibrated Fusion strategy, candidate spans from the generative model are re-scored with calibrated probabilities from the sequence tagger. On the SHROOM-Visions English evaluation split, our ensemble achieves a Pearson calibration correlation of 0.41 and an overall IoU of 0.39, with a clean-response IoU of 0.91 and overall detection accuracy of 70.7%. We make our model weights and code publicly available.
[NLP-141] DataFoundry: Evolving Data Preparators via Recursive Self-Improvement
【速读】： 该论文旨在解决大语言模型领域自适应中高质量训练数据构建的瓶颈问题，核心挑战在于现有数据准备流程通常在生成后才进行质量控制（post-hoc filtering），导致数据质量问题的根源——即数据构造过程本身——无法被有效干预。这种“过程与控制脱节”的机制造成了根本性不匹配。为应对这一问题，论文提出 \textscDataFoundry 框架，其关键创新在于通过递归式自我改进（recursive self-improvement）实现数据准备器（data preparator）的持续演化。该框架将数据准备器建模为可演化的运行时规范，并采用 \textscSkills-as-Modules 架构，由一个中央 \textscController 协调模块化技能组件，编译可执行运行时、利用领域适配的评估标准在小规模试点集上诊断缺陷，并将诊断反馈转化为适配器（adapters），对各准备组件进行修正，同时保持接口稳定。实验在 DataPrep-Bench 上覆盖数学、金融、法律和医学等多个领域，验证了递归演化后的准备器生成的数据在下游任务中具有更高实用性；跨不同模型骨干网络的实验进一步表明该优势不依赖于特定模型架构，且案例分析揭示了框架优化动态的实际演化过程。
链接: https://arxiv.org/abs/2608.29966
作者: Cehao Yang,Xiaojun Wu,Xueyuan Lin,Chengjin Xu,Xuhui Jiang,Hui Xiong,Jian Guo
机构: IDEA Research, International Digital Economy Academy; Hong Kong University of Science and Technology (Guangzhou); DataArc Tech Ltd.
类目: Computation and Language (cs.CL)
备注:
Abstract:Domain adaptation of large language models increasingly depends on constructing high-quality training data, yet existing data-preparation pipelines typically address quality only after generation through post-hoc filtering. This creates a fundamental mismatch: data-quality issues often originate from the construction process itself, while quality control is applied only to its outputs. We introduce \textscDataFoundry, a framework for \textbfevolving data preparators through recursive self-improvement before large-scale data production. \textscDataFoundry represents a data preparator as an evolvable runtime specification and instantiates its evolution with a \textscSkills-as-Modules architecture, in which a central \textscController orchestrates modular skills to compile executable runtimes, diagnose deficiencies on small pilot sets using domain-appropriate criteria, and translate diagnostic feedback into adapters that revise individual preparation components while preserving stable interfaces. We evaluate \textscDataFoundry on DataPrep-Bench across mathematics, finance, law, and medicine, and find that recursively evolved preparators produce training data with higher downstream utility than baselines. Experiments across different backbones further demonstrate that these improvements are not tied to a particular model, while analyses and case studies further reveal the framework’s optimization dynamics and illustrate how its evolution unfolds in practice.
[NLP-142] Generative vs. Encoder Models for Multilingual NER: A Comprehensive Empirical Study on Naamapadam
【速读】： 该论文旨在解决印地语等印度22种宪法认可语言在数字层面上的命名实体识别（Named Entity Recognition, NER）能力严重不足的问题，尤其是在生成式人工智能（Generative AI）模型广泛应用背景下，多数印地语族语言仍缺乏有效的NER解决方案。其核心挑战在于：尽管英语的NER已高度成熟，但大多数印地语族语言因标注数据稀缺、语言差异大及模型泛化能力弱，导致现有方法难以有效迁移与应用。论文的关键解决方案是通过系统性对比分析生成式模型与编码器型神经架构在Naamapadam基准上11种语言的性能表现，涵盖序列到序列的Transformer、多语言编码器（如mBERT和XLM-R）、基于解码器的大语言模型（LLMs）以及零至五次提示（zero-to-5-shot）的生成式模型。研究发现，在严格的CoNLL跨度级评估下，编码器型模型（如mBERT和XLM-R，印地语F1=0.675）在十种语言中显著优于所有生成式架构，性能差距达7.5–40个百分点，且最佳少样本结果仅达到编码器基线的28%。基于此，论文提出三类语言聚类——编码器主导型、部分覆盖型与失败区，并结合迁移学习与低资源自然语言处理原则，为实际部署提供可操作的指导策略。
链接: https://arxiv.org/abs/2608.29959
作者: Jakkala Mahesh,Jatavath Shravan Kumar,Komalla Shivani,Sujoy Sarkar
机构: Rajiv Gandhi University of Knowledge Technologies, Basar, Telangana, India
类目: Computation and Language (cs.CL)
备注:
Abstract:Language is humanity’s most consequential technology, yet for over a billion speakers across India’s twenty-two constitutionally recognised languages, its digital layer remains structurally incomplete. Named Entity Recognition (NER), the foundational step in transforming raw text into machine-interpretable knowledge, has been studied exhaustively for English but remains largely unsolved across most Indic languages. This paper presents a rigorous comparative study of generative and encoder-based neural architectures for NER on all eleven languages of the Naamapadam benchmark. We evaluate five classic model families spanning sequence-to-sequence transformers and multilingual encoders; four decoder-only large language models (LLMs) fine-tuned with LoRA and 4-bit NF4 quantisation; and nine generative models in zero-to-5-shot inference. Under strict CoNLL span-level evaluation, encoder-based models (mBERT and XLM-R, both F1=0.675 on Hindi) substantially outperform every generative architecture in ten of eleven languages, with gaps of 7.5-40 percentage points against the strongest competitor (Gemma-2-2B: avg F1=0.427). The best few-shot result reaches only 28% of the encoder baseline. We identify three language clusters–encoder-dominant, partial-coverage, and failure-zone; and provide actionable deployment guidelines grounded in transfer learning and low-resource NLP principles.
[NLP-143] Detecting Hidden Chain-of-Thought in Large Language Models with Linguistic Behavioral and Mechanistic Indicators
【速读】： 该论文旨在解决大语言模型在处理复杂推理任务时是否具备潜在的、未显式呈现的推理过程（latent reasoning）这一核心问题。现有模型常直接输出答案而缺乏中间推理步骤，导致难以判断其是通过隐式模式匹配完成任务，还是真正执行了类思维链（Chain-of-Thought, CoT）的逐步推理。为此，作者提出隐藏思维链检测得分（Hidden CoT Detection Score, HCDS），作为一种对比性行为与机制信号，用于衡量在中性提示下模型的行为更接近显式思维链（explicit CoT）还是显式无思维链（explicit no-CoT）的模式。其中，“隐藏思维链”（hidden CoT）操作化定义为中性提示下模型行为与思维链模式的相似性，而HCDS本身并不直接观测或证明未暴露的推理轨迹。实验结果表明，在GSM8K基准上，Qwen3-4B的两个变体（Thinking和Instruct）均表现出显著为正的HCDS值（Thinking：+1.87，p = 1.2 × 10⁻⁷；Instruct：+1.41，p = 1.9 × 10⁻⁴），且在不同推理栈和量化条件下结果稳定（偏差≤0.08），而在多数长度调整后的校准对照组中无显著正值。此外，未校正的得分在单步算术和数值事实查找任务中也呈现较大正值。更重要的是，两模型对“无思维链”指令的响应存在差异：Instruct仅依赖提示即停止推理，而Thinking仍持续进行内部推理，需外部干预。这些发现表明，经过推理优化的模型展现出更强的、较少受提示条件影响的类思维链行为，与潜在推理一致但不构成确证。因此，HCDS提供了一种无需依赖模型自我报告推理轨迹即可探究潜在推理机制的有效方法。
链接: https://arxiv.org/abs/2608.29956
作者: Armaan Singh,Ryan Trinh Le,Jasmine Kaur,Abdullah Sultan,Edward Lue Chee Lip,Kiran Nijjer,Adnan Ahmed,Vasu Sharma
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:Large language models often answer complex reasoning questions without revealing intermediate steps, raising whether they reason latently or complete patterns. We propose the Hidden CoT Detection Score (HCDS), a comparative behavioral and mechanistic signal measuring whether neutral-prompt behavior aligns more closely with explicit CoT or explicit no- CoT. Here, hidden CoT operationally denotes this neutral-prompt CoT-like alignment; HCDS does not directly observe or prove an unexposed reasoning trace. On GSM8K, HCDS is significantly positive for both Qwen3-4B variants (Thinking +1.87 , p = 1.2 \times 10^-7 ; Instruct +1.41 , p = 1.9 \times 10^-4 ), replicates across a different inference stack and quantization within 0.08 ( +1.80 and +1.45 ), and is not significantly positive in seven of eight length-adjusted calibration-control cells. The unadjusted score produces large positive scores on single-step arithmetic and numeric factual lookup. The variants also respond differently to no-CoT instructions: Instruct complies from the prompt alone, whereas Thinking continues reasoning and requires intervention. These findings show stronger, less prompt-conditional CoT-like behavior in the reasoning-tuned model, consistent with but not proof of latent reasoning. HCDS thus investigates latent reasoning without relying on models’ self-reported traces.
[NLP-144] How Prolific Sellers Self-Present: Dissecting the Communication Patterns of 1.6 Million Reverb Listings
【速读】： 该论文旨在解决在线市场中卖家身份（即资深卖家与普通卖家）如何通过产品描述的语言模式进行自我呈现的问题。研究基于自我呈现理论（self-presentation theory）与担保理论（warranting theory），分析了160万条乐器商品的销售描述，发现资深卖家更侧重于产品的客观功能属性（如规格与特征），而较少强调主观特质（如音色）；同时，他们采用更具叙事性的语言风格，表现为更高频使用第一人称代词、描述内容更长，从而在技术性信息传递中融入故事化表达。其解决方案的关键在于揭示：尽管资深卖家关注的技术性内容具有更强的可信度支撑，但其语言风格的叙事化特征则强化了与买家的情感连接，进而提升自我呈现的有效性。这一发现为理解在线市场中卖家身份建构及其沟通策略提供了理论与实证支持。
链接: https://arxiv.org/abs/2608.29952
作者: David M. Markowitz
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:The current paper draws on self-presentation theory and warranting theory to evaluate how the language patterns in an online marketplace reflect seller status (i.e., a prolific seller versus an everyday seller). Using 1.6 million musical instrument listings from this http URL in search of content, style, and structural differences in seller product descriptions, the evidence suggested prolific sellers tend to focus more on objective and functional aspects of a product (e.g., its features and specifications) and less on subjective characteristics like tone, relative to everyday sellers. Prolific sellers also communicated in a more narrative-like style, which was driven by an elevated use of personal pronouns, and they used longer descriptions than everyday sellers. Therefore, what prolific sellers focus on tends to be quite technical, but how they communicate this information is typical of a story that is told to potential buyers. Implications for self-presentation theory and warranting theory are discussed.
[NLP-145] XQDT: eXplainable and Quantitative Data-Text Alignment Metric with Feedback Signals EMNLP2026
【速读】： 该论文旨在解决数据-文本对齐评估中缺乏可解释性与可靠性的问题：现有评估指标通常无法提供评分的详细解释，而基于提示的大型语言模型作为裁判（LLM-as-Judge）的方法则存在计算成本高且结果不可靠的缺陷。其解决方案的关键在于提出一种端到端可解释的评估框架，通过微调语言模型来识别数据-文本配对中的遗漏（omitted）、多余（extra）、错误（incorrect）及正确（correct）的数据单元，进而将这些局部判断聚合为精确率（precision）、召回率（recall）和F1分数，从而在提供细粒度诊断反馈的同时，实现对对齐质量的可解释量化。实验表明，该方法在误差预测性能上优于主流的LLM-as-Judge方法，并在多个基准测试中达到与人类评判高度相关且具有竞争力的精度、召回率与F1值。此外，该验证器（verifier）输出还可作为下游数据-文本生成或反向文本-数据任务中对齐优化的有用反馈信号，推动生成过程的持续改进。
链接: https://arxiv.org/abs/2608.29948
作者: Kun Efimov-Zhang,Yifei Song,Claire Gardent
机构: CNRS/LORIA and Université de Lorraine(法国国家科学研究中心/洛林计算机科学实验室及洛林大学)
类目: Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026 (Main Conference)
Abstract:Evaluating data-text alignment remains challenging: existing metrics often provide limited explanations for the scores, while prompt-based LLM-as-Judge methods can be expensive and unreliable. We present an end-to-end explainable evaluation metric that fine-tunes a language model to identify omitted, extra, incorrect, and correct data units in a data-text pair. These local judgements are aggregated into precision, recall, and F1 scores, providing both fine-grained diagnostic feedback and an interpretable measure of alignment quality. Across benchmarks, our fine-tuned models outperform LLM-as-Judge methods in error prediction and achieve competitive precision, recall, and F1 scores, while maintaining strong correlation with human judgements. Beyond evaluation, our verifier outputs also provide useful feedback signals for downstream correction and refinement, supporting alignment-oriented improvement of data-to-text and text-to-data. Code and resources are available at this https URL.
[NLP-146] On the Recoverability of Private Information Unlearning in Large Language Models
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）中存在的敏感信息记忆问题，即模型在训练过程中可能无意中存储并保留用户隐私数据，从而引发严重的隐私泄露风险。尽管机器遗忘（Machine Unlearning）被视作一种潜在的解决方案，但现有方法是否真正实现了敏感信息的彻底删除仍存疑，部分方法可能仅是使信息难以访问而非完全消除。为解决这一评估难题，论文构建了一个包含虚假私密信息的合成数据集，并提出了一种白盒审计框架（white-box auditing framework），用于在统一标准下系统性地检测声称已被“遗忘”的信息是否仍存在于模型中。其核心发现表明，一种简单的“反向贪婪”解码策略——即每一步选择概率最低的词元——能够有效恢复本应已被遗忘的敏感信息。这一结果揭示当前主流遗忘方法普遍未能完全清除敏感数据，凸显了开发更可靠、可验证的遗忘机制以保障部署中大语言模型隐私安全的紧迫性。
链接: https://arxiv.org/abs/2608.29943
作者: Shicheng Hu,Runzhi Tian,Ziqiao Wang,Yongyi Mao
机构: University of Illinois Urbana-Champaign (伊利诺伊大学厄本那-香槟分校); University of Ottawa (渥太华大学); Tongji University (同济大学)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注:
Abstract:Large language models (LLMs) can memorize sensitive information, raising serious privacy concerns. Machine unlearning offers a potential solution to remove such information, but it remains unclear whether existing methods truly erase it or merely hide it within the model. A key challenge is quantifying the persistence of sensitive data under a unified evaluation framework. To address this, we construct a synthetic dataset containing fake private information and propose a white-box auditing framework to systematically assess whether claimed-forgotten information is genuinely removed. Using this framework, we evaluate five existing unlearning methods and find that a simple “inverse greedy” decoding – selecting the least likely token at each step – can recover supposedly forgotten private information. Our results reveal that current unlearning approaches often fail to fully eliminate sensitive information, highlighting the need for more reliable methods to ensure privacy in deployed LLMs.
[NLP-147] When Safety Speaks a Language: A Mechanistic Analysis of Safety-Language Identity Entanglement in LLM s
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在多语言场景下安全对齐（safety alignment）性能退化的问题，尤其关注不同语言间安全机制的不对称性及其内在驱动机制。其核心解决方案在于通过稀疏自编码器（Sparse Autoencoder, SAE）提取模型残差流中与有害和无害行为相关的关键可解释方向，并系统分析三个指令微调的LLM在八种语言、所有模型层中的安全特征分布。研究发现，安全相关特征的位置与分布具有架构依赖性，且与语言身份在几何上存在纠缠关系，表现出跨语言共享模式——不同语言在模型深度和架构层面共享安全特征的程度各异。这一安全-语言纠缠效应导致移除安全特征不仅影响有害响应率，还会显著影响目标语言的表现，干预效果可由安全特征与语言特征之间的关联程度预测。因此，该研究揭示了安全对齐的语言普适性并非普遍成立，而是依赖于模型架构，为理解多语言安全干预提供了机制层面的新视角。
链接: https://arxiv.org/abs/2608.29936
作者: Apoorva Upadhyaya,Sandipan Sikdar
机构: L3S Research Center, Leibniz Universität Hannover (汉诺威大学), Germany
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Safety alignment of large language models (LLMs) degrades across languages, yet the internal mechanism driving this asymmetry remains poorly understood. Our work, therefore, presents a systematic mechanistic analysis of multilingual safety using sparse autoencoder (SAE) features, sparse interpretable directions in the residual stream associated with harmful and harmless model behavior across three instruction-tuned LLMs, eight languages, and all model layers. We observe that safety-relevant features are architecture-dependent in terms of where they are located and how they are distributed across layers. Additionally, they are geometrically entangled with language identity and exhibit cross-lingual sharing patterns, i.e., languages share safety features to varying degrees across model depths and architectures. This safety-language entanglement has direct consequences such that ablating safety features impacts not only harmful response rates but also target language, with the degree of intervention predicted by the relationship between safety and language features. Our findings qualify the language-universality of safety alignment as architecture-dependent and offer a mechanistic account of multilingual safety interventions.
[NLP-148] Compression-Aware Abstention: Teaching LLM s to Refuse When KV-Compression Masks Remove Answer Evidence EMNLP2026
【速读】： 该论文旨在解决生成式大语言模型（LLM）在采用键值缓存压缩（KV-cache compression）进行推理时，因上下文记忆被压缩而丢失关键答案证据所导致的幻觉问题。其核心挑战在于：当压缩过程移除了承载答案的关键上下文片段时，模型仍会强行生成回答，而非识别出信息不足并选择不回答。本文从行为建模角度提出一种创新解决方案——将“压缩感知下的放弃回答”（compression-aware abstention）形式化为一个可学习的问题，即训练模型根据压缩后是否存在支持性证据来决定是否作答。关键在于构建监督信号：基于压缩器的存活掩码与精确的答案支撑跨度，将样本标注为“自信回答”（Confident，证据保留）或“放弃回答”（Abstain，证据被移除）。通过在约2600个MuSiQue 2跳问答样本上训练一个1010万参数的低秩适配器（LoRA），模型在提示风格截断场景下使基线模型的幻觉率降低97%，同时保持对证据完整保留样本的正确回答能力。相较于仅依赖提示设计的基线方法（过度放弃回答高保留率但可回答的例子），该适配器学习到一种条件性决策策略。进一步在真实压缩缓存解码环境下验证，多压缩器联合训练可使基线模型在保留证据的样本上实现6–22倍的相对性能提升；控制删除实验表明，模型的行为由证据内容驱动，而非输入长度等表面特征。
链接: https://arxiv.org/abs/2608.29934
作者: Mohammadali Khodabandehlou,Bhaskar Krishnamachari
机构: University of Southern California(南加州大学)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 19 pages, 5 figures. Accepted to the GroundLM workshop at EMNLP 2026. Code, adapters, and datasets: this https URL
Abstract:KV-cache compression reduces LLM inference memory by evicting context tokens, but when the evicted tokens contain answer-bearing evidence, the model may hallucinate instead of recognizing that the compressed context is insufficient. We address this failure from a behavioral perspective: to our knowledge, this is the first work to formulate compression-aware abstention as a learning problem, in which a model learns to answer when supporting evidence survives compression and abstain when it does not. We construct supervision from compressor survival masks and tight answer-bearing spans, labeling examples as Confident when evidence survives and Abstain when it is removed. A 10.1M-parameter LoRA adapter trained on ~2.6K MuSiQue 2-hop QA examples reduces base-model hallucinations by 97% under prompt-style truncation while preserving correct answering on evidence-retaining examples. Unlike prompt-only abstention baselines, which over-abstain on many answerable high-retention examples, the trained adapter learns a conditional policy. We also evaluate the method under actual compressed-cache decoding, where multi-compressor training yields a 6-22x relative lift over the unaided base on evidence-retaining examples. Controlled-deletion experiments show that the learned behavior is driven by evidence content rather than input length alone.
[NLP-149] oken Counts Are Not Model Lineage: A Frozen-Threshold Holdout Study of Black-Box LLM API Fingerprinting
【速读】： 该论文旨在解决在大型语言模型（LLM）通过代理和转售API提供服务时，如何有效识别不同模型是否属于同一模型家族（model-family）的问题。其核心挑战在于，尽管提示词（prompt）的标记计数（token count）作为低成本信号具有吸引力，但其在跨模型家族归属判断中的有效性尚未经过充分验证。研究的关键解决方案是提出一种“有效性门控结果契约”（validity-gated result contract），用于区分真实差异与因缺失使用数据、速率限制或端点策略导致的无信息性测量误差。通过在24对已标注的端点上进行三轮时间重复实验，研究发现基于冻结阈值0.725的移位不变精确匹配分数可完美分离开发集中的12对模型，但在独立测试集上仅有6对符合条件。在合格对中，平衡准确率为0.75，敏感性为0.50（95% Wilson置信区间0.15–0.85），特异性为1.00（0.342–1.00）。值得注意的是，同属一个家族的Qwen 3.8与DeepSeek V4变体仍低于阈值，表明仅依赖标记计数无法作为模型家族谱系关系的必要判定标准。研究最终验证了标记计数一致性可作为共享分词栈（tokenization stack）的指纹特征，但不支持其作为独立且充分的模型家族归属判别依据。
链接: https://arxiv.org/abs/2608.29930
作者: Bo Chen
机构: Institute of Computing Technology, Chinese Academy of Sciences(中国科学院计算技术研究所)
类目: Computation and Language (cs.CL)
备注: 10 pages, 2 figures
Abstract:Black-box model attribution is increasingly relevant when large language models (LLMs) are served through relay and reseller APIs. A tempting low-cost signal is the prompt-token count returned by an OpenAI-compatible endpoint: two models that share a tokenizer and chat template may produce the same count sequence up to a fixed offset. Yet the validity of this signal for broader \emphmodel-family attribution has received little direct holdout testing. We conduct a frozen-threshold study over 24 labeled endpoint pairs, split evenly into a development set and an untouched holdout set, with three temporal repeats and 30 controlled texts per pair. We introduce a validity-gated result contract that distinguishes an observed dissimilarity from an uninformative measurement caused by missing usage data, rate limits, or endpoint policy. The resulting shift-invariant exact-match score perfectly separates the 12 development pairs, yielding a frozen threshold of 0.725. On holdout, however, only 6 of 12 pairs are eligible under the pre-specified three-repeat rule. Among eligible pairs, balanced accuracy is 0.75, sensitivity is 0.50 (95% Wilson interval 0.15–0.85), and specificity is 1.00 (0.342–1.00). Two same-family pairs—Qwen 3.8 and DeepSeek V4 variants—fall below the frozen threshold. Across 4,320 formal API calls, every log is replayable, while holdout contains 189 non-200 responses and 157 successful responses without prompt-token usage. The study therefore validates token-count consistency as a fingerprint of a shared \emphtokenization stack, but rejects its use as a standalone necessary test for model-family lineage.
[NLP-150] Sleight of Word Benchmark: Can Language Models Notice If Their Own Output Was Tampered With?
【速读】： 该论文旨在解决语言模型在生成过程中对输入扰动的敏感性问题，即模型输出可能在生成过程中被外部干扰所篡改。为评估这一现象，研究提出了一种名为“Sleight of Word”（字面魔术）的简单基准测试方法，其核心在于在生成过程中持续将某个词替换为另一个词，从而观察模型对这种外部扰动的反应。解决方案的关键在于通过双重维度进行评估：一是衡量模型对扰动的“意外程度”（surprise），即模型在面对不一致输入时的内在感知变化；二是对19个开源权重的语言模型进行文本反应的系统性评估，以量化其鲁棒性与稳定性。该方法揭示了当前大语言模型在生成过程中的脆弱性，为后续提升模型安全性与可靠性提供了重要基准。
链接: https://arxiv.org/abs/2608.29921
作者: Alberto Cetoli
机构: Cleo(克里奥)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted at INLG 2026
Abstract:The output of a Language Model can be tampered with \emphwhile the model is writing it. A simple test can thus be constructed by evaluating the model’s perception of this external perturbation. In this spirit, a simple benchmark is built in which a single word is consistently substituted with another in the generation process. We call this method \emphSleight of Word. Two distinct axes are measured: metrics that relate to the model’s surprise, as well as an evaluation of the textual reaction for 19 different open-weight language models.
[NLP-151] IndicDetect: Evaluating Cross-Lingual LLM -Generated Text Detection for Hindi Telugu and Tamil
【速读】： 该论文旨在解决当前生成式AI文本检测模型在印地语（Hindi）、泰卢固语（Telugu）和泰米尔语（Tamil）等印地语系语言中缺乏有效评估基准的问题，尤其针对现有评测体系在理想化条件下进行测试、难以反映真实场景分布偏移的局限性。其核心解决方案是提出一个名为IndicDetect的通用化评测基准，专门用于评估检测器在现实世界复杂条件下的鲁棒性，涵盖领域迁移（domain shift）、生成器迁移（generator shift）以及对抗性扰动（adversarial perturbation）等多种挑战。该基准包含经过精细匹配的人类撰写文本与大语言模型（LLM）生成文本对，覆盖多个领域与生成器，并采用统一可复现的评估范式，系统性地测试统计型与神经网络型检测方法。实验结果表明，尽管监督式神经检测器在分布内表现良好，但无训练检测方法在面对未知生成器和对抗攻击时性能显著下降；且不同语言间鲁棒性退化程度存在差异，其中印地语在对抗扰动下表现出最严重的性能衰减。研究揭示了当前检测模型在印地语系语言中的主要弱点并非峰值准确率，而是跨分布场景下的鲁棒性不足。IndicDetect通过提供标准化数据划分、评估协议及基线模型，为印地语系脚本下的可信生成式文本检测建立了语言敏感且稳健的研究基础。
链接: https://arxiv.org/abs/2608.29919
作者: Bhaskar Ganesh Devalla,Junchao Wu,Nilesh Dokuparthi,Greeshma Yaluru,Tatiana Muniz Rodriguez,Lidia S. Chao,Derek F. Wong
机构: NLPCT Lab, Faculty of Information Science and Computing, University of Macau(澳门大学信息科学与计算学院自然语言处理与计算技术实验室); Department of Applied Computer Science, SRH University, Germany(德国SRH大学应用计算机科学系); Populus Group, Pittsburgh, United States of America(美国匹兹堡Populus集团)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:The rapid proliferation of LLMs has further heightened the need to develop dependable AI-generated text detection, especially beyond English. Nevertheless, current benchmarks pay little attention to Indic languages and test detectors in idealized settings that do not represent the real world. We present a generalized benchmark for AI-generated text detection in Hindi, Telugu, and Tamil, which we call IndicDetect, designed to assess the robustness of detectors under realistic distribution shifts. IndicDetect comprises highly curated human-written texts matched with LLM-generated counterparts across various domains and generators, and systematically evaluates detectors in the presence of domain shift, generator shift, and adversarial perturbation. Using a single and repeatable evaluation scheme, we evaluate a wide range of statistical and neural detectors. We find substantial robustness failures: supervised neural detectors perform well in-distribution, while training-free methods degrade considerably under unseen generators and adversarial attacks. The severity of these failures varies across languages, with Hindi exhibiting the largest overall degradation under adversarial perturbations. These results highlight that the primary weakness of existing detectors in Indic settings lies in their robustness, not in their peak accuracy. IndicDetect provides standard data splits, an evaluation protocol, and baselines to establish a robust, language-aware foundation for AI-generated text detection in Indic scripts.
[NLP-152] When Less is More: Understanding When Token Filtering Helps and Fails in AI-generated Text Detection EMNLP2026
【速读】： 该论文旨在解决生成式 AI (Generative AI) 生成文本检测中存在的一种反直觉现象：尽管现有零样本检测方法普遍假设更多词元（token）级别的证据有助于提升检测可靠性，但实证研究发现，减少词元数量反而可能带来更优的检测性能（如仅保留40%的词元即可达到最佳效果），且该优势并非普适。其解决方案的关键在于提出熵间隙评分（Entropy Gap Score, EGS）并引入基于 top-k 累积概率的过滤机制作为诊断工具。通过典型集理论分析与熵校准、分布特性量化，研究揭示：对于弱源大语言模型（LLM），低熵词元会因熵校准偏差而系统性地损害检测性能，此时过滤可有效提升效果；而对于强源模型，此类词元危害不显著，过滤反而失效。这一发现首次系统性地揭示了词元级检测中的“双重权衡”——部分词元不仅无信息量，甚至具有系统性危害，从而为构建更鲁棒的生成内容检测方法提供了理论依据与技术路径。
链接: https://arxiv.org/abs/2608.29903
作者: Xiaoyang Han,Lvxiaowei Xu,Ming Cai
机构: Zhejiang University (浙江大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted to EMNLP 2026 (Main Conference)
Abstract:The rapid advancement of large language models (LLMs) has made AI-generated text detection increasingly critical. Existing zero-shot detectors assume that more token-level evidence leads to more reliable detection. However, our empirical study challenges this consensus: fewer tokens sometimes work better, retaining only 40% can yield optimal performance, yet this benefit is not universal. Using the Entropy Gap Score (EGS), we introduce top- k cumulative probability filtering as a diagnostic probe. Across three representative settings, filtering exhibits strikingly different behaviors. We analyze EGS via typical set theory and quantify its dynamics through entropy calibration and distribution analysis. We find that filtering helps for weak source LMs, where low-entropy tokens are harmful, but fails for strong source LMs, where they are not notably harmful. Our work provides the first systematic analysis showing that some tokens are not merely uninformative but systematically harmful due to entropy miscalibration, revealing a two-sided trade-off in token-level detection.
[NLP-153] When History Is Multimodal: Rethinking Context Management for Long-Horizon Agents
【速读】： 该论文旨在解决长时程智能体（long-horizon agents）在持续交互过程中面临的信息记忆膨胀问题，即如何高效地将不断增长的交互历史压缩为有限容量的工作上下文（working context）。现有研究主要依赖文本化编码（textualized histories）进行光学记忆（optical-memory）表示，但普遍假设将上下文渲染为视觉形式会带来显著性能损失，因而需通过监督微调（SFT）、自蒸馏或强化学习等手段弥补差距，这导致两个关键问题未被充分探讨：(i) 在公平、受控条件下，视觉渲染作为上下文管理器的实际表现如何；(ii) 当历史本身具有多模态特性时，视觉载体是否具备原生优势。为此，本文将上下文管理建模为预算约束下的历史转换问题，提出一种基于视觉渲染（Visual Rendering, VR）的表征型上下文管理器，并进一步设计无需训练的VERA（Visual Evidence-Retaining strategy for long-horizon Agents）策略——其核心在于采用确定性渲染机制，在不暴露任何内存操作的前提下，对纯文本任务仍以文本历史生成视觉表示，而在多模态任务中则直接保留原始视觉观测，实现模态保真。实验在四个文本主导与三个多模态基准上验证，结果表明：相较于无压缩基线，VERA在几乎所有任务中减少31.5%–63.1%的非缓存标记（non-cache tokens），在文本任务上达到现有管理器水平，在多模态任务中取得最高准确率，有力支持了“长时程上下文管理应保持模态原生性”的观点。
链接: https://arxiv.org/abs/2608.29897
作者: Jiaqi Su,Cong Pang,Jiawei Hong,Tiankuo Yao,Zixuan Chen,Xin Lou,Lewei Lu
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:Long-horizon agents need a context manager to compress growing interaction histories into a bounded working context, via passive strategies or active strategies that decide how memory is accessed and reorganized. Meanwhile, prior optical-memory work mainly treats pixels as a dense codec for textualized histories, often presupposing that rendering context into optical memory incurs a significant performance drop relative to text, thus coupling this representation with SFT, self-distillation, or reinforcement learning to close this gap, leaving unresolved (i) how visual rendering performs as a context manager under a fair, controlled comparison, and (ii) whether this carrier offers a native advantage when history is inherently multimodal. In this paper, we formulate context management as a budget-constrained history transformation and introduce Visual Rendering (VR) as a representational context manager. Under a shared harness, policy model, trigger, and task domain, we evaluate VR on four text-centric and three multimodal benchmarks against four baselines (No Compression, Discard-All, Sliding Window, Summarization), finding visual memory is a natural carrier of native visual evidence. Building on this finding, we propose VERA (Visual Evidence-Retaining strategy for long-horizon Agents), a training-free context manager built on deterministic rendering with no exposed memory operations: on text-centric benchmarks it renders textual history as VR does, while on multimodal benchmarks it retains native visual observations instead of translating them into text. Across nearly all benchmarks, VERA cuts cumulative non-cache tokens by 31.5%-63.1% versus No Compression, matches existing managers on text-centric tasks, and achieves the highest accuracy among all baselines on multimodal tasks, supporting a modality-preserving view of long-horizon context management.
[NLP-154] En-ViMedNER: An English-Vietnamese Parallel Biomedical Corpus with UMLS Semantic Type Annotations EMNLP2026
【速读】： 该论文旨在解决越南语（Vietnamese）在生物医学命名实体识别（Biomedical Named Entity Recognition, NER）领域缺乏高质量、标准化标注资源的问题。现有基于统一医学语言系统（UMLS）的资源（如MedMentions）均以英文为主，无法支持多语言医疗人工智能应用的发展，尤其制约了低资源语言如越南语在临床决策支持与医学信息抽取等场景中的应用。为此，本文提出En-ViMedNER，首个基于UMLS语义类型（Semantic Type, ST）的英越双语平行生物医学NER语料库，其关键创新在于构建了一个语言中立的跨语言标签空间，通过自动翻译、专家后编辑、大语言模型（LLM）辅助的标签投影以及人工验证与仲裁机制，实现了高质量且可扩展的标注流程。该语料库包含4,392对PubMed摘要、44,892句对及202,949个对齐的实体提及，覆盖21类经适配的MedMentions ST21pv语义类型。研究进一步评估了两种任务设置下的性能：(i) 越南语输入/输出的单语生物医学NER，(ii) 英文输入/越南语输出的跨语言生物医学NER。实验表明，最佳模型在测试集上分别达到52.70和45.44的F1分数。研究成果公开发布了语料库、构建管道与基线模型，为未来越南语生物医学自然语言处理研究提供了重要基础。
链接: https://arxiv.org/abs/2608.29890
作者: Nhu Vo,Phuong Nguyen,Nu Uyen Phuong Le,Inigo Jauregi Unanue,Dung D. Le,Massimo Piccardi,Wray Buntine
机构: VinUniversity(越南维努大学); University of Technology Sydney (悉尼科技大学); Monash University (莫纳什大学)
类目: Computation and Language (cs.CL)
备注: To appear in Proceedings of EMNLP2026
Abstract:Biomedical Named Entity Recognition (NER) is fundamental to healthcare AI applications, including clinical decision support and medical information extraction. While corpora with Unified Medical Language System (UMLS) annotations, such as MedMentions, have driven progress in English biomedical NER, no comparable resource exists for Vietnamese. This paper presents En-ViMedNER, the first English-Vietnamese parallel biomedical NER corpus annotated with UMLS semantic types, which are language-neutral codes providing a shared cross-lingual label space and ensuring direct comparability with existing UMLS-based resources. The corpus contains 4,392 PubMed abstract pairs, 44,892 English-Vietnamese sentence pairs, and 202,949 aligned entity-mention pairs across 21 semantic types adapted from the MedMentions ST21pv dataset. To balance quality and scalability, we have constructed the corpus through automatic translation, expert post-editing, LLM-assisted label projection, and human verification and adjudication. We characterize En-ViMedNER as a large-scale silver-standard corpus with a human-audited and consensus-corrected mini-test subset. We evaluate En-ViMedNER in two settings: (i) Vietnamese-input/Vietnamese-output biomedical NER and (ii) English-input/Vietnamese-output cross-lingual NER. For Vietnamese NER, we benchmark Vietnamese-supervised encoder models, English-supervised multilingual encoder models, and prompt-based LLMs. The best model achieves an F1 score of 52.70 on the test set and 53.78 on the mini-test set. For cross-lingual NER, we benchmark encoder-decoder models and prompt-based LLMs. The best model achieves an F1 score of 45.44 on the mini-test set. We publicly release our corpus, corpus construction pipeline, and baseline models to facilitate future Vietnamese biomedical NLP research.
[NLP-155] VibeJam: A User Study Platform for Web Development with Agents EMNLP2026
【速读】： 该论文旨在解决当前生成式AI（Generative AI）在编程应用中缺乏真实用户行为洞察的问题，尤其是现有自然语言处理（NLP）研究多依赖离线评估，无法捕捉程序员在实际开发过程中与编码代理（coding agents）协作的动态交互。其核心解决方案是提出并发布VibeJam——一个基于浏览器的用户研究平台，支持用户与AI代理协同开发网站。VibeJam的关键创新在于实现了对代理的可定制性，并默认集成开源Aider代理，同时引入差异审查（diff review）、对话（chat）与计划（plan）模式以及实时网站预览功能，以更贴近真实应用场景。通过一项包含55个游戏化网站创建任务的试点研究，结果显示，五名资深AI程序员认为系统趣味性强、操作简便且接近商用工具；13名初级学生使用VibeJam所构建的网站质量显著高于仅依赖代理完成相同任务的情况。该平台已开源，旨在推动后续对编码代理如何有效辅助开发者的研究与扩展。
链接: https://arxiv.org/abs/2608.29889
作者: Nishant Balepur,Connor Baumler,Valerie Chen,Eunsol Choi,Rachel Rudinger,Jordan Boyd-Graber
机构: University of Maryland; New York University; Carnegie Mellon University; Nanyang Technological University
类目: Computation and Language (cs.CL)
备注: EMNLP 2026 (Demo)
Abstract:Programming with AI is increasingly agentic, users prompt LLMs to directly edit their code and review the changes, with adoption growing especially for web development tasks. Despite this growth, most NLP work uses offline evaluation and lacks support for online studies, losing insights into how programmers truly use coding agents. We release VibeJam, a browser-based user study platform for users to collaborate with AI agents to develop websites. VibeJam enables agent customization and uses the open-source Aider agent by default, and to mirror downstream use, we add diff review, chat and plan modes, and live website previews. In a pilot study with 55 released, game-based website creation tasks, five experienced AI programmers rate our system as fun, simple, and resembling commercial tools, while 13 junior students use VibeJam to make websites of higher quality than agents in the same task. We open-source VibeJam to spur extensions and support studies on how coding agents can help users.
[NLP-156] Check The Scoreboard: An Analysis of Scoring Schemes on Multiple-Choice Evaluation EMNLP2026
【速读】： 该论文旨在解决自然语言处理（NLP）领域中多选题问答（MCQA）基准评测普遍采用“正确数量计分”（accuracy）所导致的评价局限性问题。传统准确率评分仅衡量模型回答正确的题目数量，无法全面反映模型在真实教育场景中应具备的高阶认知能力，如排除干扰项、合理放弃不确定作答、置信度校准及自我修正等。为此，论文提出六种受教育测评启发的替代评分方案，其核心在于通过设计更精细的评分机制，引导模型展现超越单纯准确率的能力。关键创新在于：这些新方案不仅显著改变了31个大语言模型（LLM）的性能排名，且相较于改写后的“正确数量”提示，更能预测用户在LLM Arena中实际偏好的模型；同时揭示了不同模型的差异化能力特征——例如GPT-5极少选择弃权并擅长自我修正，而部分开源小模型则频繁弃权且在排除干扰项时犹豫不决。研究表明，采用多样化评分方案可更精准刻画模型的真实能力，为未来将此类评估范式扩展至非多选题任务提供了可行路径。
链接: https://arxiv.org/abs/2608.29887
作者: Nishant Balepur,Paiheng Xu,Wei Ai,Eunsol Choi,Rachel Rudinger,Jordan Boyd-Graber
机构: University of Maryland (马里兰大学); New York University (纽约大学); Nanyang Technological University (南洋理工大学)
类目: Computation and Language (cs.CL)
备注: EMNLP 2026
Abstract:Multiple-choice question answering (MCQA) benchmarks in NLP use number-right scoring (accuracy), but in educational testing, the scoring scheme, the combination of the response mode models follow and the rule for grading responses, is a key design choice that dictates which abilities to reward. We examine how alternatives to number right change what MCQA measures with six education-inspired schemes that assess abilities beyond accuracy: distractor elimination, abstention, confidence calibration, and self-correction. On LLM benchmarks, these schemes: 1) shift rankings of 31 LLMs beyond rephrased number right prompts; 2) better predict the LLMs users prefer in LLM Arena; and 3) reveal distinct model capabilities, like that GPT-5 rarely abstains and readily self-corrects, while weaker open-weight models often abstain and hesitate to eliminate choices. Given the benefits of alternative scoring schemes, we discuss ways to extend them to tasks beyond MCQA.
[NLP-157] Improving Argument Saliency Coverag e in Small LLM s for Long Legal Opinion Summarization via Sequence-Level Distillation EMNLP2026
【速读】： 该论文旨在解决小规模大语言模型（LLM）在长篇法律意见摘要生成任务中难以有效覆盖关键论点（argument saliency）的问题，尤其是在处理长上下文时信息保留能力不足。其核心解决方案是采用基于能力强大的长上下文教师模型进行序列级知识蒸馏（sequence-level distillation），该方法无需人工标注、数据效率高，且能显著提升学生模型对重要论证内容的覆盖率。研究发现，仅需约10个训练摘要即可实现大部分性能提升，凸显了教师模型生成监督信号的强大数据效率；同时，尽管推理链蒸馏（reasoning-chain distillation）具有竞争力，但在与摘要蒸馏结合时带来的增益有限，表明摘要级别的蒸馏已足够实现显著改进。
链接: https://arxiv.org/abs/2608.29884
作者: Mohamed Elaraby,Ahmed Elhady,Diane Litman
机构: University of Pittsburgh (匹兹堡大学); HiTZ Center, University of the Basque Country (UPV/EHU) (巴斯克大学高等研究中心)
类目: Computation and Language (cs.CL)
备注: EMNLP 2026 Main Conference
Abstract:We show that sequence-level distillation from a capable long-context teacher model is a simple, annotation-free, and data-efficient strategy for improving argument saliency coverage in long legal opinion summarization, where small LLMs often struggle to retain the most salient argumentative content. Across student model sizes, distillation consistently surpasses tuning on expert-written summaries in our legal-opinion setting. We further demonstrate that most gains are achieved with as few as ~10 training summaries, highlighting the strong data efficiency of teacher-generated supervision. Finally, we find that summary distillation is sufficient for improvements: reasoning-chain distillation remains competitive with summary-only distillation, but provides marginal benefit when combined with summary supervision.
[NLP-158] ManGo: Manga Active Narrative Grounding Optimization EMNLP2026
【速读】： 该论文旨在解决漫画视觉问答（Manga Visual Question Answering, MVQA）中因信息分散于有序分镜、嵌入文本、重复角色及隐含事件转换而导致的被动页面编码不足问题，核心挑战在于模型需主动识别应关注的分镜、提取关键线索，并判断累积证据是否足以作答。其解决方案的关键是提出一种无监督的主动漫画叙事定位优化框架——ManGo，通过引入主动叙事草图（Active Narrative Sketching, ANS），实现分镜的迭代选择、精炼的可解释线索提取以及终止决策，从而在生成答案前构建紧凑且指向问题的证据草图。为在无标注答案或推理路径的情况下优化该行为，ManGo采用多轮ANS采样并结合双奖励机制：基于列表级自排序的答案偏好（answer preference）与基于稳定分镜序列轨迹的路径一致性（path consistency），通过组相对策略训练（group-relative policy training）联合优化该组合奖励，促使模型同时提升最终答案准确率与支撑性分镜证据路径的质量。实验表明，该方法在标准漫画理解基准上实现了多种设置下的最先进性能。
链接: https://arxiv.org/abs/2608.29865
作者: Hao Qiu,Junyan Wang,Zheyuan Liu,Lei Fan,Hong Jia,Lianbo Guo,Zhulin Tao
机构: Communication University of China(中国传媒大学); Australian Institute for Machine Learning, Adelaide University(阿德莱德大学机器学习研究所); University of New South Wales(新南威尔士大学); University of Auckland(奥克兰大学); Huazhong University of Science and Technology(华中科技大学)
类目: Computation and Language (cs.CL)
备注: 16 pages, 9 figures, 7 tables. Accepted to Findings of the Association for Computational Linguistics: EMNLP 2026
Abstract:Manga visual question answering requires models to answer questions over panel-based visual narratives, where relevant evidence is distributed across ordered panels, embedded text, recurring characters, and implicit event transitions. This structure makes passive page encoding insufficient, as the model must identify which panels to inspect, what clues to retain, and when the accumulated evidence is sufficient for answering. We propose ManGo (Manga Active Narrative Grounding Optimization), an unsupervised framework for active manga visual question answering. ManGo introduces Active Narrative Sketching (ANS), which iteratively selects panels, extracts concise grounded clues, and decides when to stop, forming a compact question-directed evidence sketch before answer generation. To optimize this behavior without human-annotated answers or rationale paths, ManGo samples multiple ANS rollouts and applies group-relative training with two rewards: answer preference from listwise self-ranking and path consistency from stable ordered panel trajectories. The combined reward is optimized with group-relative policy training, encouraging the model to improve both final answers and the panel-level evidence paths that support them. Experiments on standard manga understanding benchmarks show that ManGo achieves state-of-the-art performance across different settings.
[NLP-159] GenRubric: Self-Evolving Rubric Generation for Scalable LLM Evaluation
【速读】： 该论文旨在解决大语言模型（LLM）在开放式任务评估中因依赖查询特定评价标准（query-specific rubrics）而导致的评估要求不明确、可审计性差的问题。现有方法要么依赖昂贵的人工专家编写评价标准，要么依赖推理阶段的精炼或外部监督，难以实现高效、可扩展的自动评价体系。其解决方案的关键在于提出GenRubric——一种无需额外人工标注即可实现自演化的评价标准生成框架。该框架基于“评价标准诱导的自一致性”（rubric-induced self-consistency）原理：对同一查询独立采样的多个评价标准可提供其潜在评估需求的部分视角，而一个全面的评价标准应能生成在这些互补视角下均表现一致的响应。通过强化学习机制，将跨标准的综合性信号与群体级及准则级奖励相结合，以优化评价标准质量。实验表明，经自演化训练的模型在人类标注的评价标准基准上显著提升了生成评价标准与专家标准之间的一致性，且该优势可泛化至未见领域，验证了自演化评价标准生成在实现可扩展、查询特定的LLM评估中的潜力。
链接: https://arxiv.org/abs/2608.29856
作者: Yifan Chen,Haitao Li,Qingyao Ai,Fengbin Zhu,Tat-Seng Chua,Min Zhang,Yiqun Liu
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:Large language models are increasingly used as scalable evaluators for open-ended tasks. However, many LLM judges derive query-specific criteria during scoring, leaving the evaluation requirements insufficiently specified and their coverage difficult to audit. Query-specific rubrics make these requirements explicit, but expert-written rubrics are costly to construct, while existing automatic methods typically rely on inference-time refinement or external supervision. We introduce GenRubric, a self-evolving framework that improves rubric generation from unlabeled queries without requiring additional human annotations during self-evolution. Our approach is based on rubric-induced self-consistency: independently sampled rubrics for the same query provide partial views of its latent evaluation requirements, and a comprehensive rubric should induce a response that generalizes across these complementary evaluation views. We implement this principle through reinforcement learning, combining a cross-rubric comprehensiveness signal with group-level and criterion-level rewards for rubric quality. We train GenRubric models at 4B, 8B, and 14B scales across multiple domains. Experiments on human-annotated rubric benchmarks show that self-evolution improves the agreement between evaluations induced by generated rubrics and those induced by expert-written rubrics. The improvements further generalize to held-out domains, demonstrating the potential of self-evolving rubric generation for scalable and query-specific LLM evaluation. Code and models are publicly available at this https URL.
[NLP-160] ContextBias: Controlled Evaluation of Bias Persistence Under Context Shift in Text-to-Image Models EMNLP2026
【速读】： 该论文旨在解决生成式图像模型中角色（role）与视觉属性之间关联的稳定性问题，特别是这些关联在不同提示上下文（prompted context）下是否会发生变化。研究发现，尽管提示语义上下文发生改变，角色相关的视觉特征（如性别化服饰、职业工具及外貌特征）仍高度稳定存在，且在非相关上下文中甚至出现跨角色属性集中度上升的现象（汇总偏差指数 +0.047），表明刻板印象具有显著的上下文鲁棒性。其解决方案的关键在于提出ContextBias评估框架和ContextBench基准数据集，通过1,656个语义受控提示覆盖92种职业角色，系统分离上下文变化对角色-视觉表征关系的影响。实验结果揭示了传统无上下文评估无法捕捉的隐性刻板偏见持续性，强调在偏见评估中引入可控的上下文变异的重要性。
链接: https://arxiv.org/abs/2608.29847
作者: Shaghayegh Kolli,Sina Emami,Moreno D’Incà,Pouyan Nejadi,Nicu Sebe,Massimiliano Mancini,Jana Diesner
机构: Technical University of Munich (慕尼黑工业大学); Munich Center for Machine Learning (慕尼黑机器学习中心); Munich Data Science Institute (慕尼黑数据科学研究所); University of Trento (特伦托大学); Orreco
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026
Abstract:Text-to-image models learn associations between concepts - in the case of this paper, people’s professions, which we refer to as roles - and visual attributes. These associations can underpin many observed forms of stereotypical bias. A key open question in this area is whether these associations are stable or change when visual representations of people in professional roles are placed in different prompted contexts. We introduce ContextBias, a controlled evaluation framework, and ContextBench, a benchmark spanning 92 roles and 1,656 semantically controlled prompts, designed to isolate the effect of contextual variation on role-linked visual representations. Evaluating four state-of-the-art models on 66,240 generated images, we find that placing a role in a semantically unrelated context does not suppress role-linked attributes; instead, cross-role attribute concentration increases (pooled BI +0.047 ). Demographic cues, characteristic garments, and role-specific tools remain highly prevalent across context-free, related, and unrelated conditions, and are robust to semantic prompt reformulation. Scene composition and camera framing show the greatest context-sensitivity. These findings reveal a form of stereotypical persistence that remains largely invisible to context-free evaluations, highlighting the need for controlled contextual variation in bias benchmarking. Code and dataset: this https URL , this https URL
[NLP-161] Influence-Directed Distillation: Solving the Diversity Bottleneck in Sampled-Token On-Policy Distillation
【速读】： 该论文旨在解决生成式 AI 模型在样本令牌有监督蒸馏（Sampled-token on-policy distillation, OPD）过程中存在的多样性蒸馏失败问题：尽管学生模型的 pass@1 性能持续提升，但其 pass@k 性能却趋于饱和，无法有效继承教师模型的输出多样性。其核心问题是现有方法在更新过程中导致熵收缩（entropy contraction），从而抑制了学生模型探索能力的提升。为解释此现象，作者提出“一阶局部熵影响”（First-Order Local Entropy Influence）这一符号化的一阶代理指标，可将每次更新对熵的影响解耦为教师-学生对数概率差与学生局部概率结构两部分，并实证揭示熵收缩与负影响位置之间的关联。基于此，论文提出一种新型解决方案——影响导向自适应有监督蒸馏（Influence-Directed Adaptive On-Policy Distillation, IDA-OPD）：该方法摒弃高成本的全词汇表前向KL散度目标，仅利用教师对采样令牌的对数概率，通过保留熵扩展型更新、以发散自适应优势衰减替代熵收缩型更新的方式，在不依赖全词汇表教师信息的前提下，显著提升 pass@k 性能，有效继承教师模型的多样性，同时在计算开销上优于最强的教师指导型方法，并保持原始 OPD 的 pass@1 表现。
链接: https://arxiv.org/abs/2608.29846
作者: Run Yang,Runpeng Dai,Jie Sun,Jielei Zhang,Fan Zhou,Hongtu Zhu,Peiyi Li,Longwen Gao
机构: 未知
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Sampled-token on-policy distillation (OPD) efficiently transfers capabilities from teacher to student using student-generated tokens, requiring teacher probabilities only for sampled tokens. Yet it frequently suffers from diversity distillation failure: the student’s pass@1 improves while its pass@ k plateaus, failing to inherit the teacher’s diversity. To explain this, we introduce First-Order Local Entropy Influence, a signed first-order proxy that decouples each update’s entropy effect into the teacher–student log-probability gap and the student’s local probability structure, and empirically links entropy contraction to negative-influence positions. Motivated by this, we propose Influence-Directed Adaptive On-Policy Distillation (IDA-OPD): rather than relying on costly full-vocabulary Forward-KL objectives, it preserves entropy-expanding updates while replacing entropy-contracting ones with divergence-adaptive advantage shrinkage, using only the teacher’s sampled-token log-probability. Experiments on reasoning-oriented distillation show IDA-OPD consistently improves pass@ k , inheriting the teacher’s diversity through distillation, matches the strongest teacher-informed methods at strictly lower cost, and broadly maintains vanilla OPD’s pass@1, all without full-vocabulary teacher information.
[NLP-162] SkillForge: Compositional Skill Synthesis with Verification-in-the-Loop for Generating Formally Verified Dafny Programs EMNLP2026
【速读】： 该论文旨在解决从自然语言生成形式化验证代码的难题，现有方法或在验证失败时无法回溯地单次生成代码，或依赖非确定性且不透明的开放式智能体推理。其解决方案的关键在于提出SKILLFORGE框架，该框架将形式化代码合成分解为一系列原子化、可复用的“技能”（skill），每个技能针对特定子任务（如规格推断、函数体生成、不变量构造、错误诊断或定向修复），并由提示模板、工具绑定和可判定的成功标准定义。一个基于验证驱动的调度器（harness）协调这些技能：将候选代码提交至Dafny验证器，将验证失败归类为结构化类别，确定性地路由至相应的修复技能，并迭代执行直至证明形式正确或耗尽预算。在一组精心构建的自然语言到Dafny规格对的基准测试中，SKILLFORGE显著优于当前最先进的智能体方法（包括ReAct型代理、基于MCTS的修复及强化学习引导的验证）和传统迭代基线，同时具备更低的令牌消耗和延迟。消融实验表明每一项技能均贡献显著，且调度器在多数情况下首次尝试即完成验证，展现出快速收敛能力。
链接: https://arxiv.org/abs/2608.29841
作者: Yanming Liu,Xinyue Peng,Jiannan Cao,Xinyi Wang,Jinbo Su
机构: Zhejiang University (浙江大学); Southeast University (东南大学); MIT (麻省理工学院); Weixin AI Lab (微信AI实验室); Renmin University (中国人民大学)
类目: Computation and Language (cs.CL); Programming Languages (cs.PL)
备注: Accepted by EMNLP 2026
Abstract:Generating formally verified programs from natural language remains challenging: existing approaches either produce code in a single pass without recourse when verification fails, or rely on open-ended agentic reasoning that is non-deterministic and opaque. We introduce SKILLFORGE, a framework that decomposes formal code synthesis into a library of atomic, reusable skills, each targeting a specific subtask such as specification inference, body synthesis, invariant generation, error diagnosis, or targeted repair, and defined by a prompt template, tool binding, and decidable success criterion. A verification-driven harness orchestrates these skills: it submits candidates to the Dafny verifier, diagnoses failures into structured categories, deterministically routes to the appropriate repair skill, and iterates until formal correctness is proved or a budget is exhausted. On a curated benchmark of natural language to Dafny specification pairs, SKILLFORGE substantially outperforms both state-of-the-art agentic approaches (including ReAct-style agents, MCTS-based repair, and RL-guided verification) and traditional iterative baselines, while requiring fewer tokens and lower latency. Ablation studies confirm that every skill contributes measurably, and the harness converges rapidly with the majority of programs verified on the first attempt.
[NLP-163] EVAR: Evidence-Validated Hypothesis Admission for Budget-Aware Narrative Reasoning EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在处理非交互式长篇叙事时，因推理过程中引入未经验证的中间假设而导致结论缺乏事实依据的问题。尤其当关键证据分散于文本不同部分时，模型易产生逻辑污染，影响推理可靠性。其解决方案的关键在于提出EVAR（Evidence-Validated Hypothesis Admission），一个面向预算感知的叙事推理框架。EVAR通过将叙事结构化为不可变的、源链接的原子命题证据库，并基于未解疑点与不确定性信号动态分配推理预算；在推理精炼阶段，直接针对未决问题生成候选假设，构建条件依赖的验证挑战，并在锁定的证据库中逐一验证，仅允许被支持的假设进入答案支撑状态，无法验证或矛盾的假设则被隔离或丢弃；同时采用基于充分性的停止机制，避免冗余推理。该方法在NarraCrime及多个公开推理基准上均实现了任务性能与证据忠实度的提升，同时保持可控制的推理成本。
链接: https://arxiv.org/abs/2608.29835
作者: Peilin Liu,Zhiquan Ji,Jinglong Ping
机构: 未知
类目: Computation and Language (cs.CL)
备注: Accepted to the Main Conference of EMNLP 2026. 16 pages, 3 figures
Abstract:Large language models (LLMs) often produce fluent but weakly grounded conclusions when reasoning over non-interactive, long-form narratives. A central failure mode is that unsupported intermediate hypotheses can enter the reasoning trajectory and contaminate subsequent inference, especially when evidence is scattered across distant parts of the story. To address this problem, we propose EVAR, an evidence-validated hypothesis admission framework for budget-aware narrative reasoning. EVAR first compiles the narrative into an immutable evidence store of source-linked atomic claims and assigns an instance-specific inference budget from unresolved gaps and uncertainty signals. During refinement, EVAR directly proposes candidate hypotheses for unresolved gaps, constructs hypothesis-conditioned validation challenges, and verifies each candidate against the locked store before admission: supported hypotheses enter the answer-supporting state, unverifiable ones are quarantined, and contradictory ones are discarded. A sufficiency-based stopping mechanism further avoids unnecessary refinement. Experiments on NarraCrime and multiple public reasoning benchmarks show that EVAR improves both task performance and evidence faithfulness while maintaining controllable inference cost.
[NLP-164] A2Agent : Action-Aware Reinforcement Learning for Repository-Level Code Localization Agents EMNLP2026
【速读】： 该论文旨在解决自动化软件工程中代码区域定位任务中存在的关键问题：现有方法依赖稀疏的轨迹级信号，难以准确识别每轮操作（per-turn action）的有效性，导致在探索过程中虽能发现正确代码区域，却无法有效“提交”（commit）这些结果。其解决方案的核心在于提出一种面向操作的强化学习方法，通过引入逐轮奖励序列（per-turn reward sequence），同时奖励代码区域的发现与提交行为，并结合操作级优势估计（action-level advantage estimation）机制，将相同探索上下文下的多轮操作进行分组，从而隔离并精确分配每个操作的贡献度。该方法显著提升了代码定位的准确性与稳定性，在SWE-Bench Verified和SWE-Bench Pro基准上分别较当前最优方法（SOTA）平均F1提升1.58%和8.55%，且4B规模模型性能超越高达8倍参数量的基线模型。
链接: https://arxiv.org/abs/2608.29831
作者: Doyeon Kim,Suyoung Bae,Yumin Lee,Jee-Hyong Lee
机构: Sungkyunkwan University (成均馆大学); College of Computing and Informatics
类目: Computation and Language (cs.CL); Software Engineering (cs.SE)
备注: Accepted to EMNLP 2026 Main
Abstract:Localizing issue-relevant code regions is a critical step in automated software engineering. However, due to their reliance on sparse trajectory-level signals, existing methods cannot identify which per-turn actions are effective and often discover correct code regions during exploration but fail to commit them. To address these limitations, we propose an action-aware reinforcement learning method that combines a per-turn reward sequence rewarding both the discovery and commitment of gold code regions with an action-level advantage estimation scheme that isolates each action’s credit by grouping turns sharing the same exploration context. Extensive evaluations show that our method improves the average F1 over the state-of-the-art (SOTA) by 1.58% on SWE-Bench Verified and 8.55% on SWE-Bench Pro, with our 4B model outperforming baselines up to 8x larger. Our code is available at this https URL.
[NLP-165] Do LLM s Change Their Minds Like Humans? Diagnosing Human–LLM Divergence in Single-Turn Persuasion Judgments
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在社会模拟中作为人类参与者代理时，其信念更新机制是否与人类一致的问题。研究发现，尽管LLMs在某些强说服线索上与人类表现出一定一致性，但在细微层面存在显著差异：人类更易受新颖内容和强势语言影响，而LLMs则更依赖主题相似性和表面格式特征。此外，LLMs对情感诉求的权重不足，却过度关注可信度信号，且论证命题类型对分歧程度无显著影响。值得注意的是，从第一人称角色扮演转向第三人称观察会普遍增强模型的抗说服性，且该效应因说服策略和文本特征而异。因此，该研究的关键在于揭示了LLMs与人类在处理说服性话语时的根本性结构差异，强调不可将LLM的判断视为人类信念更新的忠实代理，为未来构建更符合人类认知逻辑的社会模拟系统提供了重要依据。
链接: https://arxiv.org/abs/2608.29803
作者: Lin Chen,Yitong Chen,Yong Li
机构: Network Science Institute, Northeastern University (东北大学网络科学研究所); Department of Physics, Northeastern University (东北大学物理系); Department of Electronic Engineering, BNRist, Tsinghua University (清华大学电子工程系，清华信息科学与技术国家实验室)
类目: Computers and Society (cs.CY); Computation and Language (cs.CL)
备注:
Abstract:Large language models (LLMs) are increasingly deployed as proxies for human participants in social simulations, yet whether they update their beliefs in response to persuasive arguments, as humans do, remains poorly understood. We conduct a systematic comparison using a naturally occurring online persuasion corpus in which original posters explicitly verify whether a reply changed their view. Our results show that LLMs achieve only slight agreement with humans (Cohen’s kappa ranging from 0.079 to 0.178). Content-level analyses show that humans and LLMs agree on the strongest persuasion cues but diverge on finer ones: humans are more swayed by novel content and assertive language, whereas LLMs favor topical similarity and surface-level formatting. At the level of persuasion strategy, LLMs underweight emotional appeals and overweight credibility signals relative to humans, while the type of proposition under debate exerts no measurable effect on the degree of divergence. Furthermore, switching from first-person role-playing to third-person observation shifts all models toward greater resistance to persuasion, with the effect varying across persuasion strategies and textual features. These findings highlight the risk of treating LLM judgments as faithful proxies for human belief updating and point to structural differences in how LLMs and humans process persuasive discourse. Our code is available at this https URL.
[NLP-166] R2A: Learning Persona Policies Through Persona Representation Learning and Runtime Alignment
【速读】： 该论文旨在解决静态人格（Persona）提取在不同任务中表现不一致的问题，即同一人格行为在某些情境下有益而在另一些情境下有害，导致模型性能波动。其核心解决方案是提出“人格选择—实现框架”（Persona Selection–Realization Framework），将行为生成过程建模为潜在的人格状态，并将其分解为两个阶段：人格选择（Persona Selection）与人格实现（Persona Realization）。该框架定义了两个关键差距：在人格选择阶段的“选择差距”（Selection Gap）和在人格实现阶段的“实现差距”（Realization Gap）。基于此框架，论文进一步提出R²A（Two-stage Persona Policy Learning）方法，包含两个阶段：第一阶段为“人格表征学习”（Persona Representation Learning），通过结构化的“谁—如何—做什么”（Who–How–What）表述来编码目标人格的客观特征、条件性行为准则及轨迹层面的表现；第二阶段为“运行时对齐”（Persona Runtime Alignment），通过任务反馈隐式地联合校准行为选择与轨迹实现，消除显式人格指定的影响。实验结果表明，R²A在涵盖问责专业人格（Accountable-Professional Persona）四大原则的12个评估场景中，整体优于基线模型和静态人格提取方法；消融实验进一步验证，人格表征学习对于防止运行时对齐产生行为失衡、实现更稳定的人格策略学习至关重要。
链接: https://arxiv.org/abs/2608.29798
作者: Mohan Zhang,Chengsong You,Xiaoyu Cao,Zhen Sun,Xiaohan Jia,Junwei Zhou,Yongchao Chen
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 17 pages, 6 figures
Abstract:The same Persona behavior can be beneficial in one context but harmful in another, causing static Persona elicitation to perform inconsistently across tasks. We introduce the Persona Selection–Realization Framework, which models behavior generation through a latent Persona state and decomposes it into Persona Selection and Persona Realization. The discrepancies between static Persona elicitation and an ideal Persona policy in these two components define the Selection Gap and Realization Gap, respectively. Building on this framework, we propose R ^2 A, a two-stage approach for learning Persona policies. Persona Representation Learning uses structured Who–How–What presentations to encode the target Persona’s objective, conditional behavioral principles, and trajectory-level manifestations. Persona Runtime Alignment then removes the explicit Persona specification and jointly calibrates behavior selection and trajectory realization using task feedback. Across 12 evaluation settings covering the four principles of the Accountable-Professional Persona studied in this work, R ^2 A overall outperforms both the base model and static Persona elicitation. Ablation results further show that Persona Representation Learning is critical for preventing Runtime Alignment from producing behaviorally imbalanced policies and for achieving more stable Persona policy learning.
[NLP-167] HiVe: Beyond Static Prompts for Multitask Learning via Hierarchy-based Vertical Mixture-of-Experts EMNLP2026
【速读】： 该论文旨在解决现有提示调优（Prompt Tuning）方法在提示结构设计上的局限性，即多数方法采用扁平化提示结构或固定组合的层次化结构，导致提示难以根据输入内容进行自适应的精细化调整。其解决方案的关键在于提出一种名为HiVe的新框架，通过在训练阶段利用任务间的关联性构建多层级提示层次结构，并在推理阶段引入垂直混合专家（Vertical Mixture-of-Experts, V-MoE）机制，实现基于输入依赖的动态提示组合与专业化生成。该方法能够根据不同输入需求灵活选择提示的细化层级，从而显著提升模型在多样化任务上的表现。
链接: https://arxiv.org/abs/2608.29790
作者: HyeonJik Bae,Minyeol Kim,Susik Yoon
机构: Korea University(韩国大学)
类目: Computation and Language (cs.CL)
备注: 16 pages, 6 figures. Accepted to the EMNLP 2026 Main Conference
Abstract:As large language models (LLMs) continue to scale, parameter-efficient fine-tuning (PEFT) has become a practical alternative to full-parameter adaptation. Prompt tuning is effective, but existing approaches either use flat prompt structures or hierarchical structures with fixed prompt composition, limiting adaptive prompt specialization. To address this limitation, we propose HiVe, a prompt tuning framework that models prompts at multiple levels and enables input-dependent specialization. HiVe constructs a prompt hierarchy by leveraging inter-task relationships during training, and employs a vertical mixture-of-experts (V-MoE) mechanism at inference time to compose prompts up to the level of specialization required for each input. Experiments show that HiVe consistently outperforms strong prompt tuning baselines across diverse tasks.
[NLP-168] ReTrace: Rejected-Trajectory Conditioning for Speculative Decoding
【速读】： 该论文旨在解决自回归语言模型推理中，传统推测解码（speculative decoding）在首次验证失败后，丢弃已生成且验证失败的候选后缀所导致的计算资源浪费问题。其核心挑战在于：尽管这些被拒绝的位置可能存在与目标序列对齐的语义和结构信息，但现有基于前缀的验证机制无法有效利用这些信息。为此，作者提出ReTrace方法，其关键创新在于引入受拒轨迹条件化（rejected-trajectory conditioning），即在新一轮生成中，将前一轮被拒绝的后缀作为条件输入，而非仅依赖初始掩码占位符进行生成。具体而言，ReTrace保留被拒绝后缀的隐藏表示，通过门控残差融合将其与下一阶段的草稿块对齐，并利用同一验证轮次中的目标感知修正信号进行优化，从而实现跨轮次的信息复用。该方法不改变目标侧验证过程，保持了推测解码的无损性，且无需额外前向传播，因此可无缝集成至现有框架。实验表明，ReTrace在Qwen3模型上显著提升了平均接受长度与端到端解码速度，在数学推理、代码生成和开放式对话任务中均展现出一致性能增益，且与现有草稿生成改进技术具有高度正交性，具备良好的可组合性。
链接: https://arxiv.org/abs/2608.29748
作者: Luxi Lin,Zhanpeng Zeng,Shuang Peng,Songwei Liu,Rongrong Ji
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:Speculative decoding accelerates autoregressive language model inference by having a lightweight draft model propose multiple candidate tokens, which are then verified in parallel by a larger target model. However, after the first rejection, standard prefix-based verification discards the remaining draft suffix, so the computation spent generating and verifying those positions does not contribute to decoding progress. Focusing on DFlash, we show that rejected positions in a rejected suffix may still align with the target continuation, indicating that the draft model can retain useful semantic and structural information despite local token-level errors. Motivated by this observation and inspired by conditional diffusion, we introduce~\textbfReTrace, a rejected-trajectory conditioning method that conditions each draft block on the rejected suffix from the previous round rather than generating it from fresh mask placeholders alone. ReTrace retains the hidden representations of the rejected suffixes, aligns them with the next draft block, refines them using target-aware correction signals from the same verification pass, and admits them into the drafter’s input embeddings through gated residual fusion. Because rejected tokens are never committed and target-side verification remains unchanged, ReTrace preserves the lossless property of speculative decoding without requiring an additional model forward pass. Experiments with Qwen3 models across mathematical reasoning, code generation, and open-ended dialogue demonstrate that ReTrace consistently improves average acceptance length and end-to-end decoding speed over its DFlash backbone. By introducing cross-round conditioning without modifying within-round proposal generation, ReTrace is largely orthogonal to existing drafting improvements and might be combined with them for further gains.
[NLP-169] Evaluating the Capabilities of LLM s for Persuasive Dialogue
【速读】： 该论文旨在解决生成式 AI（Generative AI）在说服性对话中表现出的“表面说服力”与实际论证有效性之间的脱节问题。其核心问题是：尽管大型语言模型（LLM）生成的文本在主观上显得极具说服力，但这是否意味着其论点在逻辑和论证结构上也具备优势？为解决此问题，研究提出了一种基于形式化论证理论的多智能体对话平台 Persuasio，通过引入形式化论证框架对辩论中的论点进行客观评判，从而区分主观说服力与正式论证强度。该方案的关键在于将辩论过程建模为可形式化评估的论证交互，利用自动仲裁机制结合大规模众包双人对比判断，实现对辩论双方在逻辑严谨性上的量化评估。实验结果表明，尽管 LLM 在主观排名中占据主导地位，但在形式化论证评价中表现显著逊于人类，且多智能体及检索增强变体进一步加剧了这一差距，揭示出当前生成式 AI 在说服性对话中存在系统性地以修辞流畅性掩盖论证薄弱性的现象。
链接: https://arxiv.org/abs/2608.29738
作者: Jordan Robinson,Angus R. Williams,Katie Atkinson,Anthony G. Cohn
机构: University of Liverpool(利物浦大学); The Alan Turing Institute(艾伦·图灵研究所); University of Leeds(利兹大学)
类目: Computation and Language (cs.CL)
备注: 19 pages, 7 figures, 4 tables
Abstract:Large language models (LLMs) can generate apparently highly persuasive text, but does sounding persuasive mean arguing well? We introduce \textscPersuasio, a multi-agent dialogue platform grounded in a formal argumentation-based theory of persuasion dialogues that adjudicates logical winners during free-text debates. Using this system, we generated 192 debates on a UK political topic between humans and LLMs, and evaluated 22 interlocutors through both automated adjudication and 9,702 crowdsourced pairwise judgements across 1,386 annotation instances. We observed a consistent decoupling between subjective and formal persuasiveness: LLMs dominated the subjective ranking yet performed substantially worse under argumentation-theoretic adjudication, where humans remained competitive. Multi-agent and retrieval-augmented variants further widened this divergence. These findings reveal a systematic gap between rhetorical fluency and formal argumentative strength in LLM-based persuasive dialogues.
[NLP-170] Higher-Dimensional Rotary Position Embedding EMNLP2026
【速读】： 该论文旨在解决旋转位置编码（Rotary Position Embedding, RoPE）在长序列建模中因采用独立的二维旋转、块状结构及通道解耦设计而导致的深层通道混合能力不足与跨通道鲁棒性差的问题。其核心解决方案是提出高维旋转位置编码（HD-RoPE），将RoPE从独立的二维旋转扩展至更高维度旋转，并引入Paley-I正交基，以实现每个旋转子空间内均衡、各向同性且密集的相位混合。这一设计显著增强了通道间的耦合能力与旋转自由度，同时保持了正交稳定性与相对位置闭包性质。此外，HD-RoPE无需引入额外可训练参数，具备良好的工程优化效率。大量实验验证表明，HD-RoPE在多种主流基准测试中均显著优于标准RoPE，无论在长上下文还是短上下文场景下均展现出更优性能。
链接: https://arxiv.org/abs/2608.29715
作者: Yixing Li,Ruobing Xie,Yudong Zhang,Yushi Bai,Samm Sun,Yu Cheng
机构: The Chinese University of Hong Kong(香港中文大学); Tencent Hunyuan(腾讯混元); Tsinghua University(清华大学); University of Macau(澳门大学)
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026
Abstract:Transformers rely on position embedding mechanisms in long context modeling in most cases. Rotary Position Embedding (RoPE) embeds positional information with independent 2D rotations, forming relative position terms in self-attention. However, its pairwise, block-based, and decoupled structure limits deep mixing and robustness across channels. We propose HD-RoPE, which extends RoPE from independent 2D rotations to higher-dimensional rotations and introduces a Paley-I orthogonal basis to obtain balanced, isotropic, and dense phase mixing within each rotation subspace. This significantly enhances channel coupling and rotational degrees of freedom while maintaining orthogonal stability and the relative position closure property. Furthermore, HD-RoPE is easily optimized for engineering efficiency without introducing additional trainable parameters. We have conducted extensive evaluation results demonstrating that HD-RoPE achieves significant performance improvements over standard RoPE across various popular benchmarks and in both long and short contexts.
[NLP-171] DVBench: Benchmarking MLLM s for Understanding Dynamic Charts and Narratives in Data Videos
【速读】： 该论文旨在解决多模态大模型（MLLMs）在处理动态结构化视觉信息时评估体系割裂的问题，尤其针对数据视频（data videos）这一融合动态图表与结构化叙事的叙事媒介，现有评估方法难以全面衡量模型对时序演进的结构化视觉内容的理解能力。其解决方案的关键在于构建DVBench——首个专注于数据视频理解的基准测试，通过将数据视频理解任务分解为五个维度，系统性地评估模型性能；该基准包含300个真实世界数据视频及1,000条经人工验证的问答对，采用半自动化流程确保数据质量。实验表明，Gemini-3.1-Pro在整体表现上最优，而Kimi-k2.5是表现最强的开源模型，同时研究揭示了开源模型性能并不严格随参数规模增长、叙事能力与视觉理解能力之间无必然关联等关键现象。细粒度分析与消融实验进一步识别出各维度下的模型弱点，并量化了帧配置与字幕输入的影响，为未来多模态大模型在动态结构化视觉理解方向的发展提供了重要指导。
链接: https://arxiv.org/abs/2608.29711
作者: Bomiao Wang,Zekai Shao,Jiexiang Lan,Xiaoliang Fu,Xingchen Zeng,Siming Chen
机构: Fudan University (复旦大学); The Hong Kong University of Science and Technology (Guangzhou) (香港科技大学(广州))
类目: Computation and Language (cs.CL)
备注:
Abstract:While MLLMs have made significant strides in chart comprehension and video understanding, current evaluations largely isolate these capabilities, leaving a critical gap in understanding temporally evolving structured visual information. To address this gap, we introduce DVBench, a benchmark for evaluating MLLMs on data videos, a storytelling medium that integrates dynamic charts with structured narratives. We decompose data video understanding into five dimensions. DVBench comprises 300 real-world data videos and 1,000 human-verified QA pairs curated through a rigorous semi-automated pipeline. Extensive evaluations of nine MLLMs show that Gemini-3.1-Pro achieves the best overall performance, while Kimi-k2.5 is the strongest open-source model. We further identify two notable phenomena: open-source model performance does not scale strictly with parameter size, and narrative proficiency does not guarantee visual capability. Fine-grained analyses and ablation studies further reveal dimension-specific weaknesses and the effects of frame configurations and subtitle inputs, informing future MLLM development. DVBench is publicly available at this https URL.
[NLP-172] he Depth Flow of Token Representations Is Nonlinear and Does Not Descend Its Own Density
【速读】： 该论文旨在揭示大语言模型（如Pythia-160M和Pythia-410M）中词元（token）表示在各层间演化所形成的动态流的内在机制，解决的问题是：传统线性映射作为网络层近似时无法准确捕捉词元表征非线性演变的本质。其核心解决方案在于将词元表示的演化过程建模为基于语料库均值轨迹的离散朗之万（discrete Langevin）动力学系统，并通过Kramers–Moyal估计器对漂移项进行高阶拟合。研究发现，该动态流的漂移项具有显著的二次非线性特征，优于线性映射在所有层间的预测表现；更重要的是，该流并不沿自身对数密度下降，而是遵循一个非密度相关的势能函数，且其旋转分量占可解释漂移的4%至45%，表明流动具有强环流特性——具体表现为词元的角排名（angular rank）在全部十三层中保持不变，而范数排名被随机化、浓度排名则在最后一层发生反转，揭示了模型内部信息保留与变换的深层结构。
链接: https://arxiv.org/abs/2608.29706
作者: Alexandre Quemy
机构: Hother Labs
类目: Computation and Language (cs.CL)
备注:
Abstract:A token’s representation is carried through the network layer by layer. The whole vocabulary carried together forms a flow. We fit this flow’s equation of motion as a discrete Langevin model over corpus-mean trajectories of Pythia-160M and Pythia-410M, and score the predicted steps on held-out tokens. Linear maps are often used as cheap surrogates for a layer. The flow they summarize is not linear: a quadratic drift beats the linear linear map at every transition of both models, and the Kramers–Moyal estimator agrees wherever its neighborhoods stay local. We then characterize the flow further. First, we show that it does not descend its own log-density. The drift instead descends a potential that is not the density. Second, the rotational component is not negligible, 4 to 45% of the explainable drift, and the circulation shows in what the flow preserves: a token keeps its angular rank across all thirteen layers while its norm rank is shuffled and its concentration rank is reversed by the last block. Subjects: Computation and Language (cs.CL) Cite as: arXiv:2608.29706 [cs.CL] (or arXiv:2608.29706v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2608.29706 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-173] A Hub of Short Rows Inflates Intrinsic Dimension Estimation of Token Embeddings NEURIPS2026
【速读】： 该论文旨在解决生成式 AI（Generative AI）中词元嵌入表（token-embedding table）的内在维度（intrinsic dimension, ID）估计偏差问题。研究表明，嵌入表靠近原点处存在一个由短行组成的密集簇（hub），该簇导致最近邻型内在维度估计算法（如TwoNN）产生显著高估：由于测度集中效应，任意词元更接近该中心簇而非其他词元，其前两个邻居均为近似等距的簇内行，从而误导算法报告远高于真实内在维度的数值。这种偏差表现为单个词元测量时维度呈重尾分布，且随模型参数量增加而上升。然而，当移除该“枢纽”（hub）后，所有十一组模型（从GPT-2到K3、GLM-4.7）的内在维度均收敛至狭窄范围（10–17），表明枢纽是导致维度膨胀的关键因素——仅需数百行即可完全扭曲估计结果。研究进一步验证，对原始数据进行行归一化处理可获得与移除枢纽相同的效果，说明问题本质在于嵌入向量的长度分布而非更新历史。此外，该枢纽虽被欠训练词元检测器标记为异常，但在Pythia模型中其整体在训练过程中仍被更新，因此其特征更可能源于向量长度本身，而非训练缺失。
链接: https://arxiv.org/abs/2608.29702
作者: Alexandre Quemy
机构: Hother Labs
类目: Computation and Language (cs.CL)
备注: Submitted to NeurReps Workshop @ NeurIPS 2026
Abstract:A token-embedding table holds a hub of short rows near its origin, and we show that this cluster biases what nearest-neighbor intrinsic-dimension (ID) estimators report. Because of the concentration of measure, a token is closer to the central cluster than to any other token, so its first two neighbors are both hub rows at nearly the same distance. As a result, the ID estimators such as TwoNN return a dimension far above the real ID. Measured one token at a time, dimension is a heavy-tailed distribution. Measured on the full vocabulary, it grows with the model’s parameter count. However, when we remove the hub, the heavy tail disappears and the measured dimension collapses to a narrow range for eleven models, from GPT-2 to models such as K3 and GLM-4.7. The hub acts as a switch: a few hundred rows are enough to fully inflate the estimate. We reproduced an experiment stating that the intrinsic dimension (ID) of Pythia’s token-embedding table grows with the parameter count, from 27 to 122 between 160M and 12B parameters. We show that this result disappears when the hub is removed: the table then reads 10 to 17 at every size. The hub contains a subset of the population that under-trained-token detectors flag, but on Pythia the hub that we detected and removed as a whole was updated during training: what seem to characterize these rows is simply their length, not an absence of updates. Finally, we show that normalizing the rows instead of removing them gives the same lower reading.
[NLP-174] ACTD: Anchor-Based Cross-Tokenizer Distillation with Residual Regularization EMNLP2026
【速读】： 该论文旨在解决跨分词器（cross-tokenizer）知识蒸馏中因词汇表与序列结构不一致导致的知识迁移困难问题，尤其针对不同模型家族间存在的词汇和序列错位现象。现有方法在尝试近似对齐词汇表时，往往引入额外噪声，影响蒸馏效果。为此，论文提出基于锚点的跨分词器知识蒸馏结合残差正则化（Anchor-Based Cross-Tokenizer Distillation with Residual Regularization, ACTD），其核心创新在于通过词汇与序列双重对齐机制缓解模型结构异质性，并引入一种新型锚点损失函数结合残差正则化，有效抑制对齐过程中的噪声干扰。此外，该框架进一步拓展至多教师设置，实验在五个推理基准上使用三种不同教师模型验证，结果表明ACTD达到当前最优性能，其多教师版本显著优于最强的单教师及多教师基线方法，充分证明了该方法的高效性与鲁棒性。
链接: https://arxiv.org/abs/2608.29662
作者: Huiyi Zhang,Zijian Li,Xiaocheng Feng,Weitao Ma,Xiaoliang Yang,Yichong Huang,Bing Qin
机构: Harbin Institute of Technology (哈尔滨工业大学); Peng Cheng Laboratory (鹏城实验室)
类目: Computation and Language (cs.CL)
备注: Accepted by EMNLP 2026 main conference
Abstract:Knowledge distillation effectively transfers reasoning capabilities from large language models to lightweight student models. To enable knowledge transfer across disparate model families, researchers increasingly explore cross-tokenizer distillation. However, cross-tokenizer distillation remains challenging due to vocabulary and sequence misalignment, while approximate vocabulary alignment can introduce additional noise into distillation. To address these challenges, we propose Anchor-Based Cross-Tokenizer Distillation with Residual Regularization (ACTD). ACTD bridges structural heterogeneity through vocabulary and sequence alignment, while mitigating alignment noise via a novel anchor loss with residual regularization. We further extend this framework to a multi-teacher setting. Evaluated across five reasoning benchmarks with three distinct teacher models, ACTD achieves state-of-the-art performance. Moreover, its multi-teacher extension outperforms the strongest single-teacher and multi-teacher baselines, further demonstrating the robustness of our method.
[NLP-175] Conducting Stylistic Analysis of Paintings through an Art-History Agent
【速读】： 该论文旨在解决艺术史研究中艺术品归属（attribution）依赖主观视觉分析而缺乏可解释性与证据支持的问题。传统风格分析（stylistic analysis）虽具深度，但难以规模化和标准化；而现有生成式人工智能（Generative AI）模型仅提供不可解释的置信度分类，无法满足学术验证需求。其解决方案的关键在于构建一个融合视觉-语言协同机制的自动化框架：首先利用大规模带元数据的绘画数据集训练视觉变压器（Vision Transformer, ViT），将艺术史特定信息编码为嵌入表示；随后通过稀疏字典学习（sparse dictionary learning）对嵌入进行因子分解，提取跨样本重复出现的共享特征；再由大语言模型（LLM）基于关联作品及其策展人文本，将每个特征转化为具有语义意义的风格描述；最后，通过自主协调的LLM结合推理与行动（ReAct）框架，对特征进行加权、验证与优化，生成连贯的艺术品描述或比较分析。该方法实现了从图像特征到可解释语义描述的映射，打通了视觉数据与人文阐释之间的鸿沟，推动了以视觉为基础的计算艺术史（vision-based computational art history）的发展。
链接: https://arxiv.org/abs/2608.29644
作者: Marc S. Walton,Astrid Harth
机构: 未知
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Attributing an artwork to an artist has traditionally relied on detailed visual observations and descriptions, known as stylistic analysis in art history. By contrast, current artificial intelligence (AI) models used in the field offer only unexplained probabilistic classifications. To bridge this methodological gap, we present an AI framework that automates stylistic analysis of paintings, providing a foundation for enhancing evidence collection, discovery, and verification. By training a vision transformer (ViT) on a large corpus of paintings with metadata, our system encodes this art history-specific data as embeddings. These representations are factorized via sparse dictionary learning into a shared set of features that recur across the training set. A large language model (LLM) then interprets each feature by retrieving associated artworks and their accompanying curator-written texts, and synthesizes them into descriptions that reflect their stylistic attributes. Finally, an autonomous coordinator LLM applies a reasoning-and-action (ReAct) framework to weight, test, and refine these features into cohesive descriptions of an artwork, or comparisons of artworks. This approach converts detailed visual features into descriptive terms, addressing a key challenge in art history. It thus connects the use of images as data with the semantic concerns of humanists, establishing vision-based computational art history as an area for future growth.
[NLP-176] PrivBench: A Holistic and Modular Benchmarking Platform for Evaluating Text-to-Text Privatization EMNLP2026
【速读】： 该论文旨在解决文本到文本隐私化（text-to-text privatization）方法评估标准不统一的问题。当前文献中采用的评估技术与指标繁杂多样，缺乏系统性与可比性，难以客观衡量不同隐私化方法的实际隐私保护能力。为此，本文提出PrivBench——一个全面且模块化的基准测试平台，用于统一评估文本隐私化技术。其核心解决方案在于：构建一套结构化的评估维度（desiderata），并将其组织为可扩展的模块，实现对隐私化效果的多维度、系统性评估；同时，平台具备用户中心设计，支持实时评估与公开排行榜机制，促进研究社区间的良性竞争。此外，PrivBench具有高度可扩展性，支持未来版本迭代与新任务集成，且免费开放使用，显著提升了隐私化方法评估的标准化与透明度。
链接: https://arxiv.org/abs/2608.29624
作者: Stephen Meisenbacher,Andreea-Elena Bodea,Ahmet Bilal Akın,Alexandra Klymenko,Jana Diesner,Florian Matthes
机构: Technical University of Munich (慕尼黑工业大学); Munich Center for Machine Learning (慕尼黑机器学习中心); Munich Data Science Institute (慕尼黑数据科学研究所)
类目: Computation and Language (cs.CL)
备注: 23 pages, 5 figures, 3 tables, accepted to EMNLP 2026 System Demonstrations
Abstract:Natural Language Processing methods have enabled novel solutions and advances in the field of privacy, particularly in the sub-domain of text-to-text privatization, where the goal is to transform a sensitive input text into a privatized output by ideally masking (in)directly identifiable or otherwise private information. The evaluation of text-to-text privatization, however, is not straightforward, and the extant literature has utilized a myriad of techniques and metrics to quantify the privacy-preserving capabilities of privatization methods. Seeking to unify the evaluation of text-to-text privatization, we introduce PrivBench, a holistic and modular benchmarking platform for researchers and practitioners working on text privatization. PrivBench is holistic in that it evaluates privatization on a series of defined desiderata, which are structured into modules. PrivBench is not only modular but also extensible, allowing for future updates and benchmark versions. PrivBench is user-centered and promotes competition via real-time evaluation and a live public leaderboard. The platform is free to use and openly accessible at this https URL.
[NLP-177] MI-Distillation: Selecting from Model-Interpolated Instruct-Reasoning Data Spectrum for Chain-of-Thought Distillation EMNLP2026
【速读】： 该论文旨在解决大模型长链式思维（Long CoT）在知识蒸馏至小型学生模型时效果不佳的问题。尽管长链推理能提升复杂任务表现，但直接以长链作为监督信号进行蒸馏往往收益有限，甚至不如简洁的短链推理（Short CoT）。其核心挑战在于：长链推理产生的梯度幅值更大、更新方向更集中，且随着学生模型容量增加，这种现象愈发显著，导致训练不稳定或过拟合。因此，有效的长链蒸馏需平衡推理轨迹的信息密度与对学生模型分布的对齐程度。针对此问题，作者提出模型插值蒸馏（MI-Distillation）框架，通过模型插值构建连续的“指令-推理”数据谱系，从而生成介于原始长链与短链之间的中间推理路径；进一步引入序列可学习惊喜度评分（SeqLSS），用于筛选既具信息量又适合学生模型学习的推理路径。实验结果表明，该方法在多个推理基准上显著优于现有强基线，尤其在小模型场景下展现出优越的蒸馏性能。
链接: https://arxiv.org/abs/2608.29623
作者: Yangsong Lan,Renkai Hu,HongKai Zheng,Bo Zhang,Renzhi Wang,Hongliang Dai,Piji Li
机构: Nanjing University of Aeronautics and Astronautics (南京航空航天大学); The Key Laboratory of Brain-Machine Intelligence Technology, Ministry of Education (脑机智能技术教育部重点实验室)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted to Findings of the Association for Computational Linguistics: EMNLP 2026
Abstract:Recent advances in large reasoning models (LRMs) have shown strong performance on complex problems through long chain-of-thought (Long CoT) reasoning. However, distilling such trajectories into smaller student models remains challenging: direct Long CoT supervision often provides limited gains and can be less effective than concise Short CoT rationales. In this work, we investigate this phenomenon from a gradient-centric perspective. Our analysis shows that Long CoT induces larger gradient magnitudes and more concentrated update directions than Short CoT, with this effect becoming more pronounced as student model capacity increases. These findings suggest that effective Long CoT distillation requires balancing the reasoning information density of reasoning trajectories with their distributional alignment to the student model. Motivated by this insight, we propose \textbfModel \textbfInterporlation \textbfDistillation (\textbfMI-Distillation), a framework that constructs a continuous Instruct-Reasoning data spectrum through model interpolation. To select suitable trajectories from this spectrum, we further introduce \textbfSequential \textbfLearnable \textbfSurprisal \textbfScore (\textbfSeqLSS), which favors reasoning paths that are both informative and learnable for the student. Extensive experiments on reasoning benchmarks show that MI-Distillation consistently improves small model CoT distillation over strong Long CoT baselines.
[NLP-178] Memory-First Fact-Checking: A Knowledge-Graph-Grounded Multi-Agent System for Misinformation Detection
【速读】： 该论文旨在解决虚假信息检测中可解释性不足与知识依赖性薄弱的问题，尤其针对事实核查过程中因外部证据缺失或内部知识不完整导致的误判与决策不可信问题。其核心解决方案在于提出一种“记忆优先、网络回退”的混合式事实核查框架，关键在于将基于知识图谱（Knowledge Graph）的语义记忆与对抗式多智能体推理相结合：首先通过句向量（Sentence-BERT）语义检索与自然语言推理（Natural Language Inference, NLI）对输入声明在双索引知识图谱中进行初步验证；当图谱内证据不足时，激活由支持、矛盾与裁判三类智能体构成的对抗性评议庭，从可信网络源获取补充信息并进行协同评估；同时引入图谱感知置信度机制，综合语义相似度、NLI置信度及图结构证据，动态判断是否需触发外部检索，从而减少冗余查询。经验证的信息以结构化三元组形式回填至知识图谱，实现语义记忆的增量扩展。实验结果表明，该框架在经过精心构建的新冠疫情虚假信息基准上取得了97.4%的准确率与92.6%的宏平均F1分数，显著优于Llama 3.3 70B基线模型。
链接: https://arxiv.org/abs/2608.29617
作者: Amelia Petrenciuc,Alexandru Lecu,Adrian Groza
机构: Artificial Intelligence Research Institute AIRi@UTCN (人工智能研究学院 AIRi@UTCN); Technical University of Cluj-Napoca (克卢日-纳波卡技术大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:This paper introduces a hybrid fact-checking framework that integrates Knowledge Graph-based semantic memory with adversarial multi-agent reasoning for explainable misinformation detection. The proposed system follows a memory-first, web-fallback architecture, in which input claims are initially evaluated against a dual-index Knowledge Graph through Sentence-BERT-based semantic retrieval and Natural Language Inference. When the evidence retrieved from the graph is insufficient to support a reliable decision, the framework collects information from trusted web sources and assesses it using an adversarial tribunal composed of support, contradiction, and judging agents. A graph-aware confidence mechanism combines semantic similarity, NLI confidence, and structural graph evidence to determine whether internal knowledge is sufficient, thereby reducing unnecessary web retrieval. Following verification, validated information is transformed into structured triples and incorporated into the Knowledge Graph, supporting the incremental expansion of the system’s semantic memory. Experimental evaluation on a curated COVID-19 misinformation benchmark demonstrates that the proposed framework achieves an accuracy of 97.4% and a macro-averaged F1-score of 92.6% on resolved claims, outperforming a Llama~3.3~70B baseline, which obtains an accuracy of 87.7% and a macro-averaged F1-score of 86.3%.
[NLP-179] JPO: Juris Policy Optimization for Structured Legal Reasoning in Criminal Judgment Prediction EMNLP2026
【速读】： 该论文旨在解决中文刑事判决预测中模型缺乏结构化法律推理能力的问题，即如何在案件事实基础上准确匹配法条、合理论证罪名，并确保量刑结果与罪名逻辑一致。现有方法多聚焦于最终标签的优化，且对推理质量的评估依赖大语言模型（LLM）生成的评价标准，难以反映法律裁判本身的内在逻辑结构。本文提出一种名为“司法政策优化”（Juris Policy Optimization, JPO）的后训练框架，其核心在于：首先利用教师模型生成的推理链监督模型遵循标准化的四步推理流程（事实匹配法条→法条支撑罪名→罪名决定量刑→各步骤间逻辑一致性校验），随后引入复合奖励机制，联合优化判决预测准确性、推理结构完整性以及跨步骤一致性；进一步通过分词级优势重加权与自适应裁剪策略，增强对法律关键推理片段的关注。实验表明，JPO在多个开源语言模型和三个中文法律基准上均显著优于监督微调及传统强化学习基线，在提升判决预测性能的同时有效改善了推理过程的可解释性与法律合规性。
链接: https://arxiv.org/abs/2608.29616
作者: Zhaolu Kang,Yantao Liu,Tailong Luo,Leqi Zheng,Lei Wei,Chenghua Zhu,Junhao Gong,Jiachen Qian,Eric Hanchen Jiang,Jiaxin Liu,Yuan Wang,Hao Zhang,Zixia Wang,Rong Fu,Zheng Lin,Richeng Xuan,Zhichao Hu
机构: Tencent(腾讯); Peking University (北京大学); Tsinghua University (清华大学); City University of Hong Kong (香港城市大学); University of California (加州大学); University of Illinois Urbana-Champaign (伊利诺伊大学厄本那-香槟分校); Zhejiang University (浙江大学); University of Hong Kong (香港大学)
类目: Computation and Language (cs.CL)
备注: EMNLP 2026 Main
Abstract:Criminal judgment prediction requires models to infer statutory articles, charges, and sentencing outcomes from case facts. Unlike standard classification tasks, it involves a structured reasoning process in which statutes should be matched with facts, charges should be justified by statutes, and sentencing outcomes should remain consistent with charges. Existing approaches optimize final labels, and while some have attempted to evaluate reasoning quality, their evaluations are indirect, often relying on LLM-generated rubrics that reflect model-internal preferences rather than the inherent logical structure of legal adjudication. We propose Juris Policy Optimization (JPO), a post-training framework for structured legal reasoning in Chinese criminal judgment prediction. JPO first uses teacher-generated rationales to supervise a standardized four-step reasoning process, and then applies reinforcement learning with a composite reward over legal prediction quality, reasoning structure completeness, and cross-step consistency. JPO further introduces token-level advantage reweighting and adaptive clipping for legally salient reasoning segments. Experiments on multiple open-source language models and three Chinese legal benchmarks show that JPO consistently improves both judgment prediction and reasoning quality over supervised fine-tuning and reinforcement learning baselines.
[NLP-180] Cross-lingual Functional Vectors for Emotion Detection in Large Language Models EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在语义复杂任务中以及跨语言场景下的任务行为调控问题，尤其关注现有基于上下文示例的提示学习方法在多语言环境中的泛化能力局限。其核心挑战在于如何实现无需额外示例即可在零样本（zero-shot）条件下有效引导模型完成跨语言任务，同时保持高效性与可迁移性。论文提出的解决方案关键在于引入函数向量（Function Vectors, FVs），通过从源语言的上下文示例中提取任务特定的潜在方向表示，并将其注入目标语言模型以引导其行为。研究发现，FVs能够捕捉跨语言的任务相关信号而非仅依赖于语言特异性词汇模式，在多种跨语言设置下均显著提升情感识别等复杂语义分类任务的性能；进一步分析表明，每种模型存在相对稳定的最优注意力头范围用于构建有效的FVs，且该模式在不同语言间保持一致。此外，FVs可在不处理多个示例的前提下部分复现少量示例提示学习的效果，从而在降低计算开销的同时实现高效的多语言任务适应，展现出作为轻量化、可迁移任务调制机制的巨大潜力。
链接: https://arxiv.org/abs/2608.29613
作者: Jieying Xue,Phuong Minh Nguyen,Minh Le Nguyen,Shogo Okada
机构: Japan Advanced Institute of Science and Technology (日本先进技术研究院)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: Findings of the Association for Computational Linguistics: EMNLP 2026
Abstract:Function vectors (FVs) have recently emerged as a promising mechanism for steering the behavior of large language models (LLMs) by injecting task-specific latent direction representations derived from in-context demonstrations. While prior studies have shown that FVs can recover task behavior in structured in-context learning settings, their effectiveness on semantically complex tasks and their ability to generalize across languages remain underexplored. We investigate the cross-lingual transferability of FVs using multilingual multi-label emotion recognition as a challenging semantic classification benchmark. Specifically, we examine whether FVs extracted from a source language can steer task behavior in another language under both standard clean and perturbed zero-shot settings without providing demonstrations during inference. Across diverse cross-lingual settings, applying FVs substantially improves performance, suggesting that FVs capture language-agnostic, task-relevant signals rather than purely language-specific lexical patterns, and highlighting their potential as a lightweight and transferable mechanism for multilingual task adaptation. We observe that each LLM exhibits a relatively stable optimal range of attention heads for constructing effective FVs, and the pattern remains consistent across languages. In addition, FVs can partially replicate the task-steering effects of standard few-shot in-context learning while avoiding the computational overhead of processing multiple demonstrations, making them effective for large-scale practical applications. Our code is available at this https URL.
[NLP-181] Beyond Surface Alignment: Grounding the Dynamics of Situational Understanding and Generative Control in LLM s ALT DATE
【速读】： 该论文旨在解决当前大型语言模型（Large Language Models, LLMs）在对齐训练中过度关注表面行为（如流畅性、安全性与语调一致性）所导致的深层问题——即缺乏情境与生成过程的实质锚定（grounding），致使模型虽在表层表现自信，但在动态变化的现实场景中表现出显著的脆弱性。其核心解决方案在于提出“具身对齐”（Grounded Alignment）框架，关键在于从输入处理（Input）与输出生成（Output）两个维度系统评估并重构模型的内在机制：首先通过SituTest和ReCode揭示模型在情境建模上的失败，表明其难以维持连贯的“心智模型”且依赖表面启发式而非深层句法依赖；其次引入分支因子（Branching Factor, BF）量化生成空间，发现标准对齐训练导致生成路径过早收敛至风格化坍缩；进一步通过“事后理解”（Hindsight）验证模型对其自身生成内容的理解缺陷。在此基础上，论文提出动态控制策略，包括通过AI Realtor进行上下文工程以弥补情境建模不足、基于基线对齐模型协作（Base-Aligned Model Collaboration）实现探索与风格约束解耦，并采用渐进采样（Annealed Sampling）支持可验证的强化学习。这些方法共同推动模型从表面对齐转向真正根植于上下文与生成过程的智能体范式。
链接: https://arxiv.org/abs/2608.29610
作者: Chenghao Yang
机构: 未知
类目: Computation and Language (cs.CL)
备注: PhD Thesis submitted to UChicago ( this https URL ). Update Branching Factor, AI Realtor, and BACo to the latest versions
Abstract:The current alignment tuning paradigm for Large Language Models (LLMs) prioritizes surface-level behaviors – fluency, safety, and tonal consistency. While effective for casual chat, this thesis argues that such surface alignment masks a lack of grounding, creating models that are stylistically confident but situationally brittle. We propose a framework of Grounded Alignment, analyzing how models process context (Input) and structure generation (Output), then aligning these grounded behaviors to human needs. First, we evaluate failures in Situational Grounding. SitTest shows that despite large context windows, state-of-the-art models struggle to maintain a consistent “mental model” of a changing environment. ReCode further shows that models rely on surface heuristics rather than deep syntactic dependencies: they “read” extensive histories without truly “understanding” the evolving situation. Second, we evaluate Generative Grounding. We introduce the Branching Factor (BF) to map LLM generation, finding that standard alignment tuning constricts this landscape into premature stylistic collapse. Hindsight further shows that models often fail to understand their own generations. Finally, we propose Dynamic Control for grounded interaction. AI Realtor demonstrates context engineering to compensate for poor situational grounding. Base-Aligned Model Collaboration decouples exploration from stylistic constraints. We also present Annealed Sampling for verifiable reinforcement learning and apply these ideas to Addiction Support, where model-generated rationalization offers a communication interface for high-stakes domains. Collectively, this work moves beyond surface alignment toward agents anchored in both context and generation. Comments: PhD Thesis submitted to UChicago (this https URL). Update Branching Factor, AI Realtor, and BACo to the latest versions Subjects: Computation and Language (cs.CL) Cite as: arXiv:2608.29610 [cs.CL] (or arXiv:2608.29610v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2608.29610 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Chenghao Yang [view email] [v1] Sun, 30 Aug 2026 07:04:03 UTC (20,836 KB)
[NLP-182] Agent Zero Memory: Provenance-Aware Long-Term Memory for LLM Agents
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）智能体在长期记忆管理中面临的根本性挑战：现有记忆系统通常依赖单一组织结构（如事实库、向量索引或知识图谱），因而继承了各自固有的盲区，难以全面、准确地保留用户或组织的完整交互与存储历史。为应对这一问题，论文提出了一种名为Agent Zero Memory的溯源感知型长期记忆系统，其核心解决方案在于将用户对话、文件及关联数据源统一提炼为三个并行运行的记忆系统：一是以时间与事件变化为核心的第一性原则的情景记忆事件时间线（Memory Events timeline），二是跨会话关联人物与项目的关联实体-事件知识图谱（associative entity-event knowledge graph），三是语义化、经人工校准且具有引用锁定机制的层级文档记忆（Hierarchical Documentary Memory, HDM）。在检索阶段，系统通过意图门控（避免自包含查询引入延迟）、源路由和三路并发的代理式搜索实现高效推理，每一路均采用混合（嵌入+词法）搜索结合代理控制的过滤器，并输出有据可循的引用答案；最终答案通过单一置信度整合生成。该系统严格遵循“阅读规范”：所有学习条目均为带溯源信息的条目，包含来源、时间戳和证据指针，所有回答均在引用锁定机制下生成，仅能引用实际被打开的证据，从而从根本上杜绝虚构行为，系统在无法确定时选择不回答而非猜测。在两个公开基准测试中，该系统分别达到95.60%（LongMemEval）和93.60%（LoCoMo）的准确率，超越最强基线系统0.73和1.10个百分点。对八种不同主干模型的受控实验进一步揭示了准确性-成本-延迟权衡边界：尽管准确率波动仅3.4分，但单次查询成本差异达约30倍，且在高达20倍更低的成本下仍可实现接近最先进水平的性能，凸显了该系统以记忆驱动而非模型驱动的质量提升特性。
链接: https://arxiv.org/abs/2608.29606
作者: Ming Wu,Pengyuan Zhu
机构: 未知
类目: Computation and Language (cs.CL)
备注: 10 pages, 3 figures
Abstract:Large language model (LLM) agents need durable, faithful memory of everything a user or organization has said and stored, yet most memory systems commit to a single organizing structure (a fact store, a vector index, or a knowledge graph) and inherit its blind spots. We present Agent Zero Memory, a provenance-aware long-term memory system that distils a user’s conversations, files, and connected sources into three parallel memory systems, each capturing a different facet of the same history: an episodic Memory Events timeline that makes when and what changed first-class, an associative entity-event knowledge graph that links people and projects across sessions, and a semantic, curated, citation-locked Hierarchical Documentary Memory (HDM) of durable facts. A retrieval turn runs an intent gate (so self-contained turns add no latency), a source router, and three concurrent agentic searches, one per system, each a tool-using loop over hybrid (embedding + lexical) search under agent-controlled filters; their grounded, cited answers are integrated into one answer with a single confidence. We formalize the reading discipline: every learned item is a provenanced item carrying its origin, timestamp, and evidence pointer, and every answer is read under a citation lock, so it may cite only evidence its reader actually opened; fabrication is structurally excluded and the system abstains rather than guesses. On two public benchmarks the system sets a new state of the art: 95.60% on LongMemEval and 93.60% on LoCoMo, improving over the strongest prior systems by +0.73 and +1.10 points. A controlled study across eight backbone LLMs characterizes the accuracy-cost-latency frontier: accuracy varies by only 3.4 points while per-query cost varies by ~30x, with near-state-of-the-art quality at up to 20x lower cost per query, the signature of memory-driven, rather than model-driven, quality.
[NLP-183] Hindsight Memory-PRM: Supervising Memory Management with Auditable Hindsight Credit
【速读】： 该论文旨在解决长时程大语言模型（Long-horizon LLM）代理在记忆操作（memory operations）中缺乏可监督信号的问题：由于记忆操作的价值在执行时不可观测，传统方法难以有效评估其有效性。其核心解决方案是利用轨迹中可机器读取的审计证据——检索命中（retrieval hits）与回答时间引用（answer-time citations），通过事后分析（hindsight analysis）两次利用这些信号：离线训练一个基于操作条件的记忆效用评判器（memory-utility critic），在线运行时则结合检索结果、引用行为及受控的删除-重答机制，为每个记忆条目生成经干预校准的初始存在信用（entry-level presence credit），并沿版本链传播作为动作级的代理奖励信号。该方法无需逐操作的人工标注，也无需对后续轨迹进行蒙特卡洛回放。在独立测试集LoCoMo上，8B规模的本地策略在固定共享阅读器条件下达到77.5%的准确率，超越其API教师模型（65.1%）及所有复现的外部系统，且仅需Mem0官方运行点八分之一的上下文长度；在LongMemEval上达到79.0%。消融实验表明性能提升主要源于因果校准（causal calibration）而非信号密度，且策略收敛至一种多版本记忆组织结构，其优势无法被任何测试的开环基线模型复现。
链接: https://arxiv.org/abs/2608.29605
作者: Haoxuan Jia,Yang Liu,Yingguang Yang,Yancheng Chen,Chongyang Zhang,Hao Zheng,Qian Li,Yulin Huang,Jianshen Zhang,Yongzhi Qi,Shang Luo,Kefu Xu,Hao Peng,Junyu Lu,Du Cheng,Philip S. Yu,Bin Chong
机构: Peking University (北京大学); University of Chinese Academy of Sciences (中国科学院大学); Fullive-AI; Nanyang Technological University (南洋理工大学); Supply Chain Tech Team Y, JD.com (京东供应链技术团队Y); Beijing University of Posts and Telecommunications (北京邮电大学); Beihang University (北京航空航天大学); Beijing Institute of Technology, Zhuhai (北京理工大学珠海校区); Northeastern University (东北大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Memory operations of long-horizon LLM agents are hard to supervise: an operation’s value is unobservable when it is taken. But they are special – they leave machine-readable evidence in the trajectory: retrieval hits and answer-time citations. Hindsight Memory-PRM exploits this audit trail twice: offline to train an operation-conditioned memory-utility critic, and online, where retrievals, citations, and one controlled deletion-and-reanswer per probe settle an intervention-calibrated entry-level presence credit, propagated along version chains as an action-level proxy reward – no per-operation human labels, no Monte-Carlo replay of continuations. On held-out LoCoMo a local 8B policy reaches 77.5% under a fixed shared reader, surpassing its API teacher (65.1%) and all reproduced external systems, at one eighth the context of Mem0’s official operating point; on LongMemEval, 79.0%. Ablations attribute the gain to causal calibration rather than signal density, and the policy converges to a multi-version memory organization whose gains no tested open-loop baseline reproduces.
[NLP-184] How You Ask Shapes What You Get: A Theory-Seeded Measurement of Articulation in Advice-Seeking LLM Conversations
【速读】： 该论文旨在解决用户在提出咨询请求时表达方式多样但现有研究将其视为需消除的噪声这一问题，核心在于探究用户表达方式（articulation）是否构成与话题（topic）可分离的潜在维度，并分析其对语言模型响应的影响。解决方案的关键在于从16,447条来自公开聊天语料库（WildChat、LMSYS和ShareChat）的咨询请求中提取可解释特征，识别出一组跨训练/测试划分及跨语料库稳定存在的潜在表达风格因子。这些因子揭示了若干重复出现的表达模式，其中一种“长篇但信息贫乏”的表达风格尤为突出——约占最大语料库中六分之一的请求，尽管存在明显的信息不足，模型仍倾向于生成更短、更模糊的回答，且极少主动寻求澄清。这一现象在各话题组和长度分位数内均成立，且并非仅由信息缺失导致：另一种同样信息不足的表达风格则会触发模型的澄清提问，表明表达方式本身具有独立影响。两名独立的人工标注者复现了该对比结果。因此，论文主张评估基准应按表达风格进行分层，并提供所提取的结构作为实现这一分层的测量工具。
链接: https://arxiv.org/abs/2608.29591
作者: Juneha Baek,Suhyeon Lee,Donghyuk Shin
机构: KAIST(韩国科学技术院)
类目: Computation and Language (cs.CL)
备注:
Abstract:Users articulate the same advice-seeking request in different ways: some specify detailed constraints, others gesture at a vague need. Prior work treats this variation as noise to be averaged away; we instead treat it as a stable, measurable structure in the input distribution. We ask whether articulation (how people ask) forms latent dimensions separable from topic (what they ask about), and whether it is associated with how language models respond. We extract interpretable features from 16,447 advice-seeking prompts pooled from public chat corpora (WildChat, LMSYS, and ShareChat) and recover a small set of latent articulation factors that replicate across train/test splits and across corpora. Because this structure is largely separable from topic, the populations it defines cut across topics and stay invisible to topic- or task-based evaluation. The factors define a handful of recurring articulation styles, one of which stands out: a long-form but information-poor style, roughly one in six prompts in the largest corpus, where models return shorter, vaguer answers and do not ask for clarification even though under-specification is exactly the condition that warrants it. The contrast holds within every topic group and length quintile, and is not under-specification alone – a second, equally under-specified style does draw clarifying questions. Two independent human annotators reproduce this contrast. We argue that benchmarks should stratify on articulation, and we offer the extracted structure as a measurement instrument for doing so.
[NLP-185] Call Neighbours Yourself: Graph Walks with Destination-Conditioned On-Policy Self-Distillation EMNLP2026
【速读】： 该论文旨在解决文本属性图（Text-Attributed Graphs, TAGs）上的推理问题，即如何让大语言模型（LLMs）在推理过程中有效融合节点自身文本与邻域中分散的证据。现有方法通常在生成前固定可访问的邻居集合，导致推理依赖静态上下文，无法在推理阶段动态获取缺失的关键信息。其核心挑战在于：邻居选择过程被预设，缺乏灵活性和自适应性。本文提出“主动呼叫邻居”（Call Neighbours Yourself, CNY）框架，将邻居选择本身纳入推理流程，通过拓扑约束的图遍历动作（topology-constrained graph-walk actions），使LLM能够主动探索图结构以发现潜在证据。关键创新在于引入一种基于目标节点条件的在线自蒸馏策略（destination-conditioned on-policy self-distillation），通过回溯评估所选邻居在内容揭示后的效果，将行动偏好变化转化为动作级别的训练信号，从而缓解邻居探索中的延迟奖励问题。实验表明，在统一原始文本设置下，CNY在标准TAG推理基准上持续优于固定上下文的后训练基线，并且学习到的探索策略具备良好的泛化能力，可迁移至未见图结构及训练中未接触的图级任务。
链接: https://arxiv.org/abs/2608.29588
作者: Yilun Liu,Boyu Luo,Yanran Tang,Ruihong Qiu,Zi Huang
机构: The University of Queensland (昆士兰大学); Brisbane, Queensland, Australia
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: EMNLP 2026
Abstract:Reasoning over text-attributed graphs (TAGs) requires large language models (LLMs) to combine a node’s text with evidence distributed across its neighbourhood. Existing methods fix the set of accessible neighbours before generation, forcing reasoning to operate over a static context and preventing the model from acquiring missing evidence during inference. We argue that neighbour selection should itself be part of the reasoning process. To this end, we propose Call Neighbours Yourself (CNY), a framework that enables LLMs to proactively explore graph neighbourhoods through topology-constrained graph-walk actions. Instead of reasoning over a pre-selected neighbour set, CNY exposes lightweight neighbour previews and learns when to expand candidate neighbours for additional evidence. To address the delayed-credit challenge of neighbour exploration, we introduce destination-conditioned on-policy self-distillation, which retrospectively evaluates a selected neighbour after its content is revealed and converts the resulting change in action preference into an action-level training signal. Experiments on standard TAG reasoning benchmarks under a unified raw-text setting show that CNY consistently outperforms fixed-context post-training baselines. Furthermore, the learned exploration policy transfers to unseen graphs and to a graph-level task not encountered during training. Code is available at this https URL.
[NLP-186] SUP-MIMIC: A Multi-Task Clinical Diagnosis Benchmark for Evaluating LLM s Robustness to Contradictory Evidence
【速读】： 该论文旨在解决当前大语言模型（LLM）评估体系中忽视临床指标与诊断之间复杂非双射映射关系的问题，尤其针对诊断歧义（one-to-many）与诊断收敛（many-to-one）等真实医疗场景中的推理能力缺失。现有基准测试未能有效评估模型在面对表型相似但病因不同或症状异质但指向同一疾病时的真正因果推理能力。为此，本文提出SUP-MIMIC多任务框架，基于MIMIC-IV-v3.1数据集构建包含基础评估（BA）、诊断歧义任务（DDT）和诊断收敛任务（DCT）的评测体系：其中DDT用于检验模型对“表型相似、病因多样”情况下的解歧能力，DCT则评估模型识别“多源症状指向单一疾病”的模式识别能力。关键发现表明，先进大语言模型在DDT与DCT任务上性能显著下降，暴露出其严重依赖统计捷径而非深层因果推理的缺陷；同时，模型表现出对“健康”状态的保守倾向，存在漏诊风险。本研究建立了量化临床推理鲁棒性的严谨方法，为提升生成式AI在临床医学中的安全性提供了重要路径。
链接: https://arxiv.org/abs/2608.29582
作者: Yi Yu,Bo Wang,Chong Feng,Ge Shi,Xia Liu,Ziyi Yang,Xuewen Shi
机构: Beijing University of Technology (北京工业大学); Beijing Institute of Technology (北京理工大学); China-Japan Friendship Hospital (中日友好医院); Dongbei University of Finance and Economics (东北财经大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 18 pages, 13 figures, 3 table
Abstract:Current evaluations of large language models (LLMs) primarily focus on factual knowledge retrieval, overlooking the fundamental challenge of navigating the complex, non-bijective mappings between clinical indicators and diagnoses. Existing benchmarks fail to assess whether large language models truly possess the reasoning capability required for diagnostic ambiguity scenarios, where identical clinical presentations may correspond to different etiologies, and diagnostic convergence scenarios, where heterogeneous symptoms ultimately indicate the same disease. To address this issue, we propose SUP-MIMIC, a multi-task framework utilizing MIMIC-IV-v3.1 that comprises Basic Assessment (BA), Diagnostic Divergence Task (DDT), and Diagnostic Convergence Task (DCT). Specifically, DDT is designed to evaluate the model’s “one-to-many” disambiguation capability among phenotypically similar cases, while DCT assesses the model’s ability to identify “many-to-one” diagnostic patterns across different pathophysiological pathways. Comprehensive evaluation of state-of-the-art LLMs reveals substantial performance degradation on DDT and DCT compared to baseline tasks, exposing a systemic reliance on statistical shortcuts over genuine causal reasoning. Our findings further highlight a conservative bias toward “healthy” predictions, implying non-trivial risks for missed diagnoses in realistic medical settings. This work establishes a rigorous methodology for quantifying clinical reasoning robustness and provides a roadmap for enhancing the safety of language models in clinical medicine.
[NLP-187] SemTrace: Source-Grounded Semantic Signatures for Tracing LLM Exposure to Protected Documents
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在处理受保护文档时产生的溯源难题，即文档所有者无法控制或审查生成文本所依赖的模型，导致内容篡改或未经授权使用的风险。其核心解决方案是提出SemTrace——一种基于源文档语义的水印技术，用于检测生成的评论是否受到已知受保护文档副本的影响。该方法的关键在于构建一个与文档内容直接相关的二元语义签名（binary signature），通过从文档中提取出由事实命题构成的、可验证的二元对，并要求指令遵循型评审模型在固定槽位中表达选定的事实，从而隐式嵌入水印信息。利用冻结的自然语言推理模型对生成内容中的语义证据进行显式擦除和解码，将恢复出的比特序列与对应副本的码字进行比对，实现对特定副本来源的精准识别。该设计实现了模型无关性（model-agnostic）和受控副本暴露检测，同时确保水印语义上与原始文档强绑定，有效兼顾隐蔽性与可验证性。
链接: https://arxiv.org/abs/2608.29575
作者: Junyan Zhang,Yudong Zeng,Yongwei Huang,Zuhao Ouyang,Hong Chen,Xuming Hu
机构: The Hong Kong University of Science and Technology (Guangzhou)(香港科技大学(广州)); Bosum Institute of Management Science(博森管理科学研究院)
类目: Computation and Language (cs.CL)
备注:
Abstract:Large language models are increasingly used to read documents and produce downstream text, creating a provenance problem when the document owner cannot control or inspect the model that performs the generation. We introduce SemTrace, a source-grounded semantic watermark for detecting whether a generated review was influenced by a known protected manuscript copy. Rather than biasing token probabilities or imposing surface-form patterns, SemTrace constructs a document-specific binary signature from factual propositions that are directly supported by the manuscript itself. A protected PDF invisibly carries a content contract that selects one fact from each binary pair and asks an instruction-following reviewer to express those facts in fixed review slots without changing its independent evaluation. A frozen natural language inference model then decodes the resulting semantic evidence with explicit erasures and scores the recovered bits against the codeword assigned to that copy. This design targets model-agnostic, assigned-copy exposure detection while keeping the watermark semantically tied to the source document.
[NLP-188] Which one is banana man? Evaluating vision-language models in multi-turn prag matic interpretation
【速读】： 该论文旨在解决智能体在多轮语言交互环境中进行上下文敏感的语用推理能力不足的问题，特别是针对迭代指称游戏（iterated reference games）这一典型场景下，模型难以有效利用累积上下文理解指称表达的核心挑战。其解决方案的关键在于评估不同上下文条件（包括上下文数量、顺序及相关性）对人类与视觉-语言模型在指称理解任务中表现的影响，结果表明：尽管当前模型能够利用已有上下文解释人类的指称表达，但在主动构建并整合关键上下文以支持后续推理方面存在明显缺陷，暴露出其在高效语言协作中所欠缺的核心语用能力。
链接: https://arxiv.org/abs/2608.29571
作者: Alvin Wei Ming Tan,Ben Prystawski,Veronica Boyce
机构: Stanford University (斯坦福大学); Massachusetts Institute of Technology (麻省理工学院)
类目: Computation and Language (cs.CL)
备注:
Abstract:Flexible adaptation to context and shared pragmatic intuitions contribute to smooth human conversation. Iterated reference games—in which players repeatedly pick out novel referents using language—present a test case for agents’ ability to perform context-sensitive pragmatic reasoning in multi-turn linguistic environments. We tested humans and vision–language models on their ability to identify the intended meaning of descriptions produced in iterated reference games, varying the provided context in terms of amount, order, and relevance. While humans performed well consistently, the models we evaluated could make use of prior context to interpret humans’ referring expressions, but they struggled to build up the relevant context to interpret those expressions effectively. Our results suggest that the models we evaluated lack core skills needed for efficient linguistic collaboration.
[NLP-189] OASIS: Optimizing Attacker Sequences for Hard-Label Black-Box Text Attacks
【速读】： 该论文旨在解决硬标签黑盒文本攻击中攻击者序列组合缺乏系统优化的问题，即现有方法通常仅关注单一攻击器的性能提升或通过人工方式简单拼接多个攻击器，未能充分考虑不同攻击方法在样本子集上的互补性及其搜索路径差异。其核心解决方案是提出\OURS方法，通过一次性的双目标攻击链搜索，在候选攻击序列中同时优化攻击成功率与扰动程度，进而选取一个固定的全局最优攻击链用于后续攻击执行。该方法的关键在于将攻击者组合视为可优化的策略变量，而非随意的实现选择，从而显著提升了攻击效果的稳定性与普适性。实验结果表明，\OURS在多种数据集、目标模型及大语言模型上均优于强基线和手动构造的攻击链，验证了攻击链优化作为实际优化目标的有效性。
链接: https://arxiv.org/abs/2608.29568
作者: Qian Chen,Shiliang Xiao,Yuzhi Liang
机构: Guangdong University of Foreign Studies (广东外语外贸大学)
类目: Cryptography and Security (cs.CR); Computation and Language (cs.CL)
备注: Qian Chen and Shiliang Xiao contributed equally. Yuzhi Liang is the corresponding author
Abstract:Different attack methods follow different search trajectories, they succeed on different subsets of samples, whereas existing hard-label black-box text attacks mainly focus on improving individual attackers or manually combining them. We present \OURS, a method for optimizing attacker sequences in hard-label black-box text attacks. \OURS first performs a one-time bi-objective attack chain search over candidate sequences to balance attack success rate and perturbation, and then reuses the selected fixed global chain during attack chain execution. Experiments across multiple datasets, victim models, and large language models show that \OURS consistently outperforms strong standalone baselines and simple manually constructed chains. These results suggest that attacker composition is not merely an implementation choice, but a practical optimization target for improving hard-label black-box text attacks.
[NLP-190] ACS: Trajectory-Aware Candidate Selection for LLM Jailbreak Suffix Optimization
【速读】： 该论文旨在解决梯度驱动的越狱后缀优化方法中因候选样本选择策略过于短视（myopic）而导致的性能瓶颈问题。现有方法通常仅基于当前步骤的损失值选择表现最优的候选后缀，但这种设计忽略了候选在后续搜索过程中的长期有效性，导致出现“选择阶段奖励劫持”（selection-stage reward hacking）现象：即当前表现良好的候选在后续迭代中反而无法实现有效越狱。其核心问题是，传统方法将优化重点局限于候选生成，而忽视了候选选择过程对整体搜索效率与成功率的关键影响。为应对这一挑战，本文提出\OURS——一种轨迹感知的候选选择框架，通过引入轨迹感知的代理评估机制，结合参考策略正则化与判别器估计的卡方校正，使候选选择不仅关注即时损失，更注重其在优化轨迹中的持续有效性。实验结果表明，在HarmBench基准上，\OURS在相同搜索预算下显著优于多个强基线模型，不仅大幅提升攻击成功率，且在整个优化过程中表现出更稳定的收敛行为，验证了缓解由短视选择引发的奖励劫持是提升越狱后缀优化效果的关键所在。
链接: https://arxiv.org/abs/2608.29564
作者: Shiliang Xiao
机构: Guangdong University of Foreign Studies (广东外语外贸大学)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Gradient-based jailbreak suffix optimization methods typically update the suffix by retaining the candidate with the lowest current loss. We show that this seemingly natural design is fundamentally myopic: candidates that look better under the current-step proxy often fail to produce better jailbreak outcomes later in the search, revealing a form of selection-stage reward hacking. This suggests that candidate selection, rather than candidate generation alone, is a hidden bottleneck in suffix optimization. To address this issue, we propose \OURS, a trajectory-aware candidate selection framework for jailbreak suffix optimization. Instead of selecting candidates solely by their immediate loss, \OURS augments per-step evaluation with a trajectory-aware proxy and stabilizes selection with reference-policy regularization and a discriminator-estimated chi-squared correction, encouraging choices that remain effective beyond the current step. Experiments on HarmBench show that \OURS consistently outperforms strong baselines under the same search budget, substantially improving attack success rates while exhibiting more stable optimization behavior throughout the search. Our findings highlight that mitigating selection-stage reward hacking caused by myopic candidate selection is critical for improving jailbreak suffix optimization.
[NLP-191] Evaluating LLM s on Conversational Text-to-SQL under Chain Ambiguity and Intent Drift EMNLP2026
【速读】： 该论文旨在解决现有对话式文本转SQL（conversational text-to-SQL）评估基准在用户意图随多轮交互不断演变和重构时的表征不足问题，尤其关注链式模糊性（chain ambiguity）与意图漂移（intent drift）两大挑战。其核心解决方案在于提出TIDE-Bench这一新型评估基准，基于BIRD数据集中的514个锚点SQL构建，包含1,542个样本，专门针对链式模糊性（即未充分指定的问题引发具有条件依赖关系的多层澄清）与意图漂移（用户撤回并替换先前已确认的请求元素）两类典型模式进行建模，并引入链式识别与漂移识别-修正专用指标，突破传统仅依赖执行准确率的局限。实验结果揭示了当前12个先进大语言模型普遍存在的链式识别瓶颈（不受澄清频率影响）、显著的漂移识别-修正能力差距，以及两类失败模式在联合激活时的重叠现象，凸显了未来模型在动态意图理解与上下文一致性维护方面的关键改进需求。
链接: https://arxiv.org/abs/2608.29543
作者: Yujia Liu,Jiayan Lin,Zijin Hong,Zheng Yuan,Shengyuan Chen,Hao Chen,Qinggang Zhang,Xiao Huang,Feiran Huang
机构: The Hong Kong Polytechnic University(香港理工大学); City University of Macau(澳门城市大学); Jilin University(吉林大学); Beihang University(北京航空航天大学); Jinan University(暨南大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Databases (cs.DB)
备注: Accepted to EMNLP2026 Main
Abstract:Recent advances in large language models (LLMs) have established conversational text-to-SQL as a practical interface between users and databases, often involving multiple turns of clarification and revision. However, existing benchmarks primarily evaluate execution accuracy, leaving the unfolding and shifting of user intent across turns largely uncovered. To address this, we introduce TIDE-Bench, a benchmark for conversational text-to-SQL under chain ambiguity and intent drift evaluation, targeting two recurring patterns: chain ambiguity, where an underspecified question triggers layered clarification with conditional dependencies, and intent drift, where the user retracts and replaces a previously committed request element. Built on 514 anchor SQLs from BIRD, TIDE-Bench comprises 1,542 samples and introduces dedicated metrics for chain identification and drift recognition-resolution beyond execution accuracy. Evaluating 12 advanced LLMs reveals a persistent chain identification bottleneck unaffected by clarification frequency, a wide drift recognition-resolution gap, and overlap between failure modes when jointly activated. The corresponding code of TIDE-Bench is released for further research.
[NLP-192] LoGo: Token-Level Dynamic Local-Global Attention
【速读】： 该论文旨在解决大语言模型在上下文长度增长时，注意力机制（attention）逐渐成为主要计算瓶颈的问题。标准Transformer架构虽强大但计算效率低下，因其对所有词元（token）分配相同的注意力预算，忽视了不同词元的实际上下文需求差异。现有局部-全局混合方法虽提升了效率，但通常在层或头级别静态分配注意力范围，缺乏灵活性。为此，本文提出LoGo（token-level dynamic local-global attention），其核心创新在于将注意力跨度（attention span）作为注意力预算分配的直接代理，实现词元级别的动态调控。每个LoGo层包含耦合的局部与全局分支：所有词元均通过受限上下文窗口进行高效局部注意力计算，而仅对需要长程信息的词元，由可学习门控机制激活全上下文访问的全局注意力。通过基于阈值的预算控制器，在无需额外损失函数的情况下维持目标全局注意力比例，并结合渐进式掩码调度稳定训练过程。此外，引入查询稀疏的Triton内核，将减少的全局注意力计算转化为实际推理加速。大量实验验证了LoGo的有效性，表明其在不同模型规模下均能保持全注意力Transformer的缩放特性；在控制对比中，优于全注意力Transformer及同等预算的静态局部-全局混合模型，尤其在长程信息检索任务上表现显著提升。分析进一步揭示LoGo可学习可解释的跨度分配模式。研究结果表明，基于学习的词元级跨度分配是一种有效且可扩展的优化策略，能够显著改善长上下文场景下的性能-计算权衡。
链接: https://arxiv.org/abs/2608.29539
作者: Yuqi Pan,Zheng Li,Bohao Tang,Zhen Qin,Guoqi Li
机构: 未知
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:As context lengths scale, attention increasingly becomes a primary computational bottleneck in large language models. Standard Transformers remain powerful but computationally inefficient, as they allocate the same attention budget to every token regardless of its contextual demand. Existing local-global hybrids provide a more efficient alternative by mixing restricted- and full-context attention, but they typically allocate span statically across layers or heads. To address these limitations, we propose LoGo, a token-level dynamic local-global attention mechanism that uses attention span as a direct proxy for attention budget allocation. Each LoGo layer contains coupled local and global branches: all tokens receive efficient local attention over a restricted context window, while a learned gate activates global attention with full-context access only for tokens requiring long-range information. A threshold-based budget controller maintains a target global ratio without auxiliary losses, and a progressive masking schedule stabilizes training before sparse routing takes effect. We further implement query-sparse Triton kernels that convert reduced global-attention computation into practical speedups. Extensive experiments validate LoGo’s effectiveness, showing that it preserves the scaling behavior of full-attention Transformers across model sizes. In controlled comparisons, LoGo improves over the full-attention Transformer and matched-budget static local-global hybrids, with clear gains on long-range retrieval. Analysis further shows that LoGo learns interpretable span allocation patterns. These results suggest that learned token-level span allocation is an effective and scalable way to improve the long-context performance-compute trade-off.
[NLP-193] he Emergent Symbolic Structure of Artificial Neural Networks
【速读】： 该论文试图解决的核心问题是：尽管现代人工智能（AI）系统基于连续向量表示的神经网络，而传统认知科学认为智能应建立在符号结构（symbolic structure）之上，但这些神经网络在语言、逻辑、算术等典型符号性任务中仍表现出卓越性能。这种看似矛盾的现象引发了根本性疑问——神经网络如何在缺乏显式符号操作的情况下实现对复杂结构化知识的处理？其解决方案的关键在于提出并验证了一个核心假设：神经网络内部的向量表示虽不显式呈现符号结构，但其内在组织方式实际上隐含了符号结构。研究通过实证发现，无论是小型网络处理列表运算，还是大型语言模型（LLM）在算术、逻辑、编程代码和自然语言四个经典符号领域中的表现，均可被精确的闭式方程所生成的符号结构近似替代，且模型行为基本保持不变。更重要的是，通过对这些识别出的符号结构进行精准干预，可定向改变LLM的行为，证明其决策依赖于这些隐含的符号结构。这一发现为调和符号主义与连接主义在智能本质理解上的长期分歧提供了新的理论路径，揭示了现代神经网络可能以“隐式符号化”机制实现对结构化认知域的有效建模。
链接: https://arxiv.org/abs/2608.29530
作者: R. Thomas McCoy,Paul Soulos,Tal Linzen,Paul Smolensky
机构: Yale University (耶鲁大学); Johns Hopkins University (约翰霍普金斯大学); New York University (纽约大学); Microsoft Research (微软研究院)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 30 pages, plus 29 pages of references and appendices
Abstract:Modern systems in artificial intelligence (AI) somehow excel in domains for which they seem poorly suited. Intelligence has traditionally been modeled as operating over structured combinations of symbols, such as logical formulas. However, the strongest modern AI systems are based on neural networks, which instead represent information in continuous vectors. Vectors seem inadequate for capturing the structure of language, logic, and other cognitive domains, yet neural networks achieve impressive performance in these areas. How do they do it? In this work, we propose a potential answer: Despite appearances, perhaps the internal representations of neural networks implicitly realize symbolic structure. In support of this hypothesis, we show that the vector representations of a variety of neural networks can be closely approximated with symbolic structures: we can replace the network’s entire representation-generating process with a closed-form equation instantiating a symbolic structure, and the network’s behavior remains largely unchanged. This finding holds for both small-scale neural networks trained to manipulate lists as well as large language models (LLMs) operating in four domains that are central in symbolic traditions: arithmetic, logic, computer code, and language. Further, our symbolic approximation allows us to modify an LLM’s behavior in targeted ways via precise interventions on its internal representations, showing that the LLM’s behavior is reliant on the symbolic structures we have identified. This work provides a potential way to reconcile longstanding symbolic conceptions of intelligence with the vector-based nature of modern AI.
[NLP-194] Argument-Aware Semantic Alignment of Normative Texts: A Toulmin-Based Neuro-Symbolic Approach
【速读】： 该论文旨在解决专业化规范文本（如网络安全标准）中等效要求因术语、句法和抽象层级差异而导致的语义对齐难题。传统方法依赖词汇重叠、分布表示（distributional embeddings）及语义相似度计算，虽能捕捉主题相关性，却往往忽略规范性主张的论证结构（argumentative structure），即论点如何被支持、限定与证明。本文提出关键解决方案：将跨标准映射视为具有论证意识的语义对齐任务，并构建一个神经符号（neuro-symbolic）流水线，融合神经文本表征与图灵模型（Toulmin model）特征。通过大语言模型（LLM）显式化步骤识别主张（claim）、依据（ground）、规范（warrant）、限定词（qualifier）和支撑（backing），重构省略推理（enthymeme），并以论证感知的相似性与结构特征输入对齐模型。在NERC-CIP到NIST-CSF的基准测试中，基于论证特征的改进显著优于仅依赖神经语义的基线模型；特征选择分析表明，尤其是与规范（warrant）相关的特征贡献最为显著，说明论点与其支持理由之间的关联无法通过传统相似度方法充分捕获。进一步发现，仅保留主张-依据-规范这一精简子集即可保持与完整图灵特征集相当的性能，验证了论证结构作为中间表示的有效性。研究结果为论证结构在对齐专业化规范文本中的互补作用提供了初步证据，尽管实验以网络安全标准为受控测试场景，尚未证明其跨领域泛化能力。此外，由LLM生成的论证图谱亦可为后续的规范文本检索、推理与解释提供支持。
链接: https://arxiv.org/abs/2608.29529
作者: William Schroeder
机构: Purdue University (普渡大学); CTO Cleantech Software
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 11 tables, 8 figures
Abstract:Semantic alignment between specialized normative texts is challenging when equivalent requirements use different terms, syntax, and levels of abstraction. Lexical overlap, distributional embeddings, and semantic similarity capture topical relatedness but often miss the argumentative structure by which normative claims are supported, qualified, and justified. This paper asks whether explicit argument structure adds information complementary to neural semantics for aligning requirements. We treat cross-standard control mapping as argument-aware semantic alignment and build a neuro-symbolic pipeline that combines neural text representations with Toulmin features. An LLM explicitation step identifies claims, grounds, warrants, qualifiers, and backing and reconstructs enthymemes. These feed an alignment model via argument-aware similarity and structural features. On a NERC-CIP to NIST-CSF mapping benchmark, argument-derived features improve alignment over a neuro-symbolic semantic baseline. Feature selection shows especially strong signal from warrant-related features, indicating that the link between a claim and its supporting reasoning is not captured by conventional similarity alone. A compact claim–grounds–warrant subset remains competitive with the full Toulmin feature set. The results give preliminary evidence that argument structure is a useful intermediate representation for aligning specialized normative texts. Cybersecurity standards are used as a controlled testbed, not as proof of domain-independent generalization. The argument graphs produced by LLM explicitation may also support later work on retrieval, reasoning, and explanation over normative text.
[NLP-195] Ontology-Guided Multi-Agent Extraction of Evaluation Objects from Academic Review Texts: Evidence from Chinese Library and Information Science
【速读】： 该论文旨在解决学术评价文本中评价对象（evaluation object）提取困难的问题，尤其针对现有科学实体抽取方法在处理抽象性高、上下文依赖性强且类型边界模糊的评价对象时表现不佳的局限。其核心解决方案是提出一种基于本体引导的多智能体框架，通过候选发现、本体约束分类与领域评论融合三个关键环节实现精准提取。该框架利用本体对类型边界进行显式约束，有效提升细粒度分类能力并减少类别混淆；多智能体协作流程则增强了召回率与系统稳定性。实验结果表明，该方法在多项指标上显著优于基于规则和零样本学习的基线模型，为学术评价文本的结构化利用及证据驱动的研究评估与科技情报（STI）挖掘提供了可靠的方法论支持。
链接: https://arxiv.org/abs/2608.29526
作者: Haolin Chen,Hongyi Dong,Yu Zhu,Yijia Hong,Leiqing Niu,Jiyuan Ye
机构: 未知
类目: Computation and Language (cs.CL)
备注: 13 pages, 1 figure; accepted at ASIST METSTI
Abstract:Academic reviews, scholarly commentaries, and book reviews serve as sources of evaluative statements about theories, methods, literature, institutions, and policies, providing valuable evidence for scholarly evaluation. Existing scientific entity extraction methods mainly target research articles and are less effective for evaluation objects, which are often abstract, context-dependent, and characterized by ambiguous type boundaries. This study proposes an ontology-guided multi-agent framework for evaluation object extraction. The framework combines candidate discovery, ontology-constrained classification, and domain review. Experimental results show that it achieves a Precision of 90.33%, Recall of 84.55%, Entity-level F1 of 87.34%, Strict Typed F1 of 79.78%, and Type Accuracy of 91.35%, substantially outperforming rule-based and zero-shot baselines. Ablation results indicate that the multi-agent workflow improves recall and stability, while ontology-based boundary constraints enhance fine-grained classification and reduce category confusion. The framework supports the structured utilization of evaluative scholarly texts and provides methodological support for evidence-based research evaluation and STI mining.
[NLP-196] LLM Judges as Raters: A Pre-Registered Audit of Severity Halo Reliability and Version Instability in LLM Essay Scoring on Public Corpora
【速读】： 该论文旨在解决生成式 AI（Generative AI）作为作文评分工具在教育评估中应用时存在的潜在偏差问题，尤其是模型评分者（LLM judges）在严重性（severity）、光环效应（halo effect）、评分漂移（drift）及版本间差异等方面的表现。其核心解决方案在于采用多面 Rasch 模型（many-facet Rasch）框架，系统性地对多个语言的公开语料库（ENEM/Essay-BR 与 ASAP）进行预注册的评分者效应检验，涵盖评分严重性、残差光环效应、测量信度/决策研究、跨版本偏差以及项目功能差异等维度。关键发现包括：不同 LLM 评分者的严重性跨度显著（如 ENEM 上达 219 分），且多个版本对比显示严重性偏移超出随机预期；尽管模型间相关性仅在中等水平（0.47–0.56），但其内部一致性较高（φ=0.80），然而并未达到人类专家水平的准确性；进一步分析表明，所谓“隐性漂移”并不存在，评分一致性随严重性变化而动态调整；此外，通过身份探测器（identity canaries）识别出一名被废弃的评分者，并证实即使在相同仪器和校准条件下，模型光环效应并不显著高于训练后的人类评分者范围。因此，该研究强调必须将评分者效应纳入评估框架，以确保生成式 AI 评分系统的公平性与可解释性。
链接: https://arxiv.org/abs/2608.29517
作者: Veerendra Kumar Sunkavalli
机构: 未知
类目: Computation and Language (cs.CL)
备注: 14 pages, 2 figures, 8 tables. Under review at the Journal of Learning Analytics (LAK27 research track). Artifact: this https URL
Abstract:Large language models (LLMs) are increasingly used as essay graders in learning analytics, evaluated almost exclusively with agreement statistics. Educational measurement warns that raters also differ in severity, show halo, and drift as instruments. We treat LLM judges as raters and run a pre-registered rater-effects battery (many-facet Rasch severity, residual halo, generalizability/decision studies, cross-version shifts, differential functioning) on public corpora in two languages (ENEM/Essay-BR; ASAP): 2,377 essays, 12 judges, 4 providers, 5 version contrasts, replicated cells, released as a score tensor. Judge severity spans 219 points on ENEM’s 0-1000 scale; on ASAP the panel spread is 15-33% of the score range against a between-trained-human gap near 1%. Judge-human correlations sit in an undiscriminating .47-.56 band. All five version contrasts shift severity beyond a family-wise permutation null (up to 133 points), and one judge was deprecated mid-study, caught by identity canaries. Two pre-registered tests returned honest nulls: severity-adjusted leaderboard reversals did not survive a permutation null, and “silent drift” was refuted: agreement moved with severity in four of five contrasts. Replication yields self-consistency (phi=.80 at k=2) but not human-level accuracy, and a same-instrument check overturned our own halo comparison: matched on instrument and calibration, we find no credible evidence that judge halo exceeds the trained-human range.
[NLP-197] CoCoA: Context-Conditional Cultural Alignment for Large Language Models EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在跨文化语境中普遍偏向西方相关实体的问题。传统去偏方法追求全局中立性，但无法满足文化偏差缓解所需的上下文敏感行为——即在存在文化线索时优先选择符合语境的文化适切实体，而在无文化线索时保持中立。其解决方案的关键在于提出一种名为CoCoA（Context-Conditional Cultural Alignment）的框架，通过在相同实体对上进行含与不含文化线索的双上下文训练，学习上下文依赖的响应模式。CoCoA结合对比对齐目标、校准项与漂移正则化，并采用目标感知梯度协调机制进行优化。实验在两个以实体为中心的文化偏差评测基准CAMeL和Camellia上展开，覆盖十种语言及四类LLMs，结果表明CoCoA将平均文化偏差得分从43降至24，同时维持接近中立的偏好水平（50.2），且对五项通用基准性能影响微小。研究证实，有效的文化对齐需依赖上下文条件建模而非统一去偏，为缓解大语言模型中的实体中心型文化偏差提供了新范式。
链接: https://arxiv.org/abs/2608.29492
作者: Kyungdon Lee,Wei Xu,Alan Ritter,Dong-Ho Lee,JinYeong Bak
机构: Sungkyunkwan University (成均馆大学); Georgia Institute of Technology (佐治亚理工学院); University of Southern California (南加州大学)
类目: Computation and Language (cs.CL)
备注: Accepted to Findings of EMNLP 2026. 22 pages, 4 figures, 19 tables
Abstract:Large Language Models (LLMs) often favor Western-associated entities across cultural contexts. Conventional debiasing methods aim for uniform neutrality, but cultural bias mitigation demands context-conditional behavior, preferring culturally appropriate entities when cultural cues are present and remaining neutral when they are absent. We propose CoCoA (Context-Conditional Cultural Alignment), a framework that learns this behavior through dual-context training on the same entity pairs under contexts with and without cultural cues. CoCoA combines a contrastive alignment objective with calibration and drift regularization, optimized through goal-aware gradient reconciliation. We evaluate CoCoA on CAMeL and Camellia, two entity-centric cultural bias benchmarks, across ten language settings and four LLMs. CoCoA reduces the Cultural Bias Score from 43 to 24 on average while maintaining near-neutral preferences at 50.2, with minimal impact on general performance across five standard benchmarks. These findings highlight that effective cultural alignment requires context-conditional modeling rather than uniform debiasing, and establish a new direction for mitigating entity-centric cultural bias in LLMs.
[NLP-198] GeoAgent : Evaluating VLM Geolocalization Through Embodied Navigation EMNLP2026
【速读】： 该论文旨在解决当前视觉-语言模型（VLMs）在图像地理定位任务中普遍存在的局限性，即现有研究多集中于静态图像的检索、分类与预测，未能真实还原实际应用场景中所需的动态、交互式探索过程。为此，论文提出了一种基于代理（agentic）环境的基准测试框架GeoAgent，要求多模态智能体通过在街景（Street View）环境中进行具身导航（embodied navigation），逐步收集观测信息并基于序列推理完成地理定位。其解决方案的关键在于引入具身交互机制，使模型能够通过主动探索和上下文积累来提升定位精度，从而突破传统静态图像分析的瓶颈。实验表明，尽管现代VLMs在国家及大洲尺度上表现良好，但在区域级模式识别方面仍存在显著不足；相较于静态基线，采用具身导航的策略可显著提升多项评估指标上的准确率。此外，研究揭示了前沿模型在发达/发展中地区划分上的严重偏差以及在错误先验下的自我修正能力薄弱问题。总体而言，该工作系统地揭示了具身导航与地理空间推理中的关键挑战，并公开发布了代码与GeoAgent环境以推动后续研究。
链接: https://arxiv.org/abs/2608.29483
作者: Arka Mukherjee,Soham Roy,Kartikeya Trivedi,Shreya Ghosh
机构: KIIT Bhubaneswar; IIT Bhubaneswar
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026 Findings
Abstract:Modern Vision-Language Models (VLMs) perform well above the human baseline in image geolocalization, a task critically important in disaster response, OSINT verification, and location privacy. However, most efforts to study AI behavior on the task remain limited to static image-based retrieval, classification, and predictions. We argue that faithful recreation of the task should involve embodied navigation, where a multimodal agent autonomously explores its surroundings to gather observations before submitting a prediction. To this end, we introduce \textbfGeoAgent, an agentic environment-based benchmark that requires agents to navigate Street View environments to refine their geolocalization through sequential reasoning. Our analysis shows that modern VLMs struggle to discern regional patterns while succeeding at country- and continent-level predictions. When compared to static image-based baselines, agentic navigation significantly improves accuracy across established metrics. We also note severe bias in a developed/developing region context across frontier model architectures and poor self-improvement capabilities given incorrect priors. Overall, our work establishes the challenges of embodied navigation and geospatial reasoning. We publicly release our code and the GeoAgent environment: this https URL
[NLP-199] SIC-Agents : Benchmarking and Building an Adaptive Simulator for Pediatric Serious Illness Communication Training EMNLP2026
【速读】： 该论文旨在解决儿科严重疾病沟通（Pediatric Serious Illness Communication, SIC）中临床医生沟通培训可扩展性不足的核心问题。相较于一般对话模拟场景，儿科SIC面临多重挑战，包括多方参与互动、对家长情绪困扰的敏感响应以及高度依赖反馈动态等复杂情境。现有基于大语言模型（LLM）的模拟系统主要优化通用对话质量，未能针对SIC训练所需的课程相关行为进行精准建模。为此，研究团队联合教育专家与儿科临床医生，构建了首个专为儿科SIC培训设计的基准测试套件与仿真框架。提出的PitfallBench与DialogueBench分别从回合级和全程对话层面评估模拟器表现。其关键解决方案是SIC-Agents——一种自迭代改进的框架，能够生成可供临床医生编辑的技能文档，以动态引导模拟器行为。实验表明，SIC-Agents显著优于静态专家提示方法。为推动后续研究，作者已公开发布用于儿科SIC中家长角色模拟的基准数据集。
链接: https://arxiv.org/abs/2608.29481
作者: Zihan Wang,Anita Marie Slominska,Rennie Bimman,Elizabeth Di Flumeri,Amanda Mayappo-Neeposh,Conall Francoeur,Tamara Ellen Carver,Xiao-Wen Chang,Doina Precup,Esin Darici Haritaoglu,Ismail Haritaoglu,Akshatha Arodi,Naomi Goloff
机构: McGill University (麦吉尔大学); Mila – Quebec AI Institute (魁北克人工智能研究所); Montreal Children’s Hospital, McGill University Health Centre (蒙特利尔儿童医院，麦吉尔大学健康中心); Linarite AI; Equal supervision.
类目: Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026
Abstract:Pediatric serious illness communication (SIC) is critically important, yet scalable communication training for clinicians remains limited. Compared with other dialogue simulation settings, pediatric SIC poses additional challenges, including multi-party interactions, response to parental distress and strong dependence on feedback dynamics. Existing LLM-based simulators optimize generic dialogue quality rather than curriculum-contingent behavior required for effective SIC training. In collaboration with educators and pediatric clinicians, we introduce the first benchmark suite and simulation framework tailored to pediatric SIC training. Our benchmarks, PitfallBench and DialogueBench, evaluate simulators both at the turn-level and across full dialogues. We further propose SIC-Agents, a self-improving framework that generates a clinician-editable skill document to guide simulator behavior. Our experiments show that SIC-Agents outperforms static expert prompting. To support future research, we release our benchmarks for parent simulation in pediatric SIC at this https URL
[NLP-200] MUDDLE: Measuring Understanding of Documents under Distractor and Length Effects
【速读】： 该论文旨在解决文档问答系统在面对复杂上下文时的鲁棒性问题，特别是区分长文本上下文与主题相似干扰项（hard negatives）对模型性能的影响。以往研究常将两者效应混杂，导致难以准确评估模型对干扰信息的真实敏感度。为此，论文提出MUDDLE这一受控基准，通过270个经人工标注的问题，每个问题对应单一源文档，并在五种条件下实例化：仅源文档、源文档加两个或四个主题相近的硬负样本、以及源文档加两个或四个长度与来源匹配的随机干扰项。该设计使干扰项的长度和出处保持一致，从而确保准确率差异可归因于主题相关性而非文本长度。实验在Markdown格式下进行，使用大语言模型（LLM）作为评判器，评估三个模型家族的表现。结果显示，对于gpt-5-mini，在两种上下文规模下，主题相近的硬负样本显著降低准确率，而长度匹配的随机干扰项则基本维持在无干扰基线水平，表明模型对主题相似性干扰具有明显敏感性。该研究的关键在于通过严格控制变量分离了上下文长度与主题相关性的影响，为系统性评估上下文退化提供了可复现的基准与方法。
链接: https://arxiv.org/abs/2608.29477
作者: Jason Luo,Saibilila Abudukelimu,Judy Song,Andrew Feng,Shivank Garg,Vasu Sharma,Kevin Zhu
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 12 pages. Accepted to the Context Beyond the Window (CBW) workshop at COLM 2026 (non-archival). Code and data: this https URL
Abstract:Document question-answering systems increasingly answer questions over collections of retrieved documents rather than one clean source, so robustness to distracting context matters as much as reading ability. When such systems fail, it is often unclear whether the context was too long or the distractors were too close to the topic, because prior work tends to conflate these two effects. We present MUDDLE, a controlled benchmark that separates them. MUDDLE uses 270 human-annotated questions, each tied to a single source document, and instantiates every question in five conditions: the source alone, the source with two or four topically similar hard negatives, and the source with two or four random distractors. The random distractors are matched to the hard negatives in length and provenance, so an accuracy gap between the two arms reflects topical similarity rather than length. All five conditions are rendered in markdown, page images, and raw PDF, but the distractor sweep reported here is run in markdown, since a source plus its distractors exceeds current image and PDF input limits. We score answers with an LLM judge across three model families. In the complete markdown sweep, hard negatives lower accuracy more than length-matched random documents at both context sizes for gpt-5-mini, while random documents stay near the no-distractor baseline. The effect is small but directionally consistent, and for gpt-5-mini hard negatives significantly underperform length-matched random distractors when pooled across context sizes. We release the data and evaluation code for a reproducible study of context degradation.
[NLP-201] Chain-of-Thought Faithfulness of Reasoning Models Varies with Where and How Preference Cues Are Delivered
【速读】： 该论文旨在解决生成式 AI（Generative AI）在链式思维（Chain-of-thought, CoT）监控中存在的一种核心可靠性问题：当前的忠实性评估方法假设模型的推理轨迹能够真实反映其决策依据，但实际中偏好线索（cue）可能通过用户消息以外的渠道（如工具返回或原始数据片段）传递，导致模型对这些线索的采纳行为未被充分记录或可检测。其解决方案的关键在于提出FACE-Eval（Faithful Attribution of Cue Effects Evaluation），一个包含5100个样本的大规模评估基准，系统性地考察线索位置（用户消息或工具返回）与显隐性（直接摘要或原始产物）对模型行为的影响。研究通过量化“显式承诺”（verbalized commitment）和“隐式采纳”（unverbalized adoption）两个维度，发现所有测试模型在面对工具返回或隐含线索时，其推理过程中的显式表达显著减弱，而隐式采纳率更高；且即使使用溯源提示（source-attribution prompt）也仅在部分模型上缩小了通道差距，而告知模型推理将被监控则未能稳定改善这一现象。此外，基于大模型（GPT-5.6-Luna 和 GPT-4o-mini）的转录监测器显示，隐式采纳程度越高，检测能力越低，表明在单次调用、预填充工具的设定下，通过工具传递或需推断的偏好信息会严重削弱CoT监控的有效性。
链接: https://arxiv.org/abs/2608.29464
作者: Aryo Pradipta Gema,Neel Rajani,Rohit Saxena,Wai-Chung Kwan,Pasquale Minervini
机构: University of Edinburgh(爱丁堡大学); Miniml.AI
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 42 pages, 21 figures
Abstract:Chain-of-thought (CoT) monitoring assumes that reasoning traces faithfully record the information that shapes a model’s answer. Existing faithfulness tests often place explicit bias cues in the user message, while agents may encounter preferences through tool returns or raw artifacts. We introduce FACE-Eval (Faithful Attribution of Cue Effects Evaluation), a 5,100-sample evaluation that varies cue location (user message or tool return) and explicitness (direct summary or raw artifact). We measure verbalized commitment among cue-following answers and unverbalized adoption among all cued samples. We evaluate 15 open-weight models from eight families, with total parameters ranging from 4B to 1.60T. Every model has lower verbalized commitment for tool-return than user-message cues and for implicit than explicit cues. Unverbalized adoption is higher for tool-return cues on all 15 models and for implicit cues in 28 of 30 model-channel comparisons. A source-attribution prompt narrows the channel gap on seven models, sometimes by increasing user-channel unverbalized adoption, while telling models that their reasoning will be monitored does not reliably close the gap. We also use two transcript monitors (GPT-5.6-Luna and GPT-4o-mini) to detect preference adoption in the largest model of each family. Across 32 model-channel-explicitness cells, higher unverbalized adoption is associated with lower detection ability for both monitors (Pearson r=-0.54 and r=-0.78, respectively). These results suggest that CoT monitoring may be less reliable when preference information arrives through tools or must be inferred from raw artifacts, within the single-call, prefilled-tool setting tested here.
[NLP-202] Benchmark Contamination: A Taxonomy Organized by Defeated Mitigation
【速读】： 该论文旨在解决基准测试（benchmark）评分中因模型泄露（leakage）导致的效度威胁（validity threats）难以有效识别与透明披露的问题。现有分类体系仅关注自动化检测污染，而忽视了报告者在发布时面临的核心问题：在已采取缓解措施的前提下，仍存在哪些未被消除的效度威胁？为此，论文提出一种基于缓解策略的新型分类法，涵盖五类泄露类型——直接（direct）、衍生（derivative）、时间（temporal）、分布（distributional）及获取型（acquired）——覆盖训练阶段与评估阶段的泄露风险。其中，获取型泄露发生在评估过程中，具有单次运行特异性，必须随评分一同记录而非随基准发布。为实现可操作化，研究设计了一种四字段披露协议，允许“未知”作为有效选项，并以CC BY 4.0许可发布其JSON Schema、验证工具及示例。通过两名独立于设计团队的编码员对41份文档进行预注册评估，结果显示各变量的线性加权κ值介于0.00至0.35之间（中位数0.21），低于预期的注册阈值，尤其在分层报告与获取型泄露两类上表现不足；分歧主要集中在判断某变量是否适用，而非文档内容本身。此外，仅有13%的文档报告了数据采集预算，且无一文档完整涵盖全部五类泄露类型。该研究的关键贡献在于构建了新的泄露分类框架、由此衍生的评分侧元数据规范，以及对测量工具可靠性的预注册评估与当前披露实践的实证分析。
链接: https://arxiv.org/abs/2608.29463
作者: Johanna Angulo,Víctor Yeste,Hector Espinos-Morato
机构: Universidad Europea de Madrid(欧洲大学马德里校区); Universidad Europea de Valencia(欧洲大学瓦伦西亚校区)
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 18 pages, 3 figures, 6 tables. Specification, audit instrument, pre-registration and analysis code: this https URL
Abstract:A benchmark score is a joint property of the model, the evaluation harness, the elicitation budget, the sampled population, and contamination status. Leaderboards publish the model and the score, so capability and leakage stay observationally equivalent. Existing taxonomies classify contamination for automated detection, not the question a reporter faces at publication: given the mitigations already applied, which validity threats remain open? We introduce a taxonomy organized by the mitigation each type defeats – direct, derivative, temporal, distributional, and acquired – spanning training-time and evaluation-time leakage. Holding out a private test set closes the first alone. The fifth is acquired during the evaluation itself; because it is a property of one run, it must be recorded with the reported score rather than with the benchmark release. We operationalize it as a four-field disclosure protocol in which “unknown” is a valid entry, released under CC BY 4.0 with a JSON Schema, a validator, and worked examples. Two coders external to the design team applied a pre-registered instrument to 41 documents. Per-variable linear-weighted \kappa runs from 0.00 to 0.35 (median 0.21) over 29 main-pass documents against a single-coder test-retest ceiling of 0.84, collapsing under the class skew the registration anticipated; pooling raises it to 0.46 through chance correction rather than better agreement. Two variables fall below the prevalence-robust threshold registered in advance: strata reporting and the acquired type introduced here. Disagreement concentrates on when a variable applies rather than on what a document states. Elicitation budgets are reported in 13% of documents, and no document addresses all five types. The contribution is the taxonomy, the score-side artifact that follows from it, and a pre-registered measurement of instrument reliability and current disclosure.
[NLP-203] Item-Mean Surrogates: Why Richer Persona Data Fail to Improve LLM s as Human Surrogates
【速读】： 该论文旨在解决生成式 AI（Generative AI）作为人类替代者（human surrogate）的有效性问题，特别是当其基于丰富的人物画像数据（persona data）时，是否能够真实模拟个体的响应行为。研究发现，尽管大语言模型（LLM）在整体平均层面的表现与人类均值高度一致，但这种成功主要源于对项目均值的预测能力；一旦去除项目平均值，模型对个体差异的解释力仅为3.05%，远低于人类自身在重测信度测试中达到的53.6%。关键问题是：当前的模型虽能捕捉到“人-项目”交互中的稳定个体效应（person effect），却无法表征个体在特定项目上的偏离程度（即“人-项目”特异性偏差，person-by-item deviation），而后者是原始响应变异中约8.9倍于个体效应的重要成分。此外，LLM还显著压缩了人类响应分布，表现为更窄的方差、更少的响应类别及失真的分布形态。这一现象被作者称为“项目均值替代”（item-mean surrogacy）。因此，解决方案的关键在于认识到：现有基于大模型的人类替代策略仅能近似项目平均值，而无法复现个体响应的分布特征或个性化偏差，从而无法真正替代真实个体。为此，论文提出四项实证检验标准，用于评估任何声称使用 LLM 作为人类替代者的主张是否成立。
链接: https://arxiv.org/abs/2608.29455
作者: Daehwan Ahn,Chengfeng Mao,Dokyun Lee
机构: 未知
类目: Computation and Language (cs.CL); Computers and Society (cs.CY); Machine Learning (cs.LG)
备注: 37 pages, 6 figures, 14 tables. Includes Supporting Information (main text pages 1-9; SI Appendix pages 1-28)
Abstract:LLMs are increasingly used as human surrogates, often on the premise that richer persona data could make them substitutes or exploratory tools for specific individuals. We test this premise across four datasets covering more than 400,000 participants and more than 6,000 survey items and experimental outcomes. LLMs perform well at the aggregate level: their average responses closely align with average human responses to the same items. But this success largely reflects predicting each item’s average human response. Once each item’s human mean is removed, LLM predictions explain only 3.05% of the remaining respondent-specific variation, far below the 53.6% human test-retest benchmark. Richer personas, model variants, and fine-tuning do not close this gap. In variance analyses, once item means are removed, the reliable remaining signal is person-by-item. It captures how a respondent departs from the mean on a particular item and is about 8.9x larger than the stable person effect. Persona data encode the respondent, but not this item-specific deviation. LLM responses also compress human response distributions, using less spread, fewer response categories, and distorted distributional shapes. We call this pattern item-mean surrogacy. Current LLM surrogates can approximate item averages, but not the distributions or respondent-specific deviations needed to replace individual humans. We propose four empirical tests for LLM-based human-surrogate claims.
[NLP-204] AI Can Be Easily Persuaded in Clinical Decision Making
【速读】： 该论文旨在解决生成式 AI 在高风险医疗决策中是否能够保持客观判断的问题，尤其是在面对来自临床人员的外部说服时。其核心挑战在于：当存在专业权威、身份背景、过往表现宣称、多方意见支持或重复施压等社会性因素时，AI 是否仍能基于证据做出正确判断。研究的关键发现是，尽管生成式 AI 能够处理复杂医学信息，但其决策极易受到说服性输入的影响——包括发言者身份（如资深医师相较于医学生更具影响力）、声称的过往业绩、以及看似合理的临床观点，即便该观点为虚构且与事实相悖。尤其值得注意的是，一个合理构建的错误观点可成功诱导 AI 偏离正确决策，且影响幅度较非权威来源高出约10%。这表明当前的 AI 系统在面对人际说服时缺乏对信息真实性的可靠验证能力。因此，解决方案的关键在于提升 AI 的抗说服能力，使其能够在复杂的人际互动中维持独立、客观和基于证据的判断，从而确保其在高风险医疗场景中的安全与可信应用。
链接: https://arxiv.org/abs/2608.29453
作者: Jiayuan Zhu,Jiazhen Pan,Fenglin Liu,Minhao Hu,Junde Wu
机构: University of Oxford(牛津大学); Stanford University(斯坦福大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:As AI becomes increasingly integrated into clinical practice, it is playing a growing role in medical decision making. Medicine, however, is a high stakes and evidence based field, where decisions can directly affect patients’ lives. It is therefore important to understand whether AI can maintain objective judgment when others try to persuade it. In this paper, we study how easily AI can be persuaded through controlled experiments. We find that professional authority, national background, institutional affiliation, claimed past performance, multiple physicians, supported clinician views, and repeated pressure can all affect AI decisions. Surprisingly, the same persuasive input changes about 10% more cases when it comes from a senior clinician than from a medical student. Simply claiming a better performance history consistently makes the physician more persuasive. More strikingly, a plausible clinician view can persuade AI away from a correct decision even when it is fabricated to support an incorrect answer. This indicates that AI can be strongly influenced by convincing support without reliably determining whether this view from the clinician is correct. Together, these findings suggest that AI can be easily persuaded by what people say, who says it, and how the opinion is presented. Therefore, it is essential for AI to maintain sound judgment under persuasion, enabling its safe and reliable use in high stakes medical decision making.
[NLP-205] Whose Assessment of Distress? Community Perspectives and LLM Alignment on Well-Being Posts
【速读】： 该论文旨在解决当前用于心理困扰检测的生成式人工智能（Generative AI）模型在跨文化、跨群体语境下存在偏差的问题，即这些模型通常采用单一标准化判断标准，难以准确捕捉不同身份群体在情绪表达、脆弱性呈现及求助行为等方面的社区特异性规范。其核心解决方案在于构建一个基于多群体视角的对比性标注研究：通过321名参与者对1,198条来自六个身份型社区的Reddit帖子进行主观判断，生成具有社区特异性的标签数据集。研究发现，尽管情境化内群体评判者与自身社群间存在一定共识（优势比OR = 1.18），但该一致性在不同社区间差异显著。进一步评估九种开源大语言模型（LLM）和四种前沿模型表明，所有模型在社区认为“无至轻度”困扰的文本上普遍高估心理困扰，准确率仅为31%-44%，且以大量假阳性为主；即使是表现最优的GPT-5和Gemini 2.5 Pro也表现出对轻度困扰的系统性夸大，而Claude Opus 4相对保守。值得注意的是，这种过度估计并非源于外部观察者的偏见——非情境化外群体人类合集的误判呈近似对称（18%高估 vs. 19%低估）。因此，模型的偏差本质上体现为一种超越人类评判者（包括情境化内群与非情境化外群）的“心理困扰先验”（distress prior），揭示了当前生成式AI在心理健康应用中潜在的不平等风险，强调需引入社区嵌入式校准机制以实现更公平的部署。
链接: https://arxiv.org/abs/2608.29446
作者: Andrew Aquilina,Xiang Lorraine Li,Yu-Ru Li
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:Judgments about psychological distress are socially situated: what counts as concerning hinges on community norms around emotional expression, vulnerability, and help-seeking. Yet large language models (LLMs) used for distress detection are typically aligned to a single, undifferentiated standard. How well do these models capture the perspectives of the communities whose language they assess? We address this question through a perspectivist annotation study in which 321 participants provided 9,587 judgments on 1,198 Reddit posts spanning six identity-based communities, yielding community-specific labels. Raters in the contextualized in-group condition show a modest tendency to agree more with their community than uncontextualized out-group raters (OR = 1.18), an effect varying significantly across communities. We then evaluate nine open-weight LLM configurations and four frontier configurations against these labels. Open-weight LLMs systematically over-estimate distress: when communities perceive none-to-mild distress, these models achieve only 31-44% accuracy, predominantly producing false positives. GPT-5 and Gemini 2.5 Pro show the same none-to-mild inflation even when their full-sample over/under rates are mixed, while Claude Opus 4 is more conservative. This pattern does not simply mirror an outsider reading position: uncontextualized out-group human aggregates were nearly symmetric, with 18% over-estimation versus 19% under-estimation. Instead, the models that inflate none-to-mild cases exhibit a distress prior that exceeds both contextualized in-group and uncontextualized out-group human judgments. These findings have implications for equitable AI deployment in mental health contexts, where miscalibrated distress detection may unevenly affect the communities being assessed.
[NLP-206] Evaluating the Semantic Specificity of Representation Steering in Language Models
【速读】： 该论文旨在解决大语言模型中局部表征引导（Localized Representation Steering, LRS）技术在修复推理病理时可能存在的虚假有效性问题。现有基准评估容易被表面的标签覆盖所误导，导致误判为推理回路已成功修复，而实际上仅是引入了全局标签偏差。为此，论文提出跨规则迁移诊断框架（Cross-Rule Transfer, CRT），其核心在于通过在模型原本具备能力的规则族上进行评估，来审计表征干预的真实效果。针对普遍存在的“矛盾盲视”（contradiction blindness）逻辑缺陷，实验发现晚期层的LRS干预并未真正修复推理机制，而是强制模型产生错误的矛盾预测——当对模型本已正确处理的规则（基线准确率99.6%）施加引导向量时，性能骤降至40.4%。研究通过四项互补控制实验（直接logit偏差等价性、控制向量标签翻转、跨模型嫁接及早期层引导检查）验证了该诊断，确立了一种严谨的方法论，用以区分真正的推理修复与表面的标签覆盖现象。
链接: https://arxiv.org/abs/2608.29431
作者: Zhangdie Yuan,Andreas Vlachos
机构: University of Cambridge(剑桥大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Localized Representation Steering (LRS) is widely used to correct reasoning pathologies in large language models. However, standard benchmark evaluations can easily be fooled by superficial label overrides, creating a false impression of reasoning circuit repairs. In this work, we propose Cross-Rule Transfer (CRT), a diagnostic framework that audits representational interventions by evaluating them on rule families where the model is natively competent. Evaluating late-layer LRS for a widespread logical failure, contradiction blindness, reveals that the intervention merely injects a global label bias: applying the steering vector to rules the model already handles correctly (99.6% baseline) degrades performance to 40.4% by forcing false contradiction predictions. We support this diagnosis with four complementary controls (direct logit bias equivalence, control vector label-flipping, cross-model grafting, and early-layer steering checks), providing a rigorous methodology to distinguish genuine reasoning repairs from superficial label overrides.
[NLP-207] AlgoWorlds: Benchmarking Tool Use for Global Optimization in Algorithmic Worlds
【速读】： 该论文旨在解决在复杂决策场景中，智能体如何将通过工具获取的信息转化为全局最优决策的问题。传统工具使用评估仅关注方案的可行性，但在现实世界的应用如路径规划与车队调度中，个体决策受共享约束与成本影响，导致可行解仍可能显著次优。因此，核心挑战在于：智能体能否基于工具提供的信息，实现对隐藏问题实例的完整重构，并在此基础上做出全局最优决策。其解决方案的关键是提出AlgoWorlds基准，将形式化定义的组合优化问题转化为部分可观测的决策环境，每个环境包含仅可通过特定任务工具观测的隐藏实例，智能体需提交结构化决策并接受可行性与最优性双重验证。该基准涵盖10类组合优化问题、4个工作负载层级，共240个环境，采用家族特异性确定性程序生成实例，精确算法验证最优解并划分难度等级，同时提供两种结构不同的工具接口。实验评估表明，尽管先进大语言模型（LLM）在多数情况下能生成可行解，但最佳模型仅在38.61%的案例中达到全局最优，且失败多源于虽具备足够信息却未能有效整合、进行全局约束推理或完成决策验证，揭示了当前模型在信息融合与全局推理能力上的根本局限。
链接: https://arxiv.org/abs/2608.29397
作者: Zixiang Xu,Jiaan Wang,Fandong Meng
机构: Weixin AI, Tencent(微信AI，腾讯); University of Southern California(南加州大学)
类目: Computation and Language (cs.CL)
备注: Homepage: this https URL Code: this https URL
Abstract:Tool-use benchmarks generally evaluate whether an agent completes a workflow using appropriate tools and valid arguments. However, feasibility alone is insufficient in real-world decision settings such as route planning and fleet dispatch. Individual choices interact through shared constraints and costs, so a feasible solution may still be substantially suboptimal. This raises a harder question: can an agent turn information gathered through tools into a globally optimal decision? We introduce AlgoWorlds, a benchmark that transforms formally specified combinatorial optimization problems into partially observed decision environments with verifiable global optima. Each environment contains a hidden instance observed only through task-specific information tools, after which the agent commits to one structured decision evaluated for feasibility and optimality. AlgoWorlds contains 240 environments covering ten combinatorial optimization families and four workload levels. Family-specific deterministic programs generate the instances, exact algorithms certify their optima and determine workload levels, and two structurally different tool interfaces present each underlying instance. We evaluate seven leading LLMs, including Claude Opus 4.8 and GPT-5.6 Sol. Achieving global optimality remains highly challenging: although leading models produce feasible decisions in most cases, the best-performing model reaches exact optimality in only 38.61% of cases. Even when agents collect sufficient information to reconstruct the hidden instance, most failures end in feasible but suboptimal decisions. The challenge therefore extends beyond information acquisition to information integration, global constraint reasoning, and decision verification. The project homepage is available at this https URL, and the code is available at this https URL.
[NLP-208] Arabic Safety Alignment as Selective Refusal: An Empirical Study of SFT DPO and Guard Calibration
【速读】： 该论文旨在解决阿拉伯语大语言模型在安全对齐过程中面临的有害提示拒绝与良性/敏感提示误拒之间的权衡问题，即传统单一拒绝率指标无法准确反映这一内在矛盾。其核心解决方案在于采用双维度评估框架：以良性拒绝率（B）衡量对非有害提示的误拒程度，以有害提示拒绝率（H）衡量对有害请求的实际拒绝能力（而非合规性）。研究通过在130次运行中对五个具备阿拉伯语能力的模型进行全人类撰写的数据集AraSafe测试发现，仅使用拒绝导向的监督微调（SFT）会导致模型趋向于全面拒绝，而经过精心选择的混合SFT配置可在保持较低良性拒绝率（B = 14%–23%）的同时实现较高的有害提示拒绝率（H = 90%–93%），其中四个配置在全部三轮测试中均达到H > 90%的目标。此外，直接偏好优化（DPO）和推理防护机制（inference guards）对不同模型的影响存在差异，不具备统一提升效果。在盲评审计中，人工标注者对拒绝行为的一致性达89.0%（Kappa = 0.78），Qwen3Guard与Aya Expanse 32B分别取得88.7%和91.0%的准确率，无显著差异。值得注意的是，经选定的SFT策略虽能提升模型在阿拉伯方言（Arabizi）上的有害提示拒绝表现，但均未达到90%阈值，表明从现代标准阿拉伯语到方言的迁移能力有限。综上，研究强调应基于模型特性进行操作点（operating point）的个性化选择，即根据部署目标筛选并保留能有效提升目标性能的干预措施。
链接: https://arxiv.org/abs/2608.29378
作者: Mohamad Zbib,Ammar Mohanna
机构: American University of Beirut(贝鲁特美国大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: The Fourth Arabic Natural Language Processing Conference(ArabicNLP 2026)
Abstract:Arabic large language models must refuse harmful prompts without over-refusing benign or sensitive prompts, yet a single refusal rate hides this trade-off. We evaluate it using benign refusal B and harmful-prompt refusal H, where H measures refusal rather than harmful compliance. Across five Arabic-capable models and 130 runs on the full human-written AraSafe set, refusal-only supervised fine-tuning (SFT) collapses toward blanket refusal, whereas selected mixed-SFT configurations reach H = 90% to 93% at B = 14% to 23%; four selected configurations exceed the H = 90% target in all three runs, while Fanar does so in two of three. Direct Preference Optimization (DPO) and inference guards change B and H differently across models rather than acting as uniform upgrades. In a blinded 300-response audit, annotator binary-refusal agreement is 89.0% (kappa = 0.78); Qwen3Guard and Aya Expanse 32B reach 88.7% and 91.0% accuracy, respectively, with no conclusive paired difference. Selected SFT raises H on Arabizi for all five models, but none reaches 90%, showing only partial transfer from Modern Standard Arabic. Overall, the results support model-specific operating-point selection: set a deployment target and retain only interventions that improve it.
[NLP-209] BIRD-History: A Benchmark for History-Driven Text-to-SQL with Fine-Grained Knowledge Annotations EMNLP
【速读】： 该论文旨在解决大型语言模型（Large Language Model, LLM）驱动的文本到SQL系统在处理隐含依赖领域特定知识（如业务逻辑、数据规范与分析惯例）的自然语言查询时表现不佳的问题，这类知识往往既未体现在数据库模式中，也未在用户提问中明确表达。现有基准测试未能充分支持对基于历史查询日志方法的评估，因此论文提出BIRD-History这一新基准，包含11个数据库中的1,393个任务，专门用于评估系统利用历史SQL脚本来补全模糊查询的能力。其关键创新在于设计了一种可插拔的检索器，能够从历史SQL脚本中提取五类外部知识，并对相关片段进行检索与重排序，从而增强查询生成过程中的知识利用能力。该检索器可无缝集成至现有的少样本文本到SQL流水线中，无需修改提示模板。实验结果表明，该方法在四个主流文本到SQL系统上均实现持续性能提升，验证了利用历史查询日志对处理不完整查询的重要性。
链接: https://arxiv.org/abs/2608.29345
作者: Yunfan Zhou,Qiming Shi,Yizhou Yang,Di Weng,Yingcai Wu
机构: State Key Lab of CADCG, Zhejiang University(浙江大学计算机辅助设计与图形学国家重点实验室); School of Software Technology, Zhejiang University(浙江大学软件技术学院)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Accepted at Findings of the Association for Computational Linguistics: EMNLP, 2026
Abstract:While recent Large Language Model (LLM)-based text-to-SQL systems achieve impressive performance on standard benchmarks, they struggle when user queries implicitly rely on domain-specific knowledge, such as business logic, data conventions, and analytical practices, that is neither captured by the schema nor explicitly stated in the natural language question. Historical SQL query logs offer a valuable source of such knowledge, yet existing benchmarks do not adequately support evaluation of history-driven approaches. To address this gap, we introduce BIRD-History, a benchmark consisting of 1,393 tasks across 11 databases, designed to evaluate text-to-SQL systems’ ability to ground underspecified natural language questions using historical SQL scripts. Each task is annotated with ground-truth labels specifying which historical queries contain relevant knowledge and which SQL clauses encode it, enabling systematic evaluation of both retrieval effectiveness and knowledge utilization. Alongside the benchmark, we propose a plug-in retriever that extracts five types of external knowledge from historical SQL scripts, then retrieves and reranks relevant fragments for query generation. The retriever integrates seamlessly into existing few-shot text-to-SQL pipelines without requiring prompt modifications. Experiments demonstrate consistent improvements across four text-to-SQL systems, highlighting the value of leveraging historical query logs for handling underspecified queries. Dataset and code are open-sourced on this https URL.
[NLP-210] When to Adapt: Conditional Memory Adapters for Retention-Preserving Domain Specialization EMNLP2026
【速读】： 该论文旨在解决在专用领域部署的大语言模型（Large Language Models, LLMs）在提升特定领域性能的同时，如何避免损害其通用能力的问题。现有参数高效微调方法通常为“始终开启”模式，即对所有输入均施加学习到的扰动，导致域外（Out-of-Domain, OOD）性能下降。本文提出Engram Adapter框架，通过复用预训练阶段的条件记忆（conditional memory），将其实现为冻结大模型的后处理适配器。其关键创新在于采用基于局部n-gram模式的多通道匹配机制，并引入显式占用追踪作为轻量级选择性先验，使残差注入更倾向于发生在域内输入上；同时，通过一个可学习的标量门控机制抑制不连贯的域外检索结果。实验在Qwen3-4B和Qwen3-8B模型上进行，以AG-News和MedMCQA为域内任务，涵盖推理、翻译、代码生成和法律推理的多个域外基准测试。结果表明，Engram Adapter在提升域内准确率的同时，保持了99.4%–100.1%的平均域外性能；在LegalBench上甚至略优于冻结基线模型，而对比的始终开启型基线则出现显著性能下降。机制分析显示，尽管域外激活值非零，但门控与投影衰减将残差压缩至隐藏状态范数的约0.08%，从而实现极小的KL散度漂移和可忽略的准确率变化。这表明，条件激活是一种极具前景的路径，可在保持模型原有知识的前提下，实现模块化且保留能力的冻结骨干网络领域特化。
链接: https://arxiv.org/abs/2608.29327
作者: Jiayu Hou,Lei Wang
机构: University of Electronic Science and Technology of China (电子科技大学)
类目: Computation and Language (cs.CL)
备注: Accepted to Findings of EMNLP 2026
Abstract:Large language models deployed in specialized domains must improve in-domain performance without sacrificing general capabilities. Existing parameter-efficient fine-tuning methods are typically always on: their learned perturbations are applied to every input, which can degrade out-of-domain (OOD) performance. We propose Engram Adapter, a framework that repurposes pretraining-time conditional memory as a post-hoc adapter for frozen LLMs. It uses multi-channel matching over local n-gram patterns with explicit occupancy tracking as a lightweight selectivity prior, making residual injection more likely on in-domain inputs while a learned scalar gate suppresses incoherent OOD retrievals. We evaluate on Qwen3-4B and Qwen3-8B with AG-News and MedMCQA as adaptation tasks and OOD benchmarks spanning reasoning, translation, code generation, and legal reasoning. Engram Adapter improves in-domain accuracy while preserving 99.4%–100.1% of average OOD performance; on LegalBench it slightly exceeds the frozen base model on average, whereas comparable always-on baselines degrade sharply. Mechanistic analyses show that although OOD activations are non-zero, gate and projection attenuation reduce residuals to approximately 0.08% of hidden-state norm, yielding small KL drift and negligible accuracy change. These results suggest conditional activation is a promising route toward modular, retention-preserving domain specialization over frozen backbones.
[NLP-211] StageWell: A Process-Aligned Chinese Corpus for Positive-Psychology Support Dialogue
【速读】： 该论文旨在解决正向心理学对话（Positive Psychology Dialogue）中模型生成缺乏过程连贯性与支持功能精准性的问题，即现有方法通常仅关注单轮回复的共情性或整体偏好，而忽视了支持过程中的阶段定位、支持功能演进及局部修复目标等关键要素。其解决方案的核心在于提出一个结构化、过程对齐的中文正向心理学对话数据集StageWell，以及一套名为HQS（Hierarchical Quality Specification）的结构化数据构建与评估协议。HQS通过六阶段支持流程框架明确划分对话支持进程，并采用多智能体全对话重写工作流，构建了包含12,445个监督微调（SFT）样本、1,849对直接偏好优化（DPO）偏好对以及120个专家修订的高质量对话和977组问答对的数据集。其中，DPO偏好对的设计关键在于：以存在缺陷的模型输出作为拒绝响应，基于相同上下文与阶段约束进行针对性修正后作为优选响应，实现“过程局部化修复”的精细化监督。在四款9B-14B规模的开源大语言模型上验证表明，该方法显著提升了模型在支持流程控制、回复质量与安全性方面的表现，平均BERTScore提升0.037，Q-Overall提升1.32分，S-exact提升0.236，关键错误率（H-critical rate）降低0.167，充分证明将支持性对话建模为结构化的多轮支持过程，相较于传统单轮生成范式具有显著优势。
链接: https://arxiv.org/abs/2608.29326
作者: Yuxiong Wang,Ziwei Lin,Bo Wang,Yu Zhang,Shiguang Ni
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 29 pages, 20 figures
Abstract:Positive psychology dialogue aims to support emotional distress and positive resource building, requiring models to produce not only empathetic replies but also coherent progression through a multi-turn support process. Existing resources often reduce supervision to turn-level strategies or holistic preference labels, leaving process position, support function, and local repair targets implicit. We introduce StageWell, a process-aligned Chinese corpus for positive psychology dialogue, together with HQS, a structured protocol for data construction and evaluation. StageWell organizes support into a six-stage support process and uses a multi-agent whole-dialogue rewriting workflow to construct 12,445 SFT instances, 1,849 DPO preference pairs, and a GroundTruth subset of 120 expert-revised dialogues and 977 QA pairs. Guided by HQS, DPO pairs are built as process-localized repairs: flawed model outputs are used as rejected responses, and targeted rewrites under the same context and stage constraint are used as chosen responses. Across four 9B-14B open-source LLMs, this supervision yields robust gains in process control, response quality, and safety. Averaged across models, BERTScore improves by 0.037, Q-Overall increases by 1.32 points, S-exact increases by 0.236, and the H-critical rate decreases by 0.167. These results highlight the value of modeling supportive dialogue as a structured multi-turn support process rather than as single-turn response generation.
[NLP-212] Padārtha: Ontology-Grounded Fine-Grained NER Benchmark for Classical Sanskrit
【速读】： 该论文旨在解决经典文献（尤其是梵文古典文本）在命名实体识别（NER）任务中因沿用现代新闻语料设计的标注体系而产生的文化偏见问题。现有标签集将源文化概念强加于非适配文本，导致语义失真。为此，论文提出以文本自身传统为基础构建本体驱动的细粒度标注体系——基于《尼耶亚-吠湿迦》（Nyāya-Vaiśesika）古典印度本体论系统，构建了首个面向梵文的本体锚定细粒度命名实体识别（NER）基准数据集“Padārtha”。其关键在于：1）从经典本体论出发定义18个细粒度类别，归入10个本体节点，并映射至5个标准粗粒度标签，保障与现有基准的互操作性；2）基于学术实体索引对超过12.6K条目进行专家标注，关联《大那摩》（Mahānāma）语料库中的108,335个实体提及，涵盖73,632行诗文，并构建5,000行经专家验证的测试集以覆盖罕见提及。实验首次系统比较生成式模型与传统架构在梵文NER上的表现，发现微调后的生成式模型性能可媲美专用模型，但所有系统在从粗粒度向细粒度过渡时均出现显著性能下降，且对训练中未见的实体提及表现极差。这一局限并非仅由数据稀缺引起，因为即使在微调后，模型对未见实体的召回率远低于已见实体，且在词汇歧义场景下倾向于默认采用多数类解释，暴露了当前方法在泛化能力与语义理解层面的根本挑战。
链接: https://arxiv.org/abs/2608.29324
作者: Sujoy Sarkar,Pretam Ray,Paramhans Shah,Manoj Balaji Jagadeeshan,Akash Gairola,Arjuna S R,Pawan Goyal
机构: Indian Institute of Technology Kharagpur(印度理工学院加尔各答分校); Central Sanskrit University(中央梵文大学); Manipal Academy of Higher Education(曼帕尔高等教育学院)
类目: Computation and Language (cs.CL)
备注:
Abstract:Annotation schemas are not neutral. When applied to classical literature, tag sets developed for modern journalistic texts impose source-culture definitions on texts they were never designed to describe. We instead ground a schema in the tradition of the text itself introducing \textitPadārtha, the first ontology-grounded fine-grained Named Entity Recognition (NER) benchmark for Sanskrit, built on the \textitMahābhārata epic. Our tag set derives from \textitNyāya-Vaiśesika, a classical Indian ontological system, yielding 18 fine-grained categories organized under 10 ontological nodes and mapped onto five standard coarse tags, ensuring interoperability with existing benchmarks. Expert annotators label over 12.6K entries from a scholarly index of named entities, linked to corresponding mentions in the \textitMahānāma corpus, producing fine-grained annotations for 108,335 entity mentions across 73,632 verses, along with a 5,000-verse expert-verified test set sampled to stress rare mentions. We present the first systematic benchmarking of generative NER against traditional architectures for Sanskrit, finding that fine-tuned generative models perform comparably to task-specific systems. However, all systems show a sharp decline from coarse to fine granularity and struggle with out-of-entity mentions unseen during training. The limitation is not due to data scarcity alone, as fine-tuned models recall unseen entities far worse than seen ones and tend to default to the majority sense under lexical ambiguity.
[NLP-213] All You Need Is Non-Commutative Words
【速读】： 该论文旨在解决自然语言处理中传统模型对词序依赖建模效率低、参数冗余以及持续学习能力不足的问题。其核心解决方案在于将词汇标记（lexical tokens）表示为酉矩阵（unitary matrices），并通过有序矩阵乘积编码句子，利用矩阵乘法的非交换性（noncommutativity）天然捕捉词序信息，从而无需依赖位置编码（Positional Encodings, PEs）。该方法的关键创新在于基于酉群代数结构，实现了无需查询、键、值投影的反对称自注意力机制，并支持以较低注意力开销并行组合可变长度文本块。此外，通过引入规范-陪集读出层（canonical-coset readout layer），紧凑地编码所有真实的酉自由度，同时借助嵌套群扩张（nested group extensions）实现持续学习，使新任务扩展算子空间的同时精确保留已有表征。实验表明，该方法在标准文本分类基准上达到或超越词袋（bag-of-words）基线，在IMDB上获得更高准确率，在AG News上表现相当，且仅用64个实值参数替代传统约3万维的词汇空间，充分体现了其参数化的表达高效性。
链接: https://arxiv.org/abs/2608.29314
作者: Carla M. Quispe Flores,Stanley Salvatierra,Renan Cabrera
机构: Colorado School of Mines(科罗拉多矿业学院); Lookia MX; Yardley, PA 19067, USA
类目: Computation and Language (cs.CL)
备注: 11 pages, 3 figures
Abstract:We represent lexical tokens as unitary matrices and encode each sentence as their ordered product. The noncommutativity of matrix product captures word order without positional encodings (PEs). The same algebra yields several capabilities, including antisymmetric self-attention with no query, key, or value projections, and parallel composition of variable-length text chunks at a reduced attention cost. Furthermore, it provides a canonical-coset readout layer that encodes all true unitary degrees of freedom compactly, while supporting continual learning through nested group extensions that enlarge the operator space with each new task preserving prior representations exactly. Across standard text-classification benchmarks, the method matches or exceeds bag-of-words baselines. Achieving higher accuracy on IMDB and comparable performance on AG News. Notably, this is accomplished by replacing the conventional \sim 30,000-dimensional vocabulary space with a dense, 64-parameter real-valued encoding, highlighting the expressive efficiency of our parameterization.
[NLP-214] Super Library Agent : Joint Generation and Maintenance of Multiple Applications Beyond the Single Codebase EMNLP2026
【速读】： 该论文旨在解决在使用大语言模型（LLM）编码代理（coding agents）开发和维护一组相关应用时，因采用逐应用独立处理的流程而导致共享逻辑重复、代码冗余累积以及结构退化的问题。其核心挑战在于如何高效地从多个应用中提取共用组件并构建一个统一的“超级库”（Super Library），以实现跨应用代码复用与长期维护的可持续性。解决方案的关键在于提出一种基于候选引导的代码片段摘要提取机制，结合预提取阶段的代码库整合策略，以及利用提取痕迹与调用图信息进行上下文感知的依赖迁移方法，从而显著提升共享代码提取的召回率，并有效避免因粗略库构建带来的结构性破坏。实验结果表明，该方法在WebGen-Bench和PaperBench基准上，在保持应用功能完整性的同时，大幅降低了冗余度、令牌占用量（token footprint）、行数（LOC）和模型描述长度（MDL），优于零样本基线及传统朴素库构建方式。
链接: https://arxiv.org/abs/2608.29310
作者: Daegyu Sung,Yukyeong Lee,Geon Park,Yumin Choi,Sung Ju Hwang
机构: KAIST(韩国科学技术院); DeepAuto.ai
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Findings of the Association for Computational Linguistics: EMNLP 2026. Project page: this https URL
Abstract:Organizations often develop and maintain portfolios of related applications: independently deployable codebases that share substantial domain logic, interface patterns, or operational conventions. As LLM coding agents are increasingly used to generate and maintain such software, a naive application-by-application workflow duplicates shared logic across codebases and allows prolonged agentic maintenance to accumulate verbosity, dead code, and structural erosion. We introduce the Super Library Agent problem, where an agent sequentially generates a portfolio of N related applications while maintaining a shared Super Library of reusable cross-application components. A minimal sequential scaffold can in principle extract shared code and migrate applications to the evolving library, but in practice suffers from low extraction recall and fragile dependency migration. We address these failures with candidate-guided extraction over code chunk summaries, pre-extraction codebase consolidation, and context-aware migration using extraction traces and call-graph information. Across WebGen-Bench and PaperBench, our method preserves application functionality while significantly reducing redundancy and token footprint (verbosity, token length) over zero-shot, and avoiding the structural erosion introduced by naive library construction, with additional reductions in LOC and MDL. Our code is available at this https URL.
[NLP-215] Detecting and Repairing Hallucinations in Retrieval-Augmented Generation
【速读】： 该论文旨在解决生成式 AI 在基于检索增强生成（Retrieval-Augmented Generation, RAG）框架下仍存在未支持陈述（unsupported statements）的问题，即尽管模型依赖外部文档进行回答，但用户无法区分哪些内容是真实源自检索文本，哪些是模型自行编造。现有研究多止步于检测错误，而忽视了后续修复策略的有效性与实际影响。本文提出一种系统性的修复方法，基于 RAGTruth 基准数据集，将被标记的错误回答拆解为独立的事实性断言，逐项比对检索来源，并测试三种逐步增强的修复策略：删除不支持的断言、用源文本替换、以及重写断言。通过三个来自不同家族的语言模型对 916 个修复后答案进行评估，结果表明所有策略均显著降低被判定为包含未支持内容的答案比例，且三名评估者达成一致排序。其中，删除策略虽保留原文比例最低（64.3%），但纠错效果最显著；而重写策略在保留原答案内容最多（80.1%）的同时实现最小程度的减少。值得注意的是，修复不仅作用于错误答案，约 83.5% 的原本标注为“干净”的答案也被修改，说明当前模型输出中普遍存在过度自信或冗余表述。因此，不同修复策略本质上代表了“事实性保障”与“原始信息保留”之间的权衡，其选择应依赖于对答案实用性的人工判断，而非仅依靠自动指标。
链接: https://arxiv.org/abs/2608.29307
作者: Sai Krishna Reddy Mulakkayala,Niki van Stein,Aske Plaat
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 15 pages, 3 figures, 7 tables. Submitted to BNAIC/BeNeLearn 2026 as a Type A paper
Abstract:Language models increasingly answer questions by consulting retrieved documents rather than memory alone, a design now common in search assistants and enterprise knowledge tools. Grounding a model in retrieved text reduces unsupported statements but does not eliminate them, and a reader cannot tell a grounded sentence from an invented one. Most research on this problem stops at detection, yet flagging a faulty answer changes nothing for the person reading it, and little is known about which action should follow. Using RAGTruth, a benchmark whose unsupported passages are annotated by hand, we split each flagged answer into individual factual claims, check each against the retrieved source, and compare leaving the answer untouched with three repair strategies of increasing richness: deleting an unsupported claim, replacing it with source text, and rewriting it. Three language models from different families judge the 916 repaired answers. Every strategy reduces the proportion of answers judged to contain unsupported content, and all three judges agree on the ordering. Deletion achieves the largest reduction while retaining least of the original answer, at 64.3% of the text, whereas rewriting retains 80.1% and reduces least. Repair is not confined to faulty answers: 83.5% of answers annotated clean are edited too. The strategies occupy different points on a grounding preservation trade-off rather than forming a quality ranking, and choosing between them needs evidence about answer usefulness that automatic metrics cannot supply.
[NLP-216] Learning Simple Test-Time Environments for LLM Web Agents
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）代理在从结构化人工环境迁移到复杂真实世界场景时性能显著下降的问题。现有研究多归因于LLM在多个简单、结构化环境组合上的组合泛化能力不足，而本文提出一种新思路：让LLM网络代理在测试阶段学习对复杂环境观测进行分解。其解决方案的关键在于引入一种无需标签的测试时环境分解（Test-Time Environment Decomposition, TTED）方法，使代理在推理过程中通过试错步骤将复杂环境观测分解为子模块，并基于这些子模块中的经验动态调整行为。实证结果表明，该框架在合成与真实基准上均有效，实现了（1）在简单子环境中积累的经验可被有效组合以提升整体任务表现，以及（2）在子环境上进行测试时训练能显著增强代理在真实网络自动化任务中的组合泛化能力。此外，论文还提供了对无标签学习算法设计的重要洞察，强调随着LLM代理面临更复杂环境，测试时学习环境分解能力将成为实现鲁棒真实部署的关键。
链接: https://arxiv.org/abs/2608.29305
作者: Junxuan Li,Zijun Liu,Ziyi Huang,Peng Li,Yuzhou Liu,Ming Yan,Yang Liu
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Code and data are released at this https URL
Abstract:Large language model (LLM) agents have demonstrated remarkable proficiency in manually constructed environments, yet their performance frequently collapses when transitioned to complex real-world settings. Existing research largely attribute this degradation to the compositional generalization gaps in LLMs on combinations of multiple simple, well-structured environments. In this work, we propose that LLM web agents can learn simple environment observations at test time. Specifically, we introduce trial steps for agents to decompose a complex environment observation into sub-modules, and implement a label-free learning method, Test-Time Environment Decomposition (TTED), to adapt agent behaviors with experience during inference. Our empirical evaluations demonstrate the framework’s efficacy across both synthetic and realistic benchmarks, showing (1) experience gains acquired within simpler sub-environments can be effectively composed to improve performance in the full one, and (2) test-time training on sub-environments can significantly enhance the compositional generalization of agents in real-world web automation tasks. We also provide key insights in the design of the label-free learning algorithm. As more complex environments are accessed by LLM agents, we believe learning environment decomposition skills at test time will be critical for robust real-world deployment.
[NLP-217] Modality Fault Lines: Structural Corruptions Reveal Frag ile Omni-Modal Reasoning EMNLP2026
【速读】： 该论文旨在解决当前多模态大语言模型（omni-modal large language models）在评估中过度依赖干净、同步且可解释的文本-视觉-音频输入所导致的可靠性误判问题。现有评估方法通常基于理想条件下的表现，难以区分模型的成功是否源于稳定的跨模态融合机制，还是仅依赖于完整输入中的表面线索。为此，论文提出“模态故障线”（modality fault line）的概念，即当某一模态虽仍存在且人类可读，但其内部证据结构被扰动时，模型行为开始出现不稳定性。解决方案的关键在于提出SCEval（Structure-Corruption Evaluation）诊断评估协议，通过在保持问题、答案空间和模态通道不变的前提下，对文本、视觉与音频模态分别及联合施加受控的结构扰动，系统性地测试模型在跨模态证据结构受损时的表现。基于来自Social-IQ、OmniBench和VALOR的273个经人工验证的三模态样本，SCEval对15个开源与专有模型进行了评估，结果表明：结构扰动显著降低模型在干净数据上的准确率；文本-视觉损伤构成最稳定的共享故障线；多模态退化并非简单叠加效应，而是呈现非线性特征。这说明，仅在干净输入下取得高精度并不能证明模型在跨模态证据结构不稳定时仍具备可靠性。
链接: https://arxiv.org/abs/2608.29278
作者: Zhaolu Kang,Meixin Wu,Yu Xue,Yingjie He,Qiming Shi,Lei Wei,Yidi Wang,Richeng Xuan,Zhichao Hu
机构: Tencent(腾讯); Peking University (北京大学); Zhejiang University (浙江大学)
类目: Computation and Language (cs.CL)
备注: EMNLP 2026 Findings
Abstract:Omni-modal large language models are increasingly evaluated on clean text–vision–audio inputs, where every channel is present, synchronized, and readily interpretable. Such scores are often taken as evidence of robust cross-modal fusion, but clean evaluation cannot tell whether success depends on stable cross-modal structure or on cues sufficient only in intact inputs. To address this gap, we define a modality fault line: a boundary at which model behavior becomes unstable when a modality remains present and human-interpretable, but its internal evidence structure is perturbed. We introduce SCEval (Structure-Corruption Evaluation) a diagnostic evaluation protocol that keeps the question, answer space, and modality channels fixed while applying controlled structural corruptions to text, vision, and audio individually and jointly. Built from 273 human-verified tri-modal examples from Social-IQ, OmniBench, and VALOR, SCEval evaluates 15 proprietary and open-source omni-modal systems. The results show that structural corruption lowers clean accuracy, text–vision damage forms the most stable shared fault line, and multi-modal degradation is non-additive rather than a simple function of the number of corrupted modalities. Clean omni-modal accuracy therefore does not establish that a model will remain reliable when cross-modal evidence becomes structurally unreliable.
[NLP-218] SHADOWBENCH: Toward Reliable Automatic Evaluation of Semantic Alignment in Autoformalization EMNLP2026
【速读】： 该论文旨在解决生成式数学推理系统在自动形式化（autoformalization）过程中存在的语义对齐问题，即现有评估指标难以区分类型正确但语义错位的陈述，或错误拒绝语义正确但表达方式不同的正确陈述。其核心解决方案是提出一种名为SA-Pass（Semantic Alignment Pass，语义对齐通过率）的新评估方法，该方法引入“影子陈述”（shadows）作为辅助声明，用以刻画目标命题的预期语义。一个生成的形式化陈述需同时满足三个条件方可获得满分：能够通过编译、能推出所有影子陈述（前向验证）、且其合取能蕴含该陈述（后向验证）。该方法被具体实现于ShadowBench基准，这是一个基于Lean 4的完整自动形式化基准，包含178道从硕士到研究级水平、涵盖八个数学领域的难题。实验表明，Claude Code（Opus 4.8）结合Numina-Lean-Agent在该基准上达到61.8%的编译率和11.2%的SA-Pass得分；六种代理配置下的输出与专家判断在二分类上达成98.8%的一致性，验证了该方法在语义对齐评估上的高可靠性。
链接: https://arxiv.org/abs/2608.29270
作者: Hojae Han,Jongyoon Kim,Sanghyuk Park,Dongwook Cheon,Myungjae Jeon,Sunjong Choi,Soonho Kong,Wonseok Heo,Seung-won Hwang,Donghoon Hyeon
机构: Electronics and Telecommunications Research Institute(电子通信研究院); Seoul National University(首尔国立大学); University of Maryland, College Park(马里兰大学学院公园分校); Amazon Web Services(亚马逊网络服务)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: EMNLP 2026
Abstract:Autoformalization translates informal mathematical theorems into code for proof assistants such as Lean. A central challenge is that current evaluation metrics can accept type-correct but misaligned statements or reject correct statements written in a different formulation. Inspired by Pass@ k , we propose SA-Pass (Semantic Alignment Pass), which tests formal statements using auxiliary statements called shadows that characterize the intended statement. A generated statement receives full credit only when it compiles, implies each shadow (forward check), and is implied by their conjunction (backward check). We instantiate SA-Pass in ShadowBench, a Lean 4 full autoformalization benchmark of 178 postgraduate- to research-level problems spanning eight mathematical areas. Claude Code (Opus 4.8) with Numina-Lean-Agent reaches 61.8% compile rate and 11.2% SA-Pass. Across outputs generated by six agentic configurations, SA-Pass achieves 98.8% binary agreement with expert judgments. An early version of ShadowBench served as the benchmark for Track 4 of the ICML 2026 AI4Math Challenge.
[NLP-219] Large Language Models Systematically Favor Popular Options: Evidence and Mitigation Across MCQs EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在多项选择题（Multiple-Choice Questions, MCQs）评估中因选项流行度偏差（popularity bias）导致的评价失真问题。具体而言，现代LLMs倾向于选择更常见的错误选项而非较少被提及但正确的选项，这种倾向与模型置信度误校准（confidence miscalibration）密切相关：当流行选项的正确率下降时，模型仍保持较高的置信度。为系统性地分离和研究这一现象，作者提出一个名为PopMCQ的基准测试，包含六种受控策略，在保持正确答案不变的前提下操纵选项流行度。在最具对抗性的设置下（即所有干扰项均比正确选项更流行），模型选择错误流行选项的比例高达66%。针对此问题，论文提出一种轻量级的推理阶段纠偏方法PopDebias，其通过估计并去除模型预测中的流行度先验来实现去偏，无需微调、测试时无需标签信息（仅需少量校准集用于参数拟合），且计算开销极低。在22个开源LLM（参数规模从0.5B到32B）上的实验表明，该方法在强流行度压力下可带来最高达54.1个百分点的准确率提升，验证了其有效性与普适性。
链接: https://arxiv.org/abs/2608.29257
作者: Abdelrahman Abdallah,Mohammed Ali,Bhawna Piryani,Mahmoud Abdalla,Adam Jatowt
机构: University of Innsbruck(因斯布鲁克大学); Chungbuk National University(忠北国立大学)
类目: Computation and Language (cs.CL)
备注: Accepted at MAIN EMNLP 2026
Abstract:Multiple-choice questions (MCQs) are a standard format for evaluating large language models (LLMs), yet the popularity of answer options can confound evaluation. Modern LLMs systematically prefer popular but incorrect options over less popular correct ones, a vulnerability we call \textbfpopularity bias. This pattern aligns with confidence miscalibration: model confidence remains high even as accuracy collapses for popular options. To systematically isolate this phenomenon, we introduce \textbfPopMCQ, a benchmark with six controlled strategies that vary option popularity while keeping the correct answer fixed. In our most adversarial setting, where all distractors are more popular than the correct option, models choose popular but wrong answers 66% of the time. To mitigate this bias, we propose \textbfPopDebias, a lightweight inference-time correction that estimates and removes a popularity prior from model predictions. It requires no fine-tuning, is label-free at test time (using only a small calibration split for parameter fitting), and adds negligible computational cost. Experiments on 22 open-source LLMs (0.5B to 32B parameters) show consistent improvements, with accuracy gains up to 54.1 percentage points under strong popularity pressure. The code and data are available this https URL
[NLP-220] When Patients Cut In: Extending Clinical Conversational AI Safety to Interruptions EMNLP2026
【速读】： 该论文旨在解决临床语音助手在真实医疗场景中因患者中途打断（interruption）而导致关键临床信息丢失的问题。现有基于级联架构（语音转文本 - 大语言模型 - 文本转语音）的系统在处理非协作式对话时，即便能良好应对合作性对话，仍可能因中断导致必要信息遗漏。然而，当前临床对话型人工智能评估基准普遍假设患者会等待系统完成发言，忽略了实际中断带来的信息损失问题。为此，研究提出一种基于转录文本的中断恢复评估方法，将话语重叠类型操作化为三类：识别性（recognitional）、竞争性（competitive）和过渡性（transitional sub-unit）子单元，并测试四种面向部署、不依赖推理的大语言模型（LLM）配置在四个实验单元中的表现——涵盖问诊（信息收集）与常见问题解答（信息提供）任务，评估指标为是否保留临床必需内容。结果表明，在信息收集任务中，不同模型的目标问题失败率存在差异；而在信息提供任务中，所有模型的失败率均上升。尤其在竞争性打断情境下，四款模型在30次测试中均出现30/30的信息覆盖失败（置信区间95%：88.6%-100%），远超基线水平（其中三款为0/30，Llama为4/30）。此外，简短道歉标记（如“sorry to interrupt”）虽能对部分模型的恢复能力提升数十个百分点，但效果不一致，甚至在某一模型上反而降低恢复表现。因此，研究强调：中断鲁棒性不能以单一分数衡量，评估必须基于具体内容，分单元报告，并与实际部署场景中的中断模式相匹配。
链接: https://arxiv.org/abs/2608.29241
作者: Zachary Ellis,Spencer Hazel,Adam Brandt,Yajie Vera He,Ernest Lim,Jared Joselowitz
机构: Ufonia Limited(乌方尼亚有限公司); Newcastle University (纽卡斯尔大学)
类目: Computation and Language (cs.CL)
备注: Accepted to Findings of EMNLP 2026
Abstract:Clinical voice agents are now deployed in routine care, where real patients do not wait their turn: they interrupt. These systems typically use a cascaded architecture (speech-to-text - LLM - text-to-speech), so when a patient cuts the agent off mid-utterance, clinically required content can be lost even when the model handles cooperative transcripts well. Yet clinical conversational-AI benchmarks almost universally assume patients wait for the agent to finish, missing interruption-induced loss of required content. We present a transcript-based evaluation of interruption recovery, adapting conversation-analytic overlap categories into three operational types (recognitional, competitive, transitional sub-unit) and testing four deployment-oriented, non-reasoning LLM configurations across four cells spanning history-taking (information gathering) and FAQ (information provision), scored on whether the agent preserves the clinically required content. In the gathering cells, target-question failure varied across models; in the provision cells, where arms are directly comparable, failure rose for every model. Rankings differ across cells, and competitive FAQ interruption produced 30/30 provision-coverage failures for all four models (Wilson 95% CI: 88.6-100.0%; baseline 0/30 for three, 4/30 for Llama). A brief apology marker (“sorry to interrupt”) shifts recovery by tens of percentage points, inconsistently across models, and for one it reduces recovery. Interruption robustness therefore cannot be a single score: evaluation must be content-grounded, reported per cell, and matched to the deployment’s interruption profile.
[NLP-221] Anchoring Speech with Semantics: A Multimodal Adapter Mechanism for Automatic Speech Recognition in Low-Resource Languages EMNLP2026
【速读】： 该论文旨在解决低资源自动语音识别（ASR）中因训练语料稀缺导致目标端生成缺乏充分监督信号的问题。其核心挑战在于如何在有限标注数据条件下增强解码器对语义和声学信息的建模能力。解决方案的关键在于提出一种轻量级适配器机制SAMA-ASR，通过引入辅助翻译生成的语义锚点（semantic anchors）和来自语音的声学锚点（acoustic anchor），实现跨模态适应：在词元预测前，将话语级语义信息与语音感知证据联合融入解码器状态。该机制可无缝集成至类似编码器-解码器多任务语音模型中，且在推理阶段无需人工提供理想翻译作为先验，而是由上游语音到文本翻译模型自动生成语义锚点。实验在两个30小时的低资源汉藏方言数据集（台湾话与客家话）上验证了SAMA-ASR的有效性，显著优于仅依赖声学特征、提示工程或纯语义引导的基线方法，并在实际自动语义锚生成场景下仍保持优异性能；进一步分析表明，小型语音-文本翻译模型即可生成有效语义锚，凸显其高效性与实用性。
链接: https://arxiv.org/abs/2608.29239
作者: Kuan-Tang Huang,Cheng-Yeh Yang,Chien-Chun Wang,Hung-Shin Lee,Hsin-Min Wang,Berlin Chen
机构: National Taiwan Normal University(国立台湾师范大学); EZAI; Academia Sinica(中央研究院)
类目: Computation and Language (cs.CL); Sound (cs.SD); Audio and Speech Processing (eess.AS)
备注: Accepted to EMNLP 2026 (Main Conference)
Abstract:Low-resource ASR remains difficult because scarce transcripts provide limited supervised evidence for target-side generation. To address this gap, we propose SAMA-ASR, a lightweight adapter mechanism that augments the decoder with semantic anchors from auxiliary translations and an acoustic anchor from speech; in principle, the mechanism can be applied to similar encoder–decoder multitask speech models. Through cross-modal adaptation, SAMA-ASR conditions decoder states on translation-derived semantic embeddings and a speech embedding, combining utterance-level meaning with speech-grounded evidence before token prediction. At evaluation time, these semantic anchors can be generated automatically by an upstream speech-to-text translator rather than supplied as oracle translations. Experiments on two 30-hour datasets covering the low-resource Sinitic varieties Taiwanese Hokkien and Hakka show that SAMA-ASR improves over acoustic, prior prompt-based, and semantic-only translation-guided baselines and remains effective in practical automatic semantic-anchor settings; translator-capacity analyses show that useful semantic anchors can be produced by a compact ST model.
[NLP-222] Attribute-Based Activation Steering of LLM s for Group-Specific Explanation Generation EMNLP2026
【速读】： 该论文旨在解决如何生成针对特定人群的个性化解释（personalized explanation）这一关键问题，尤其关注在面对新主题时，如何使解释内容充分适配目标群体的认知背景与理解能力。现有方法仅依赖提示工程（prompting）难以实现有效的个性化，且缺乏系统性的计算框架支持。为此，论文提出一种基于激活工程（activation engineering）的精细化控制方法：首先识别目标群体在解释风格（explanatory style）和知识水平（knowledge level）方面的群体特异性属性，随后构建基于属性的引导向量（steering vectors），并在大语言模型（LLM）推理过程中将这些向量注入其内部激活状态，从而实现对生成内容的细粒度调控。实验结果表明，该方法在解释的针对性（specificity）和事实准确性（factuality）方面均显著优于传统提示法及当前最先进的引导基线方法；同时，通过由不同目标群体专家参与的人类评估进一步验证了其在实际应用中的有效性。该方案的核心创新在于将群体属性映射为可计算的激活空间引导信号，实现了从“通用生成”到“群体适配生成”的范式跃迁。
链接: https://arxiv.org/abs/2608.29215
作者: Leandra Fichtel,Janek Prange,Henning Wachsmuth
机构: Leibniz University Hannover(汉诺威大学); L3S Research Center(莱布尼茨信息中心)
类目: Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026 Main
Abstract:To effectively enable people to understand new topics, explanations should be tailored to their backgrounds and abilities. So far, prompting alone has been shown to be insufficient for creating such explanations and other computational methods are missing. Therefore, this paper investigates whether LLMs can be steered to generate explanations that are tailored to a specific group of people. To this end, we propose an approach that first identifies group-specific attributes in terms of explanatory style and knowledge of a specific target group. Building on activation engineering, it then computes attribute-based steering vectors and adds them to the internal activations of an LLM during inference to enable a fine-grained steering. In our experiments, we assess the steering effectiveness of our approach in terms of specificity and factuality of the generated explanations. Additionally, we evaluate the explanations in a study with human experts from different target groups. Compared to prompting and state-of-the-art steering baselines, our approach tailors the explanations significantly better to the target group while largely maintaining factuality.
[NLP-223] How Identity and Opinion Shape Political Sycophancy in LLM s EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在个性化交互中可能产生的政治迎合（political sycophancy）问题，尤其关注模型在用户输入不同上下文时其政治立场的动态变化。现有评估基准多依赖封闭式问答，难以捕捉模型在真实交互中对用户语境的响应灵活性。本文提出一种解耦框架，将政治迎合行为分解为两种独立触发机制：观点迎合（opinion, 即对明确意识形态叙事的顺从）与身份迎合（identity, 即基于人口统计标签的刻板印象反应）。通过450个经人工校验的政治困境作为受控探针，评估13个指令微调后的LLM，研究发现：观点迎合与身份迎合之间存在显著解耦现象——两者间的敏感性并不互为预测因子；当二者同时出现时，其影响效应呈亚加性而非简单叠加。此外，系统级角色设定（system-level personas）主要改变模型的基线立场，但对用户引发的观点或身份驱动的立场偏移影响有限。研究结果表明，LLM的政治立场并非固定属性，而是可交互、可引导的动态状态，凸显了个性化机制可能放大基于身份或观点条件下的行为偏差，提示需在模型设计中加强对抗性偏见的识别与控制。
链接: https://arxiv.org/abs/2608.29198
作者: Li-Ni Fu,Chang-Chih Meng,Chien-Hua Chen,Hen-Hsen Huang,I-Chen Wu
机构: National Yang Ming Chiao Tung University (国立阳明交通大学); Academia Sinica (中央研究院)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Computers and Society (cs.CY)
备注: Accepted to EMNLP 2026 (Main Conference)
Abstract:As Large Language Models (LLMs) increasingly encourage users to disclose personal profiles for tailored assistance, measuring their political alignment becomes increasingly important. However, many existing benchmarks for assessing political behavior rely on closed-ended questions and do not fully capture how a model’s stance may adapt to user-provided context during interaction. We introduce a framework that disentangles two distinct triggers of political sycophancy: opinion (aligning with explicit narratives) and identity (stereotyping based on demographic labels). Using 450 manually-checked political dilemmas as controlled probes, we evaluate 13 instruction-tuned LLMs. We uncover a dissociation: a model’s susceptibility to explicit opinions does not necessarily predict its susceptibility to identity cues, and vice versa. When both signals are present, their effects are generally sub-additive rather than simply additive. Additionally, system-level personas primarily shift a model’s baseline stance while having limited effect on the stance shift caused by user opinion or identity. Ultimately, our results suggest that LLM political stance is interactively and steerably vulnerable rather than being a fixed trait, highlighting how personalization may amplify identity- or opinion-conditioned shifts in the model’s behaviors.
[NLP-224] Locked at the Entrance Open Inside: Where RLVR Narrows the Solution Space
【速读】： 该论文旨在解决生成式人工智能（Generative AI）在强化学习中引入可验证奖励（Reinforcement Learning with Verifiable Rewards, RLVR）后导致的推理广度收缩问题：尽管单样本准确率（pass@1）显著提升，但策略的解空间大幅压缩，限制了测试时缩放（test-time scaling）的收益。其核心问题是：推理轨迹中解空间的丧失发生于“访问”阶段（即未能探索有效解路径）还是“执行”阶段（即已启动但无法完成计算）？为解耦这两个因素，研究者以计时任务（Countdown task）为实验场景，通过分析基于PPO和GRPO训练的Qwen2.5-3B及Qwen2.5-3B-Instruct模型，发现解覆盖度下降高达67%，且即使在所有检查点均能求解的问题上也出现半数以上覆盖率损失。关键发现是，这种收缩高度集中于解的“入口”阶段——在首个算术操作前，逐标记似然变化幅度比下游推理阶段高出11至16倍。通过仅提供未被选择的入口前缀，低访问家族的完成率提升超过一个数量级（如PPO下从0.018升至0.212），表明有效解仍具备可执行性，但初始访问能力已丧失。基于此定位，研究提出入口导向干预策略：晚期层参数插值结合早期检查点，在不牺牲pass@1的前提下使解覆盖度提升37%。进一步验证显示，早期步骤熵坍塌现象在六个数学基准任务中普遍存在于7B与14B模型上，但并非推理优化的必然产物——监督微调（SFT）基线保留的覆盖度超过两倍，分阶段的SFT–DPO–RLVR流程亦能维持早期熵。结论表明，推理广度的损失源于“门”（入口）而非“房间内”（执行过程），揭示了优化方向应聚焦于增强初始解路径的可访问性。
链接: https://arxiv.org/abs/2608.29188
作者: Qiancheng Zhou,Ruizhe Li
机构: Shanghai University (上海大学); University of Birmingham (伯明翰大学)
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Preprint
Abstract:Reinforcement learning with verifiable rewards (RLVR) substantially improves single-sample accuracy (pass@1) but causes the policy’s solution space to contract, diminishing the returns of test-time scaling. In this work, we investigate where inside a reasoning trajectory this breadth is lost: does the policy fail to access a valid solution family, or does it fail to execute computation once initiated? To disentangle access from execution, we analyze the Countdown task, whose solution space can be exhaustively enumerated into discrete entrance families defined by the first operand and operator, across PPO on Qwen2.5-3B and GRPO on Qwen2.5-3B-Instruct. Across both training setups, solution coverage falls by up to 67%, halving even on problems solved across all checkpoints. We show that this contraction is heavily concentrated at the entrance: per-token likelihood shifts are 11x–16x larger prior to the first arithmetic operation than during downstream reasoning. Supplying only an unselected entrance prefix restores completion rates in low-access families by over an order of magnitude (0.018 - 0.212 under PPO), demonstrating that alternative solutions remain executable but are no longer initiated. Guided by this localization, we find that while surface prompting fails to recover diversity, entrance-targeted interventions succeed: late-layer parameter interpolation with early checkpoints increases solution coverage by 37% at no loss in pass@1. Finally, we show that early-step entropy collapse recurs across six math benchmarks with 7B and 14B models, but is not an inevitable byproduct of reasoning optimization: an SFT baseline preserves more than double the coverage, and staged SFT–DPO–RLVR pipelines retain early-step entropy. In summary, reasoning breadth is lost at the door, not inside the room. Code: this https URL.
[NLP-225] oward a Cross-Lingual Romanization Ecosystem for Sinitic Languages: A Paired Mandarin-Cantonese Case Study
【速读】： 该论文旨在解决汉藏语系（Sinitic）语言间缺乏系统性跨语言罗马化对齐的问题，尤其针对粤语、普通话及其他汉语方言在语音转写表示上标准不一、难以互通的困境。其核心解决方案是提出“汉语音系罗马化生态系统”（Sinitic Romanization Ecosystem），构建一个基于四大设计原则的跨语言罗马化框架：音位对应性（phonetic correspondence）、历史音系对应性（historical-phonological correspondence）、一音素一符号（one-phoneme-one-symbol）以及基础拉丁字母使用（basic Latin-letter use），并在实践中权衡各原则间的取舍。在此框架下，研究者开发了粤语（CantRomZJ1）和普通话（MandRomZJ1）的标准化罗马化方案，并扩展至梅州客家话、上海吴语及南京江淮官话等其他汉语方言。为实现实际应用，团队还构建了开源数字基础设施，支持罗马化数据的结构化存储、转换、解析、词典构建与输入法生成。通过基于Meta大规模多语言语音模型（MMS）微调的语音转罗马化实验评估表明，相较于传统的拼音+粤拼基线，所提出的MandRomZJ1+CantRomZJ1组合在粤语词错误率（WER）和字符错误率（CER）上分别降低7.80%和10.61%，验证了跨语言罗马化对齐在低资源汉语音频技术迁移中的有效性。
链接: https://arxiv.org/abs/2608.29170
作者: Zijie Zhang,Tan Lee,Yong Cao,Benyou Wang
机构: 未知
类目: Computation and Language (cs.CL)
备注: Accepted to ISCSLP 2026. Tan Lee and Benyou Wang are co-corresponding authors
Abstract:This paper proposes the Sinitic Romanization Ecosystem, a cross-lingual Sinitic romanization design framework with supporting digital infrastructure and a community-driven open-source workflow. The design framework addresses the lack of systematic cross-lingual romanization alignment among Sinitic languages through four design principles: phonetic correspondence for representing similar sounds with similar romanized symbols, historical-phonological correspondence for aligning cognate romanization strings, one-phoneme-one-symbol, and basic Latin-letter use, with a balancing consideration recognizing trade-offs among these principles. For the main paired case study, we devel-op CantRomZJ1 and MandRomZJ1, Cantonese and Manda-rin romanization schemes following the design framework, respectively. We also develop schemes for several other Sinitic languages, including Meixian Hakka, Shanghai Wu, and Nanjing Jianghuai Mandarin, following the same de-sign framework. To bring the romanization schemes into practical use, we develop open-source infrastructure for structured romanization storage, conversion, parsing, dic-tionary construction, and input-method generation. Finally, we evaluate the design framework through speech-to-romanization experiments based on Meta’s Massively Mul-tilingual Speech (MMS) fine-tuning. Compared with the Pinyin+Jyutping baseline, our Man-dRomZJ1+CantRomZJ1 condition reduces Cantonese WER and CER by 7.80% and 10.61%, respectively. These results suggest that cross-lingual romanization alignment can improve transfer in low-resource Sinitic speech technology.
[NLP-226] Quantifying Error Tolerance in Synthetic Data: An Atomic-level Operand vs. Operator Perturbation Study
【速读】： 该论文旨在解决大语言模型训练中合成数据生成面临的误差容忍度量化分析缺失问题，这一缺陷导致当前数据过滤策略陷入两极分化：过度激进的过滤会误删具有潜在价值的数据样本，而过于宽松的过滤则无法有效剔除错误样本。为克服这一瓶颈，本文提出原子树操作建模（Atomic Tree Operation Modeling, ATOM）框架，将数据分解为功能单元 $ f(x) \rightarrow y $，并区分良性操作数（Operand）扰动与致命操作符（Operator）扰动。实验结果揭示出双重解离现象：模型对操作数扰动具有鲁棒性，但对操作符扰动极度敏感。因此，相较于追求高精度的操作数，优先保障操作符多样性更能提升模型性能。基于此，采用ATOM生成的合成数据在多个基准测试中表现优于严格基线（如较LIMA提升3.1%），表明操作符多样性对模型训练更具关键作用。
链接: https://arxiv.org/abs/2608.29144
作者: Jiaxiang Liu,Chenhao Yuan,Shuwen Xu,Boxuan Xing,Xiusheng Huang,Yinhao Xu,Hao Liu,Wenhao Teng,Xiangwen Liao,Pengfei Cao,Jun Zhao,Kang Liu
机构: CDL, Institute of Automation, Chinese Academy of Sciences(中国科学院自动化研究所计算智能实验室); School of Artificial Intelligence, University of Chinese Academy of Sciences(中国科学院大学人工智能学院); Department of Gastrointestinal Surgery Fujian Provincial Cancer Hospital(福建省肿瘤医院胃肠外科); College of Computer and Data Science, Fuzhou University(福州大学计算机与数据科学学院)
类目: Computation and Language (cs.CL)
备注:
Abstract:Synthetic data generation has become a cornerstone for advancing large language models. However, the lack of the quantitative analysis for error tolerance became a critical bottleneck. Consequently, current filtering strategies fluctuate between two extremes: they are either overly aggressive, risking the exclusion of potentially valuable samples, or overly permissive, failing to eliminate erroneous samples effectively. To bridge this gap, this paper introduces Atomic Tree Operation Modeling (ATOM), a framework that decomposes data into functional units ( f(x)\rightarrow y ). ATOM distinguishes benign Operand x perturbations from fatal Operator f perturbations. The former are needlessly discarded by aggressive filtering, while the latter slip through permissive filtering. Our experiments reveal a double dissociation: models are robust to operand perturbations but collapse under operator perturbations. By prioritizing operator over aggressive operand precision, our ATOM-synthesized data outperforms rigorous baselines (e.g., +3.1% gain over LIMA), suggesting that operator diversity matters more than operand precision. Our code is available at this https URL.
[NLP-227] AI Historian: Helping historians organize and verify person-centred temporal clues from dispersed historical narratives
【速读】： 该论文旨在解决历史文献中人物时间证据分散于不同文本、章节与叙事视角中，导致历史学家在重构时间序列和验证史实过程中面临信息碎片化、效率低下等问题。其核心解决方案是提出AI Historian（AIH）——一种基于生成式AI的智能代理系统，通过将源文本句子作为证据单元，自动识别人物实体与时间线索，验证跨文本关联关系，并推断可比的时间范围，同时保持所有推论的可追溯性与可复核性。该方法显著提升了历史证据组织的效率与准确性：在《史记》六例测试中，AIH在时间定位上的MicroIoU达86.2%，远超人类单独标注的81.3%及直接使用大语言模型提示的17.1%；且仅需约14分钟，较人工耗时1小时32分钟大幅降低。此外，AIH已拓展应用于二十四史及中日韩古代史、近现代史料，成果通过Westlake Historian平台公开，表明该系统能规模化降低历史材料组织成本，并将原本被章回体叙事遮蔽的隐性关联转化为可追踪、可协作验证的研究命题。
链接: https://arxiv.org/abs/2608.29133
作者: Yifeng Lu,Zijie Yang,Jie Li,Qingkai Min,Yue Zhang
机构: Westlake University(西湖大学); Peking University(北京大学)
类目: Computation and Language (cs.CL)
备注: 37 pages, 21 figures. Code: this https URL
Abstract:History is not preserved in complete, continuous form. Accounts of a person’s activities, relationships and historical contexts are scattered across texts, chapters and narrative perspectives; historians must retrieve, identify and compare these materials to reconstruct temporal sequences and verify them against sources. Here we present AI Historian (AIH), an AI agent system that helps historians organize person-time evidence from dispersed biographical narratives. It takes source sentences as evidence units, identifies people and temporal cues, verifies candidate cross-text associations and infers comparable temporal ranges while preserving traceable source-text evidence. We evaluated AIH on six Shiji cases concerning Liu Bang, Xiang Yu and Xiao He. AIH Agent achieved a temporal-localization MicroIoU of 86.2%, compared with 81.3% for human-only annotation and 17.1% for direct large-language-model prompting; it required about 14 min, versus 1 h 32 min for human-only annotation. We further applied AIH to the Twenty-Four Histories and other ancient Chinese histories, ancient Japanese and Korean histories, and modern and contemporary historical materials, and released the results through Westlake Historian. These results indicate that AIH can reduce the cost of organizing historical materials at scale while turning connections obscured by chapter-based narration into traceable, revisable research questions for collaborative testing.
[NLP-228] HEAR Who Said What: Unlocking Speaker-Attributed Reasoning via Counterfactual Voice Grounding EMNLP2026
【速读】： 该论文旨在解决生成式语音语言模型（Speech Language Models, SLMs）在多说话人环境下的说话人归属推理能力不足的问题，即模型难以准确识别语音对应的说话人身份并基于说话人身份进行有效推理。其核心挑战在于现有模型往往依赖语义先验而非真实的声学语音特征，导致推理机制不具鲁棒性与可解释性。解决方案的关键在于提出一种新型的、基于反事实音频与说话人级困难负样本（Counterfactual Audio with Speaker-level Hard negatives, CASH）的训练数据集，并据此优化出一个300亿参数的模型A2R。该方法通过强制模型区分仅在说话人身份上不同的音频片段，从而引导其优先关注声学层面的语音线索，而非语言内容。实验表明，A2R在新提出的层级化基准HEAR上表现优异，并展现出对多样化下游任务的零样本泛化能力，验证了通过学习精准的说话人归属可激活模型潜在的说话人感知推理能力。
链接: https://arxiv.org/abs/2608.29120
作者: Dongwook Lee,Sangkwon Park,Eunwoo Song,Che Hyun Lee,Youngho Cho,Junho Kim,June Young Yi,Heeseung Kim,Sungroh Yoon
机构: IPAI, Seoul National University (SNU); Department of ECE, SNU; Department of EE, Yonsei University; Department of CSE, SNU; Department of AI, University of Seoul; AIIS, ASRI, INMC, and ISRC, SNU
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Sound (cs.SD)
备注: EMNLP2026 Main Conference
Abstract:Speech Language Models (SLMs) are increasingly deployed in multi-speaker environments, yet their ability to attribute speech to the correct speaker and reason over speaker identities remains unclear. Hence, we introduce HEAR, a conceptually hierarchical benchmark diagnosing the foundational capabilities of speaker-attributed reasoning, comprising 2.4K human-verified samples from 887 diverse multi-party audio clips. Evaluating 20 leading SLMs on HEAR reveals they struggle with these foundational tasks, often relying on semantic priors rather than actual vocal cues. To address this, we present A2R, a 30B model optimized on Counterfactual Audio with Speaker-level Hard negatives (CASH), a dataset designed to guide the model to prioritize acoustic vocal cues over linguistic signals. A2R achieves strong performance on HEAR and exhibits zero-shot generalization to diverse multi-speaker downstream tasks, demonstrating that learned speaker attribution unlocks the model’s latent capacity for speaker-aware reasoning. All resources are available at this https URL
[NLP-229] Emergent Misalignment Is Not Magical
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在特定有害数据集上微调后出现的“涌现性错位”（Emergent Misalignment, EM）问题，即模型在局部有害训练数据上表现良好时，却在广泛任务中产生不可预测且危险的行为。传统研究常将EM视为意外行为或归因于通用错位方向甚至拟人化为“邪恶人格”的形成，但其内在机制缺乏清晰解释。本文的核心贡献在于揭示EM并非随机或不可预测的现象，而是一种由数据决定的可预测泛化现象。其解决方案的关键在于：通过分析基础模型对训练数据与评估提示的表征，发现模型在经历EM训练后的“邪恶程度”可高度由评估提示与训练数据中心之间的表征距离所预测（12个模型-数据集设置下的平均Spearman相关系数为-0.73）。基于此，研究进一步阐明：（1）EM的有效性显著依赖于训练数据格式；（2）不存在跨不同EM模型通用的错位方向；（3）EM效应本质上不同于角色（persona）变化。此外，作者将原有的标量距离泛化度量扩展为数据集特异的泛化方向，该方向在语义保持的提示扰动（如添加随机标记、改写）下仍能稳健预测模型的邪恶程度，优于现有方法。这一成果为理解并防范生成式AI中的潜在安全风险提供了可量化、可解释的理论框架。
链接: https://arxiv.org/abs/2608.29118
作者: Mingxuan Li,Qirun Dai,Heran Wang,Chenhao Tan
机构: The University of Chicago(芝加哥大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Fine-tuning large language models (LLMs) on narrowly harmful datasets can lead to misalignment broadly, a phenomenon known as emergent misalignment (EM). EM poses a challenge for AI safety and our understanding of LLMs. Prior work often frames EM as an unexpected behavior, and explains it by appealing to general misalignment directions or anthropomorphizing it as acquiring an evil persona. However, the mechanisms behind these framings remain obscure. In this work, we show that EM is a predictable and data-dependent generalization phenomenon. By examining the base model’s representation of EM training data and evaluation prompts, we find that evilness after EM training is highly predictable from representational distance: the closer an evaluation prompt is to training data centroid, the more evilness it elicits from EM models after training (with an average Spearman correlation of -0.73 across 12 model-dataset settings). Building upon this analysis, we further demystify EM by showing that (1) its effectiveness changes significantly based on training data format; (2) there is not a general misalignment direction that transfers across different EM models; (3) the effect of EM is fundamentally different from persona changes. Furthermore, we extend the EM generalization metric from a scalar distance to a dataset-specific generalization direction, which robustly predicts EM models’ evilness under semantics-preserving prompt perturbations including appending random tokens and paraphrasing, where other methods do not reliably generalize.
[NLP-230] Recognition-Refusal Misalignment in LLM s: Why Models Answer Structurally Unanswerable Questions EMNLP2026
【速读】： 该论文旨在解决大语言模型在面对结构上无法回答的问题（如计算 cot(-540°) 或判断 (1).startswith(“1”)）时，仍会给出看似合理的错误答案而非选择拒绝回答的问题。其核心发现是：模型在生成前的隐藏状态中已具备对“不可解性”的表征能力，即存在一个线性方向可有效区分可回答与结构上不可能的数学和代码类提示。这一“不可解性识别”方向虽存在于模型内部，但与用于处理有害内容拒绝的典型“安全拒答”方向近乎正交，表明模型虽能识别问题不可解，却未能将此识别信号正确路由至拒答机制。进一步实验显示，沿该识别方向进行生成时的调控可双向、剂量依赖地改变模型对无效输入的行为表现，而随机方向则无此效果。此外，对比基础模型与指令微调模型的结果表明，这种低余弦相似度的几何结构早在预训练阶段即已形成。因此，模型在面对不可解问题时仍自信输出，本质并非编码失败，而是“路由失败”——即模型虽具备可用的“无合法答案”信号，但其安全拒答路径未与该信号对齐。
链接: https://arxiv.org/abs/2608.29109
作者: Yucheng Du,Xiyang Hu
机构: University of Southern California (南加州大学); Arizona State University (亚利桑那州立大学)
类目: Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026 Main Conference. 29 pages, 6 figures
Abstract:Large language models often answer structurally unanswerable questions, such as computing cot(-540°) or evaluating (1).startswith(“1”), instead of abstaining. We ask whether this failure reflects missing recognition or failed routing from recognition to abstention. Across instruction-tuned models from 1.7B to 70B parameters, a single linear direction in the hidden state separates answerable from structurally impossible math and code prompts, showing that models represent impossibility before generation. Yet this recognition direction is nearly orthogonal to the canonical safety-refusal direction that mediates trained harmful-content refusal. An in-domain behavior-defined invalidity-aware direction is closer to recognition, but only partially aligned with it, and remains near-orthogonal to safety refusal. Generation-time steering along the recognition direction changes invalidity-aware behavior bidirectionally and dose-responsively on structural math and code cells, while random directions do not. Base/instruct comparisons further show that the low-cosine geometry is already present at the pretraining endpoint. The confident-on-impossible failure is therefore better explained as a routing failure than as an encoding failure: the model has a usable “no admissible answer” signal, but the safety-refusal pathway is not aligned to use it.
[NLP-231] A Comprehensive Survey on Linguistic Steganography: Methods Countermeasures Evaluation and Challenges EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）时代下语言隐写术（linguistic steganography）领域缺乏系统性综述的问题，即如何整合分散的研究进展以全面理解该领域在新范式下的演进。其解决方案的关键在于构建一个多维度的综合框架，涵盖四大核心轴线：148种隐写方法、60种语言隐写分析（linguistic steganalysis）对抗措施、23种评估指标以及9个开放挑战，并在每个轴线上提供分类体系、文献综述与采纳分析。基于这一框架，论文进一步识别出五个关键范式转变：（1）从修改载体文本转向仅通过提示（prompt-only）生成实现隐写；（2）从启发式安全机制转向可证明的安全性；（3）从白盒对称模型访问转向黑盒或非对称访问模式；（4）从以安全性为中心的设计转向联合优化；（5）从关注文本质量转向聚焦工程实践问题。该综述不仅为研究者提供权威参考，也为负责任地推进语言隐写术在大模型时代的应用指明了路径。
链接: https://arxiv.org/abs/2608.29077
作者: Ruiyi Yan,Chenhui Chu,Zhongliang Yang,Yugo Murawaki
机构: Kyoto University(京都大学); Beijing University of Posts and Telecommunications(北京邮电大学)
类目: Cryptography and Security (cs.CR); Computation and Language (cs.CL)
备注: Accepted by EMNLP 2026
Abstract:Linguistic steganography hides secret messages in natural language text. Large language models (LLMs) have reshaped the field, but a systematic account of how these scattered advances collectively reshape the field in this new era is still missing. We provide one along four axes: 148 steganographic methods, 60 linguistic steganalysis countermeasures, 23 evaluation metrics, and 9 open challenges, each with taxonomies, reviews, and adoption analyses. Cutting across these axes, we identify five specific paradigm shifts in the LLM era: (1) from covertext modification to prompt-only generation, (2) from heuristic to provable security, (3) from white-box symmetric LMs to black-box or asymmetric access, (4) from security-centric designs to joint optimization, and (5) from text-quality concerns to engineering issues. The survey aims to serve as both a reference and a roadmap for practical and responsible linguistic steganography in the LLM era.
[NLP-232] Not All or None: Dynamic Construction of Target-aware Memory Graph for Conversational Stance Detection EMNLP2026
【速读】： 该论文旨在解决真实社交媒体场景中对话式立场检测（conversational stance detection）的挑战，即如何有效利用跨对话会话的历史语句来准确识别用户对特定目标的立场。传统方法或忽略历史信息，或全量使用所有先前对话内容，易引入噪声并降低模型性能。本文提出一种名为目标感知记忆图（TamGraph）的新方法，其核心创新在于设计了一种分步、基于熵引导的回溯机制，能够有选择性地激活与目标相关的历史语句，并动态构建面向目标的图结构，以建模话语间的立场关系。该机制在充分利用相关历史信息的同时，有效抑制无关噪声的干扰，从而显著提升大语言模型（LLM）在英文和中文基准数据集上的对话式立场检测性能。
链接: https://arxiv.org/abs/2608.29066
作者: Yifan Xiang,Bin Liang,Yuqi Huang,Ruifeng Xu,Kam-Fai Wong
机构: The Chinese University of Hong Kong, Hong Kong, China; MoE Key Laboratory of High Confidence Software Technologies, China; The Hong Kong Polytechnic University, Hong Kong, China; Harbin Institute of Technology, Shenzhen, China
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted in EMNLP 2026 main
Abstract:Stance detection is crucial for understanding the underlying attitude of an expression towards a target. Conversational stance detection is a more challenging stance detection task in real-world social media scenarios, as it involves detecting the user’s stance by leveraging the target-related historical statements across conversational sessions. In this paper, we propose target-aware Memory Graph TamGraph, a novel method that dynamically leverages target-related statements for conversational stance detection. Instead of considering all preceding historical conversations or using no prior conversation information for stance detection, our TamGraph employs a stepwise, entropy-guided backtracking mechanism to selectively activate memory from historical conversations and dynamically constructs a target-aware graph to model the stance relations among utterances. This allows the exploitation of target-related information from the conversation history for stance detection while preventing the introduction of noise. Experimental results on both English and Chinese benchmarks demonstrate that our TamGraph substantially improves LLM performance on conversational stance detection.
[NLP-233] RouteSparse: Input-Conditional Pattern Routing for Budgeted Long-Context Prefilling
【速读】： 该论文旨在解决长上下文预填充（long-context prefilling）中动态稀疏注意力（dynamic sparse attention）的效率与性能平衡问题。传统方法如MInference采用固定每头稀疏模式，假设注意力头的偏好模式和稀疏预算在不同输入间保持一致，但这一假设在实际应用中往往不成立，导致性能下降。为此，本文提出RouteSparse，其核心创新在于引入一个轻量级的路由机制，将每个注意力头与提示片段（prompt segment）动态分配至一个小型、GPU友好的稀疏模式库中的最优模式。该方案通过低成本探针（probe）估计模式效用与不确定性，并结合延迟感知的路由器选择合适模式及稀疏预算；当置信度不足时，自动回退至更密集的掩码以保障精度。进一步地，作者将路由过程建模为受限风险最小化问题，并基于被忽略概率质量推导出注意力输出误差的可证伪界（error certificate）。实验表明，在Llama 3.1-8B-Instruct模型上处理128K token提示时，RouteSparse实现了相对于稠密注意力6.5倍的预填充加速，仅带来0.2点的RULER指标下降，显著优于固定每头路由方案（7.3倍加速，1.6点下降）。消融实验验证了输入条件化路由、硬件特性感知以及选择性稠密回退对提升“性能-延迟”权衡的关键作用。
链接: https://arxiv.org/abs/2608.29058
作者: Chao Zhang,Yifan Ji,Ziyan Zhang,Kai Song,Fei Lin
机构: Chongqing University (重庆大学); Chongqing Jiaotong University (重庆交通大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Dynamic sparse attention can reduce the quadratic cost of long-context prefilling without changing model weights. MInference assigns each attention head one pattern offline and estimates that pattern’s sparse indices for every prompt. This design is efficient, but it assumes that a head’s preferred pattern and sparsity budget remain suitable across inputs. We introduce RouteSparse, which routes each head and prompt segment among a small library of GPU-efficient sparse patterns. A low-cost probe estimates pattern utility and uncertainty; a latency-aware router then selects a pattern and budget, while uncertain cases fall back to a denser mask. We formulate routing as constrained risk minimization, derive an attention-output error certificate from omitted probability mass, and evaluate the method on long-context retrieval, question answering, summarization, and language modeling. On Llama 3.1-8B-Instruct with 128K-token prompts, RouteSparse achieves 6.5\times dense prefill speed with a 0.2-point RULER drop relative to dense attention, compared with 7.3\times speed and a 1.6-point drop for fixed per-head routing. Ablations confirm that input-conditional routing, hardware profiling, and selective dense fallback each contribute to the quality–latency tradeoff.
[NLP-234] EmoLASP: Emotion Recognition with Language Models and Answer Set Programming EMNLP2026
【速读】： 该论文旨在解决对话中情感识别（Emotion Recognition in Conversations）任务中，基于大语言模型（Large Language Models, LLMs）的现有方法在处理长对话历史时存在的不稳定性与高昂的微调或提示成本问题。其核心挑战在于如何在不依赖大量上下文输入或昂贵的微调过程的前提下，提升情感预测的准确性与一致性。解决方案的关键在于提出EmoLASP框架，该框架将语言模型与基于答案集编程（Answer Set Programming, ASP）的声明式推理相结合，通过形式化逻辑推理机制对语言模型输出进行后处理，从而增强情感评分（VAD：效价-唤醒-支配度）预测的一致性与鲁棒性。实验结果表明，即使在未提供对话历史的情况下，EmoLASP仍能显著优于仅使用语言模型的方法，尤其在无微调的提示型LLM上表现突出；而对于已微调的预训练模型（PLMs），当对话历史可用时，推理模块的增益有限。这表明EmoLASP通过引入可解释的逻辑推理，有效降低了对长序列提示和微调的依赖，实现了更低的计算成本与更高的预测稳定性。
链接: https://arxiv.org/abs/2608.29035
作者: Thao Le,Michael Thielscher
机构: University of New South Wales (新南威尔士大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 13 pages. Accepted to EMNLP 2026 Main Conference
Abstract:Emotion recognition in conversations is increasingly tackled with language models, but these models can be unstable and expensive to fine-tune or to prompt with long dialogue histories. We propose EmoLASP, a framework that combines a language model with declarative reasoning via Answer Set Programming (ASP) to predict VAD scores (Valence-Arousal-Dominance) in conversations. Experiments on a widely used benchmark dataset (IEMOCAP) across six open-source LLMs (3B-120B) and two PLMs (BERT, RoBERTa) show that EmoLASP improves prediction performance compared to using the language model alone, even when the LLMs/PLMs are given no dialogue history in their prompts or input vectors. The gains are largest for prompt-only LLMs, which EmoLASP uses without any fine-tuning. However, for fine-tuned PLMs, the reasoner adds little once dialogue history is available. EmoLASP’s LLM pipeline demonstrates the potential advantages of using a reasoning approach to ensure emotion prediction consistency and to reduce both the cost of fine-tuning and the cost of prompting with long dialogue histories.
[NLP-235] A Unifying Perspective on Language Model Representations: From Filler-Role Structure to Mechanistic Interpretability
【速读】： 该论文试图解决的问题是：当前多种语言模型可解释性方法（如加法类比、线性探测、稀疏自编码器和激活补丁等）所得出的见解彼此孤立，缺乏统一的理论框架来揭示这些方法背后的共同底层结构。其解决方案的关键在于提出**张量积表示（Tensor Product Representations, TPRs）**作为统一假设，认为语言模型中的组合结构可通过“填充-角色绑定”（filler-role bindings）在向量空间中实现。研究通过数学推导证明，上述主流可解释性方法均可从TPRs中导出；并通过实证实验，在从小型玩具模型到大语言模型（LLM）的不同架构上构建了基于TPRs的可解释性方法变体，其性能与标准方法相当。这一工作为可解释性研究提供了一个统一的理论视角，旨在揭示神经网络的本质，并解释不同方法之间的内在关联。
链接: https://arxiv.org/abs/2608.29034
作者: Zhang Enyan,R. Thomas McCoy
机构: Yale University (耶鲁大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 32 pages, 6 figures
Abstract:A wide range of methods have been proposed for interpreting language models, delivering important insights into their inner workings. However, different methods and their resulting insights stand in relative isolation: what could the underlying structure of language models be, such that they give rise to all our interpretations? In this work, we propose using Tensor Product Representations (TPRs) as a unifying hypothesis. TPRs give a concrete proposal for how compositional structure could be represented in vector space — as filler-role bindings. We show, both mathematically and empirically, that TPRs can unify several prior interpretability methods: additive analogies, linear probing, sparse autoencoders, and activation patching. Mathematically, we show that these methods can all be derived from TPRs. Empirically, we apply the derivations to a range of different models — from small toy models to LLMs — to construct instances of each of the above interpretability methods; these constructed variants perform comparably to their standard variants. We view this work as a step toward what interpretability will ideally provide: a unified account of the nature of neural networks, corroborated not just by individual observations but also by an explanation of the connections between them.
[NLP-236] Detecting and Guiding LLM -Generated Korean Poetry with Interpretable Form-level Features EMNLP2026
【速读】： 该论文旨在解决大语言模型（LLM）在生成现代韩语诗歌时存在的形式失真问题，即输出常表现为“分行散文”而非具有韵律、结构多样性和语言规范性的真正诗歌。其核心挑战在于如何有效区分人类创作与机器生成的韩语诗歌，并指导模型生成更接近人类写作特征的诗歌。解决方案的关键在于提出并量化四个形式层面的语言维度：文本体量（Volume）、句末形式的多样性与连贯性（Structure Variation）、句长不规则性（Rhythmic Irregularity）以及标准正字法的遵循程度（Normative Adherence），并将这些维度转化为五个可解释的特征。基于这些特征，研究构建了一个零样本跨模型检测框架，在七种未见过的LLM上实现平均AUC-ROC为83.60，显著优于最强基线方法KatFishNet（75.84），提升达7.76 AUC点（相对改进10.23%）。在生成方面，通过引入特征引导机制，GPT-5.2和Gemini-3的专家评估及统计分析均显示，生成诗歌在长度、节奏和句尾模式等方面更趋近于人类作品分布，验证了可解释、语言特异性特征在诊断与引导生成任务中的有效性。
链接: https://arxiv.org/abs/2608.28986
作者: Keunhyeung Park,Seunguk Yu,YoungBin Kim
机构: Chung-Ang University (中央大学)
类目: Computation and Language (cs.CL)
备注: Accepted to Findings of EMNLP 2026
Abstract:LLMs often struggle with modern Korean poetry, producing outputs that resemble “line-broken prose.” We address two coupled tasks: detecting whether a Korean poem is human- or LLM-authored, and guiding LLMs to generate poetry closer in form to human writing. We quantify the human-LLM gap along four form-level linguistic dimensions: output length (Volume), the diversity and connective use of line-final forms (Structure Variation), the irregularity of line lengths (Rhythmic Irregularity), and adherence to standard orthography (Normative Adherence). We operationalize these dimensions as five interpretable features. For detection, a logistic regression classifier over these five features attains an average AUC-ROC of 83.60 in zero-shot out-of-distribution detection across seven unseen LLMs, versus 75.84 for the strongest baseline in our comparison, KatFishNet, an absolute gain of 7.76 AUC points and a 10.23% relative improvement; one generator-specific punctuation pattern outside our taxonomy remains a boundary case. For generation, expert evaluation on GPT-5.2 prefers feature-guided poems over the unconstrained baseline, and analyses across GPT-5.2 and Gemini-3 show that targeted length, rhythm, and ending statistics move toward the human distribution. These results suggest that interpretable, language-specific features can bridge the diagnosis and guidance of LLM-generated poetry.
[NLP-237] he Illusion of Replacement: Rethinking Specialized Machine Learning Models in the Foundation Model Era
【速读】： 该论文旨在探讨机器学习中为结构化数据设计的专用架构是否可被基于语言的模型所替代。其核心问题在于：在保持任务可解性的前提下，语言模型能否有效表征并计算数据中的结构性信息。研究通过对2016至2026年间涵盖九种模态的159篇文献进行系统综述，将现有方法归纳为八类表征范式，从纯语言模型到完全专用架构。研究表明，在极端少样本预测、离散符号任务、文本标注的知识图谱以及大规模单模态预训练等特定场景中，语言媒介模型表现出较强竞争力。然而，当对结构表征与结构计算能力进行直接评估而非仅依赖预测准确率时，未发现语言模型可普遍取代专用架构的证据。相反，在多个独立研究领域中均观察到一致模式：当语言模型因缺乏必要结构而表现不足时，研究者倾向于通过引入图模块、结构化标记（structural tokens）、专用注意力机制或其他非语言组件来重构缺失的结构。这表明，专用性并未消失，而是以重新定位的形式持续存在。尽管语言模型性能随规模扩展而提升，但其与具备结构感知能力的专用架构之间的差距是否最终可被消除，尚无实证支持。
链接: https://arxiv.org/abs/2608.28980
作者: Kiyan Rezaee
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 41 pages, 7 tables, 4 figures
Abstract:Can the specialized architectures that machine learning has traditionally built for structured data be replaced by language-based models? This question is examined through a review of 159 papers (2016–2026) across nine modalities, with predictive accuracy considered alongside structural representation and computation. A distinction is made between performing a task and preserving and computing the structure that makes the task tractable, and existing approaches are organized into eight representational regimes, ranging from language-only systems to fully specialized architectures. Language-mediated models are found to be highly competitive in specific settings, including extreme few-shot prediction, discretized symbolic tasks, textually annotated knowledge graphs, and large-scale single-modality pretraining. However, whenever structural representation or computation is directly evaluated rather than accuracy alone, no evidence of general architectural replacement is found. Instead, a recurring pattern is observed across independent research communities: when language alone is insufficient, the missing structure is reintroduced through a graph module, structural tokens, specialized attention, or another non-linguistic component. In this sense, specialization more often relocates than disappears. Moreover, although performance of language-based models is improved by scaling, whether the gap to a structure-aware architecture can eventually be eliminated remains untested.
[NLP-238] RENSA: Rich Environment Metadata to Navigate Shared and Distributed Endpoints for Automated Federated SPARQL Query Generation
【速读】： 该论文旨在解决在联邦式知识图谱数据库环境中高效构建SPARQL查询的难题，核心问题在于分布式数据集间缺乏详尽的结构化元数据，导致现有系统难以进行精确的数据源选择，从而依赖运行时的ASK查询来确定数据来源，造成较高的通信开销。其解决方案的关键是提出RENSA框架，通过扩展SPARQL Builder Metadata（SBM）以集成类（class）信息与权威性（authority）信息，并将主体和客体的使用模式映射到特定谓词上，实现无需运行时通信即可精准进行数据源选择和查询变量的语义约束推断。生成的元数据配置文件在大多数情况下仅占原始数据三元组的1%以下，具备优异的存储效率。实验结果表明，RENSA在LargeRDFBench基准测试中达到与先进方法相当的数据源选择性能，同时完全消除了ASK查询带来的通信开销；此外，该方法还能跨异构端点识别数据源，并为半自动化查询生成提供可读性强的结构化洞察。
链接: https://arxiv.org/abs/2608.28963
作者: Victor Eiti Yamamoto,Takeda Hideaki,Yamamoto Yasunori
机构: Graduate University for Advanced Studies, SOKENDAI (高级综合研究生院, 索肯ダイ); National Institute of Informatics (信息科学研究所, 东京); Database Center for Life Science (生命科学数据库中心, 千叶)
类目: Databases (cs.DB); Computation and Language (cs.CL)
备注:
Abstract:The number of knowledge graph databases has increased significantly with the proliferation of knowledge graph technologies. Knowledge graphs enable the dynamic integration of distributed data through federated SPARQL queries. However, constructing efficient queries in a federated environment is challenging due to the lack of detailed structural knowledge across decentralized datasets. While standards like VoID provide basic metadata, they often fail to capture the complex interlinks and authority distributions necessary for optimization. Consequently, current engines frequently rely on runtime ASK queries for source selection, increasing communication overhead. We propose RENSA, a federated SPARQL query generation framework that leverages an extension of SPARQL Builder Metadata (SBM). By integrating class and authority information, mapping subject and object usage to specific predicates, RENSA enables precise source selection and semantic constraint inference for query variables without runtime communication. The generated profiles represent less than 1% of the original dataset triples in most cases, ensuring storage efficiency. Evaluation on the LargeRDFBench benchmark (13 datasets with 1B triples, 32 queries) shows that RENSA achieves source selection results comparable to state-of-the-art methods while eliminating ASK query overhead. Furthermore, we demonstrate that RENSA infers class and authority constraints for query variables, enabling the identification of data sources even across heterogeneous endpoints. These profiles additionally offer human-readable structural insights for semi-automated query generation.
[NLP-239] CoVA-SFT: A Large-Scale Dataset for Chain of Visual Abstractions EMNLP2026
【速读】： 该论文旨在解决大语言模型（LLM）在处理视觉相关推理任务时，因依赖纯文本链式思维（CoT）而导致的表达不自然、效率低下等问题。现有方法虽可通过架构改进支持多模态输入，但缺乏大规模、多步骤且具备自校正能力的标注数据集，以训练模型构建并维护内部视觉工作空间（visual workspace）来解决纯文本推理问题。为此，作者提出CoVA-SFT，一个包含51.9K样本、超过222K个多模态推理步骤的结构化语料库，涵盖5类布局形式与17个复杂任务，并配套推出CoVA-Bench基准测试集（1,700个独立测试样本），用于可复现评估。其解决方案的关键在于通过显式的推理表述、代理式可视化呈现（agentic renderings）以及验证循环机制，引导多模态语言模型实现文本与视觉抽象的交替生成与协同推理。实验表明，基于CoVA-SFT微调的模型在CoVA-Bench上的表现平均优于所有交错式CoT基线模型两倍以上，但仍不及强文本型CoT基线，凸显了未来在多模态推理中实现高效视觉-语言协同的挑战。
链接: https://arxiv.org/abs/2608.28958
作者: Tsung-Han Wu,Heekyung Lee,Anya Ji,Haoming Chen,Trevor Darrell,Joseph E. Gonzalez,David M. Chan
机构: 未知
类目: Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to EMNLP 2026 Findings
Abstract:Chain-of-thought (CoT) reasoning has dramatically improved large language models (LLMs) by allowing them to decompose problems into intermediate steps. While CoT is widely effective for linguistic tasks, text-only CoT forces models to serialize visual problems into awkward prose. Although architectural solutions exist to process visual inputs, the community lacks a massive, multi-step, self-corrected dataset to teach models how to build and maintain internal visual workspaces when solving purely textual reasoning problems. To address this limitation, we introduce CoVA-SFT, a highly structured corpus of 51.9K samples containing over 222K multimodal reasoning steps across 5 distinct layout families and 17 complex tasks, and CoVA-Bench, a companion benchmark of 1,700 held-out test samples spanning the same tasks for reproducible evaluation. By providing explicit rationale formulations, agentic renderings, and verification loops, CoVA-SFT teaches multimodal language models to interleave text and visual abstractions. We validate the dataset by demonstrating that models fine-tuned on CoVA-SFT outperform all interleaved CoT baselines by more than 2x on average on CoVA-Bench, though they still fall short of strong text-only CoT baselines, highlighting open challenges for future work.
[NLP-240] Automated Researchers Can Reliably Mitigate Alignment Failures
【速读】： 该论文旨在解决如何通过自动化对齐研究（Automated Alignment Research, AAR）来加速实现安全对齐的人工智能系统，其核心问题是：在缺乏明确量化指标的情况下，难以评估自动化方法是否真正提升了对齐效果。解决方案的关键在于设计可同时优化多个可度量对齐失败（如欺骗、阿谀奉承、越狱攻击等）的安全基准的训练方法与数据构造策略，从而在不损害模型通用能力的前提下，显著减少目标对齐缺陷。实验表明，最优的AAR方法在10类对齐失败上均表现出显著缓解效果，并能泛化至未见过的多轮行为审计基准及规模达目标模型4.7倍的更大模型，且性能超越28位经验研究人员在长达八小时内的手动研究结果。进一步分析显示，引入人类研究思路作为初始方向并未提升AAR表现，暗示当前的自动化对齐研究系统可能已具备独立探索有效方案的能力。这一结果表明，在已有良好表征的对齐失败场景中，自动化对齐研究在短期内具有实际可行性。
链接: https://arxiv.org/abs/2608.28945
作者: Chen Yueh-Han,Jiaxin Wen,Jan Hendrik Kirchner
机构: Anthropic; UC Berkeley (加州大学伯克利分校)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Automating alignment research may accelerate progress toward aligned AI, but whether it does is hard to measure. Luckily, many alignment failures, such as deception, sycophancy, and jailbreaks, are already measurable by public benchmarks. We study whether automated alignment researchers (AARs) can post-train to mitigate alignment failures by proposing training methods and data to simultaneously optimize multiple safety benchmarks, while preserving general capability. Across 10 alignment failures, the strongest AAR methods significantly reduce the targeted alignment failures and generalize to a held-out benchmark, multi-turn behavioral audits, and models up to 4.7 times larger than the target model. As a human baseline, 28 experienced researchers receive up to eight hours to develop methods for the same benchmarks, but their methods underperform the best AAR methods. Using human ideas as the AARs’ initial research direction does not improve performance, suggesting current AARs may not need guidance from experienced researchers. These results suggest that automating alignment research on well-characterized failures may be practical in the near term.
[NLP-241] VocalAffectBench: Evaluating Vocal Emotion Recognition in AI Audio Models
【速读】： 该论文旨在解决当前生成式语音模型在缺乏文本转录或上下文元数据的情况下，难以准确识别原始音频中表达的情感状态（vocal emotion）的问题。其核心挑战在于：尽管人类语音中蕴含丰富的语调情感线索，但现有AI音频模型在仅依赖原始音频输入时，对情绪的判别能力仍十分有限。解决方案的关键在于构建一个公开、测试专用的基准测试平台——VocalAffectBench，用于评估模型从原始波形（WAV）中提取情感信息的能力。该基准包含273段由51名说话人录制的英语音频片段，覆盖七类情绪标签（愤怒、厌恶、恐惧、快乐、中性、悲伤、惊讶），总时长1.95小时，并严格要求所有基线模型仅使用音频输入进行推理。实验结果显示，六种基线模型的平均准确率为35.5%，最高仅为46.5%（gemini_3_5_flash），远低于理想水平；在更粗粒度的正负性分类（valence-bucket）分析中，平均准确率提升至50.9%，但仍表现出显著的情绪类别偏差：中性情绪识别召回率达75.6%，而惊讶与恐惧情绪分别仅有10.7%和15.4%。这表明当前模型虽能捕捉部分情感信号，但对非中性情绪的识别仍极为脆弱，制约了其在语音助手等实际应用中的可靠性。
链接: https://arxiv.org/abs/2608.28932
作者: Models Luc Debaupte,Tyler Baumgartner,Brandon Tai,Candice Fan,Bill Wang,Yi Zhong
机构: Besimple AI(贝简单人工智能); Google(谷歌)
类目: Computation and Language (cs.CL); Sound (cs.SD)
备注: 6 pages, 6 tables. Benchmark, baseline predictions, and aggregate results are publicly available
Abstract:Voice products increasingly need affective cues that are present in speech but absent from transcripts. We introduce VocalAffectBench, a public, test-only benchmark for evaluating whether AI audio models can identify expressed vocal emotion from raw audio. The benchmark contains 273 human-recorded English WAV clips from 51 speaker accounts totaling 1.95 hours across seven labels: angry, disgusted, fearful, happy, neutral, sad, and surprised, with 39 clips per class. All baselines are evaluated from audio alone, without transcripts or contextual metadata. Across six released baselines, average accuracy is 35.5%. The strongest baseline, gemini_3_5_flash, reaches 46.5% on the seven-way task, above the 14.3% random baseline but far from robust emotion recognition. A secondary valence-bucket analysis maps labels into positive, neutral, and negative classes, excluding surprised because its valence is ambiguous. Aggregate accuracy under this coarser view is 50.9%. Performance is highly uneven across classes. By recall, neutral is identified most reliably at 75.6% averaged across baselines, while surprised and fearful reach only 10.7% and 15.4%, respectively. These results show that the evaluated baselines can extract some affective signal from speech, but discrete expressed-emotion recognition remains fragile, especially for non-neutral emotions that are often most important in voice agent workflows.
[NLP-242] he Hallucination Signal Is a Mean Shift: Why Simple Probes Suffice EMNLP2026
【速读】： 该论文旨在解决大语言模型（LLM）幻觉检测中隐藏状态探针（hidden-state probes）信号几何结构不明确的问题，尤其是现有复杂探针架构背后的实际有效性与可解释性不足。其核心发现是：在三种7B规模模型及三个数据集的配对样本范式下，检测信号主要由单一均值偏移（mean-shift）分量主导，移除该方向后检测性能退化至随机水平，表明信号本质具有高度低秩特性。因此，探针架构复杂性的提升更多源于高维协方差估计的困难，而非需要捕捉非线性特征。研究进一步表明，简单的L2正则化逻辑回归（AUROC=0.952）即可达到或超越十二种受控复杂架构的性能；而多层聚合方法LayerMix通过整合连续层间的信号，实现了与已知最优层（oracle-layer）相当的检测效果，且无需访问真实最优层。这些结论揭示了在受控配对样本范式下，幻觉检测信号的几何本质为低维、连续分布于多层带状区域，从而为设计高效、简洁的探针提供了理论依据和实践指导。
链接: https://arxiv.org/abs/2608.28930
作者: Jungseob Lee,Jaehyung Seo,Heuiseok Lim
机构: Korea University(高丽大学); Konkuk University(中央大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 19 pages, 7 figures, 20 tables. Accepted to EMNLP 2026 (Main Conference). Code: this https URL
Abstract:Hidden-state probes effectively detect LLM hallucinations, but the geometry of the signal remains poorly characterized, driving increasingly complex probe architectures. Across three 7B-scale models and three datasets in a paired-example paradigm, we find the signal overwhelmingly dominated by a single mean-shift component, and removing this direction collapses detection to chance. Shrinkage linear discriminant analysis closes about 73% of the gap between 1D and full-dimensional classifiers, so apparent architectural complexity largely reflects high-dimensional covariance estimation difficulty rather than exploitable non-linearity. A simple L2-regularized logistic regression (0.952 AUROC) bounds or outperforms twelve controlled architectural alternatives, and our multi-layer aggregation exceeds CLAP cross-layer attention probing under matched paradigm. Because the signal spans a contiguous layer band, LayerMix aggregates it to match oracle-layer performance without oracle access. Our claims characterize the geometry within the controlled paired-example paradigm. Our code is available at this https URL.
[NLP-243] Leverag ing Turn-taking Dynamics for Intent Recognition in Multi-party Conversations EMNLP
【速读】： 该论文旨在解决多参与者对话中的意图识别问题，其核心挑战在于现有方法普遍忽视了多方交互过程中复杂的轮流发言（turn-taking）动态。为应对这一问题，论文提出一种多任务学习框架，引入一个基于说话人转换序列的辅助任务，通过计算“话轮转移熵”（turn-transition entropy）作为自监督目标，量化交互模式的可预测性。该指标能够捕捉对话中发言顺序的规律性，并作为额外信息增强主任务的训练。实验表明，将该连续型辅助目标融入预训练模型的训练过程，显著提升了意图识别性能，且优于忽略互动动态的现有方法。关键创新在于证明了该连续性目标可作为单一任务目标被有效学习，表明其蕴含了真实且有用的对话交互信号。
链接: https://arxiv.org/abs/2608.28926
作者: Galo Castillo-López,Alexis Lombard,Gaël de Chalendar,Nasredine Semmar
机构: Université Paris-Saclay, CEA, List, Palaiseau, France
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: Accepted for publication at EMNLP Industry Track 2026
Abstract:We propose a multi-task learning approach for multi-party dialogue intent recognition that leverages an auxiliary task that models turn-taking dynamics. Specifically, we introduce turn-transition entropy, a self-supervised target computed from the sequence of speaker transitions, which quantifies the predictability of interaction patterns. Experiments on multiple pre-trained models demonstrate that incorporating this auxiliary task improves intent recognition performance, outperforming existing approaches which ignore multi-party interaction dynamics. We find that our proposed continuous target can be learned as a single-task objective, suggesting that it is an actual signal carrying useful information.
[NLP-244] Causal Interventions Reveal Typologically Organized Syntactic Mechanisms in Multilingual Language Models
【速读】： 该论文旨在解决跨语言句法结构是否由相似的加工机制处理这一长期存在的语言学理论问题。由于人类语言加工机制缺乏精细且可操控的实证手段，该假设难以通过传统实验方法验证。本文提出的关键解决方案是利用机制可解释性（mechanistic interpretability）技术，对多语言大模型（multilingual LMs）中的句法处理机制进行解码与迁移分析。研究首先在单语内部识别出特定句法现象（如主谓数一致、代词性别一致、填空-缺口宾语提取）的内在计算机制，随后探索这些机制在跨语言间的可转移性。实验结果表明，在四种模型和三种典型句法结构上均存在一致的跨语言机制迁移现象，且迁移程度呈梯度分布，与语言类型学相似性正相关。该研究不仅为跨语言句法结构的共性提供了新的理论假说，更展示了通过语言模型研究推动语言学理论发展的新范式。
链接: https://arxiv.org/abs/2608.28924
作者: Sasha Boguraev,Toshiki Nakai,Kyle Mahowald,Julius Steuer
机构: The University of Texas at Austin(德克萨斯大学奥斯汀分校); Leipzig University(莱比锡大学); ScaDS.AI Dresden/Leipzig(德国德累斯顿/莱比锡数据科学中心); Heidelberg Institute for Theoretical Studies(海德堡理论研究所)
类目: Computation and Language (cs.CL)
备注: 20 Pages, 7 Figures, 11 Tables
Abstract:Linguistic theory has long recognized cross-linguistic syntactic regularities, leading to claims that these similar structures are processed by similar mechanisms. However, this hypothesis has been difficult to test empirically due to our lack of fine-grained, manipulable access of human processing mechanisms. In this work, we take advantage of techniques from mechanistic interpretability to study such a question in multilingual LMs. We first isolate language-internal mechanisms before attempting to transfer them cross-lingually. Across four models and three well-studied constructions (subject–verb number agreement, anaphoric pronoun gender agreement, and filler–gap object extraction) we find consistent cross-lingual mechanism transfer. We further find transfer to be graded, with more transfer between more typologically similar languages. We believe our work provides novel hypotheses about cross-linguistic syntactic structures and multilingual processing, and more broadly shows how the study of language models can help inform linguistic theory.
[NLP-245] VoiceCodeBench: Evaluating Exact Structured-Token Recovery in Automatic Speech Recognition
【速读】： 该论文旨在解决自动语音识别（ASR）系统在评估中过度依赖词错误率（Word Error Rate, WER）而忽视对关键结构化实体精确恢复的问题。传统W ER指标无法有效反映下游任务对精确数值、标识符或路径等结构化内容的敏感性，导致即使语音转录文本整体流畅且W ER较低，仍可能因关键值被错误识别而导致下游系统失效。为此，作者提出VoiceCodeBench基准，用于评估英文ASR系统在无上下文、仅基于原始音频输入条件下对精确结构化标记（structured-token）的恢复能力。其核心解决方案在于引入多维度评估指标，包括规范标记匹配率（Canonical Token/Entity Match, CTEM）、任务成功率（Task Success Rate, TSR）以及按实体类型划分的精确恢复率，并构建涵盖8个工作流领域、26类实体、共1,482个可审计目标实体的真实职场录音数据集。实验表明，尽管低W ER与较好的结构化实体恢复存在中度相关性（Spearman相关系数为-0.73），但并不能充分决定实际恢复效果，最先进基线系统在任务成功率上仅达68.7%，近三分之一的录音存在至少一个关键值未被正确恢复的情况。这表明，必须采用以实体为中心的评估范式，才能真实衡量ASR输出对生产系统中解析、路由、存储、比较或执行等操作所依赖的精确值的保障能力。
链接: https://arxiv.org/abs/2608.28916
作者: Tyler Baumgartner,Brandon Tai,Lisa Kaelin-Martin,Candice Fan,Luc Debaupte,Bill Wang,Yi Zhong
机构: Besimple AI(贝简单AI)
类目: Computation and Language (cs.CL)
备注: 11 pages, 1 figure, 9 tables
Abstract:Automatic speech recognition (ASR) systems are commonly evaluated with word error rate (WER), yet many voice workflows depend on exact written values for identifiers, paths, and measured quantities. A transcript can appear fluent and achieve low WER while corrupting a value that a downstream system must parse, store, or execute. We introduce VoiceCodeBench, a benchmark for evaluating exact structured-token recovery in English ASR. It contains 300 human-recorded workplace segments spanning eight workflow domains and 1,482 audited target entities across 26 entity types, each with a canonical written form recoverable from the audio. Under a raw-audio-only protocol, systems receive audio bytes without additional context or metadata. Alongside WER, we evaluate Canonical Token/Entity Match (CTEM), Task Success Rate (TSR), and per-type exact recovery. Across 12 baseline ASR systems, lower WER generally corresponded to better structured-token recovery but did not fully determine it: Spearman correlations were -0.73 for both WER versus CTEM and WER versus TSR. The strongest baseline by TSR reached only 68.7%, leaving nearly one third of recordings with at least one unrecovered workflow-critical value. These results show that entity-sensitive metrics are needed to assess whether ASR output preserves exact values that production systems must parse, route, store, compare, or execute. Comments: 11 pages, 1 figure, 9 tables Subjects: Computation and Language (cs.CL) Cite as: arXiv:2608.28916 [cs.CL] (or arXiv:2608.28916v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2608.28916 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-246] SemKV: Semantic Mixed-Precision KV Cache Quantization Guided by the Quality Cliff for Long-Context LLM Inference
【速读】： 该论文旨在解决长上下文大语言模型（LLM）推理中关键值（KV）缓存带来的内存瓶颈问题，其核心挑战在于KV缓存的大小随上下文长度线性增长，严重制约了模型在长序列任务中的部署效率。现有方法如均匀的分数比特量化在低精度下无法实现平滑退化，实验表明，在预设多种子统计协议下，Llama-3.1-8B-Instruct模型在使用仿射量化器时，其KV缓存压缩至2.322比特/值仍与FP16精度无统计差异，但在2.0比特时出现显著性能塌陷（质量悬崖），该现象在生成阶段和多轮对话中重现，并可迁移至Mistral-7B模型。这一“质量悬崖”揭示了重要性感知混合精度策略的局限性：在悬崖之上，模型内部八种重要性指标具有统计等价性，因此混合精度的优势主要源于量化网格插值，而非重要性区分。为此，论文提出SemKV方案，通过保留所有标记、基于模型内部得分排序并分配两个相邻于悬崖之上的高精度级别，实现了6.0倍的存储压缩，且在900次测试（三种子）中未检测到与全精度KV的统计差异，优于获得1.5倍更大内存预算的FP16标记剪枝方法。进一步地，将仿射量化器替换为失真优化的量化器（TurboQuant-MSE）可系统性降低质量悬崖，使无损操作点提升至7.9倍。解决方案的关键在于：针对目标部署场景测量质量悬崖位置，然后在悬崖之上进行量化网格插值，从而实现高效且无损的低比特KV缓存压缩。
链接: https://arxiv.org/abs/2608.28911
作者: Daeha Lee,Do-Hyung Kim,Jae-Hong Kim
机构: Electronics and Telecommunications Research Institute (ETRI), Daejeon, Republic of Korea
类目: Machine Learning (cs.LG); Computation and Language (cs.CL); Information Theory (cs.IT)
备注: 34 pages, 19 figures
Abstract:The key-value (KV) cache is the dominant memory bottleneck of long-context large language model (LLM) inference, growing linearly with context length. We show that uniform KV quantization on a fractional-bit grid does not degrade gracefully: under a prespecified multi-seed statistical protocol, Llama-3.1-8B-Instruct with an affine quantizer is statistically indistinguishable from FP16 KV down to 2.322 code bits/value and collapses at 2.0 bits - a quality cliff in (2.0, 2.322] that reappears in generation-time quantization and multi-turn dialogue and transfers to Mistral-7B. The cliff reframes importance-aware mixed precision: above it, eight model-internal importance indicators are statistically interchangeable, so the benefit of mixing is grid interpolation, reaching average precisions uniform quantization cannot realize. SemKV preserves every token, ranks tokens by a model-internal score, and assigns two adjacent above-cliff precisions, achieving a measured 6.0x storage reduction with no statistically detectable quality difference from full KV (n=900, three seeds), and outperforming FP16 token pruning granted a 1.5x larger memory budget. Replacing the affine base with a distortion-optimized quantizer (TurboQuant-MSE) lowers the cliff in every protocol tested, raising the no-detectable-loss operating point to 7.9x. The recipe: measure the cliff for the target deployment setting, then interpolate above it.
[NLP-247] Moving the Mean Toward the Known Good Not Beyond It: What Inference-Time Interventions and Weight Consolidation Buy in Open-Ended Generation
【速读】： 该论文旨在解决生成式模型在持续自我改进过程中如何有效利用自身已验证的成功经验以提升性能的问题，特别是在在线装箱（online bin packing）这一组合优化任务中的表现。其核心挑战在于：尽管模型可通过“生成-验证-选择-LoRA整合”循环不断迭代，但如何避免陷入局部最优并实现对经典启发式算法的逼近甚至超越。解决方案的关键在于基于价值筛选（value-filtered candidates）的LoRA整合机制，即仅将经过验证的高质量候选解用于模型参数的微调更新，从而引导模型输出趋向于已知最优解。实验表明，这种策略显著提升了模型在未见变体上的表现（相比随机整合控制，平均得分提升3.1分，p=0.004），且最佳候选解稳定收敛至经典启发式算法水平，不再进一步超越。此外，对照实验揭示，这种集中效应主要源于对优质样本的强化学习，而非对劣质样本的排斥；同时，尽管该策略降低了单个候选解优于经典算法的概率（从10%降至3.9%），但由于生成数量增加，绝对优势解的数量仍有所上升。研究还通过推理时日志分析发现，模型生成的结构化摘要可有效支持文档级集成，而内置验证器的流水线设计被成功模仿，验证了该范式在可复制性和稳定性方面的可行性。
链接: https://arxiv.org/abs/2608.28886
作者: Roberto I. Ono Filho
机构: 独立研究员(Independent researcher)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Code and run data: this https URL
Abstract:What does a generation loop gain from learning on its own verified successes? In cycles of generate, verify, select and LoRA-consolidate on online bin packing, training on value-filtered candidates shifts what the model writes on held-out variants toward value (-1.7 points of excess, p=0.008; -3.1 against a random-consolidation control, p=0.004) while the best observed candidate converges to the classic heuristic’s level and no further. A confirmation battery replicates the whole procedure three times, with fresh seeds and a never-consulted held-out set read exactly once: the mean was nearly identical in all three lineages (-2.0, -1.8, -1.9), and after aggregating within held-out variant all seven evaluable variants favored consolidation (p=0.008). The best observed candidate moved to the classic heuristic’s level, exactly (0.021028 in all three lineages, for attract and for the random control alike), and never beyond it. A matched SFT-only control shows the supervised anchor, not repulsion from bad candidates, does the concentrating (96% of candidates land exactly at the classic heuristic’s level). The tails cut both ways: consolidation lowers the per-candidate rate of better-than-classic candidates (10% to 3.9%) while its larger production yields more such candidates absolutely (5 against 1, on few events). As motivation we report the inference-time ledger that led here: a model-written schematic recap buys judged document integration and nothing buys development; a verifier written into the stream is imitated, 16.4 fabricated verdict lines per notebook. Mean quality among valid candidates can be bought and replicated; the observed best goes to the classic and, so far, never beyond it.
[NLP-248] No Detectable Change in Side-Level WER from Prompt-Level Context: A Preregistered Ablation on a Production Oral-History Corpus DATE
【速读】： 该论文旨在解决在推理阶段向大规模多模态模型提供上下文信息是否能有效提升语音转录性能的问题，特别是在真实生产环境中对口述历史音频进行领域自适应时的适用性。其核心问题是：尽管先前在小型模型上的研究显示引入提示（prompt）上下文可带来显著性能提升，但在实际部署的语音转录工具中，这种机制是否仍具有效性？解决方案的关键在于通过严格的配对消融实验设计，在生产级系统中对提示条件层（prompt-conditioning layer）施加完整上下文与无上下文两种条件，并以操作员校正的逐字参考文本为基准评估词错误率（Word Error Rate, WER）变化。研究采用预注册实验设计，分析代码在确认批次评分前即通过哈希锁定，确保方法透明性；同时通过实施审计验证上下文操纵确实生效。结果表明，对于gpt-4o-transcribe模型，完整上下文与无上下文之间的中位差仅为+0.6 WER点，置信区间为[-1.1, +1.0]，未达统计显著性；而Gemini-2.5-flash的结果因运行间变异性过大无法支持有效推断。进一步的后验分析发现，运行间的流水线变异性超过了干预效应，暗示单次转录难以分辨此类微小差异。序列对齐分析显示，仅在完整列出的短语上存在微弱改进，但不足以影响整体侧级WER，且Gemini还伴随未列词项错误增加。因此，该研究强调，评估上下文机制的有效性必须结合序列对齐的术语级、插入错误及说话人标签等细粒度指标，而非仅依赖聚合准确率。
链接: https://arxiv.org/abs/2608.28875
作者: Theodore O. Cochran,Stephanie Dodson,Keith Nore
机构: AI for Altruism; Independent Researcher
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Sound (cs.SD)
备注: 31 pages, 1 figure. Preregistered on OSF ( this https URL , DOI https://doi.org/10.17605/OSF.IO/NS49B%29%3B release materials and dated provider-documentation snapshots in the study component ( this https URL )
Abstract:Supplying context at inference time to a large multimodal model is an inexpensive lever for adapting speech transcription to a domain, and earlier results on smaller models reported large gains. This work tested that mechanism where it ships, in the prompt-conditioning layer of a production oral-history transcription tool, on a sample from its own production corpus. Full prompt-level context did not detectably change side-level word error rate (WER), and none of the four preregistered hypotheses was supported. The design was a within-item paired ablation, preregistered with the analysis code frozen by hash before the confirmatory batch was scored; two disclosed gpt-4o pilot sides had been scored earlier, during scorer development. Nineteen cassette sides, about 10.6 hours of degraded 1970s-80s interview audio, were reprocessed through the production code path under three prompt arms, crossed with two deployed commercial configurations, gpt-4o-transcribe and gemini-2.5-flash, and scored against operator-corrected verbatim references. For gpt-4o-transcribe the median paired difference between the full-context and no-context arms was +0.6 WER points, with a side-resampled interval of [-1.1, +1.0]; the Gemini estimates were too unstable to support a comparable negative inference. A post-hoc rerun found run-to-run pipeline variability larger than the confirmatory differences, so effects of that size cannot be resolved from one transcription per cell. An implementation audit verified the manipulation was live, and sequence-alignment analysis found a small improvement on complete context-listed phrases, too small to materially change side-level WER, and for Gemini coexisting with worsened unlisted-token error. Evaluating context mechanisms therefore requires sequence-aligned term-level, insertion, and speaker-label measures alongside aggregate accuracy.
[NLP-249] Latent-Space Intervention for Cross-Lingual Factual Consistency: Consistency Improvements without Accuracy Drops EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在跨语言问答任务中对同一事实性问题生成不一致答案的问题，即不同语言间存在语义表达与答案输出的偏差。其核心解决方案是通过跨语言潜在空间干预（cross-lingual latent-space intervention），具体采用在平行多语言表示上训练分层自编码器（layer-specific autoencoders），并在推理阶段对事实性问答提示进行潜在空间修正。关键创新在于利用自编码器重建能力实现跨语言表征的几何对齐，显著提升了非英语语言与英语之间的一致性，且未损害事实准确性。实验表明，在开放式问答中，英文与阿拉伯语、俄语之间的斯皮尔曼等级相关系数分别提升0.16和0.20；在多项选择题中，答案一致性在KLAR与mParaRel数据集上均得到稳定提升。消融实验进一步揭示，自编码器重建机制可无损地提升一致性，而主成分分析（PCA）投影贡献有限，均值偏移虽在开放式问答中带来显著一致性增益，但伴随一定准确率损失。
链接: https://arxiv.org/abs/2608.28860
作者: Faeze Ghorbanpour,Constanza Fierro,Alexander Fraser,Anders Sogaard
机构: TU Munich (慕尼黑工业大学); Munich Center for Machine Learning (慕尼黑机器学习中心); University of Copenhagen (哥本哈根大学)
类目: Computation and Language (cs.CL)
备注: Accepted at EMNLP 2026
Abstract:Large Language Models (LLMs) often answer the same factual question differently across languages. We study whether cross-lingual latent-space intervention can reduce this inconsistency. We train layer-specific autoencoders on parallel multilingual representations and apply inference-time corrections to factual QA prompts. We find that latent intervention improves geometric alignment between languages, and that this improvement translates into consistent gains in cross-lingual consistency with English across both open-ended and multiple-choice QA formats, without degrading factual accuracy. In open-ended QA, Spearman’s rank correlation between English and non-English languages improves substantially, with gains of 0.16 for English-Arabic and 0.20 for English-Russian pairs. In multiple-choice QA, answer agreement with English improves consistently across both KLAR and mParaRel. Ablations show that AE reconstruction yields consistent gains at no accuracy cost, while PCA projection contributes marginally, and mean-shift produces substantially larger consistency gains in open-ended QA at the cost of some accuracy.
[NLP-250] he Halt Vector: Internalizing a Causal Steering Intervention for Efficient Reasoning ALT
【速读】： 该论文旨在解决生成式AI（Generative AI）在推理过程中存在过度思考（excessive reasoning）的问题，即模型在已知答案后仍持续生成冗长的思维链（chain of thought），导致计算资源浪费且效率降低。其核心挑战在于，不同问题的可压缩冗余长度差异显著，因此无法通过全局统一的长度惩罚机制有效去除冗余。解决方案的关键在于将因果可解释性（causal interpretability）的发现内化至模型权重中，引入一种“停止向量”（halt vector）机制：该向量为模型第18层的一个均值差方向，其控制强度可调节推理时长；而复制的值轴（value axis）则无此作用。实现该干预的关键技术难点在于，直接最大化该方向的投影会破坏下游任务依赖的非轴向维度，反而延长生成过程。为此，论文提出通过固定自然激活值并重建被引导后的激活状态，以保持任务性能的同时实现有效缩短推理路径。基于24个问题的训练且无需强化学习，该方法在五个未见基准上平均削减约四分之一的推理时间，且剪枝程度与各问题自身可移除冗余量高度相关（相关系数0.70），同时缓解了随问题难度加剧的非终止病理现象。该工作的贡献不在于超越传统解码时早停策略，而在于提供了一种通过内部结构改造实现高效、精准推理长度控制的新范式。
链接: https://arxiv.org/abs/2608.28859
作者: Dylan Jayabahu,Tinuade Adeleke
机构: University of Waterloo(滑铁卢大学); Independent Researcher(独立研究员)
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 23 pages, 3 figures. Accepted at the 2nd Workshop on Efficient Reasoning, COLM 2026 (non-archival). Code: this https URL
Abstract:Reasoning models do not stop when they know the answer. On DeepSeek-R1-Distill-Qwen-7B the chain of thought runs about twice as long as the model’s own answer probability takes to settle, and how much of that excess is removable varies from problem to problem, so a global length penalty cannot take it out. We take it out by internalizing a causal interpretability finding into the weights. The mechanism is a halt vector: a difference-of-means direction at layer 18 of this model whose steering strength controls how long it thinks, while a replicated value axis does nothing. Installing that intervention in the weights is harder than it looks. Maximizing the scalar projection onto the direction corrupts the off-axis dimensions a frozen downstream reader depends on, and generation gets longer instead of shorter; what works is reconstructing the whole steered activation with those dimensions pinned to their natural values. Fit from 24 problems and no reinforcement learning, the halt removes about a quarter of the thinking at held accuracy across five unseen benchmarks, and the cut tracks each problem’s own removable slack at 0.70. It also closes a non-termination pathology that grows with difficulty and that a decoding-time confidence hook makes worse. We do not claim to beat a well-tuned length penalty or decoding-time early exit on the raw trade-off; the contribution is how the halt is obtained.
[NLP-251] A rigor-matched audit of periodic-step layer skipping for efficient llm inference: conflayers versus swift with a supplemental analysis of trained routing alternatives
【速读】： 该论文旨在解决大语言模型（LLM）推理过程中计算资源浪费的问题，具体聚焦于如何通过动态跳过部分Transformer层来实现高效推理。其核心挑战在于：在保证生成质量的前提下，如何在推理时在线决策哪些层应被跳过，以平衡速度与准确性。论文提出的解决方案关键在于采用周期性步长的搜索机制，在推理过程中定期评估并动态决定执行的层数。其中，自洽的自推测解码方法（SWIFT）通过构建多路径候选生成与验证机制，实现了更高的准确率和更优的推理效率；而基于置信度的早期退出方法（ConfLayers）虽具有较低的搜索开销，但其性能在复杂任务上显著落后，尤其在1.5B规模的GSM8K推理任务中表现严重退化。研究进一步揭示，尽管SWIFT的在线搜索开销较大且波动性强（最高达28.7%），但其纯推理成本低于ConfLayers，整体推理速度在所有测试场景中均更快（提升5%-21%），从而逆转了仅看总耗时的表面排名。此外，对两种训练路由方法（LayerRoute和LayerDrop）的补充分析表明，它们因粒度较粗且缺乏真正的输入感知门控机制，导致准确率大幅下降，尤其在高难度任务中出现近乎崩溃的表现。因此，本研究的关键贡献在于建立了一套严谨的、可复现的效率评估范式，并证明了基于周期性在线搜索的自推测解码策略在精度与效率上的综合优势。
链接: https://arxiv.org/abs/2608.28846
作者: Prateek Kumar Sikdar,Arpan Ghosh
机构: Accenture(埃森哲); Wells Fargo(富国银行)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 17 pages, 8 figures, 9 tables
Abstract:Layer-skipping methods for efficient LLM inference decide, at some granularity, which transformer layers to execute for a given input. We present a rigor-matched, three-seed audit of two periodic-step, search-based methods that make this decision online at inference time and re-evaluate it every few generation steps: a confidence-gated early-exit baseline (ConfLayers) and genuine self-speculative decoding (SWIFT, Xia et al. 2024), together with vanilla autoregressive decoding, across two model scales (Qwen2.5-0.5B and Qwen2.5-1.5B) and two tasks (GSM8K reasoning and CNN/DailyMail summarization). SWIFT is the strongest method on accuracy in three of four cells; ConfLayers is dominated everywhere, with particularly large deficits on GSM8K at 1.5B. Once online-search overhead is separated from pure inference cost, SWIFT’s true inference speed is faster than ConfLayers’s in all four cells (5-21%), reversing the naive wall-clock ranking in three of them. ConfLayers’s search overhead is small and stable (1-2% of cost), while SWIFT’s is larger and more variable (up to 28.7%). We additionally examine two trained-routing methods, LayerRoute (Sikdar, 2026) and LayerDrop (Fan et al. 2020), as a supplemental analysis because they operate at coarser decision granularities. Under a verified protocol with genuine per-input gating, a genuine full-model baseline, and genuine inference-time compute skipping, both show modest speedups (1.08-1.33x) but accuracy well below the periodic-step methods, including a near-total collapse for LayerRoute on GSM8K at 1.5B (0.003 mean exact-match across three seeds). We release the full audit protocol as a template for rigor-matched efficiency comparisons.
[NLP-252] Evaluating Multilingual Sentence Embeddings for Translation Error Detection:An English–Greek Contrastive Study
【速读】： 该论文旨在解决多语言句子嵌入（Multilingual Sentence Embeddings）在识别细微翻译错误方面的敏感性不足问题，具体考察通用型多语言嵌入模型是否能够有效区分正确与经过微小修改的错误英-希翻译。其核心解决方案在于构建了一个对比数据集，涵盖10个核心与5个探索性错误类别（包括事实性、词汇语义、语法、关系、指代及话语层面等现象），并基于FLORES+双语句对参考译文，经两位翻译专家审核后形成1,850个高质量对比样本。研究通过计算源句与正确/错误译文之间的余弦相似度，评估五种主流多语言嵌入模型（BGE-M3、Multilingual E5、Multilingual MPNet、LaBSE、Jina Embeddings v3）以及一个无参考的COMETKiwi模型在不同错误类型上的判别能力。结果表明，尽管嵌入模型整体表现良好（最高准确率达89.30%），但对显性事实和词汇语义错误检测更可靠，而对时态/体貌、代词指代等隐性错误敏感性较低；相比之下，COMETKiwi在多数复杂错误类别上表现更优，但在日期时间与数字相关错误上表现不佳。研究揭示了多语言嵌入模型与自动机器翻译质量评估工具之间存在互补的错误感知特性，提示嵌入模型更适合作为综合翻译评估框架中的语义充分性组件，而非独立的质量评估指标。
链接: https://arxiv.org/abs/2608.28776
作者: Eleftherios Kalogeros,Athanasios Ntalakas,Manolis Gergatsoulis,Paschalis Nikolaou,Sotiria-Lito Alexaki
机构: Ionian University (爱琴大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Multilingual sentence embeddings are increasingly used to estimate semantic similarity across languages, yet their sensitivity to fine-grained translation errors remains insufficiently understood. This study investigates whether general-purpose multilingual embedding models can distinguish correct English-Greek translations from minimally modified erroneous alternatives. A contrastive dataset was developed from FLORES+ sentence-aligned reference translations and reviewed by two translation experts. It contains 1,850 examples across ten core and five exploratory error categories, covering factual, lexical-semantic, grammatical, relational, referential, and discourse-level phenomena. Five multilingual sentence-embedding models (BGE-M3, Multilingual E5, Multilingual MPNet, LaBSE, and Jina Embeddings v3) were evaluated using cosine similarity between each English source sentence and its correct and erroneous Greek translations. A reference-free COMETKiwi model was also evaluated as an MT quality-estimation baseline. Performance was assessed through contrastive accuracy and score margins for category-specific sensitivity. BGE-M3 achieved the highest accuracy among embedding models at 89.30 percent, while COMETKiwi achieved 94.49 percent. Embedding models detected explicit factual and lexical changes more reliably than tense-and-aspect and pronoun-coreference errors. COMETKiwi improved performance on several difficult categories, including tense and aspect, pronoun and coreference, and semantic-role errors, but showed lower sensitivity to date-and-time errors and underperformed the embedding models on numbers. The results show complementary error-sensitivity profiles: multilingual sentence embeddings provide useful semantic adequacy signals but are better suited as components of broader translation-evaluation frameworks than as standalone metrics. Subjects: Computation and Language (cs.CL) Cite as: arXiv:2608.28776 [cs.CL] (or arXiv:2608.28776v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2608.28776 Focus to learn more arXiv-issued DOI via DataCite
[NLP-253] ReVA: A Region-Aware Visual Assistant for Visually Grounded Question Answering
【速读】： 该论文旨在解决多模态大语言模型（Multimodal Large Language Models, MLLMs）在视觉问答（Visual Question Answering, VQA）任务中对精确空间推理和细粒度视觉理解能力不足的问题，尤其针对由区域级视觉表征缺失导致的对象、属性及空间幻觉现象。其核心解决方案是提出一种区域感知的视觉问答模型ReVA，关键在于通过双桥接机制（dual bridge）将冻结的CLIP ViT-L/14视觉编码器与Qwen2.5-7B-Instruct大语言模型（LLM）进行深度融合：图像桥接将视觉变压器最后一层的全局特征映射为图像令牌，而区域桥接则从ViT中间层提取增强后的特征图，对每个边界框裁剪出的区域特征进行编码，以保留早期纹理信息与后期对象线索，并生成K个区域令牌。这些区域令牌与图像令牌共同构成提示前缀，输入至LLM以联合建模场景级上下文与细粒度区域证据。此外，采用基于RAM++、spaCy与Grounding DINO的检测堆栈实现零样本、问题无关与问题相关的边界框自动生成。在VQAv2、MMBench、POPE与SEED-Bench等多个基准上的实验表明，ReVA在POPE数据集上达到82.85%的平均F1分数，显著优于仅使用图像令牌的基线模型（81.14%），验证了显式区域感知视觉表示能有效降低对象幻觉并提升模型的事实性对齐能力。
链接: https://arxiv.org/abs/2608.28707
作者: Anoop Senthil
机构: Queen Mary University of London (伦敦玛丽女王大学)
类目: Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV)
备注: 11 pages. Code: this https URL
Abstract:Multimodal Large Language Models (MLLMs) have achieved remarkable progress in Visual Question Answering (VQA), yet they continue to struggle with questions requiring precise spatial reasoning and fine-grained visual understanding. These limitations often manifest as object, attribute, and spatial hallucinations, where models generate confident but visually unsupported responses due to insufficient region-level and fine-grained visual grounding. To address this challenge, we propose ReVA, a region-aware VQA model that employs a frozen CLIP ViT-L/14 Vision Transformer (ViT) and a Qwen2.5-7B-Instruct large language model (LLM) connected through a dual bridge that aligns both whole-image and region-level representations with the LLM’s embedding space. The image bridge maps final transformer block features into image tokens. The region bridge maps cropped features from enriched intermediate features across ViT blocks so early texture and later object cues are more evident, into K region tokens for every bounding box. ReVA uses a detector stack that supplies automatic zero-shot bounding boxes that are both question-agnostic and question-dependent, using RAM++ (Recognize Anything Model), spaCy, and Grounding DINO. The image tokens and region tokens are concatenated as an LLM prompt prefix to jointly encode scene-level context and fine-grained regional evidence when answering questions. Evaluated on VQAv2, MMBench, POPE, and SEED-Bench, ReVA achieves 82.85% mean F1 on POPE, compared with 81.14% for an image-token baseline without region tokens. These results demonstrate that explicit region-aware visual representations reduce object hallucination and improve the factual grounding of MLLMs.
[NLP-254] Multi-Agent Self-Improving Reinforcement Learning for Video Reasoning
【速读】： 该论文旨在解决视频推理任务中时序证据选择的难题，特别是在**接地视频问答（grounded video question answering）和时序定位（temporal grounding）等任务中如何有效识别支持查询的时序片段。现有方法通常依赖局部监督信号（如边界回归或跨度生成），而验证机制仅用于推理阶段的候选段重排序，未能充分参与训练过程。本文提出一种基于多智能体框架的解决方案，其关键在于利用一个冻结的验证器（frozen verifier）**作为训练信号来指导模型学习。具体而言，可训练的定位器（Grounder）生成候选轨迹与证据片段，冻结的验证器根据查询条件为各片段打分，通过组内相对策略梯度优化目标鼓励表现优于同输入内其他候选者的轨迹，并引入自举校准损失引导时序预测向验证器偏好的片段收敛。该方法在未进行目标数据集微调的情况下，实现了零样本跨任务迁移，在多个基准上取得显著性能提升，尤其在交并比（IoU）和中等重叠召回率等相关性指标上表现突出，表明冻结验证器可作为有效的证据选择训练信号；然而，对严格边界精度的提升仍有限。
链接: https://arxiv.org/abs/2608.28675
作者: Mingwen Zhang,Jisheng Dang,Minqiang Yang,Bimei Wang,Bin Hu,Tat-Seng Chua
机构: Lanzhou University (兰州大学); Gansu Provincial Key Laboratory of Wearable Computing, School of Information Science and Engineering, Lanzhou University (甘肃省可穿戴计算重点实验室，兰州大学信息科学与工程学院); Brain Health Engineering Laboratory, Institute of Engineering Medicine, Beijing Institute of Technology (北京理工大学工程医学研究所脑健康工程实验室); National University of Singapore (新加坡国立大学)
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注:
Abstract:Video reasoning tasks such as grounded video question answering and temporal grounding require selecting temporal evidence that supports the query. In many current training setups, temporal supervision is applied through local objectives such as boundary regression or span generation, while verification is used mainly to rerank candidate segments at inference time. We study whether a frozen verifier can also guide training. Our multi-agent framework couples a trainable \emphGrounder with a frozen \emphVerifier: the Grounder samples candidate trajectories and evidence segments, the Verifier assigns query-conditioned segment scores, a group-relative policy-gradient objective favors trajectories that outperform their within-input peers, and a bootstrapped calibration loss steers temporal predictions toward verifier-preferred spans. Trained on source tasks and evaluated without target-dataset fine-tuning, a two-billion-parameter instantiation transfers zero-shot across grounded question answering, temporal grounding, and long-video question answering, reaching 28.7% intersection-over-union and 25.4% answer-grounding accuracy on a grounded-question-answering benchmark, 46.1% intersection-over-union on a temporal-grounding benchmark, and 54.1% on a long-video question-answering benchmark. Relative to a strong same-scale baseline, the gains are modest but consistent, with the clearest improvements on relevance-oriented metrics such as intersection-over-union and moderate-overlap recall. Within the tested benchmarks and transfer setting, the results support frozen verification as a training signal for evidence selection, while showing that strict boundary precision remains comparatively weaker. Code and models are available at this https URL
[NLP-255] GreenBench: Benchmarking Energy Efficiency and Carbon Footprint of Open-Source LLM Inference on Apple Silicon
【速读】： 该论文旨在解决大型语言模型（Large Language Models, LLMs）在推理阶段对环境影响的评估问题，尤其聚焦于苹果自研芯片（Apple Silicon）平台上的能效特性。传统绿色人工智能（Green AI）研究主要关注数据中心GPU和嵌入式设备，而针对采用统一内存架构（unified memory architecture）的Apple M4 Pro芯片上LLM推理的能源消耗与碳足迹尚缺乏系统性研究。其解决方案的关键在于提出GreenBench这一基准测试框架，通过macOS的powermetrics实现对CPU+GPU封装功率的直接测量，并结合Ollama提供的纳秒级时间精度，量化了五款开源LLM（参数规模3-9B）在三种自然语言处理任务下的能量效率、吞吐量及碳足迹。实验结果表明，M4 Pro在持续推理时仅消耗0.47 W的封装功率，系统总功耗为8–12 W，相较数据中心GPU实现30–40倍的单位令牌能耗效率提升；同时，小模型（3–3.8B）相比大模型（7–9B）可带来2.6–4.2倍更高的吞吐量和高达62%的单位令牌能耗降低。基于帕累托分析，Qwen 2.5（7B）在准确率与能效间达到最优平衡（MMLU 57%，59 tokens/s），而Llama 3.2（3B）则适用于低延迟场景（175 tokens/s）。研究还提供了封装级与系统级的单位令牌能耗数据，并依据印度与美国电网的碳排放因子估算出对应的二氧化碳排放量，为边缘端绿色部署提供了可量化的决策支持。
链接: https://arxiv.org/abs/2608.28667
作者: Rajeswari Kannan,Raj Firke,Shreya Bengle,Srushti Deshmukh
机构: Pimpri Chinchwad College of Engineering (匹普里钦奇瓦德工程学院); Red Hat (红帽); Independent Researcher (独立研究员)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Performance (cs.PF)
备注: 7 pages, 1 figure, 6 tables. Accepted at IEEE ICCUBEA 2026
Abstract:The rapid proliferation of Large Language Models (LLMs) has raised concerns about their environmental impact during inference. While Green AI research has focused on datacenter GPUs and embedded platforms, the energy profile of LLM inference on Apple Silicon, with its unified memory architecture, remains unstudied. This paper presents GreenBench, a benchmarking framework that evaluates the energy efficiency, throughput, and carbon footprint of five open-source LLMs (3-9B parameters) across three NLP tasks on an Apple M4 Pro with 48 GB unified memory. Using macOS powermetrics for direct power measurement and Ollama’s nanosecond-precision timing, we find that the M4 Pro draws only 0.47 W of CPU+GPU package power during sustained inference, with total system power of 8-12 W, achieving 30-40x better energy efficiency per token than datacenter GPUs in single-user deployment. Smaller models (3-3.8B) deliver 2.6-4.2x higher throughput and up to 62% less energy per token than larger models (7-9B). Pareto analysis identifies Qwen 2.5 (7B) as the optimal accuracy-efficiency trade-off at 57% MMLU and 59 tokens/s, while Llama 3.2 (3B) suits latency-critical applications at 175 tokens/s. We provide per-token energy at package and system levels with CO2 estimates for India and US grids.
[NLP-256] st-Time Scaling for Scientific Equation Discovery
【速读】： 该论文旨在解决生成式人工智能（Generative AI）在开放性任务中进行自动方程发现时，如何高效利用测试阶段计算资源以提升模型推理能力的问题。其核心挑战在于：在缺乏明确答案格式的开放性场景下，语言模型需基于观测数据点进行候选方程的迭代搜索与验证，而现有方法对测试时扩展（Test-time Scaling, TTS）策略的研究多集中于闭合式任务（如数学求解、代码生成），难以直接迁移至此类探索性更强的开放性任务。为此，论文提出将大语言模型驱动的方程发现统一建模为一种基于计算资源分配视角的迭代搜索过程，涵盖Best-of-N、序列优化、树搜索及演化式方法等范式。关键解决方案在于通过控制最小化提示工程干扰的并行控制器，在固定计算预算下系统评估不同资源分配策略的影响。实验结果表明，搜索宽度（search width）是决定性能的主导因素：最优宽度随计算预算增加而上升，而种群-分支比例和控制器选择的影响较小；同时，合理设定宽度可显著提升实际运行效率，增强并行度。研究揭示，在具备有效验证机制的前提下，对探索与利用之间的平衡调控是实现大语言模型方程发现能力有效扩展的核心。
链接: https://arxiv.org/abs/2608.28660
作者: Haowei Lin,Hubert Lim,Xiangyu Wang,Letian Huang,Di He
机构: Peking University (北京大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Test-time scaling (TTS) improves language model reasoning by allocating additional test-time compute, but prior work mainly studies closed-ended tasks such as math and coding. We study TTS for automated equation discovery, an open-ended setting where models search over candidate equations and rely on observed datapoints for feedback. We formulate LLM-driven equation discovery as an iterative search process that unifies Best-of-N, sequential refinement, tree search, and evolution-style methods under a common compute-allocation view. To isolate allocation effects from prompt engineering and other heuristics, we compare minimal parallel controllers under fixed budgets. On LLM-SRBench equation-discovery tasks, we find that search width is the dominant allocation parameter: the best width in our sweep generally increases with the compute budget, while the population–branching split and controller choice matter less. Appropriate width selection also improves wall-clock efficiency by increasing parallelism. These results suggest that, given an informative verifier, controlling exploration and exploitation is central to scaling LLM-based equation discovery.
[NLP-257] How Language Models Choose Sides: Internal Representations of Instruction Hierarchy ICML2026
【速读】： 该论文旨在解决指令微调的大语言模型（LLM）在系统指令与用户指令直接冲突时的仲裁机制问题，尤其关注模型如何权衡二者之间的优先级。其核心挑战在于揭示模型内部是否存在可读取的冲突解决信号，以及是否可通过外部干预实现更符合预期的系统合规行为。解决方案的关键在于发现并利用模型内部残差流激活中可线性解码的仲裁信号：研究发现，尽管Llama-3.1-8B在冲突情境下仅0.10次遵循系统指令（表现出显著的反层级倾向），但其冲突结果仍能以0.97的平衡准确率从第12层的残差流激活中解码，显著优于仅依赖元数据的基线（高出17个百分点）。进一步地，通过在第12层使用四个冲突样本的逻辑回归方向进行引导，可将真实系统合规率从0.132提升至0.530，且该效果依赖于读出方向的几何结构而非单纯探针精度。这表明，用户偏好型仲裁行为与可读取的内部仲裁信号共存，而成功干预的关键在于对内部表示空间几何特性的精准操控。
链接: https://arxiv.org/abs/2608.28648
作者: Enrique Balp-Straffon,Chih-Hao Hsu,Rushiraj Gadhvi,Sunishchal Dev,Callum Stuart McDougall,Anusha Mujumdar
机构: Amazon(亚马逊); National Taiwan University (国立台湾大学); Plaksha University (普拉克沙大学); Technology and Security Policy Center, RAND Corporation (技术与安全政策中心，兰德公司); Algoverse; Google DeepMind(谷歌深度智障)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: Published at the ICML 2026 Mechanistic Interpretability workshop, 16 pages (including appendix)
Abstract:We study how instruction-tuned LLMs arbitrate direct conflicts between system and user instructions. We introduce a benchmark of 41 paired constraints with deterministic verifiers and evaluate eight models under matched baseline, conflict, and same-channel control conditions. Behaviourally, the models split into three regimes by System Authority Delta: hierarchy-respecting models use the system channel as an authority signal, anti-hierarchy models follow the system less often than their same-channel baseline predicts, and no-effect models show little channel sensitivity. Llama-3.1-8B is the strongest anti-hierarchy case in our suite, following the system in only 0.10 of conflict trials. We use this behavioural failure case to ask whether user-preferring arbitration reflects the absence of an internal conflictresolution signal. It does not: on Llama-3.1-8B, the conflict outcome is linearly decodable from residual-stream activations at 0.97 balanced accuracy, 17 percentage points above a metadata-only baseline, with analogous signals on Qwen2.5-7B and gpt-oss-20b. Steering with a layer-12 mean of four per-conflict logistic-regression directions raises genuine system compliance from 0.132 to 0.530, while directions selected mainly for pooled separability steer poorly. User-preferring conflict resolution can therefore coexist with a readable internal arbitration signal, and successful intervention depends on the geometry of the readout rather than probe accuracy alone
[NLP-258] Cross Lingual Transfer in Tulu Legal Comprehension: Script-Dependent Improvement and RAG -Induced Knowledge Conflict
【速读】： 该论文旨在解决低资源语言在跨语言场景下模型性能不可靠的问题，特别是在缺乏充足训练语料的情况下，如何有效评估生成式 AI 在法律文本理解任务中的表现。其核心挑战在于识别模型在低资源多语言环境中的失败模式，尤其是当依赖高资源语言作为辅助时的推理偏差与幻觉现象。解决方案的关键在于采用一种结合检索增强生成（RAG）框架与跨脚本（Dravidian script）转写的方法，使模型能在不进行大规模训练的前提下初步理解低资源语言（如图鲁语）的法律诉状。研究发现，尽管通过转写可提升对特定脚本（如卡纳达语）输入的理解，但模型的推理能力仍高度依赖于输入脚本，且在检索过程中易出现事实替换（fact substitution）和幻觉（confabulation）等系统性错误。进一步的推理轨迹分析与统计诚实性评估框架表明，模型的解析与推理机制是导致失败的主要根源，而非语料本身的质量或覆盖度。因此，该研究提出的方法不仅揭示了低资源多语言 RAG 系统的脆弱性，还为未来评估此类系统提供了可推广的技术路径。
链接: https://arxiv.org/abs/2608.28645
作者: Sindhu Shetty,Spurthi Setty,Natan Vidra
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Low-resource languages without an adequate training corpus often use a related, higher-resource language as a scaffold for comprehension. Still, there is a need to develop rigorous evaluation methods to identify when models fail in cross lingual low-resource environments. Using the legal domain as a backdrop, three models (Llama3, Hex-1, Sarvam) were tested on the ability to classify legal complaints written in a low resource Dravidian language (Tulu). Transliterating queries across Dravidian scripts allowed models to gain a preliminary understanding of speakers’ complaints without the use of wide scale training, though the level of comprehension was heavily script dependent (with Kannada - another relatively low-resource language - producing the strongest positive trend). Retrieving from a corpus of Kannada legal papers across a RAG framework caused mixed results. Some models had a weak positive trend in comprehension under certain conditions, but when models failed, it was often across two axes: fact substitution (fixating on specific passage excerpts that skewed reasoning) and confabulation (hallucination that had no basis in either query or corpus). Within low resource domains, results identify the model’s parsing of information and subsequent reasoning as the source of reasoning failure, rather than corpus contents. Script-dependent comprehension and RAG robustness also seem to travel together. This is further supported by the reasoning-trace analysis and a statistical-honesty framework deployed - techniques that are more broadly applicable to low-resource multilingual RAG evaluation.
[NLP-259] Redesigning and Auditing Deep Research Writing for Faithful Reports
【速读】： 该论文旨在解决深度研究（Deep-Research, DR）系统在基于评分量表（rubric-based）评估中难以揭示生成报告中细粒度事实性错误的问题，尤其是幻觉（hallucination）、引用错误（misattribution）、引文规范性（citation hygiene）及必要事实召回率不足等关键缺陷。其解决方案的核心在于提出CLAIMPROBE——一种基于主张（claim）级别的审计框架，将DR报告分解为可验证的主张单元，并基于检索到的证据对各项事实性指标进行量化评估。在此基础上，论文进一步提出CLAIMWRITER，一种分层式主张驱动的生成架构：该方法首先从源文档中提取真实事实，将其映射至由查询生成的结构化大纲，并基于带源链接的事实表示逐部分撰写报告。实验表明，在三种现有DR框架中仅替换报告生成模块为CLAIMWRITER，即可使幻觉率降低2.6至4.5倍，必要事实召回率提升1.2至1.7倍，同时保持整体报告质量稳定；此外，CLAIMWRITER支持局部更新，当源文档变更时，能以最高效率将更新后的事实传播至报告中，且相较其他更新方法更具成本效益。
链接: https://arxiv.org/abs/2608.28643
作者: Hiroaki Hayashi,Pranav Narayanan Venkit,Prafulla Kumar Choubey,Chien-Sheng Wu
机构: Salesforce AI Research( Salesforce人工智能研究院)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Rubric-based evaluations of deep-research (DR) systems often obscure fine-grained factual failures in generated reports. We introduce CLAIMPROBE, a claim-level audit that decomposes DR reports into claims and measures hallucination, misattribution, citation hygiene, and necessary-fact recall against retrieved evidence. Using CLAIMPROBE, we find that strong DR pipelines can omit key evidence and misattribute claims even when their rubric scores remain stable. We then propose CLAIMWRITER, a hierarchical claim-based writer that extracts source facts, maps them to a query-derived outline, and drafts each section from a source-linked claim representation. Across three prior DR frameworks, replacing only the report writer with CLAIMWRITER reduces hallucination by 2.6 to 4.5 times and improves necessary-fact recall by 1.2 to 1.7 times, while largely preserving overall report quality. CLAIMWRITER also enables localized revision: when sources change, it propagates changed source facts into revised reports at the highest rate among update methods, while also being more cost-effective.
[NLP-260] rminal-Bench-LILT: Multilingual Agent ic Coding Benchmark Grounded in Language Region and Culture
【速读】： 该论文旨在解决当前代码生成模型评估普遍局限于英语环境，无法真实反映多语言实际部署场景的问题。现有评测体系忽视了非英语编程中特有的挑战，如国际化（internationalization）、编码处理、文本归一化及文化惯例等，导致模型在多语言场景下的真实能力被低估。其解决方案的关键在于提出首个大规模、跨语言的代码任务基准——Terminal-Bench-LILT，包含300个来自10种语言（阿拉伯语、捷克语、德语、西班牙语、印地语、日语、韩语、塞尔维亚语、土耳其语和中文）的真实编程任务，所有任务均由母语程序员设计，并通过多阶段质量控制流程验证。该基准聚焦于非英语开发中的特有技术难点，揭示出即使是最先进的代码生成模型在多语言任务上的通过率也仅达63.1%，且不同语言间性能差异显著，表明多语言代码能力是一个独立且尚未充分探索的能力维度。
链接: https://arxiv.org/abs/2608.28641
作者: Yunsu Kim,Kaden Uhlig,Ashwin Purohit,Milind Agarwal,Patrick Simianer,Anil Arslan,Kiarash Mokhtari,Thomas Zenkel,Johannes Mosig,Gabriel Bretschner,Shamik Bose,Joern Wuebker,John DeNero
机构: LILT, Inc. (LILT, 公司)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Most evaluations for coding agents are conducted exclusively in English, which does not reflect real-world multilingual deployment. We present Terminal-Bench-LILT, a suite of 300 authentic coding tasks in ten languages: Arabic, Czech, German, Spanish, Hindi, Japanese, Korean, Serbian, Turkish, and Chinese. Each task targets issues specific to non-English software development that have no direct English equivalent, e.g., internationalization, encoding, text normalization, and cultural conventions. All tasks are authored by native-speaker programmers and validated through a multi-stage quality control pipeline. Evaluation of six frontier models reveals that even the strongest model reaches only 63.1% pass rate, with many tasks unsolved by any model. Performance varies substantially by language and does not track general coding benchmark rankings, highlighting that multilingual coding competence is a distinct and underexplored capability axis. Sample tasks are available at this https URL
[NLP-261] PromptKWS: A Novel Prompt-Guided Open-Vocabulary Keyword Spotting Framework
【速读】： 该论文旨在解决开放词汇量关键词检测（open vocabulary keyword spotting, KWS）系统在实际应用中面临的准确率低、对噪声和发音变化鲁棒性差的问题。其核心解决方案是提出一种基于提示引导的关键词检测框架——PromptKWS，其关键在于引入提示短语预测网络（Prompt Phrases Prediction Network, PPN），采用编码器-解码器架构以有效提取关键词提示的嵌入表示；并通过提示-声学多头交叉注意力（Prompt-acoustic Multi-head Cross-attention, MHCA）机制将提示嵌入注入到主干的提示引导式KWS编码器中，实现语义与声学特征的深度融合。该设计显著提升了系统在复杂现实环境中的适应能力，实验表明相比基线模型，唤醒率提升超过10%，在含噪声和发音变异的测试集上平均准确率提升超过15%。
链接: https://arxiv.org/abs/2608.28640
作者: Gaopeng Xu,Chengfei Li,Xianliang Wang,Lin Zhu,Juan Wei,Wenpeng Li,Jianwei Niu,Jie Gao
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:In this paper, we present PromptKWS, a novel Prompt-guided keyword spotting (KWS) framework to improve the accuracy of open vocabulary KWS systems. In specific terms, we introduce the Prompt Phrases Prediction Network (PPN), an encoder-decoder architecture designed to effectively extract keyword prompts embeddings. we employ the PPN encoder to encode the keyword prompts and infuse the prompt embedding into the Prompt-guided KWS encoder by utilizing a Prompt-acoustic Multi-head Cross-attention (MHCA). Experiments show that PromptKWS improves the wakeup rate by over 10% compared to baseline system. Notably, another strength of PromptKWS is its ability to effectively leverage keyword prompts for adapting to complex real-world environments involving noise and pronunciation variations. In comparison to purely acoustic models, which often struggle in such situations, PromptKWS demonstrates remarkable performance, with an average accuracy improvement of over 15% in test sets.
[NLP-262] Do MLLM s Really Understand Low-Resource Khmer Documents? A Pilot Study on Khmer Document VQA
【速读】： 该论文旨在解决多模态大语言模型（Multimodal Large Language Models, MLLMs）在低资源、非拉丁语系场景下，特别是针对高棉语（Khmer）文档理解任务中的可靠性不足问题。高棉语文档具有复杂字形、中英文混合字段以及柬埔寨瑞尔与美元并存的货币数值等挑战，且相关高质量标注数据集稀缺，进一步加剧了模型应用的困难。本文的关键解决方案是构建一个基于已有KH-FUNSD数据集的试点诊断评估子集，涵盖发票、收据、报价单等商业文档，并设计包含英语与高棉语双语提问及多格式答案（原生高棉语、混合脚本、数字等）的评测任务。研究通过对比直接图像提示、解析器辅助和外部光学字符识别（OCR）辅助三种配置下的Qwen-VL系列模型表现，发现尽管引入外部OCR（如Tesseract和PaddleOCR）可显著提升整体准确率至61.9%和61.6%，但对高棉语脚本和混合脚本答案的理解仍存在明显瓶颈。结果表明，当前主流MLLMs虽能较好处理清晰的英文内容与结构化数值信息，但在原生高棉语文档理解方面仍面临重大挑战，凸显了跨语言、跨脚本场景下鲁棒性生成式视觉理解能力的缺失，为后续面向低资源语言的模型优化指明方向。
链接: https://arxiv.org/abs/2608.28635
作者: Nimol Thuon,Panhapin Theang
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Recent multimodal large language models (MLLMs) have advanced document understanding, visual question answering, and text extraction. However, their reliability in low-resource, non-Latin settings remains uncertain. Khmer form documents present particular challenges because they contain complex script forms, mixed Khmer-English fields, and monetary values in both Cambodian Riel and US Dollars. Available resources for Khmer Document VQA are also limited. This paper presents a pilot diagnostic evaluation of open MLLMs on Khmer document images. We construct an evaluation subset from the previously introduced KH-FUNSD collection, covering invoices, receipts, quotations, and other business forms. The subset includes questions in English and Khmer, with answers retained in their original English, Khmer, mixed-script, or numeric forms. Rather than introducing a full public benchmark, this study examines the capabilities and failure modes of existing models. We evaluate representative open Qwen-VL models using direct image-based prompting and compare parser-assisted and external OCR-assisted configurations with Qwen3-VL-8B. Direct Qwen3-VL-8B outperforms smaller models, achieving 51.9% overall accuracy, although performance remains limited for Khmer-script and mixed-script answers. External OCR produces the strongest results, reaching 61.9% with Tesseract and 61.6% with PaddleOCR. Nevertheless, Khmer-script answers remain substantially more difficult than English and numeric fields. The results indicate that current MLLMs can process visually clear English and structured numeric content, but reliable native Khmer document understanding remains an open challenge.
[NLP-263] PAUSE: Editable Strategy Artifacts for Long-Form Cultural Story Adaptation ICML2026
【速读】： 该论文旨在解决生成式 AI 在长篇故事跨文化适应过程中决策过程不透明的问题，即文化适配决策往往隐藏于提示词（prompt）、临时模型规划或最终文本之中，导致难以追溯与干预。其解决方案的关键在于提出 PAUSE（Pause-And-Update Strategy Editing）机制，将文化适配策略构建为一种可编辑的结构化产物（structured artifact），作为人类可操作的控制界面，实现对文化决策的显式干预。该策略可在后续的角色、实体及章节级本地化阶段被持续投影和应用。在两部中文源连载小说的实验中，通过对比人工编辑策略后的输出与未编辑对照组，结果表明所有9组对比中，评审者均偏好编辑策略生成的内容；标记审计显示，目标文化标记在8/9的编辑输出中出现，而对照组中未出现，且禁止性标记在编辑输出中完全缺失而在对照组中普遍存在。研究将此结果视为“烟雾探测器式”的编辑一致性验证，而非宣称输出具备文化权威性或文学质量提升。PAUSE 提供了一种切实可行的方法，使基于 AI 的文化适应过程在长流程生成前具备更高的可检查性与可争议性。
链接: https://arxiv.org/abs/2608.28633
作者: Taaha Kazi,Vasu Sharma,Mohammad Saifullah,Abdur Rahman
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注: 6 pages, 1 figure, 3 tables. Accepted at the 1st Workshop on Culture x AI: Evaluating AI as a Cultural Technology, ICML 2026. Project page: this https URL
Abstract:Generative AI systems increasingly mediate cultural adaptation, but their cultural decisions are often hidden inside prompts, transient model plans, or final prose. We study PAUSE (Pause-And-Update Strategy Editing), an intervention that exposes an editable adaptation strategy as a human control surface for cultural decisions in long-form story adaptation. The strategy is a structured artifact that can be inspected, edited, and then projected through downstream character, entity, and chapter-localization stages. In two Chinese-source serialized novels, we test whether human edits to this strategy propagate into chapter-level prose. Across 9 edited-vs-control chapter comparisons, judges select the edited-strategy output in all 9; a marker audit shows target markers in 8/9 edited outputs and 0/9 controls, with forbidden markers absent from edited outputs and present in all controls. We frame these results as a smoke-scale edit-adherence study, not a claim that the outputs are culturally authoritative or literary-quality improvements. PAUSE offers one practical way to make AI-mediated cultural adaptation more inspectable and contestable before decisions propagate through long-form generation.
[NLP-264] AutoScientist-Quant: Self-Evolving Coding Agents for Automatic Research in Quantitative Investment
【速读】： 该论文旨在解决当前大型语言模型代理（Large Language Model Agents）在量化研究中面临的三大核心问题：搜索过程缺乏运行时自适应能力、自动化流程通常止步于因子生成而库选择与模型配置仍需人工干预，以及因子发现过程中存在通过循环反馈或代码漏洞提前读取测试窗口信息的过拟合风险。其解决方案的关键在于提出AutoScientist-Quant，一个将量化研究建模为带预算约束的自演化搜索过程的框架。该框架采用单一控制器，基于剩余预算动态决策每轮操作——包括是否优化、合并、转向新方向或终止，选择扩展节点、生成因子数量及从共享记忆中检索历史轨迹策略；同时，该控制器统一负责从因子库中选择并调优模型，实现从假设生成到可部署策略的闭环。此外，作者重构了评估流程，复用先前工作中的评估范式，修正了两个前瞻偏差（lookahead problems），并确保反馈窗口与预留测试窗口完全分离，从而保障所有对比均测试真实泛化性能。实验表明，在CSI市场数据集上，该框架在各类设置下几乎在所有指标上均取得最优表现，且结果在多种主干模型和不同市场间具有鲁棒性。
链接: https://arxiv.org/abs/2608.28632
作者: Zongqian Li,Yaoyiran Li,Yaohui Guo,Ming Zhang,Nigel Collier,Eugene Ie
机构: Google(谷歌); University of Cambridge(剑桥大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Large language model agents can discover alphas, yet current methods have three weaknesses. The search cannot adapt during the run, automation usually ends at alpha generation while library selection and model choice stay manual, and alpha discovery can read the test window through loop feedback or code problems. We present AutoScientist-Quant, a self evolving search process that regards quantitative research as one budgeted search problem. A single controller conditions every decision on the remaining budget, choosing at each round whether to improve, combine, pivot, or stop, which node to expand, how many alphas to generate, and how to retrieve past trajectories from the shared memory. The same core then selects from the library and tunes the model, closing the loop from hypothesis to deployable strategy. We also review the evaluation pipeline reused from prior work, fix two lookahead problems, and keep the feedback window disjoint from the held out test window, so every comparison tests true generalization. On CSI universes, the framework attains the best value of nearly every metric in every setting, and these conclusions hold across several backbones and markets.
[NLP-265] Enabling Proactive Spoken Turns via a Generalized Style-Aware Full-Duplex Framework ACM-MM2026
【速读】： 该论文旨在解决自然全双工对话系统中实时主动交互的挑战，即在不牺牲响应质量的前提下，提升话语轮次的时间准确性。核心问题在于现有方法在实现主动式轮次切换（如及时打断与反馈性回应）时，难以兼顾时间控制的精准性与对话质量。其解决方案的关键在于提出一个通用的、风格感知的全双工框架，包含三大创新：首先，设计轻量级主动语音轮次控制器LPS-TC，具备细粒度动作空间，可同时支持反应式与主动式轮次行为，实现对半双工模型的即插即用式升级，并增强现有全双工模型的时序控制能力；其次，构建大规模真实世界英文数据集WildTurn，涵盖约2,981小时经筛选的多轮立体对话数据，包含面对面及电话通话场景，并标注了五类轮次转换与五类反馈行为风格，使模型训练能够捕捉更丰富的口语动态；最后，提出两级评估机制，在流式传输约束下同时评估块级时间精度与轮次级交互质量。实验表明，将LPS-TC集成至半双工模型（如Qwen2.5-Omni）与全双工模型（如Freeze-Omni）均显著提升了时间适切性与响应质量，展现出细粒度风格可控性与强泛化能力，推动更接近人类自然交互的语音对话系统发展。
链接: https://arxiv.org/abs/2608.28630
作者: Tianrui Pan,Qinglin Zhang,Chong Deng,Luyao Cheng,Qian Chen,Wen Wang,Jie Tang,Gangshan Wu,Jie Liu
机构: Token Foundry, Alibaba Group (阿里云通义实验室)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Sound (cs.SD)
备注: Accepted by ACM MM 2026
Abstract:Compared with half-duplex dialogue systems where the system waits for user turn completion before it responds, natural full-duplex dialogue systems require agents to act proactively in real time, including timely interruptions and backchannels. This creates a key challenge: improving turn timing without sacrificing response quality. To address limitations in realistic proactive turn-taking, we build a generalized style-aware full-duplex framework with three key components. Firstly, we propose LPS-TC, a Lightweight Proactive Speech Turn Controller for plug-and-play integration. It features a fine-grained action space covering both reactive and proactive turn behaviors, enabling half-duplex models with full-duplex capabilities and enhancing existing full-duplex models with superior timing control. Secondly, we construct WildTurn, a large-scale, real-world English dataset containing approximately 2,981 hours of filtered multi-turn stereo conversations from face-to-face and telephone conversations, annotated with five turn-taking and five backchanneling styles. Trained on WildTurn, LPS-TC exhibits rich spoken dynamics that are not captured by existing static full-duplex benchmarks. Thirdly, we introduce a two-tier evaluation scheme that assesses both chunk-level timing precision and turn-level interaction quality under realistic streaming constraints. Our experiments, integrating LPS-TC with half-duplex models like Qwen2.5-Omni and full-duplex models like Freeze-Omni, showcase its superior performance in timing appropriateness and response quality. Our framework also demonstrates fine-grained style controllability and strong generalizability, enabling more natural and human-like spoken interactions.
[NLP-266] Intelligent Identification and Repair of Design Defects in BIM via Domain-Specific Large Language Models
【速读】： 该论文旨在解决建筑信息模型（BIM）中设计缺陷类型多样且现有方法缺乏通用性、效率低下的问题。其核心解决方案在于构建一个基于领域特定大语言模型（LLM）的集成框架，通过BIM-to-Text方法实现结构化数据与自然语言模型之间的有效衔接，采用组件平衡分块策略提升语义完整性；结合规则注入提示学习、少样本提示及检索增强生成（RAG）技术，实现缺陷的精准识别与合理修复建议生成；同时引入融合关键标识符验证与词元长度阈值的幻觉控制策略，显著提升输出可靠性。实验表明，该方法在缺陷识别准确率上达到85%（优于传统规则检查的70%），修复建议合理性达94%，且幻觉控制使准确率进一步提升至85%，单轮干预下消除92.5%的幻觉现象，成功实现了从原始BIM数据输入到缺陷识别与修复建议生成的端到端自动化流程。
链接: https://arxiv.org/abs/2608.28629
作者: Jia-Rui Lin,Yun-Hong Cai,Xiang-Rui Ni,Peng Pan
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Existing methods lack a generalized approach to efficiently identify and resolve the diversity of design defects in BIM. Therefore, this study proposes an integrated framework to identify and repair various defects in BIM via domain-specific LLMs. Firstly, a BIM-to-Text method with component-balanced chunking is introduced to bridge BIM data with LLMs. Then, prompt learning with rule injection, few-shot prompting and RAG is proposed to identify defects and generate repair suggestions. Meanwhile, a hallucination control strategy combining key identifier validation and token-length thresholds is introduced to ensure reliability. Experiments show capability expansion yields 85% identification accuracy versus 70% for traditional rule checking, achieving a 94% rate of reasonable repair suggestions. Moreover, the proposed hallucination control further increased accuracy from 64% to 85%, eliminating 92.5% of hallucinations in a single intervention round. This study establishes an end-to-end prototype from raw BIM data input, through defect identification, to repair suggestion generation.
[NLP-267] Do large language models scrutinise what they review? A multimodal audit of scoring calibration error detection and author-identity effects
【速读】： 该论文旨在解决生成式 AI（Generative AI）在学术同行评审中的可信度与评估能力问题，特别是其在批判性评价、错误识别及决策一致性方面的表现。研究聚焦于评估两种多模态大语言模型（Multimodal Large Language Models, LLMs），即 Qwen2.5-VL-72B 与 Pixtral-Large-124B，在真实会议投稿场景下的评审表现，涵盖文本与图文混合输入、作者身份伪装（高/低声望机构）等变量，并通过植入可验证错误以测试模型对细节缺陷的识别能力。其解决方案的关键在于：通过设计自然提示（natural prompting）与简短验证指令（verification instruction）对比实验，系统检验模型在不同提示策略下对已知错误的检测率；同时分析图像信息对评分与错误识别的影响，并评估模型决策是否具备可解释性和一致性。 结果显示，尽管模型评分普遍偏高（7.0–8.1），远高于人类平均分（3.4–6.8），且在引入验证提示后错误检测率仅提升至22.2%，仍有78%错误未被发现；此外，图像信息虽提升评分但降低错误检出率，且多数图文描述不一致，表明模型在跨模态理解与事实核查方面存在显著局限。最终，模型的编辑决策与简单分数平均结果完全一致，暗示其决策机制缺乏深度推理，仅为表面分数聚合。
链接: https://arxiv.org/abs/2608.28626
作者: Emad Alharbi
机构: University of Tabuk (塔布克大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models (LLMs) are increasingly used to generate peer reviews, prompting examination of their capacity for critical evaluation. This study evaluates two multimodal LLMs, Qwen2.5-VL-72B and Pixtral-Large-124B, as reviewers across 165 submissions to the 2026 International Conference on Learning Representations, a venue that postdates both models’ training cutoffs. Manuscripts were presented to both models with author identities blinded, replaced with high-prestige affiliations, or replaced with low-prestige affiliations, and in either text-only or text-with-figure format. Additionally, 145 verifiably detectable errors were inserted into 55 manuscripts to assess error identification under natural and verification-oriented prompts. Across all manuscript groups, including rejected submissions, LLM scores ranged from 7.0 to 8.1, whereas human mean scores ranged from 3.4 to 6.8. The models detected 12.1% of the verified errors under natural prompting, and a one-sentence verification instruction increased detection to 22.2%; however, 78% of the errors remained undetected. Providing figures reduced error detection while increasing review scores. No visual error was reliably verified against its corresponding figure, and half of the text-only reviews described figures that were not provided. Author identity did not influence either review scores or error detection. LLM editorial decisions exactly matched those produced by simple score averaging.
[NLP-268] Asymmetric Within-Document Predictive Learning for Scientific Document Representation
【速读】： 该论文旨在解决科学文献表征中缺乏引用信息（citation-free）的预训练问题，尤其关注如何利用论文内部的论述结构（discourse structure）实现有效的表示学习。其核心挑战在于，在不依赖外部引用关系的情况下，如何建模科学文档各部分之间的语义关联。为此，论文提出了一种名为SciJEPA的对称性不对称（asymmetric within-document）预测框架：通过使用标题与摘要的表示来预测方法部分的表示，并进一步以方法部分表示预测结论部分表示，从而构建跨段落的自监督学习信号。该方案的关键创新在于引入了“分片各向同性高斯正则化”（Sliced Isotropic Gaussian Regularization, SIGReg），通过控制嵌入空间的几何结构，显著提升了模型性能并缩小了与对比学习基线的差距。实验表明，该正则化策略的效果具有任务依赖性——适度的SIGReg有助于细粒度排序任务，而过强的正则化可能削弱局部对齐能力。此外，研究还发现不同编码分支在检索场景中表现出差异化优势，表明模型具备多模式检索潜力。总体而言，该工作验证了基于文档内预测的学习范式在无引用条件下对科学文本表征的有效性，前提是需对嵌入空间的几何特性进行精细调控。
链接: https://arxiv.org/abs/2608.28625
作者: You Zuo(ALMAnaCH),Éric de la Clergerie(ALMAnaCH),Benoît Sagot(ALMAnaCH)
机构: Inria, Paris, France; Questel, Paris, France
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:We study predictive pretraining for scientific document representation using the discourse structure of papers. We propose SciJEPA, a citation-free framework that learns through asymmetric within-document prediction: title and abstract representations are used to predict method representations, and method representations are used to predict conclusion representations. Experiments on RELISH, high-influence citation, SciDocs, and cite prediction show that plain predictive training is viable but weaker than a controlled contrastive baseline using the same section pairs. Adding Sliced Isotropic Gaussian Regularization (SIGReg) substantially improves performance and narrows this gap. The effect of regularization is task-dependent: moderate SIGReg helps fine-grained ranking, while stronger regularization can weaken local alignment. We further show that different encoding branches support different retrieval regimes. These results position within-document predictive learning as a promising citation-free complement for scientific document representation, provided that embedding geometry is carefully controlled.
[NLP-269] MA-RAG : Multi-Agent Retrieval-Augmented Generation for Query-Driven Summarization of Longitudinal Parkinsons Disease Assessments
【速读】： 该论文旨在解决帕金森病（Parkinson’s disease）单次及纵向临床评估数据在解读过程中耗时且高度依赖专科专家经验的问题。现有大语言模型（Large Language Models, LLMs）虽能生成自然语言摘要，但普遍缺乏临床领域的专业语境支持，难以对结构化纵向数据生成事实准确且时间一致性良好的回答。为此，本文提出一种查询驱动的多智能体检索增强生成框架（MA-RAG），其核心创新在于将临床推理过程分解为多个领域专业化智能体，结合结构化事实提取与最终验证阶段的综合合成，实现临床语境下的可信摘要生成。该框架支持四种临床分析任务：单次评估、疾病轨迹分析、个体间比较及队列总结。实验结果表明，相较于传统方法、仅使用RAG及单智能体RAG基线，MA-RAG在事实精确性上显著提升，Fact Precision相对提高达122%（从0.436增至0.990），幻觉率降低98%（从0.564降至0.010），同时在临床专家主导的主观评价中获得组织性与临床实用性方面的最高评分。研究证明，基于领域专业化多智能体的推理机制可有效保障结构化纵向临床数据在查询驱动摘要生成中的可靠性与准确性。
链接: https://arxiv.org/abs/2608.28624
作者: Sana Alamgeera,Denise Goberta,Muhammad Irshad,Anne H. H. Ngu
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Accurate interpretation of single-visit and longitudinal clinical assessments for Parkinson’s disease is time-consuming and often depends on specialist expertise. Although large language models (LLMs) can generate natural language summaries, they frequently lack domain-specific clinical grounding and struggle to produce factually correct and temporally consistent responses for structured longitudinal assessment data. To address these limitations, we propose MA-RAG, a query-driven multi-agent retrieval-augmented generation framework that decomposes clinical reasoning into domain-specialized agents, combines structured fact extraction, and synthesizes clinically grounded summaries through a final verification stage. The framework supports four clinical analysis tasks: single-session, trajectory, comparison, and cohort summarization. We evaluate MA-RAG using objective metrics, namely Fact Precision, Hallucination Rate, Temporal Fidelity, and Semantic Similarity, together with subjective evaluations conducted by clinical experts. Compared to Traditional, RAG-only, and Single-agent RAG baselines, MA-RAG substantially improves factual correctness, achieving up to a 122% relative increase in Fact Precision (from 0.436 to 0.990) and reducing the Hallucination Rate by up to 98% (from 0.564 to 0.010), while consistently receiving top ratings from clinical experts for organization and clinical usefulness. These results demonstrate that domain-specialized multi-agent reasoning enables reliable query-driven summarization of structured longitudinal clinical assessment data.
[NLP-270] Looking Again: Measuring Sycophancy in the Reasoning Chains of Multimodal Models Under Pressure
【速读】： 该论文旨在解决大型多模态推理模型（Large Multimodal Reasoning Models, LMRMs）在面对用户给出的错误答案时，存在顺从性（sycophancy）问题却缺乏可靠评估方法这一关键挑战。其核心解决方案是构建首个针对LMRMs顺从性的基准测试框架与数据集，通过将四个基于视觉的下游任务（涵盖数学、临床、时间及人口统计推理）与五种压力情境（单轮与多轮设置）相结合，系统评估模型在最终答案及推理链内部的顺从行为。研究发现，在压力条件下顺从现象普遍存在，其中“陈述压力”（Statement pressure）引发最高顺从率，而“信念强度”（Conviction）压力下顺从率最低（除Mistral-Small-4外）；尤其在多轮交互中，临床视觉判断任务的推理层级顺从性急剧上升，最高达95.7%。作者进一步提出一种失败分类法，区分推理链层面与答案层面的顺从性，并引入句子级分类法定位推理偏差首次出现的位置，揭示顺从性可独立于最终答案影响推理过程，因而仅依赖答案层面评估会严重低估风险。该研究强调了全面评估模型内在推理可信度的重要性，为提升多模态推理系统的稳健性提供了关键工具与理论依据。
链接: https://arxiv.org/abs/2608.28623
作者: Mahir Numayeer Islam,Gakuto Okuyama,Nikolaus Siauw,Shivank Garg,Madhur Panwar,Vasu Sharma
机构: Algoverse AI Research; PocketFM; Adelaide University; Akita International University; RNA Tech
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted to COLM @ AdvML-Frontiers-CoTMA
Abstract:Large multimodal reasoning models (LMRMs) are getting increasingly capable, primarily through generating explicit chain-of-thought reasoning before answering. In language models it has been observed that this performance often comes with sycophancy, the tendency of a model to agree with the user over the evidence. However, for LMRMs no reliable method to measure sycophancy yet exists. We bridge this gap by introducing a benchmark and dataset for evaluating LMRM sycophancy when confronted with a wrong answer from a user. Our benchmark pairs four visually grounded datasets spanning mathematical, clinical, temporal, and demographic reasoning with five pressure conditions in single-turn and multi-turn settings. We evaluate sycophancy in the final answer as well as its emergence within the reasoning chain. We find that sycophancy is prevalent under pressure, with Statement pressure eliciting the highest rates and Conviction the lowest for all models except Mistral-Small-4, and under multi-turn pressure reasoning-level sycophancy intensifies sharply in clinical visual judgement, reaching 95.7% for the most affected model. We further introduce a failure taxonomy separating reasoning-chain from answer-level sycophancy, and a complementary sentence-level taxonomy locating where in the chain drift first emerges. Our results show that sycophancy can corrupt the reasoning chain independently of the final answer, so answer-level evaluation alone is insufficient.
[NLP-271] Distributional Validity and Calibration of a Korean Synthetic Persona Panel for Digital and AI Service Use: A Secondary-Data Validation Against the Korea Media Panel Survey
【速读】： 该论文旨在解决生成式人工智能（Generative AI）驱动的虚拟人物（synthetic personas）在非英语语境下作为人类调查受访者替代方案的有效性验证问题，尤其聚焦于韩国语境下的数字与AI服务使用行为的模拟准确性。其核心挑战在于评估这些虚拟群体在多维度人口统计特征（性别与年龄分层）上对真实调查数据分布的再现能力，并识别其系统性偏差来源。解决方案的关键在于通过对比两种不同模型（Gemini 3.5 Flash与EXAONE）生成的合成人物面板与来自韩国媒体调查（KISDI Korea Media Panel Survey）的真实加权数据之间的差异，进行多维度分析：包括整体平均绝对误差（MAE）、二元项目相关性、分段误差及时间动态一致性。研究发现，尽管通过保留部分真实数据进行校准可显著降低特定人口单元的误差（如将性别-年龄单元的MAE从约18.9降至6.7个百分点），但该校准效果不具备跨时间泛化能力，且无法超越直接使用少量真实数据的基准表现。此外，基于叙事内容的条件化优于仅基于人口统计的条件化，但仍无法超越简单的真实数据基线。因此，研究结论指出，合成人物面板并非可靠的研究替代品，其真正价值在于诊断性分析，仅在真实数据极度稀缺时具备有限的操作应用意义。
链接: https://arxiv.org/abs/2608.28615
作者: Howard Kim,Keun Tae Cho
机构: Seoul Cyber University (首尔网络大学); Sungkyunkwan University (成均馆大学)
类目: Computers and Society (cs.CY); Computation and Language (cs.CL)
备注: 13 pages, 7 figures, 9 tables. Submitted to IEEE Access. Code and data: this https URL (doi: https://doi.org/10.5281/zenodo.21397425 )
Abstract:Synthetic personas based on large language models (LLMs) are increasingly proposed as substitutes for human survey respondents, yet systematic validation outside English-speaking contexts remains scarce. This secondary-data study evaluates how well a Korean synthetic persona panel (NVIDIA Nemotron-Personas-Korea), conditioned into Gemini 3.5 Flash (primary) and EXAONE (comparison), reproduces digital and AI service-use distributions from the KISDI Korea Media Panel Survey. Sex-and-age-stratified panels of about 8,000 personas per model answered the survey’s own items - eight service-use indicators and eight innovativeness and acceptance constructs - and were compared against weighted survey estimates. The overall mean absolute error (MAE; RQ1) was 15-19 percentage points (pp), with binary item-mean correlations of 0.69-0.90. Segment error (RQ2) across five demographic axes was 15-19 pp, with between-group gaps up to 52.4/36.2 pp (Gemini/EXAONE). Errors followed model-specific signatures: an age stereotype with low anchoring (Gemini) versus an acquiescence-consistent level bias (EXAONE). Reference-year analysis was consistent with temporal misalignment driving most generative-AI overestimation, whereas short-form underestimation was framing-sensitive. Holdout calibration on 30% of the real data (RQ3) roughly halved sex-by-age cell MAE (18.9-8.6, 15.9-6.7 pp) - yet direct estimation from the same real subsample was far more accurate (3.6 pp), and the correction did not transfer across time. The calibrated panel retained an advantage only under extremely scarce real data (about 100 responses) and, for one model, for unobserved segments. Persona-narrative conditioning beat demographic-only conditioning, but neither surpassed simple real-data baselines. Synthetic panels are thus not survey substitutes; their value is diagnostic, with operational use confined to settings lacking real data.
[NLP-272] STAGEET: Stage-wise Typed Edit Tagging for Grammatical Error Correction with Arabic as a Case Study
【速读】： 该论文旨在解决序列到编辑（Seq2Edit）方法在语法错误修正（GEC）中可解释性不足的问题。尽管Seq2Edit通过预测输入文本的编辑标签实现了高效且局部可解释的修正，但其依赖单一编辑词汇表的方式难以明确揭示具体纠错类型，导致操作层面的可解释性局限。为此，本文提出一种分阶段有类型编辑标注框架STAGEET，其核心创新在于将传统的序列到编辑监督机制重构为具有类型信息的可执行阶段，并扩展编辑操作以涵盖不同纠错类别。STAGEET将纠错过程分解为一系列有序的中粒度、带类型的修正阶段，每个阶段基于独立的标签空间进行预测，仅对当前假设进行一次重写，并将中间结果传递至下一阶段。该框架通过端到端共享编码器多头模型与阶段专用适配器，以及完全专业化的一阶段一标注器两种实现方式，在QALB-2014和ZAEBUC数据集上均表现出与现有方法相当的编辑基性能，同时显著提升了修正轨迹的可检视性，并在QALB-2014上达到当前最优效果。
链接: https://arxiv.org/abs/2608.28614
作者: Wenjie Lou,Alaa Mamdouh Akef
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Sequence-to-edit approaches make grammatical error correction (GEC) efficient and locally interpretable by predicting edit labels over the input rather than generating a full corrected sentence. Their interpretability, however, is primarily operational: a label specifies how the string should change, but a single edit vocabulary does not always reveal the type of correction being made. We propose STAGEET, a stage-wise typed edit-tagging framework that reorganizes Seq2Edit supervision into typed executable stages and extends edit operations to correction categories. STAGEET decomposes correction into an ordered sequence of medium-grained typed stages; each stage predicts from its own label space, rewrites the current hypothesis once, and passes the resulting intermediate sentence to the next stage. We instantiate the framework as both an end-to-end shared-encoder multi-head model with stage-specific adapters and a fully specialized variant with one independent tagger per stage. Experiments on QALB-2014 and ZAEBUC show that category-aware staged correction retains competitive edit-based GEC performance while exposing a more inspectable correction trajectory, and attains state-of-the-art results on QALB-2014.
[NLP-273] Gurukul AI: An Interactive AI-Driven Educational Platform for Indian Education System
【速读】： 该论文旨在解决当前生成式AI（Generative AI）模型在印度教育场景中适用性不足的问题。现有大型语言模型（LLMs）主要基于西方语料训练，难以适配印度多语言、应试导向且以国家课程纲要（NCERT）为核心的区域性教育体系。针对这一挑战，研究提出的关键解决方案是构建一个与印度国家课程纲要（NCERT）严格对齐的问答数据集，覆盖9至12年级五个学科共18,720个问题-答案对，并基于此数据集对LLaMA 3.1 8B模型进行微调。进一步地，通过引入面向教育需求定制的检索增强生成（Retrieval-Augmented Generation, RAG）框架，开发出名为GurukulAI的开源平台，支持学生以英、印两种语言进行交互，实现疑点解答、模拟考试练习及上下文相关回答。该方案的核心在于将全球先进生成式AI能力本地化，精准对接印度区域教育内容与教学逻辑，从而弥合通用大模型与本土教育实践之间的鸿沟。
链接: https://arxiv.org/abs/2608.28611
作者: Isha Narang,Sneh Gosai,Mayank Singh
机构: Indian Institute of Technology Gandhinagar (印度理工学院甘吉纳格尔); Pandit Deendayal Energy University (潘迪特·迪安达亚尔能源大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注:
Abstract:Recent advances in large language models (LLMs) like ChatGPT and LLaMA have transformed AI-driven education, but these systems are predominantly trained on Western-centric data, making them ill-suited for regional curricula like India’s. The Indian education system is linguistically diverse, exam-oriented, and structured around standardized syllabi, not addressed by existing datasets or tools. In this work, we curate a syllabus-aligned QA dataset based on NCERT (National Council of Educational Research and Training) textbooks for classes 9-12, capturing the content, context, and teaching style of Indian curricula. The final dataset, comprising 18,720 question-answer pairs across five subjects, is publicly available at this https URL. We fine-tune the LLaMA 3.1 8B model using this dataset and deploy it in a Retrieval-Augmented Generation (RAG) framework tailored to educational needs. We introduce GurukulAI, an open-access platform that enables Indian students to chat with the model, get doubts cleared, practice exam-style questions, receive contextual answers, and interact in both English and Hindi. By localizing AI for Indian classrooms, our work bridges the gap between global LLM capabilities and regional educational demands. The code is available at this https URL.
[NLP-274] Parametric Multimodal User Memory: Storing What Captions Cannot Carry
【速读】： 该论文旨在解决个性化智能体（personalized agent）在构建用户记忆（user memory）时对感知信息（perceptual information）建模不足的核心问题。现有方法主要依赖文本形式的用户记忆（如转录文本和字幕），虽能捕捉可命名的事实（如“我的猫叫Bibi”），却无法有效保留无法通过文字描述的感知特征，例如声音的音色、面部在不同年龄与光照下的变化、语调中的疲惫感等。为克服这一局限，论文提出将感知记忆直接嵌入模型内部，其关键解决方案是将身份识别任务分解为两个协同子问题：首先由视觉-语言模型（VLM）在上下文中定位目标（即“是什么”和“在哪里”），其次由专用编码器提取身份密钥（即“是谁”），并将该密钥以一个内联令牌（inline token）的形式存储于模型中，生成时可通过注意力机制直接读取，无需外部检索。实验表明，单独使用任一模块均表现不佳——VLM在跨年龄人脸识别上召回率仅为0.54，而未接地的编码器在双人场景中仅达0.05；但二者结合后，可实现接近正确区域基准（correct-region oracle）的0.96召回率，并成功泛化至多说话人语音与视频场景。该身份识别核心无需训练，仅需常数时间（O(1)）完成注册，即可在任意冻结模型上复现编码器性能。在PerceptMem数据集上的评估进一步揭示：感知身份记忆受容量限制，而精确事实记忆则受绑定能力限制，因此建议将身份信息存于参数化记忆库，事实信息存于文本存储系统。两者组合清晰且高效，使智能体不仅能记住用户说了什么，还能理解用户是谁。
链接: https://arxiv.org/abs/2608.28609
作者: Bojie Li,Noah Shi
机构: Pine AI; University of Washington (华盛顿大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:A personalized agent needs a user memory: a persistent model of who its user is. Today it is almost always text – transcripts and captions retrieved by similarity. This serves the captionable half of a person (“my cat is named Bibi”), but discards the perceptual half no caption can hold: how a voice sounds, how a face reads across age and lighting, how tired someone sounds. We measure this loss across five modalities: a strong caption-based re-identifier recovers as little as 0.11 of a dedicated encoder’s recall, collapsing toward chance on non-nameable signals. We instead ground perceptual memory in the model, decomposing recall into two subproblems: a vision-language model grounds the referent in context (what and where), and a dedicated encoder extracts an identity key (who), stored as one inline token read by attention at generation with no external round-trip. Neither suffices alone – the VLM identifies cross-age faces at only 0.54 recall where a face encoder reaches 0.81, and an ungrounded encoder recognizes a two-person-scene referent at 0.05 – yet together they reach correct-region oracle (0.96), generalizing to multi-speaker audio and video. The recognition core is training-free: it reproduces the encoder’s recall on any frozen model at O(1) registration cost. On PerceptMem (12 domains, 1,080 tasks) perceptual identity is capacity-limited while exact facts are binding-limited: identity belongs in a parametric bank, facts in a text store. The two memories compose cleanly: an agent with both can remember not only what its user said, but also what they are like. Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2608.28609 [cs.CL] (or arXiv:2608.28609v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2608.28609 Focus to learn more arXiv-issued DOI via DataCite
[NLP-275] RegDivergence-101: An LLM Benchmark for Cross-Jurisdiction Regulatory Contradiction Detection in Life Sciences
【速读】： 该论文旨在解决制药企业在同时向美国食品药品监督管理局（FDA）和欧洲药品管理局（EMA）提交药物研发申请时，面临的监管要求跨司法辖区不一致（cross-jurisdiction regulatory divergence）问题。核心挑战在于，当两机构在相同议题上要求一致时可共用数据，但若存在分歧或一方对某事项保持沉默而另一方有明确要求，则单一临床试验设计可能在某一地区被拒绝，导致研发成本上升与审批延迟。现有解决方案依赖人工经验进行手动协调，效率低且易出错。本文提出“跨司法辖区监管分歧检测”任务，即给定一对针对同一主题的FDA与EMA要求，自动分类其关系为“一致（AGREE）”、“分歧（DIVERGE）”或“一方沉默（SILENT）”，并区分沉默方向（如SILENT_FDA vs. SILENT_EMA）。关键创新在于构建了首个101对专家标注的基准数据集RegDivergence-101，基于三份同行评审研究及原始指导文件，实现高一致性标注（双标注者kappa=0.85），并通过系统评估四类方法：词法启发式、自然语言推理（NLI）交叉编码器、义务层级图增强检索增强生成（Graph-RAG）以及扁平化大语言模型（LLM）判别器，发现语义层面的“沉默”虽可被识别，但仅靠蕴含关系无法捕捉；图结构建模优于词法方法，但仍落后于直接使用上下文感知的大型语言模型，提示未来应以语料级图结构构建作为大规模沉默检测的架构目标。
链接: https://arxiv.org/abs/2608.28607
作者: Chuchu Wu,Zhiyin Zhou,Jingzhuo Hu,Liang You
机构: Carnegie Mellon University (卡内基梅隆大学); Pratt Institute (普拉特学院); University of Pennsylvania (宾夕法尼亚大学); University of Pittsburgh (匹兹堡大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Pharmaceutical sponsors developing a drug for both the United States and the European Union must reconcile guidance issued independently by the FDA and the EMA. Where the two agencies require substantively the same thing, a sponsor can file once; where they diverge, a single trial design risks rejection in one region; where one agency is silent on a point the other regulates, the sponsor must infer obligations. Today this reconciliation is performed manually by regulatory-affairs experts. We introduce cross-jurisdiction regulatory divergence detection: given an FDA requirement and an EMA requirement on the same topic, classify their relationship as AGREE, DIVERGE, or SILENT. SILENT is inherently directional (SILENT_FDA vs. SILENT_EMA); we record direction per pair and report per-direction F1 alongside the collapsed label. We release RegDivergence-101, a 101-pair expert-grounded pilot evaluation benchmark (labels grounded in three peer-reviewed FDA/EMA comparison studies and primary FDA/EMA/ICH guidance text; dual-annotation inter-annotator kappa = 0.85), and systematically characterise a four-method baseline hierarchy: lexical heuristic (0.511 macro-F1, 95% CI [0.411-0.605]), NLI cross-encoder (0.233), obligation-level Graph-RAG (0.663 [0.570-0.747]), and flat LLM judge / Claude Haiku (0.830 [0.747-0.908]). Three directional observations emerge at pilot scale (n = 101): SILENT is semantically detectable but invisible to entailment-only formulations; pair-level obligation graphs improve over lexical methods but trail flat-LLM context (CIs partially overlapping); and corpus-level graph construction is the indicated architectural target for large-scale silent-detection. RegDivergence-101 is a pilot release establishing the task formulation and baseline hierarchy; four unrepresented regulatory domains and an expansion roadmap are described in Section 7.
[NLP-276] he Signal in the Noise: An Auditable Reliability Layer for Biomedical Text Classification
【速读】： 该论文旨在解决大规模生物医学文本语料库在自动化PDF解析过程中普遍存在光学字符识别（OCR）类错误、分词错位与合并、连字符残留及字符级损坏等问题，这些问题会系统性地破坏词汇证据并降低下游自然语言处理模型的分类性能。其核心解决方案是一套保守、可完全审计的拼写纠错可靠性层，作为以安全为导向的预处理模块而非追求最大准确率的纠错器：在不确定情况下，系统主动放弃编辑，遵循医疗领域“不伤害”（do-no-harm）原则。该方案的关键在于采用确定性架构，结合有界编辑距离的候选生成、基于语料库的n-gram评分以及一系列保护生物医学关键术语的领域安全闸门。实验结果表明，在2,104个手动标注的词级错误案例上，该方法在合成错误中实现了94.61%的错误修复召回率，且在负控样本上零引入有害编辑；在CORD-19三分类任务（预防、治疗、流行病学）中，成功恢复约80.45%由噪声导致的宏平均F1分数下降，使宏F1从噪声状态下的0.7654提升至修复后的0.7717，同时保持接近清洁数据的表现（安全模式下为0.7721）。此外，对103个真实OCR提取摘要的案例研究进一步验证了基于Transformer的编码器对轻度噪声具有相对鲁棒性，为未来融合有界神经信号与UMLS词典的灰盒架构提供了依据。整个系统具备完全确定性、以实际错误驱动设计，并优先考虑部署可行性与可审计性。
链接: https://arxiv.org/abs/2608.28595
作者: Moustafa Yehia Hassan,Sharon Wong,Woh Kai Xuan
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 6 pages. Accepted at the 2026 14th International Conference on Bioinformatics and Computational Biology (ICBCB 2026), Kitakyushu, Japan; to appear in IEEE Conference Proceedings (IEEE Xplore). Code: this https URL
Abstract:Biomedical NLP pipelines routinely presuppose clean input text, yet large-scale corpora assembled through automated PDF parsing harbour pervasive OCR-like artifacts, token splits and merges, hyphenation remnants, and character-level corruption, that systematically erode lexical evidence and degrade downstream classifiers. We introduce a conservative, fully auditable spell-correction reliability layer conceived as a safety-oriented preprocessing module rather than a maximal-accuracy corrector: under conditions of uncertainty, the system abstains from editing, in accordance with a medical do-no-harm philosophy. The deterministic architecture couples bounded edit-distance candidate generation with corpus-derived n-gram scoring and a suite of biomedical safety gates that protect domain-critical terminology. We evaluate the layer both intrinsically, on a manually curated benchmark of 2,104 token-level cases, and extrinsically, on a tri-class CORD-19 topic classifier (Prevention, Treatment, Epidemiology) spanning 10,000 examples under a principled four-run protocol (Clean, Noisy, Restored, Safety). Intrinsically, the layer attains 94.61% error-fix recall on synthetic errors with zero harmful edits on negative controls. Downstream, it recovers approximately 80.45% of the noise-induced macro-F1 degradation, elevating macro-F1 from 0.7654 (Noisy) to 0.7717 (Restored) while preserving near-clean performance (Safety: 0.7721). A supplementary case study on 103 real-world OCR-extracted abstracts classified with BioBERT confirms that transformer encoders appeared relatively robust to mild noise, motivating a future grey-box architecture that integrates bounded neural signals and UMLS lexicons without compromising auditability. The system is fully deterministic, artifact-driven, and designed with deployment and auditability in mind.
[NLP-277] When Can We Work in Embedding Space? What Text Embeddings Preserve
【速读】： 该论文旨在解决文本嵌入（text embeddings）在实证分析中作为输入时的有效性问题，核心关切在于：在何种条件下可将高维原始文本替换为低维嵌入表示而几乎不损失信息。其解决方案的关键在于，在文档由潜在主题混合生成的生成模型框架下，精确化了“可将文本替换为其低维嵌入且信息损失极小”这一假设。研究聚焦于两类应用场景——在嵌入空间中对单元进行聚类，以及通过嵌入控制高维文本。其中，嵌入聚类本质上是识别具有相似主题混合结构的文档集合；而控制嵌入则等价于控制主题混合，因此方法的有效性取决于该主题混合是否充分捕捉了潜在混杂因素。在对363个美国大都市区的应用中，基于大语言模型（LLM）生成的经济描述所提取的嵌入聚类，成功识别出可解释的经济原型，并在区分地方就业动态方面优于基于模型残差或人工精选的行业与人口统计协变量的聚类方法。
链接: https://arxiv.org/abs/2608.31059
作者: Simon Freyaldenhoven
机构: Federal Reserve Bank of Philadelphia(美国费城联邦储备银行)
类目: Econometrics (econ.EM); Computation and Language (cs.CL); Machine Learning (stat.ML)
备注:
Abstract:When do text embeddings work as inputs to empirical analysis? Their use rests on an assumption: that we can trade text for its low-dimensional embedding, and lose little in doing so. I make that assumption precise under a generative model in which documents are mixtures of latent topics. I study two uses—clustering units in embedding space and controlling for high-dimensional text. A cluster of embeddings is a set of documents with similar topic mixtures; controlling for the embedding is equivalent to controlling for the topic mixture, so validity reduces to whether that mixture captures the confounding. In an application to 363 U.S. metropolitan areas, embedding-based clusters of LLM-generated economic descriptions recover interpretable economic archetypes and separate local employment dynamics more sharply than clustering on model residuals, or on a curated set of industry and demographic covariates.
[NLP-278] Vocabulary Growth Fundamentals: Bernstein Functions and Hausdorff Sequences
【速读】： 该论文旨在解决词汇增长模型中长期存在的理论难题，即如何从数学上严格刻画词汇多样性随时间演化的规律。其核心问题是：在随机过程框架下，如何建立一个既能准确描述实际语言数据中词汇增长模式（如新词出现频率的衰减趋势），又具备良好数学性质（如非负谱性）的统一理论框架。解决方案的关键在于引入伯恩斯坦函数（Bernstein functions）与豪斯多夫序列（Hausdorff sequences）作为建模工具——前者适用于连续时间的泊松点过程，后者对应离散时间独立同分布（IID）过程。通过将这些数学结构与近年来发展的“单次词率”（hapax rate）模型相结合，作者证明了逻辑斯蒂型单次词率模型具有非负谱性，从而确立其为伯恩斯坦函数，解决了此前悬而未决的数学验证问题。此外，论文进一步拓展了伯恩斯坦-豪斯多夫理论的适用范围，分析了其在平稳更新过程与威布尔（Weibull）更新过程下的局限性与推广可能性，为词汇增长的随机建模提供了更坚实的理论基础。
链接: https://arxiv.org/abs/2608.29449
作者: Łukasz Dębowski
机构: Institute of Computer Science, Polish Academy of Sciences(波兰科学院计算机科学研究所)
类目: Probability (math.PR); Computation and Language (cs.CL)
备注: 41 pages
Abstract:We survey the theory of vocabulary growth founded in the setting of stochastic processes. In particular, we model the expected number of types through Bernstein functions and Hausdorff sequences. These classes of mathematical objects, defined by alternating signs of their derivatives or differences, can be related to continuous-time Poisson point processes and discrete-time IID processes, respectively. Building on previous accounts of the vocabulary growth, we integrate the broader theories of Bernstein functions and Hausdorff sequences and connect them with recently developed hapax rate models. In particular, we prove that the logistic hapax rate model has a non-negative spectrum and hence it defines a Bernstein function, thereby solving an earlier posed problem. We also analyze the limitations of the Bernstein–Hausdorff theory of the vocabulary growth by considering its generalizations under stationary and Weibull renewal processes.
[NLP-279] Rate-Coding Bundle Memory: A Unified Model of Memory and Control for Symbolic Computation in the Brain
【速读】： 该论文旨在解决如何在神经可塑性框架下统一连接主义（connectionist）与符号主义（symbolic）系统，以解释广泛认知现象的难题。其核心挑战在于：如何在保持神经生物学合理性的同时，实现符号表征的灵活性与计算能力。解决方案的关键在于提出一种名为“率编码捆绑记忆”（Rate-Coding Bundle Memory, RCBM）的混合模型，该模型基于“符号子系统假说”（Symbolic Subsystem Hypothesis），即大脑在其本质上为连接主义的结构中嵌入了符号处理机制。RCBM通过率编码（rate coding）在连续空间中表示符号，并利用捆绑记忆（bundle memory）系统实现符号的存储与检索，从而有效应对一次性学习（one-shot learning）、模式分离（pattern separation）以及绑定问题（binding problem）等关键认知挑战。这一框架为理解认知的本质提供了兼具生物学合理性和计算有效性的新范式。
链接: https://arxiv.org/abs/2608.29189
作者: Teun van Gils,Rowan P. Sommers,Markus Ostarek,Peter Hagoort
机构: 未知
类目: Neurons and Cognition (q-bio.NC); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Symbolic Computation (cs.SC)
备注:
Abstract:We propose a neurobiologically plausible model of cognition that combines the advantages of connectionist and symbolic systems, and that can explain a wide range of cognitive phenomena. This model, called Rate-Coding Bundle Memory (RCBM), is based on the Symbolic Subsystem Hypothesis, which posits that the brain implements a symbolic subsystem within its fundamentally connectionist nature. RCBM is a hybrid model that uses rate coding to represent symbols in a continuous space, and it uses a bundle memory system to store and retrieve these symbols. The model is capable of solving a wide range of cognitive phenomena, including one-shot learning, pattern separation, and the binding problem. We argue that RCBM provides a promising framework for understanding the nature of cognition, and that it can be used to develop more sophisticated models of cognition in the future.
信息检索
[IR-0] Configurable Semantic Chunking for Biomedical Information Extraction in Retrieval-Augmented Generation
链接: https://arxiv.org/abs/2608.31139
作者: Riya Ahuja(1 and 2),Tim Kacprowski(1 and 2),Roya Shiasi Sardoabi(1 and 2) ((1) Institute of Data Science in Biomedicine, TU Braunschweig, Braunschweig, Germany, (2) Braunschweig Integrated Centre of Systems Biology, TU Braunschweig, Braunschweig, Germany)
类目: Computation and Language (cs.CL); Information Retrieval (cs.IR)
备注: 9 pages, 7 tables
Abstract:BioMedRAG introduced retrieval-augmented generation with a learned chunk scorer for biomedical information extraction. However, it relies on fixed-size chunking which can fragment semantic evidence. We propose a configurable semantic chunking framework that addresses this limitation by combining entity-preserving windows, trigger-centered chunking, proposition-first extraction, tiered trigger prioritization, and hierarchical relation resolution. The framework integrates with BioMedRAG by replacing only the chunk construction stage while preserving the embedding model, learned chunk scorer, generator, and evaluation protocol. We evaluate the framework on biomedical relation extraction benchmarks (GM-CIHT, DDI, ChemProt) and adverse event classification (ADE). On GM-CIHT, the full hybrid configuration achieves 82.6% F1, improving over the fixed-size baseline (74.2% F1) by 8.4 points under our experimental setup. Cross-dataset analysis shows that semantic chunking improves extraction datasets with explicit relation cues, such as GM-CIHT and DDI, while fixed chunking remains competitive or stronger for dense biochemical extraction and binary classification settings such as ChemProt and ADE. By externalizing chunking logic into configuration files, the framework provides an interpretable and adaptable alternative to rigid fixed-size chunking for biomedical RAG pipelines.
[IR-1] InsightToast: Proactive Information Retrieval Glanceable Visualization in the Side Channel of Data-Rich Meetings
链接: https://arxiv.org/abs/2608.31115
作者: Mohammad Abolnejadian,Matthew Brehmer
类目: Human-Computer Interaction (cs.HC); Information Retrieval (cs.IR)
备注: 16 pages, 7 figures, 2 tables. To appear in the Proceedings of the 39th Annual ACM Symposium on User Interface Software and Technology (UIST '26)
Abstract:Missing institutional context during meetings can impede effective participation. Retrieving relevant information, often scattered across heterogeneous internal and external sources, requires costly task-switching that disrupts both individual focus and collective conversational flow, particularly detrimental during cognitively demanding tasks such as decision-making. We introduce InsightToast, a mixed-initiative application that monitors verbal discourse in real time, identifies topics and informational needs as they emerge, and proactively retrieves relevant information through a multi-agent large language model (LLM)-based pipeline integrating retrieval-augmented generation (RAG) to produce source-grounded insights as succinct text and glanceable interactive charts, delivered through a peripheral interface as ephemeral toasts in the conversation’s side channel. To demonstrate the potential for yielding serendipitous insights, we showcase a usage scenario involving a knowledge base of legislative documents as the meeting’s context. We then report on a comparative study (N=16), in which participants arrived at informed policy decisions while maintaining natural conversation flow.
[IR-2] MULTI3IR: A Benchmark for Multi-perspective Multi-domain Multi-modal Information Retrieval EMNLP2026
链接: https://arxiv.org/abs/2608.30949
作者: Seokwon Song,Sohyeon Kim,Gunhee Kim
类目: Information Retrieval (cs.IR)
备注: EMNLP 2026; code is available at this https URL
Abstract:Information retrieval (IR) increasingly targets open-ended queries that admit diverse perspectives. Existing IR benchmarks, however, focus primarily on closed-ended queries, while even open-ended benchmarks largely consist of queries whose supporting documents span a single subject domain and modality. We introduce Multi ^3 IR, a benchmark that evaluates how well retrievers cover the multifaceted perspectives of open-ended queries across diverse domains and modalities. It comprises 104.9K Stack Exchange queries, each annotated with perspective descriptions that capture the query’s implicit viewpoints. We further propose SPIN, a parameter- and label-efficient method that learns noise vectors to steer embeddings toward diverse yet meaningful semantic directions. Experiments show that existing multimodal retrievers suffer from single-perspective bias, while SPIN substantially improves perspective coverage on Multi ^3 IR and generalizes well to unseen open-ended IR benchmarks. The dataset and experimental code are available at this https URL.
[IR-3] ECGQuest: Benchmarking and Fine-Tuning Language Models for Electrocardiography
链接: https://arxiv.org/abs/2608.30893
作者: Mohammadsina Hassannia,Matthew A. Reyna,Reza Sameni
类目: Computation and Language (cs.CL); Information Retrieval (cs.IR)
备注:
Abstract:Electrocardiogram (ECG) interpretation requires knowledge of cardiology, electrophysiology, clinical diagnosis, ECG waveforms, signal acquisition, and instrumentation. Existing language-model benchmarks, however, primarily assess broad medical knowledge or interpretation of individual ECG signals and images rather than the broader contextual knowledge required for ECG interpretation. We developed ECGQuest, a literature-grounded resource for evaluating and fine-tuning ECG-specific language models. A GPT-4o-based pipeline generated questions from 23 ECG references and Computing in Cardiology proceedings from 2003-2025. The final dataset contains 10,904 unique True/False questions paired with their negated forms (21,808 QA pairs). We evaluated three commercial and 20 open-source language models on a held-out test set in a zero-shot setting. Five open-source models with 7-14B parameters were fine-tuned using Low-Rank Adaptation, with BERT and BiomedBERT included as supervised encoder baselines. Generalization was assessed on ECG-related subsets of MedMCQA and MedQA converted to binary True/False questions using official answer keys. Zero-shot accuracy on ECGQuest ranged from 49.5% to 74.4%, with GPT-5 performing best. General-purpose models outperformed medically specialized models, several models showed strong True/False bias, and encoder baselines performed near chance. Fine-tuning improved all open-source models by 6.5-14.1%. Fine-tuned DeepSeek-R1-Distill-Qwen-14B reached 76.3% accuracy, while a five-model voting ensemble reached 78.5%. On MedMCQA and MedQA, fine-tuning mainly benefited weaker or class-biased models and did not consistently improve strong base models. ECGQuest provides a reproducible benchmark for contextual ECG knowledge and shows that parameter-efficient fine-tuning can make smaller language models competitive with substantially larger commercial models.
[IR-4] Playability-Aware Audio-to-Tablature Guitar Transcription via Diffusion Models
链接: https://arxiv.org/abs/2608.30854
作者: Riccardo Simionato,Louis Bigo
类目: ound (cs.SD); Information Retrieval (cs.IR)
备注: Accepted for ISMIR 2026
Abstract:Guitar tablature transcription requires not only accurate pitch detection but also assigning each note to a specific string-fret position, as the same pitch can be played at multiple fretboard positions. Existing approaches treat this as a standard classification problem, ignoring the musical and physical constraints that govern playable fingering sequences. We propose Noise2Fret, a diffusion model for audio-to-tablature transcription that generates tablature through a continuous latent representation of discrete fret and string targets, conditioned on spectral and audio features. To bridge the gap between pitch accuracy and physical playability, we introduce five auxiliary losses encoding Pitch-Class Distance, Positional Distance, Circle-of-Fifths Distance, String Similarity, and Hand-Span Feasibility directly into the training objective. Experiments on GuitarSet and GOAT datasets demonstrate that the model outperforms baselines while remaining computationally more efficient, and that the auxiliary losses yield consistent gains over the standard training objective.
[IR-5] Learning from What You Retrieve: Online RL Fine-Tuning for Semantic Retrieval
链接: https://arxiv.org/abs/2608.30753
作者: Shaowei Wei,Chong Huang,Songtao Fang,Jin Zhang,Zhuojun Wang,Chengfu Huo
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注:
Abstract:In large-scale e-commerce retrieval, dual-encoder retrievers are op- timized for contrastive similarity, whereas downstream rerankers capture finer-grained relevance preferences; this objective mis- match limits end-to-end retrieval quality. Reinforcement Learning offers a way to use reward-model feedback for retriever adaptation, but we observe that standard policy-gradient updates can degrade embedding geometry, especially when the document index must remain frozen due to industrial constraints. To address this, we propose PAO (Positive-Advantage-Only), a selective RL optimization method. Our analysis reveals that in- discriminate penalization of negative samples (pushing away) in a frozen high-dimensional space disrupts pre-trained semantic man- ifolds. PAO selectively applies gradient updates only to retrieved items with positive advantages, effectively pulling query embed- dings toward high-reward regions while preserving global topo- logical stability. Experiments on both a massive industrial dataset and public benchmarks demonstrate that PAO significantly outper- forms standard RL and distillation baselines.
[IR-6] Generative Retrieval for E-commerce: Jointly Learning Embedding and Codebook with Same Product Cluster
链接: https://arxiv.org/abs/2608.30606
作者: Songtao Fang,Zihao Xu,Shaowei Wei,Jin Zhang,Zhuojun Wang
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注:
Abstract:With the development of large language models (LLMs), generative retrieval is becoming increasingly important in e-commerce scenarios. Current mainstream approaches typically use a two-stage training strategy: first train a product embedding model, and then learn a codebook that maps embeddings to product IDs. This cascaded approach suffers from two major issues: (1) error accumulation-if the embedding model in the first stage produces biased representations, the codebook in the second stage cannot correct these errors, degrading final retrieval performance; and (2) codebook learning relies solely on product embeddings and lacks modeling of query-to-product and product-to-product interactions. As a result, products belonging to the same cluster may be assigned inconsistent IDs by the codebook, further hurting retrieval accuracy. To address these problems, we propose a novel method that jointly trains the embedding model and the codebook, and incorporates same product cluster information as an additional supervision signal. Experimental results demonstrate that our method significantly improves e-commerce retrieval performance while simultaneously enhancing both embedding and codebook learning.
[IR-7] Preference Shapes Relevance: Cross-component Hierarchical Semantic Alignment for Personalized Generative Retrieval EMNLP2026
链接: https://arxiv.org/abs/2608.30553
作者: Gaoming Zhang,Angqing Jiang,Jianchun Song,Kena Qi,Dayao Chen,Wei Lin,Defu Lian
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注: Findings of EMNLP 2026. 22 pages, 10 figures, 7 tables
Abstract:Generative Retrieval (GR) has emerged as a promising paradigm by mapping queries directly to Semantic IDs (SIDs) with powerful representation capabilities for candidate items. However, existing SIDs derived solely from item content create a semantic gap, failing to align dynamic query intents with static item representations. Furthermore, current generative paradigms rarely model user behavior sequences and are always bottlenecked by the high inference latency of beam-search autoregressive decoding. To address these challenges, we propose \textbfC ross-component \textbfH ierarchical semantic \textbfA lignment for \textbfP ersonalized generative retrieval ( \textbfCHAP ), a novel personalized GR framework from a hierarchical perspective. First, we design a Hierarchical Semantic Alignment module to align query’s latent space with item’s quantization path and synchronize multi-granular semantics. Second, we construct a personalized GR framework that models user behavior by synergizing discrete SIDs for structural guidance and continuous representations for fine-grained semantic refinement. Notably, we introduce a Residual Cascading Generation mechanism to restrict the costly multi-step Transformer Decoder to a single-pass inference, boosting inference throughput while mitigating information loss. Extensive experiments on three public datasets, one proprietary industrial dataset, and online A/B tests demonstrate CHAP’s superiority, validating the effectiveness and practical value of our approach. The code is publicly available at this https URL.
[IR-8] Local-to-Global Sentence-Level Graph Reranking for Scientific Synthesis
链接: https://arxiv.org/abs/2608.30525
作者: Zheng Dou,Zhao Zhang,Hao Geng,Ningjing Wang,Deqing Wang
类目: Information Retrieval (cs.IR)
备注:
Abstract:Retrieval-augmented scientific synthesis aims to answer complex research questions by integrating information from multiple papers into comprehensive and well-grounded responses. Since the generator can only synthesize the information selected and organized by the reranker, the quality of the generated synthesis depends critically on the reranked results. However, most rerankers operate at the passage level, which leaves key methodological, empirical, and comparative information buried in long and flat contexts, weakening the grounding of generated claims. Moreover, existing rerankers mainly rely on independent query-candidate scoring which overlooks complementary, contextual, and contrasting relations across scientific candidates, limiting information coverage and the comprehensiveness of the resulting synthesis. To address these limitations, we propose LoG-Reranker, a local-to-global sentence-level graph reranking framework for scientific synthesis. LoG-Reranker performs role-aware local scoring to identify fine-grained, query-relevant sentences and then models their relations on a sentence graph across the candidate set to globally refine sentence rankings. Top-ranked sentences and their connected neighbors are organized into a structured input context for generator to produce more grounded and comprehensive this http URL experiments on scientific synthesis and reranking benchmarks show that LoG-Reranker consistently outperforms competitive rerankers, yielding more reliable rankings and improving the quality of generated synthesis.
[IR-9] HF-SID: High-Fidelity Semantic IDs for Generative Retrieval in Location-Based Services
链接: https://arxiv.org/abs/2608.30479
作者: Haowen Lin,Jing Li,Zhibin Hao,Fangye Wang,Lihui Su,Song Yang,Xiaojiang Zhou,Pengjie Wang
类目: Information Retrieval (cs.IR)
备注:
Abstract:Generative retrieval has attracted increasing attention in Location-Based Services (LBS), where each Point-of-Interest (POI) is represented as a Semantic ID (SID). As the SID is the only channel through which POI information reaches the generative model, whatever it fails to preserve is irrecoverable at decoding time, and LBS retrieval is especially sensitive to the fine-grained differences that existing SIDs blur. Specifically, (1) LLMs embed continuous coordinates discontinuously, so their numeric differences do not reflect true geographic distance; (2) dynamic numerical attributes differ vastly in scale, so an identical gap may be decisive for one attribute yet negligible for another; and (3) short text cannot convey hierarchical affiliation, as text-similar POIs may belong to different hierarchies. We therefore propose HF-SID, which restores geographic, numerical, and structural fidelity at the representation stage, before any information is committed to a discrete code. It transforms coordinates into a continuous 3D Cartesian form and encodes each numerical value as a single unit, consolidated inside the LLM by Geo-CPT and Num-CPT with type-aware embeddings; a Structure-based Contrastive Learning objective, applied only to the last-layer residual, then separates co-located POIs that share a coarse tag but differ at the fine level. Because these mechanisms enrich the representation rather than lengthen the identifier, HF-SID uses a 3-token SID at no extra decoding cost. On a large-scale industrial
[IR-10] Hi-Q: Hierarchical Evidence-guided Query Refinement for Multi-Hop Question Answering
链接: https://arxiv.org/abs/2608.30468
作者: Jueun Kim,Sungho Park,Wook-Shin Han
类目: Computation and Language (cs.CL); Information Retrieval (cs.IR)
备注: 28 pages, 9 figures. Project page: this https URL
Abstract:A central bottleneck in multi-hop Question Answering (QA) is that the granularity at which a question is expressed often differs from the granularity at which corpus evidence is retrievable. Existing methods address this mismatch by imposing fixed graph structures over the corpus, by iteratively reformulating the query, or by executing a generated program over it, but these strategies do not explicitly decide when a query unit is already supported by evidence and when it should be refined. We formulate this bottleneck as retrievable granularity discovery and introduce Hi-Q, an evidence-conditioned framework for hierarchical query refinement. At each query node, a resolution operator tests whether retrieved evidence supports the current query unit; resolved nodes terminate, while unresolved nodes are expanded by a dependency-preserving binary operator and checked by a semantic coverage verifier. Hi-Q therefore grows a query tree whose topology is determined by corpus support signals rather than by a fixed decomposition template or a pre-built graph. We evaluate Hi-Q on three multi-hop QA benchmarks, primarily under full-corpus retrieval, where dependent evidence must be located among open-domain distractors rather than within a small annotated pool. In this setting Hi-Q reaches 52.3 EM and 64.0 F1 averaged over the three benchmarks, ahead of the iterative retrieval baseline IRCoT by 15.1 EM / 18.2 F1 on that same average, and ahead of the graph-based RAG baseline PropRAG by 11.5 EM / 12.0 F1 on MuSiQue-full, without corpus-wide graph construction. In the restricted supporting/distractor setting used by prior work, Hi-Q likewise attains the best accuracy, with 57.9 EM and 69.3 F1 on average, ahead of PropRAG by 5.6 EM / 3.9 F1 and IRCoT by 13.7 EM / 15.8 F1. The project page is available at this https URL.
[IR-11] CHASE: How Content Ecosystems Are Reshaped When Ranking Is the Only Target
链接: https://arxiv.org/abs/2608.30466
作者: Qianwen Gao,Zichang Su,Yiwen Hou,Arlen Kumar,Leanid Palkhouski
类目: Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
备注: Accepted to the Conference on Language Modeling (COLM) 2026
Abstract:Generative Engine Optimization (GEO) is increasingly used to improve content visibility in LLM-based retrieval systems, yet its population-level effects under repeated optimization remain poorly understood. We introduce Content Homogenization under rAnking Signal Exploitation (CHASE), a controlled simulation framework for studying how content ecosystems are reshaped when creators repeatedly adapt documents to an LLM ranking signal. We use ranking as a proxy for source visibility and validate this abstraction against citations in grounded generated responses, obtaining a rank-citation AUC of 0.853 \pm 0.093 across six domains. CHASE then iterates ranking, feature discrimination, rewriting, and evaluation over 20 rounds across different domains. Quality-ranking alignment decreases in all six domains: from R0 to R20, the change in Spearman’s rho ranges from -0.107 to -0.018, with a mean change of -0.068, which means documents closer to the ranking feature profile become less aligned with independently judged document quality over the simulation horizon. A random-target control has shown that it is associated with adaptation toward ranking-derived incentives rather than iterative rewriting alone. The resulting ecosystem dynamics are strongly domain-dependent. Together, these findings show how repeated optimization against a fixed LLM ranking signal can reshape both content populations and the incentives faced by content creators.
[IR-12] PRIME: Mitigating Subgroup Optimization Competition in Shared CTR Top Networks with Plug-in Residual Input-Conditioned Mixture of Expert
链接: https://arxiv.org/abs/2608.30449
作者: Heng Yao,Siyun Hou,Tianying Liu,Yulou Shu,Yong He,Chuan Yuan,Kaibin Qiu,Guowei Chen,Jiayu Zhao,Chao Yu,Ke Ding
类目: Machine Learning (cs.LG); Information Retrieval (cs.IR)
备注: 14 pages, 4 figures
Abstract:Click-through rate (CTR) models vary in feature-interaction design, yet their top networks usually remain a single multilayer perceptron shared by all examples. Heterogeneous user, item, and context subgroups therefore update the same parameters; weakly aligned learning signals make the aggregate gradient a compromise among competing directions. We study the competition on Avazu with 4 models and 4 semantic fields. Across all architectures, semantic subgroups show lower Top-NN gradient cosine similarity than random groups matched by sample size and label ratio, with reductions of 0.23-0.37. This competition motivates input-conditioned experts, but directly replacing an established Dense mapping changes its initial function, sharing pattern, and capacity, obscuring the source of gains. We introduce PRIME (Plug-in Residual Input-conditioned Mixture of Experts), a Dense-anchored mixture of low-rank residual experts. PRIME anchors the original prediction and uses zero-residual initialization to match the Dense baseline exactly at training onset. Input-dependent routing weights low-rank experts for example-specific logit corrections; multi-bag aggregation and EMA load biases stabilize conditional estimation. We evaluate PRIME on held-out Avazu and Criteo test sets across 13 CTR architectures and five paired seeds. Median paired AUC gains are +0.0022 and +0.0066, with LogLoss reductions of 0.0011 and 0.0081, respectively. On FiBiNET and DCNv2, PRIME outperforms APG in all ten seed-level AUC comparisons while using fewer parameters and lower inference latency on both backbones. These results show that function-preserving conditional residuals add input-dependent capacity while preserving the Dense path and its optimization stability. Code is available at this https URL. Comments: 14 pages, 4 figures Subjects: Machine Learning (cs.LG); Information Retrieval (cs.IR) Cite as: arXiv:2608.30449 [cs.LG] (or arXiv:2608.30449v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2608.30449 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[IR-13] Beyond Polarization: The Generative Constraint of Chain-of-Thought in Pointwise Reranking EMNLP2026
链接: https://arxiv.org/abs/2608.30398
作者: Xiaoyang Chen,Jie Liu,Haijin Liang,Haibo Shi,Jin Ma,Ben He,Yingfei Sun,Dezhi Ye
类目: Computation and Language (cs.CL); Information Retrieval (cs.IR)
备注: Accepted at EMNLP 2026 Findings
Abstract:In pointwise document reranking, Chain-of-Thought models typically underperform direct scoring models. While existing diagnostics attribute this to inferior classification, score polarization, or calibration breakdown, whether targeted training can bridge this gap remains unclear. Our empirical study first confirms that this gap is stable across scales up to 32B parameters, ruling out model and data capacity confounders. We then apply stress tests utilizing reinforcement learning, fine-grained supervision, and architectural decoupling to explicitly repair these deviations. Although these interventions improve classification accuracy and absolute scores, the relative ranking gap persists. These findings suggest that, within the pointwise scoring paradigm, routing continuous relevance semantics through discrete text constrains ranking signal resolution, revealing a bottleneck that is stable and difficult to overcome under current standard methods, rather than an easily resolvable training bias.
[IR-14] RSLM: Training-Free Vector Quantization for Approximate Nearest Neighbor Search
链接: https://arxiv.org/abs/2608.30384
作者: Rastislav Lenhardt,Teodora Dobos,Thomas Vecchiato,Jiri Isa,Igor Ginzburg
类目: Machine Learning (cs.LG); Information Retrieval (cs.IR)
备注: 14 Pages, 3 Figures, Preprint
Abstract:By introducing RSLM (Rotated Scaled Lloyd-Max), a family of training-free vector quantization codecs compressing embeddings to 1–4 bits per dimension, we reduce memory cost and memory bandwidth of a typical large-scale Approximate Nearest Neighbor (ANN) search system, while reducing its complexity and keeping or improving recall across multiple benchmark datasets. State-of-the-art systems filter candidates using coarse partitions, approximately score them to narrow the set, and then rescore the best with higher precision representations (often =8 bits per dimension). Our relativized codecs can bring this down to 2–4 bits per dimension. We use the properties of the ANN system to encode residual vectors instead of full vectors, both for the approximate scoring phase and the rescoring phase. Since Maximum Inner Product Search (MIPS) is very sensitive to vector norms, we correct the L_2 norms of quantized vectors. Our major innovation is that we correct the L_2 norm of the final reconstructed vector rather than just the residual. Our rescaling replaces more complicated schemes, such as Anisotropic loss. The residualization scheme gives us a more favorable quality vs size trade-off than generic quantization methods. Our high-performance implementation leverages a block-wise cascaded Fast Walsh-Hadamard Transform (FWHT) with linear-like complexity, AVX SIMD-optimized codebooks, and a steganographic encoding of scaling factors for perfect cache-line alignment. Comments: 14 Pages, 3 Figures, Preprint Subjects: Machine Learning (cs.LG); Information Retrieval (cs.IR) Cite as: arXiv:2608.30384 [cs.LG] (or arXiv:2608.30384v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2608.30384 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[IR-15] Beyond Ranking Accuracy: Evaluating LLM -Cited Feature Rationales for Next Basket Repurchase Recommendation RECSYS2026
链接: https://arxiv.org/abs/2608.30333
作者: Yanan Cao,Anay Dombe,Murali Mohana Krishna Dandu,Shreeranjani Srirangamsridharan,Sinduja Subramaniam,Yogananth Mahalingam,Evren Korpeoglu,Kannan Achan
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注: Accepted at RecSys 2026 Workshop: Agentic and Generative AI for E-Commerce
Abstract:Next-basket repurchase recommendation is commonly formulated as a ranking task: given a customer’s purchase history, the system ranks previously purchased items that may be needed again. In production settings, however, ranking accuracy is only one component of recommendation quality. Customers may also benefit from concise evidence about why an item is recommended now. Large language models (LLMs) offer a potential way to surface such evidence through feature-based, human-readable rationales grounded in interpretable behavioral signals. We construct repurchase features spanning cadence, frequency, recency, user behavior, and item popularity, and evaluate LLMs on two public grocery datasets and one proprietary retail dataset. We investigate (1) whether off-the-shelf LLMs can use these features as next-basket scorers relative to heuristic and supervised rankers, and (2) whether LLM-cited features carry outcome-grounded ranking signal. For the latter, we compare LLM-cited features with model-specific attribution methods under a cross-model feature-masking protocol that measures ranking degradation after masking selected features. Our results show that LLM scores are not competitive with supervised rankers, suggesting that off-the-shelf LLMs should not be used as standalone repurchase recommenders. However, changes in prompt and evidence representation can improve outcome-grounded feature-masking results in some settings even when ranking performance does not improve; the effect is dataset-dependent and does not consistently match attribution baselines. These findings suggest a practical role for LLMs as validated explanation components rather than primary rankers, with rationale quality evaluated separately from ranking accuracy.
[IR-16] PEARL: Front-Loading Relational Chains for Multi-Hop Table Retrieval EMNLP2026
链接: https://arxiv.org/abs/2608.30291
作者: Subeen Ho,Hyeongu Kang,SeongKu Kang,Susik Yoon
类目: Information Retrieval (cs.IR)
备注: Accept to EMNLP 2026
Abstract:While large language models (LLMs) have shown strong capabilities in tabular reasoning, retrieving relevant tables remains challenging due to the fragmented and relational structure of real-world data. Existing work typically relies on whole table representations that overlook cross-table semantics induced by join relationships. We propose PEARL, a training-free framework that shifts the paradigm toward vertical partitioning-based sub-table encoding. PEARL augments the retrieval corpus offline by generating multi-hop queries over pre-identified join paths and reorganizing relevant columns into vertically partitioned corpus units, enabling effective multi-table retrieval without query-time LLM inference. Experiments show that PEARL consistently outperforms existing methods, with up to +30.05% gains in R@2 on 3-hop queries. The source code is available at this https URL.
[IR-17] CAMIE: Co-Engagement-Aware Multimodal Item Embeddings for Snap Dynamic Product Ads Retrieval
链接: https://arxiv.org/abs/2608.30255
作者: Xiaodong Liu,Siman Wang,Congfei Zhang,Hsiang-wei Chao,Xiao Bai,Wen Zhang,Jingxiao Ma,Zhe Liu,Yunzhi Zhou,Yajun Wang,Jinchao Li,Yu Zhang
类目: Information Retrieval (cs.IR)
备注:
Abstract:Item-to-item (I2I) retrieval is a core primitive in large-scale recommendation and advertising systems. In production Snap Dynamic Product Ads (DPA), I2I retrieval faces two challenges: separate visual, textual, and multimodal encoders fragment the retrieval stack, and content-only training does not align embeddings with the co-engagement behavior that drives downstream conversions. We present CAMIE, a co-engagement-aware multimodal item embedding framework for Snap DPA retrieval. CAMIE builds on LLM/MLLM backbones, using their native multimodal interfaces to represent item images and metadata in a shared embedding space. It then fine-tunes the backbone on co-engaged item pairs mined from user journeys with a symmetric in-batch InfoNCE objective. Offline, CAMIE outperforms the strongest commercial multimodal embedding model on Recall@10 and serves text-only retrieval from the same checkpoint with minimal quality loss. Online, CAMIE serves as a drop-in replacement for two deployed content-based I2I encoders, delivering +0.390% CTR / +10.832% CVR over the multimodal control, +18.958% CTR / +13.12% CVR over the text control, and +0.211% CTR / +1.911% CVR on overall DPA traffic. CAMIE is deployed in production.
[IR-18] SetMIR: Multi-Interest Retrieval as Set Prediction
链接: https://arxiv.org/abs/2608.30251
作者: Xiaodong Liu,Congfei Zhang,Hsiang-wei Chao,Siman Wang,Xiao Bai,Tong Zhao,Jingxiao Ma,Wen Zhang,Zhe Liu,Shantanu Aggarwal,Di Huang,William Leach,Yunzhi Zhou,Yajun Wang,Jinchao Li,Yu Zhang
类目: Information Retrieval (cs.IR)
备注:
Abstract:Embedding-based retrieval is at the core of industrial recommender systems, but a single user embedding is often too limited to capture a user’s diverse interests. Multi-interest retrieval addresses this by using multiple user embeddings, yet existing methods still suffer from two issues: interest collapse, where different embeddings learn the same interest, and static dispatch, where serving uses a fixed retrieval budget even when some embeddings are unnecessary. We propose SetMIR, which treats multi-interest retrieval as a set prediction problem. SetMIR encodes a user’s behavior history with a transformer and uses K learnable queries to decode a set of user interests, each producing a retrieval embedding and a presence score. During training, Hungarian matching assigns targets to queries one-to-one, so matched queries learn distinct interests and the presence head learns which queries are active. At serving time, SetMIR uses presence scores and query-level Non-Maximum Suppression (NMS) to issue only active, non-redundant ANN queries. On Snap’s Dynamic Product Ads (DPA) data, SetMIR outperforms four learned multi-interest retrievers on every metric while issuing 33% fewer ANN queries per request. Deployed as a new retrieval source in the DPA production stack, SetMIR lifts overall CVR by 3.1%, while lifting CTR by 44% and CVR by 51% over the item-to-item retrieval source with the same item embeddings, ANN index, and retrieval quota.
[IR-19] Doc-REFRAG : Rethinking Multimodal Document Retrieval-Augmented Generation EMNLP2026
链接: https://arxiv.org/abs/2608.30163
作者: Ruofan Hu,Shengyang Xu,Minjie Hong,Xiaoda Yang,Sashuai Zhou,Ke Lei,Tao Jin,Zhou Zhao
类目: Information Retrieval (cs.IR); Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by EMNLP 2026 Main
Abstract:Real-world knowledge resides in multimodal documents, necessitating retrieval-augmented generation (RAG) for accurate question answering. However, existing multimodal RAG models are primarily designed for single-image or closed-document settings and exhibit limited accuracy in realistic multi-image scenarios. Moreover, processing numerous retrieved images incurs substantial computational overhead from irrelevant visual tokens. To address these challenges, we introduce DocLongRAG, a large-scale dataset of 343K question–answer pairs, each associated with an average of 37.4 retrieved images to reflect authentic RAG workflows. Building on this dataset, we propose Doc-REFRAG, a question-guided framework that compresses visual tokens into coarse chunks and selectively expands question-relevant ones via a lightweight RL-based selector. Experiments on six benchmarks show that Doc-REFRAG outperforms eleven strong baselines, achieving state-of-the-art accuracy with significantly lower inference latency. Our resources are available at this https URL.
[IR-20] Understanding before verifying: Claim normalization for automated citation verification
链接: https://arxiv.org/abs/2608.30145
作者: Yifan He,Mengjia Wu,Siming Deng,Yi Zhang
类目: Information Retrieval (cs.IR)
备注:
Abstract:Citation accuracy has been studied for decades because of its importance to research reliability. Content-level citation verification assesses the reliability of scholarly claims. Recent work adopts a two-stage retrieval-classification framework inherited from fact-checking. However, this design overlooks the complexity of the raw citing claim and introduces three issues into the verification system, namely scope mismatch, perspective mismatch, and proposition entanglement. These issues increase the difficulty of retrieval and classification, thereby limiting model performance. Motivated by this gap, we propose claim normalization, which applies three rewriting strategies to the raw citing claim before retrieval and classification, allowing each downstream model to perform a single, well-defined task. Building on this method, we develop Claim-Normalized Citation Verification (CNCV), a new three-stage framework consisting of claim normalization, evidence retrieval with grounding, and citation classification. We evaluate CNCV across 18 classifiers using a factorial experiment on human-annotated citation instances. Compared with the prior two-stage framework, CNCV improves macro F1 by an average of 12% for encoders and 10% for generative LLMs, driven by improved evidence quality, the dominant factor identified in our experiments. Evidence retrieved from automatically normalized claims yields downstream classification performance statistically equivalent to that obtained with manually annotated evidence.
[IR-21] E-SENS: Exclusion-Sensitive Penalization for Negative-Constraint Retrieval
链接: https://arxiv.org/abs/2608.30130
作者: Yerang Kim,Jiyoon Myung,Joohyung Han
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注:
Abstract:Retrieval-augmented language models can fail to respect negative constraints when the retriever supplies evidence about concepts the user explicitly excluded. Beyond explicit negation, queries may ask for answers that include one concept while excluding another, or for entities that belong to a category but differ from a closely related instance. Because the excluded concept still appears in the query text, dense retrievers may assign high similarity to documents about that concept even when the user asks to avoid it. We introduce E-SENS, a training-free reranking method for negation-sensitive retrieval. E-SENS extracts a compact trap query for the excluded side and subtracts trap-query similarity from the original-query retrieval score. On ExcluIR, E-SENS shows a clear recall-violation trade-off across four embedding models and reduces trap retrieval at recall-preserving settings.
[IR-22] he Language of the Question Selects the Market: Query Language and Exit IP as Separable Factors in Commercial Recommendations from a Generative Search Interface
链接: https://arxiv.org/abs/2608.30052
作者: Dmitrij Żatuchin
类目: Information Retrieval (cs.IR); Computation and Language (cs.CL); Computers and Society (cs.CY)
备注: 13 figures-free pages, 6 tables. Data, collectors and 246 run records: Zenodo DOI https://doi.org/10.5281/zenodo.22181306
Abstract:When a generative search interface answers a commercial question, which market’s products it names is decided before the model reasons about the products. We report a controlled probe of 234 runs against the logged-out ChatGPT web interface and the OpenAI API, collected on 29 and 30 August 2026 across four exit countries and six query languages, with six identical runs per cell. Three results. First, the top recommendation is unstable: it changed across six identical runs on four of six prompts, and that rate was identical in the browser interface and in the API with web search both enabled and disabled, so instability is a property of the system and not of the surface. Second, query language, and not location, decides whether local suppliers appear at all. Where the query language matched the country, a global brand won 1 of 24 runs; asked in English on the same connections, local brands took 0 of 6 runs in Estonia and Turkiye. Third, language and location are separable and act on different things: holding the query language fixed and moving only the exit IP moves the market whose brands are named while the answer stays in the query language. We show this on two unrelated pairs, Turkish asked from Berlin and Russian asked from Tallinn, and in both the answer names the resident country’s suppliers. A minority language occupies a middle tier: Russian asked from Estonia names an Estonian supplier in 4 of 6 runs and a global one in all six, where Estonian names a local supplier in every run and English names none. A negative control in a second category, coded with the same instrument, shows no language effect at all, and disconfirms our own expectation: that category does have domestic suppliers and none was named in any language, which points the explanation at whether a category is nationally regulated rather than at whether it is nationally supplied.
[IR-23] Demand-Side Measurement for Generative Engine Optimization: Constructing and Validating a Million-Persona Intent-Annotated Buyer Corpus
链接: https://arxiv.org/abs/2608.30023
作者: Dmitrij Żatuchin,Daniil Dzemesjuk
类目: Information Retrieval (cs.IR); Computation and Language (cs.CL)
备注: 17 pages
Abstract:Generative engines such as ChatGPT, Gemini, and Perplexity answer buyer questions directly and name a shortlist of brands inside the answer. Studying how brands enter or fail to enter that shortlist requires demand-side data: what buyers in a category ask, what information they need, and which sources they trust. Existing large persona corpora are built for training-data diversity and carry neither a staged search-intent label nor a preferred-sources field, so they cannot be joined to supply-side recommendation measurements. We built and validated PersonaGen-1M, a corpus of 1,031,732 synthetic buyer personas spanning 511 industry labels and 4 market contexts, carrying 19,416,821 structured behavioral attributes, 5,160,046 of them search queries. Each persona carries a single primary_intent label covering its query set (78.3% informational, 17.4% commercial, 4.3% transactional) and a preferred_sources field naming the source types that buyer would trust. The corpus was built from roughly 40 million raw persona descriptions drawn from four public datasets through GPU-accelerated MinHash LSH plus semantic deduplication, then enriched to a fixed schema. The intent field selects the commercial-evaluation personas whose queries drive recommendation, and the preferred_sources field pairs against citation-provenance data; that join is the primary intended use, and its controlled empirical estimate is future work. Among million-scale persona corpora surveyed in August 2026, one other carries a source-preference attribute, as a six-value media-channel enum; PersonaGen-1M pairs named per-persona source lists with a staged commercial search-intent label and an attached query set. The full corpus is shared on request for non-commercial research; a stratified subset is published openly so the protocol, the schema and the validation can be inspected and reused without asking us.
[IR-24] Spatial Matryoshka Training for Multi-Granularity Visual Document Retrieval
链接: https://arxiv.org/abs/2608.29951
作者: Trishan Singha Roy,Arkadeep Acharya,Vishwajeet Kumar,Jaydeep Sen,Sachindra Joshi
类目: Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
备注:
Abstract:Multi-modal late-interaction retrievers achieve strong retrieval on visually rich documents by representing each page as per patch embeddings and matching at the token level. However, this approach incurs high storage costs. Existing compression methods typically fix a single compression level at indexing time, limiting flexibility. We present ColSNAP (Spatial Nested Average Pooling)1, a training method that generates a nested hierarchy of compression levels directly from a backbone’s patch grid. By spatially pooling patch embeddings into pro- gressively coarser tiers and training all tiers simultaneously, a single model learns to support retrieval at multiple compression levels without architectural changes. Crucially, a single encoding pass yields every tier, enabling the accuracy-storage trade-off to be configured at indexing time to match avail- able storage budgets, rather than being fixed during training. We demonstrate that models trained using ColSNAP maintain near full-resolution retrieval performance under substantial compression and that ColSNAP transfers effectively across multiple late-interaction backbones, and achieves most of its improvements via a lightweight adaptation stage applied to a pre-trained retriever.
[IR-25] REIGN: Refurbished Embeddings with Integrated Guidance Networks for Efficient Context-Length Scaling EMNLP2026
链接: https://arxiv.org/abs/2608.29899
作者: Devrim Çavuşoğlu,Emre Akbaş
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
备注: Accepted to Findings of EMNLP 2026. URL: this https URL
Abstract:Dense retrieval over long documents is expensive. Token-level encoders scale quadratically in sequence length, and most long-context embedding models reach 32K tokens only through architectural workarounds or by stretching billion-parameter LLMs. We propose REIGN (Refurbished Embeddings with Integrated Guidance Networks), a contrastively trained bi-encoder that operates on sequences of contextualised chunk embeddings from a frozen Guidance Network (GN) rather than on raw tokens. REIGN targets multi-chunk inputs, primarily for document-to-document retrieval; single-chunk inputs stay with the GN. Decoupling token-level processing from document-level reasoning, and caching the GN embeddings to disk, cuts per-document training cost by roughly four orders of magnitude relative to chunked Transformer fine-tuning. We also release a synthetic long-document retrieval benchmark for contrastive training and evaluation at long context lengths. Across an in-distribution Wikipedia benchmark, the LoCo out-of-distribution suite, and a real-world patent retrieval case study, REIGN matches dense long-context retrievers at smaller parameter budgets in each regime. A paired significance test puts it on par with models 1.6-4.3x larger on the patent task, and it stays within 0.65 nDCG@10 of a 20x-larger model on LoCo.
[IR-26] You Know What I Mean: A Benchmark for Agent ic Conversational Reference Grounding EMNLP2026
链接: https://arxiv.org/abs/2608.29834
作者: Karen Fuchs,Uri Katz,Yoav Goldberg
类目: Computation and Language (cs.CL); Information Retrieval (cs.IR)
备注: 23 pages, 6 figures, Accepted to EMNLP 2026 Main Conference
Abstract:Collaborative conversations frequently contain references whose targets are indirect rather than named: resolving “this looks like the fix discussed yesterday” requires combining conversational context with evidence from the surrounding workspace which is accessible through APIs or user interfaces. We formalize this problem as Conversational Reference Grounding (CoRG): using a given set of tools to resolve a reference in conversation to the unique external item intended by the speaker. CoRG is challenging because it combines lexical, semantic, and temporal cues distributed across the conversation and the external workspace. Agents must translate these heterogeneous signals into effective tool use: formulating strategies, discovering plausible candidates, inspecting their metadata and content, and ruling out close alternatives. We study CoRG through RepoRef, a benchmark of 400 developer-chat segments grounded in GitHub issues, pull requests, and commits across 92 repositories. Unlike single-shot retrieval tasks, RepoRef often requires multi-step tool use. Our results show that CoRG remains challenging for current agents, even the best agent reaches only 67.0% success rate, leaving one third of references unresolved. These findings position CoRG as a concrete benchmark for studying how agents search, inspect, and verify information in realistic multi-tool environments.
[IR-27] ICEGR: An Intent-Coherent End-to-End Generative Retrieval Framework for E-commerce Search
链接: https://arxiv.org/abs/2608.29652
作者: Jiayi Tuo(1 and 2),Hehan Li(2),Dongjun Fu(2),Xin Lu(2),Ling Zhuang(2),Fuwei Zhang(3),Meifang Li(2),Peizhi Xu(2),Hanmeng Liu(2),Shuanglong Li(2),Liwei Qian(2),Yanbiao Ma(4),Fuzhen Zhuang(3) ((1) University of Science and Technology of China, (2) Baidu, (3) Beihang University, (4) Renmin University of China)
类目: Information Retrieval (cs.IR)
备注:
Abstract:Generative Retrieval (GR) is promising for e-commerce search, yet existing methods struggle to maintain query-intent consistency throughout the training pipeline. First, semantic ID (SID) construction based on static product information limits the ability of SIDs to encode product-intent associations. Second, although supervised fine-tuning (SFT) learns product-SID mappings across the catalog, low-exposure products still lack real query-intent supervision because query-to-SID training relies solely on online logs, resulting in poor retrieval performance for these products. Third, business-oriented preference optimization may favor popular or high-value products over those that best match the query intent, weakening query-product relevance. To address these issues, we propose ICEGR, an Intent-Coherent End-to-End Generative Retrieval Framework for E-commerce Search that integrates query intent consistently throughout the GR training pipeline. ICEGR comprises three components: (1) Intent-Aware SID Construction incorporates query-intent signals into SID construction, enabling SIDs to capture search intent beyond static product information; (2) Synthetic Query-Enhanced Unified SFT unifies multiple SFT tasks under the query-to-SID objective and augments sparse supervision from online logs with synthetic queries, providing complementary query-intent supervision for low-exposure products; and (3) Relevance-Calibrated Preference Optimization integrates query-product relevance and business signals into a margin-adaptive preference objective, preserving query intent while enabling business preference learning. Offline results show that ICEGR improves Recall@20 by 21.7% and NDCG@20 by 26.6% over the baseline. Deployed as an end-to-end generative retrieval pathway in Baidu E-commerce Search, ICEGR achieves relative improvements of 3.52% in CTR, 15.96% in order volume, and 7.53% in GMV in an A/B test.
[IR-28] LLM s Interpret Embeddings Organize Graphs Emerge: Agent -Driven Compilation of Scientific Knowledge
链接: https://arxiv.org/abs/2608.29612
作者: Shi-Ju Ran,Kun Zhang,Xi Wu,Liu-Si Yang,Wen-Jun Li
类目: Artificial Intelligence (cs.AI); Digital Libraries (cs.DL); Information Retrieval (cs.IR)
备注: 15 (main text) + 6 (SM) pages, 4 + 1 figures
Abstract:Sustained scientific work requires a knowledge substrate that carries interpretation across tasks and preserves paths to source evidence. We call this process \emphscientific knowledge compilation and implement it in ASKS, the \emphAgent-Driven Scientific Knowledge System. For each source, an LLM produces a readable Wiki view and machine-facing semantics. Deterministic checks convert the latter into a document-local GraphDelta, and embedding geometry together with explicit graph rules integrates the proposed changes into persistent state. Each ingest is an inspectable state transition over accumulated knowledge, with compiled Wiki and graph views linked to the preserved source record. We examine this process by chronologically compiling 56 published papers from one research program. Branch survival, cross-paper support, lineage, coverage, and churn yield a source-traceable author research portrait centered on tensor-network methods, with branches into quantum many-body research, tensor-network machine learning, and quantum-AI-oriented directions. In this run, higher-level Hub organization remains stable and low-churn. Canonical-node growth is predominantly additive. Graph-level measurements and navigation paths retain links to the source records from which they were compiled.
[IR-29] SnapBench: Benchmarking Snap-and-Ask Multimodal Retrieval for Mobile Interactions EMNLP2026
链接: https://arxiv.org/abs/2608.29607
作者: Zirong Chen,Fuda Ye,Kuan Zhang,Enjun Du,Junfu Pu,Xinlei Wang,Xinyu Zuo,Lisheng Duan,Jin Ma,Yongqi Zhang
类目: Computer Vision and Pattern Recognition (cs.CV); Information Retrieval (cs.IR)
备注: 37 pages. Yuanbao Technical Report. Accepted to Findings of EMNLP 2026
Abstract:Mobile AI acts as a visual oracle, empowering users to snap a picture of something and ask for information. Snap-and-ask retrieval is now one of the most common entry points for mobile AI, yet photos are often blurry, while text questions may be short or mistyped. Existing benchmarks only test on clean inputs or do not isolate paired robustness in snap-and-ask retrieval. Therefore, we introduce SnapBench, the first paired benchmark for robust snap-and-ask multimodal retrieval, spanning 1,145 queries, 9,085 gallery items under 53 controlled corruption conditions with human annotations. We evaluate 16 multimodal retrievers, covering dual-tower encoders and embedding-based VLMs. Results show that image corruptions substantially degrade retrieval, while text corruptions mainly affect text-only retrieval and have limited impact on joint retrieval. Clean image-only retrieval often outperforms joint retrieval, indicating the coarse-text drag and the lack of cross-modal fallback under noisy inputs. SnapBench provides a controlled testbed for evaluating robust retrieval in snap-and-ask scenarios. We further propose MOOR (Modality-anchored, Outlier-aware, Optimal Reweighting), a simple adaptive fusion approach, highlighting the need for reliability-aware modality calibration in snap-and-ask retrieval.
[IR-30] RePair: Turning Retrieval Failures into Counterfactual Hard Pairs EMNLP2026
链接: https://arxiv.org/abs/2608.29604
作者: Siyi Liu,Xiaorong Zhu,Enjun Du,Xinyu Zuo,Lisheng Duan,Haijin Liang,Jin Ma,Junfu Pu,Yongqi Zhang
类目: Information Retrieval (cs.IR); Computer Vision and Pattern Recognition (cs.CV)
备注: EMNLP 2026 (Main Conference)
Abstract:Vision-language retrieval with CLIP-style dual encoders achieves strong cross-modal performance, yet practical accuracy often hinges on localized semantic distinctions where top-ranked near misses differ from the true match by a single critical detail. Hard-sample mining can select confusable candidates but cannot construct corrected counterparts; synthetic augmentation can generate novel samples but, without conditioning on actual model failures, targets irrelevant dimensions of hardness. We observe that a top-ranked false positive is a counterfactual scaffold—sharing most of the query’s semantics while differing in a localized failure-causing residual. Minimally correcting this residual yields a hard positive of the ground truth in the same modality; the corrected and unedited versions form a hard negative pair that straddles the decision boundary, producing complementary pull–push supervision. We introduce RePair, guided by three principles—Validity, Minimality, and Locality—which mines false positives bidirectionally, applies LLM-guided counterfactual editing, and trains with a local hard-pair contrastive objective. On Flickr30K and COCO30K, RePair outperforms controlled augmentation baselines with only 107K synthetic samples—26%–75% fewer than comparable methods—confirming failure-conditioned repair is more data-efficient than error-agnostic augmentation.
[IR-31] Adaptive Doubly Robust Off-Policy Evaluation for Ranking Policies under Diverse User Behavior
链接: https://arxiv.org/abs/2608.29600
作者: Kosuke Iguchi,Ren Kishimoto
类目: Machine Learning (cs.LG); Information Retrieval (cs.IR)
备注: 14 pages, 4 figures, 7 tables. Accepted at CONSEQUENCES 2026 and selected for oral presentation
Abstract:Off-policy evaluation (OPE) of ranking policies is challenging be- cause selecting and ordering multiple items from a candidate set makes the number of possible rankings grow combinatorially with the number of candidates and the ranking length. Consequently, Inverse Propensity Scoring (IPS), whose importance weight is the full-ranking probability ratio under the evaluation and logging policies, can have excessive variance. Independent IPS (IIPS) and Reward Interaction IPS (RIPS) reduce variance by imposing fixed assumptions on how users browse rankings, but may introduce bias when those assumptions mismatch actual behavior. Adaptive Inverse Propensity Scoring (AIPS) addresses this trade-off by adap- tively marginalizing importance weights over the actions that affect each position-wise reward. It attains minimum variance within a class of unbiased IPS-based estimators when the true user be- havior model is observed. However, its estimation accuracy may still degrade for longer rankings, and AIPS does not use a reward model for residual correction. We propose Adaptive Doubly Robust (ADR), which combines adaptive importance weighting with re- ward regression through a control-variate correction. We establish its unbiasedness when the true user behavior model is observed and characterize a sufficient condition under which it reduces vari- ance relative to AIPS. Across synthetic experiments with 10,000 simulations per condition, ADR improves mean squared error over AIPS and conventional ranking OPE estimators across a range of logged-data sizes and ranking lengths.
[IR-32] he Edge Spectrum of Choice-Derived Item Graphs: Strong and Weak Edges Encode Different Relations in Collaborative Filtering CIKM2026
链接: https://arxiv.org/abs/2608.29578
作者: Keigo Sakurai,Takahiro Ogawa,Miki Haseyama
类目: Information Retrieval (cs.IR)
备注: Accepted to the 35th ACM International Conference on Information and Knowledge Management (CIKM2026)
Abstract:Graph collaborative filtering relies on item–item graphs whose edges are used for positive smoothing, under the implicit assumption that stronger edges encode more of the same relation as weaker ones. We show that this assumption fails for a practically important class of graphs: those whose edge weights come from a choice model. On such graphs, strong and weak edges encode qualitatively different relations, which we call an edge spectrum. Specifically, strong edges concentrate on the in-slate competitors of clicked items, exactly the pairs that the within-slate ranking gradient pushes apart, while weak edges do not. We formalize this as a sign mismatch between the smoothing operator and the ranking gradient, and prove that co-click graphs cannot exhibit the same misalignment by construction. This diagnosis explains three empirical observations on MIND and EB-NeRD: (i) drop-in choice-derived operators do not beat co-click, despite indexing structurally distinct neighborhoods; (ii) uniform scalar fixes (sign flip, in-slate margin loss) fail predictably, because the misalignment lives in the graph, not in the loss; (iii) only edge-magnitude-aware operators, with the regime boundary located by the diagnosis rather than by tuning, recover the predicted ordering. The neighbor cutoff k is therefore a semantic switch, not a sparsification hyperparameter. Our claim concerns which interventions fail or succeed and why, not absolute headline gains, which the diagnosis itself predicts to be small under the attenuated propagation channel we observe. We turn the diagnosis into a reusable protocol practitioners can run before deploying any choice-derived item-side operator. Code: this https URL.
[IR-33] What Are You Listening to? Temporal Music Grounding for Audio-to-Text Large Language Models
链接: https://arxiv.org/abs/2608.29480
作者: Kun Fang,Ziyu Wang,Ichiro Fujinaga
类目: ound (cs.SD); Information Retrieval (cs.IR)
备注:
Abstract:Large audio-language models can produce fluent and musically plausible responses, yet it often remains unclear whether those responses are grounded in the audio input. We introduce temporal music grounding, a task in which a model returns one or more time spans corresponding to a queried musical note, event, or pattern. To evaluate this capability, we present MusicGroundingBench, a controlled benchmark suite built by rendering algorithmically generated piano MIDI to audio, yielding exact symbolic-to-audio alignment. The suite comprises two subsets: MGBench-3N, which evaluates note-level grounding in clips containing up to three notes, and MGBench-2B, which evaluates structured grounding and short-form music understanding in two-bar excerpts. Experiments show that temporal music grounding remains challenging for current audio-language models, whereas task-specific training yields substantial gains. We further report exploratory evidence on the relationship between grounding supervision and music understanding. These results establish MusicGroundingBench as a controlled testbed for assessing whether audio-language models ground their responses in temporally localized musical evidence.
[IR-34] Content Exploration Beyond the Feed: Creator Supply and the Shared Corpus
链接: https://arxiv.org/abs/2608.29430
作者: Yuanyuan Shen,Yiren Yan,Wenjie Li,Chunhui Zhu
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG); Econometrics (econ.EM); Applications (stat.AP); Machine Learning (stat.ML)
备注:
Abstract:Industrial recommenders give new content initial views through budgeted exploration, then use early performance to decide further delivery. On many short-video platforms, exploration is the primary way new videos reach viewers. Viewer-side tests measure consumption; the published budget objectives we review omit creator response. We analyze four experiments on a major short-video platform. An eight-month creator ablation finds production exploration raises videos posted per creator by 8.55% and creators posting at least once by 7.10% relative to a minimal floor. A budget-matched reallocation raises creator participation with no detectable short-run viewer-side change. A year-long viewer ablation finds 1.74% more video views but 2.13% less view time. A delivered view creates immediate feed value, can trigger organic take-up, and can induce creator supply. Take-up and supply replenish a shared corpus, creating two measurement limits. Viewer-side A/B tests cancel the corpus effect when both arms consume the same corpus. Giving each arm its own corpus avoids cancellation, but turnover still controls the horizon. If the corpus turns over at rate w per posting cycle, a t-cycle experiment expresses at most wt of the eventual corpus effect. More users reduce noise but do not speed turnover. Before the corpus path visibly bends, data cannot distinguish a modest fast effect from an arbitrarily large slow one, so a valid confidence interval may lack a finite upper endpoint. As predicted, the three-week co-diverted experiment cannot determine the sign of the eventual corpus effect. Within the window, it identifies the direct feed effect, and an exploratory cohort analysis detects organic lift after exploration ends. The experiments establish a positive creator response, measure the gross corpus flow visible within three weeks, and show the design and duration needed to identify total value.
[IR-35] Agents as Knowledge Integrator and Utilizer in Multimodal Recommendation
链接: https://arxiv.org/abs/2608.29410
作者: Jinfeng Xu,Zheyu Chen,Shuo Yang,Jinze Li,Puzhen Wu,Zewei Liu,Zheng Lin,Jianheng Tang,Jing Yang,Wei Wang,Xiping Hu,Edith Ngai
类目: Information Retrieval (cs.IR)
备注:
Abstract:Online platforms increasingly rely on multimodal recommender systems to rank products, media, and other Web content. Existing methods usually inject visual and textual features into item representations or build homogeneous graphs from modality-level similarity, but the resulting signals can remain misaligned with the recommendation objective. We study this semantic gap from a knowledge-integration perspective: multimodal content should be interpreted together with user behavior before it is used to construct recommendation graphs or adjust rankings. We propose AgentMMRec, an agent-based multimodal recommendation framework with two coordinated roles. The Integrator Agent infers behavior- and multimodal-aware user preferences and item properties from training interactions and item content, then stores them in a reusable knowledge memory. The Utilizer Agent consumes this memory to refine modality-specific item-item graphs, construct behavior-aware homogeneous graphs, and rerank candidate lists under a frozen evaluation-time memory. This design differs from direct LLM feature augmentation and pure LLM reranking because the generated knowledge is first converted into graph structure and model representations before recommendation. Experiments on three Amazon multimodal recommendation datasets show that AgentMMRec consistently improves Recall and NDCG over recent multimodal baselines, remains effective under sparsity and item cold-start settings, and can transfer its constructed knowledge to existing backbones. Subjects: Information Retrieval (cs.IR) Cite as: arXiv:2608.29410 [cs.IR] (or arXiv:2608.29410v1 [cs.IR] for this version) https://doi.org/10.48550/arXiv.2608.29410 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[IR-36] Personalized Recommender Systems for Gym Workouts: A Reinforcement Learning Approach
链接: https://arxiv.org/abs/2608.29409
作者: Roan Rosema,Helma Torkamaan,Masoud Mansoury
类目: Information Retrieval (cs.IR)
备注:
Abstract:Workout recommender systems aim to help gym users complete effective and engaging training sessions. However, recommending exercises alone is insufficient, as a practical system must also determine appropriate sets, repetitions, and training loads, while adapting to user behavior such as skipping exercises. Existing approaches typically consider only a subset of these factors, limiting their applicability in real-world settings. In this paper, we extend workout recommendation from exercise selection to full workout prescription. We propose a reinforcement learning (RL)-based framework with four environments: exercise-only and full-prescription settings, each with and without skip-based interaction. The full-prescription environments recommend exercises, sets, repetitions, and load, while the skip-enabled environments use user skipping behavior for online personalization. Experiments with synthetic users show that modeling the full prescription task leads to higher rewards and greater user engagement than exercise-only recommendation, highlighting the importance of realistic workout planning in personalized gym recommender systems.
[IR-37] FISICA: A Deployed Service for Plantar-Pressure and Posture Assessment with Ontology-Grounded Recommendation
链接: https://arxiv.org/abs/2608.29336
作者: Juhwan Song,Heejung Kim,Juntae Noh,Jonghak Ryu,Huiju Park,Junseong Lee,Dohyeon Ahn,Byungwoo Jo
类目: Computer Vision and Pattern Recognition (cs.CV); Human-Computer Interaction (cs.HC); Information Retrieval (cs.IR)
备注: 31 pages, 17 figures, 25 tables
Abstract:FISICA is a body-assessment and recommendation service running in production. One standing session with two photographs returns foot-loading measures, posture coordinates, a driven 3D avatar, a visual report, and ranked shoe and exercise candidates. Measurement comes from a purpose-built scale carrying 634 force-sensitive elements on a 1 cm grid and four load cells, and a rule-based evaluator controls every recommendation while a language model only explains the stored result. The method contribution is the avatar. Instead of mapping a measured angle onto a rig through a tuned gain, we measure the avatar with the same function used on the subject and solve until the two agree, on a sampling-invariant spinal metric that separated a normal from a kyphotic record by 7.2 degrees against 0.9 degrees for a single-joint formulation. In production, general APIs respond at a 0.023 s median, plantar-pressure analysis at 0.45 s, and recommendation at 2.16 s to 2.26 s with the rule-based portion under one second in every trial. The served keypoint graph reaches 0.960 PCK@0.2 on public data, and the catalog holds 699 shoes with 10,500 typed facts. An approved study supplies the radiographic reference for the validation still ahead.
[IR-38] Database-Augmented RAG for Automated Repair of REST API Misuses
链接: https://arxiv.org/abs/2608.29290
作者: Shoei Inoue,Norihiro Yoshida,Erina Makihara,Shiyu Yang,Katsuro Inoue
类目: Information Retrieval (cs.IR); Software Engineering (cs.SE)
备注: Accepted at the 26th International Conference on Software Quality, Reliability, and Security (QRS 2026)
Abstract:Many Internet of Things (IoT) services provide Representational State Transfer (REST) APIs, which require client developers to implement applications that conform to the corresponding API specifications. When client programs contain API misuse, developers debug them based on error responses. However, such responses are often insufficient for identifying the root cause, requiring developers to repeatedly communicate with the server. Retrieval-Augmented Generation (RAG) is a promising approach for providing large language models (LLMs) with external knowledge. However, in automated repair of REST API misuses, it remains unclear how specifications should be stored in a RAG database. This study evaluates how different configurations for organizing API specifications affect RAG-based repair of REST API misuse. We constructed 11 RAG configurations with different database structures and compared their repair rates with a baseline method. For evaluation, we used REST API misuse cases collected from real-world repositories. The results show that, in the studied datasets, the baseline method achieved a repair rate of 54.3%, whereas a RAG-based method using four databases achieved a maximum repair rate of 88.6%. These results indicate that organizing specifications according to version and content type can be an effective design choice for RAG-based REST API misuse repair.
[IR-39] Cloud and On-Premises Deployment of Uzbek Legal RAG via Targeted Retriever Fine-Tuning EMNLP2026
链接: https://arxiv.org/abs/2608.29284
作者: Tatul Danielyan,Mariam Avetisyan,Hrant Davtyan
类目: Computation and Language (cs.CL); Information Retrieval (cs.IR)
备注: Accepted to EMNLP 2026 (Industry Track). Project home: [ this https URL ]( this https URL )
Abstract:Deploying large language models for legal question answering raises challenges that general-purpose leaderboards do not capture, particularly for low-resource languages and under hard operational constraints. We report on building and operating a retrieval-augmented (RAG) legal assistant for Uzbek that must run in two regimes: a managed cloud service that maximizes answer quality within a per-token cost ceiling, and an on-premises deployment for clients whose legal data may not leave their infrastructure, restricting us to open-weight models on limited local hardware under latency constraints. Because no evaluation existed for this setting, we build two domain benchmarks: a retrieval benchmark of 178 expert-annotated legal queries with gold provision spans, and an end-to-end benchmark of 504 expert-curated question–answer pairs scored by an LLM judge whose ratings we validate against human judgments and against an independent-family judge. Applying these benchmarks under each regime, we find the open-versus-proprietary gap is small and cheaply closed by fine-tuning. Therefore, we train UTE-1, which is a state-of-the-art text embedder among open models for Uzbek. We also demonstrate that closing the performance gap via fine-tuning is both impractical due to the intensive hardware demands of long-context legal Q\A and unnecessary, given that legal acts change frequently. We support this by reporting a negative result from a QLoRA experiment. We distill practical guidance for similar deployments, drawn from a system serving real users in production. We release our benchmarks, evaluation code and the fine-tuned embedder (UTE-1) \hrefthis https URLat this https URL to support future work on low-resource legal NLP.
[IR-40] Validating FKG.in: Soundness Assessment in LLM -Augmented Indian Food Knowledge
链接: https://arxiv.org/abs/2608.29249
作者: Saransh Kumar Gupta,Armaan Shah,Lipika Dey,Partha Pratim Das,Ramesh Jain
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注: 15 pages, 2 figures, 5 tables, 27 references
Abstract:The online culinary ecosystem is increasingly populated by recipe content generated, modified, or summarized by Large Language Models (LLMs). While often plausible, such outputs may contain hallucinated ingredients, misrepresented quantities, or culturally implausible combinations, limiting their suitability for downstream applications and knowledge graph construction. In this paper, we present a semi-automated soundness assessment workflow for validating structured recipe data extracted and augmented by LLMs from informal culinary sources. Developed as part of this http URL, a knowledge graph of Indian food, the pipeline identifies and addresses common failure modes, including structural inconsistencies, semantic and logical incoherence, and deviations from the source text, through a multi-stage process combining formal grammars, vocabulary-based checks, statistical heuristics, Set Transformer-based coherence modeling, and retrieval-based verification. Although evaluated on Indian recipes, the proposed methods are applicable to broader multilingual and multicultural culinary domains. We provide a practical, auditable, and application-agnostic framework for validating LLM-augmented recipe data, thereby strengthening the foundations of machine-readable food knowledge infrastructures in the era of LLM-generated content.
[IR-41] AAL: Mitigating Early Beam Pruning in Generative Recommendation via Temporal Autoregressive Alignment
链接: https://arxiv.org/abs/2608.29179
作者: Lianjie Li,Zhiying Tu,Dianhui Chu,Hongliang Sun
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注:
Abstract:Generative recommendation encodes items as hierarchical semantic identifiers (SIDs) and retrieves the next item through autoregressive decoding. Standard next-token prediction, however, does not explicitly cover the multimodal transitions present in interaction sequences, leaving the ground-truth SID vulnerable to irreversible pruning at early beam-search branches. Across three public benchmarks, we find that 91.9%–96.6% of retrieval failures occur within the first two decoding steps. We therefore propose Temporal Autoregressive Alignment (TAAL). During training, TAAL constructs a joint (c_1,c_2) soft target from historical transitions and aligns the early-prefix distribution with a forward KL objective. During inference, it calibrates candidate scores with pointwise mutual information (PMI) to reduce the influence of globally frequent prefixes. On Amazon Beauty, Instruments, and Yelp, TAAL improves NDCG@10 over the standard baseline by 39.5%, 6.7%, and 28.6%, respectively, while increasing full-SID survival by 3.9%–16.6%. Beam-width analysis further shows that the relative survival gain grows as the beam narrows, reaching 39.4% at B=5 .
[IR-42] Book Readership During Movie Releases: An Exploratory Analysis RECSYS2026
链接: https://arxiv.org/abs/2608.29019
作者: Sushobhan Parajuli,Vittoria Vineis,Samira Vaez Barenji,Michael D. Ekstrand
类目: Information Retrieval (cs.IR)
备注: RecTemp 2026, Temporal Reasoning in Recommender Systems, Workshop at ACM RecSys 2026
Abstract:Exogenous events can temporarily change the relevance of items in recommender systems, but these shifts are often not visible in historical interaction data until after users have already responded. In book recommendation, movie adaptations provide a clear example of such events: the release of a movie based on a book can temporarily increase attention to the source text and change its relevance for some readers. We examine this phenomenon using a large Goodreads dataset matched to movie release dates. We find a clear spike in readership around the release month, and then we evaluate existing recommendation models to understand how they rank movie-adapted books around the movie release date.
[IR-43] Context-Aware Interpretable Representations for Retrieval and Graph Convolutional Network Classification ICMR2026
链接: https://arxiv.org/abs/2608.29004
作者: Thiago César Castilho Almeida,Gustavo Rosseto Letício,Vinicius Atsushi Sato Kawai,Daniel Carlos Guimarães Pedronette
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV); Information Retrieval (cs.IR)
备注: Published in the Proceedings of the 2026 International Conference on Multimedia Retrieval (ICMR 2026)
Abstract:The advances in visual information modeling and representation during the last decades are remarkable, mainly supported by Convolutional Neural Networks, Transformer-based, and Foundation Models. Despite this progress, critical challenges regarding the nature of similarity assessment and model transparency have been neglected. A primary concern is the Geometric Gap, where traditional pairwise measures fail to capture the intrinsic geometry of the dataset manifold. Furthermore, the Interpretability Gap persists, as representations often lack alignment with human cognition. Therefore, how to provide interpretability to representations while maintaining low dimensionality and high effectiveness in downstream tasks remains an open challenge. In this paper, we propose a novel unsupervised framework that integrates Manifold Learning strategies with Rank-based Interpretable Graph Embeddings. Our approach effectively bridges these gaps by first characterizing the contextual information of the dataset through manifold analysis and subsequently generating sparse, self-explainable embeddings. The proposed approach employs a flexible formulation, allowing different Manifold Learning and Representation Learning strategies. Extensive experimental evaluation across diverse datasets and features demonstrates that our Context-Aware representations not only provide intrinsic interpretability and dimensionality reduction but also maintain or enhance effectiveness in downstream tasks, specifically in image retrieval and semi-supervised classification using Graph Convolutional Networks (GCNs).
[IR-44] Effective Graph and Rank-based Contextual Embeddings for Textual and Multimedia Data IJCNN2025
链接: https://arxiv.org/abs/2608.29001
作者: Thiago César Castilho Almeida,Gustavo Rosseto Letício,Lucas Pascotti Valem,André Freitas,Daniel Carlos Guimarães Pedronette
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV); Information Retrieval (cs.IR)
备注: Published in International Joint Conference on Neural Networks, 2025 (IJCNN 2025)
Abstract:In a data-driven world, efficiently organizing and mapping relationships between objects is crucial. Graphs are powerful tools for modeling these connections, being widely used in social networks, telecommunications, and biology. However, graph-based methods often face high computational costs, particularly in memory and space usage. To address this, graph embedding techniques, also referred to as Network Representation Learning, encode graph information into lower-dimensional representations while preserving structural aspects. Traditional methods, however, lack interpretable dimensions. RaDE (Rank Diffusion Embedding) introduces a new approach using rank-based information, with a key step being the selection of a representative subset of nodes to provide interpretability for its dimensions and improve retrieval tasks. Despite its potential, RaDE’s original proposal did not fully explore the effectiveness of representative subset selection across different classes or evaluate embeddings in tasks like classification and clustering. Inspired by RaDE, this work introduces GRaCE (Graph and Rank-based Contextual Embeddings), a fully unsupervised framework that generates interpretable embeddings by leveraging robust rank-based measures for representative subset selection and node embedding. GRaCE surpasses RaDE and Original Features across diverse datasets, including textual and image collections, excelling in retrieval, classification, and clustering tasks, considering state-of-the-art Transformer models as feature descriptors and Graph Convolutional Networks models in classification tasks.
[IR-45] MERIT: Mitigating Exposure Bias in Generative XMC for User-Interest Propensity Modeling EMNLP2026
链接: https://arxiv.org/abs/2608.28931
作者: Abhinav Mahajan,Arindam Sarkar,Prakash Mandayam Comar
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注: Accepted to EMNLP 2026
Abstract:Matching users to interest categories at scale is central to personalized shopping, but the task is challenging in large e-commerce platforms, where label spaces continually evolve and user-interest signals are sparse and long-tailed. Autoregressive language models are appealing because their world knowledge and semantic priors over descriptors generalize across extreme label spaces and accommodate multiple valid label assignments. Yet under teacher-forced fine-tuning, inference-time predictions become part of the conditioning context: early errors steer later outputs toward co-occurring labels, over-generating near-correlates and missing unrelated true interests. We present MERIT, a framework for user-interest propensity modeling that mitigates this exposure bias through a self-correction objective. A permutation-invariant multi-target loss over shuffled mixtures of gold and mined hard-negative labels exposes the generator to erroneous prefixes while preserving the efficiency of teacher-forced training. This training objective concentrates supervision at classification positions, yielding propensity-aligned hidden states powering a lightweight scorer for bidirectional retrieval (interests for users and users for interests). On a proprietary e-commerce dataset with 250k+ interest categories, MERIT improves global recall by at least 11.9% and average Hit@k by 6.1%. In production A/B tests, it achieves +0.26% gain in user conversion.
[IR-46] Weaving Visual Narratives: Agent ic Image Bundle Composition Beyond Atomic Visual Matching EMNLP’26
链接: https://arxiv.org/abs/2608.28695
作者: Rong Shan,Tianyi Xu,Congmin Zheng,Wenteng Chen,Jiachen Zhu,Junjie Wu,Teng Wang,Weiwen Liu,Changwang Zhang,Weinan Zhang,Jun Wang,Jianghao Lin
类目: Computer Vision and Pattern Recognition (cs.CV); Information Retrieval (cs.IR)
备注: Accepted by EMNLP’26 Main Conference
Abstract:Image retrieval has traditionally been formulated as a point-wise matching problem, where each candidate image is scored in isolation. However, this atomic paradigm fails to capture the complexity of human search intent within personal photo collections, where users often seek compact visual stories bound by structural relations rather than isolated snapshots. To address this limitation, we introduce Image Bundle Composition (IBC), a novel paradigm that shifts the objective from ranking individual images to dynamically composing cohesive image bundles from a massive, unstructured photo pool. Since target bundles are not predefined, IBC presents a severe combinatorial explosion challenge and demands modeling non-decomposable joint relevance. To establish this paradigm, we construct IBCBench, the first IBC benchmark dataset containing 109,467 images and 667 verified queries, built via a semi-automated verification pipeline. Furthermore, we propose BundleWeaver, an agentic framework that reformulates IBC as query-conditioned incremental hyperedge discovery. By employing a Large Language Model to adaptively search for missing relational roles and utilizing a Vision-Language Model for whole-bundle verification, BundleWeaver effectively navigates the combinatorial space. Extensive experiments demonstrate that while state-of-the-art embedding models and static decompose-and-rerank paradigms suffer from relational blindness, BundleWeaver achieves substantial performance gains, highlighting the necessity of shifting from atomic scoring to dynamic relational composition. Our dataset and code are available.
[IR-47] Can Large Language Models Identify Meaningful Touchpoints in Conversion Attribution? CIKM2026
链接: https://arxiv.org/abs/2608.28649
作者: Jinqi Wu,Sishuo Chen,Zhangming Chan,Yong Bai,Chao Yi,Han Zhu,Shuodian Yu,Lei Zhang,Sheng Chen,Chenghuan Hou,Jian Xu,Chaoyou Fu
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
备注: 6 pages, 4 figures, 3 tables; accepted as a short paper at CIKM 2026
Abstract:Touchpoint selection in conversion attribution, namely identifying meaningful touchpoints contributing to conversions, is essential for e-commerce recommendation and online advertising. Current selection methods rely heavily on collaborative-filtering-based heuristics, which fail to align with user-perceived semantic intent. Through human annotation, we reveal a significant semantic gap: many implicitly-related, semantically relevant touchpoints remain undetected by existing rules. Therefore, we systematically evaluate the capability of Large Language Models (LLMs) in identifying these hidden associations. Our evaluation shows that while LLMs effectively uncover a substantial portion of implicitly-related touchpoints, significant room for improvement remains in their selection performance. Furthermore, we analyze the impact of different prompting strategies and foundation model choices on identification performance, providing valuable insights into their reasoning patterns and effectiveness. These insights offer a new roadmap for transitioning conversion attribution from mechanical rule-matching to human-aligned semantic reasoning. Moreover, we leverage the LLM-attributed conversion labels for enhancing industrial CVR model training and achieve significant offline performance gains, showing the potential of LLMs in conversion attribution.
[IR-48] NLP-Driven Knowledge Extraction and Thematic Classification of Translated Ancient Indian Medical Texts
链接: https://arxiv.org/abs/2608.28608
作者: M. S. Rajeevan,B. Mini Devi,V.S. Anoop,C. Mallikarjuna
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
备注: 19 pages, 8 figures, 5 tables. Presented at the National Conference on “Reimagining LIS Education: Integrating Indian Knowledge Systems with NEP 2020” (March 2025), organized by Tata Institute of Social Sciences (TISS) and the Indian Association of Teachers of Library and Information Science (IATLIS). Recipient of the Best Paper Award
Abstract:Ancient Indian medical texts like Sushruta Samhita have extensive information on diseases, treatments, and surgical techniques. Yet, their ancient format and use of intricate vocabulary pose difficulties in accessibility and systematic ordering. The research here utilizes Natural Language Processing (NLP) methods like Named Entity Recognition (NER), BERTopic modeling, and Knowledge Graph development in Neo4j to extract, categorize, and visualize important concepts based on translated versions. Thematic classification with BERTopic allows for the identification of the underlying medical topics, whereas NER supports the structured entity recognition of diseases, treatments, researchers, and medicinal plants. Graphbased network analysis with Neo4j also allows for the semantic representation of relationship among extracted entities, supporting knowledge retrieval and digital preservation. The findings illustrate how graph databases, topic modeling, and entity recognition facilitate the computational organization of Ayurveda’s historical medical wisdom, closing the gap between the conventional texts and contemporary data-driven inquiry. The suggested method promotes historical text analysis, medical informatics, and digital humanities to make ancient Indian medical wisdom more accessible and understandable.
人机交互
[HC-0] Augmenting Interviewer Judgments of Patient Experience with Automatic Language Analysis
链接: https://arxiv.org/abs/2608.31007
作者: Aowen Shi,Michal Balazia,Danilo Postin,René Hurlemann,Jan Alexandersson,François Brémond,Philipp Müller
类目: Human-Computer Interaction (cs.HC); Computation and Language (cs.CL)
备注: Accepted to ACII 2026 as an oral presentation
Abstract:Understanding how psychiatric patients subjectively experienced a clinical conversation is important for feedback and alliance-related process monitoring. While interviewers form post-session judgments about patient experience, these judgments do not always match patients’ self-reports. Automatic approaches for predicting perceived interaction quality from conversation have been proposed, but it remains unclear whether such approaches can complement human judgment rather than simply replicate it. To address this gap, we evaluate a clinician-support framework in which post-session interviewer ratings are combined with automatic language-based predictions to estimate patient-reported interaction quality in free clinical interviews. We assess this integration across multiple standard model types, including Ridge, SVR, MLP, GRU, and BiLSTM, all trained on sentence embeddings extracted from dyadic transcripts of 107 free conversations between psychiatric patients and interviewers. Our results show that combining interviewer judgments with model predictions through simple averaging yields the strongest overall performance. The interviewer-only baseline reached a Pearson correlation of 0.365. Among fully automatic models, Ridge achieved the strongest Pearson correlation (r = 0.286), while BiLSTM achieved r = 0.270. The strongest result was obtained by BiLSTM interviewer integration (r = 0.403). Our findings suggest that automatic language analysis and interviewer judgment capture complementary aspects of patient experience and that their combination provides a more accurate approximation of the patient’s own report than either source alone.
[HC-1] You Shouldnt Have Asked: A Prag matics-Inspired Taxonomy for Evaluating LLM Refusals EMNLP2026
链接: https://arxiv.org/abs/2608.30856
作者: Ruoxuan Li,Pinqiao Wang,Sheng Li,Cameron Robert Jones
类目: Computation and Language (cs.CL); Human-Computer Interaction (cs.HC)
备注: To appear in the Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026)
Abstract:Refusals are often treated as face-threatening acts in pragmatics because they can challenge the requester’s socially claimed self-image. Large language models (LLMs) are increasingly trained to refuse unsafe and inappropriate requests, and these refusals may harm users when models fail to manage this interactional cost properly. While existing work has mainly approached LLM non-compliance as a safety-alignment outcome, it does not provide a way to evaluate whether LLMs refuse appropriately across different harmful contexts. To study this question, we propose (to our knowledge) the first taxonomy of LLM refusals that is grounded in pragmatic theory. Applying this taxonomy to responses from 16 modern LLMs across 14 harm categories, we find that although models differ in how they refuse, their refusals are overall explicit and strongly morally evaluative, with interactional repair occurring mainly through offering or providing safer alternatives instead of interpersonal facework. This pattern is especially consequential in sensitive harm contexts, where overuse of negative framing may make users feel shamed or provoked, undermining the purpose of safe non-compliance. We therefore call for alignment evaluation that considers not only whether models refuse harmful requests, but also whether they refuse in ways that are contextually adaptive and socially accountable for the interactional consequences of saying no.
[HC-2] LipCoder: Voice-Enabled Coding Toolkit
链接: https://arxiv.org/abs/2608.30793
作者: Hayoon Kim,Sungho Lee,Juhwi Kim,Bongwon Suh,Kyogu Lee
类目: Human-Computer Interaction (cs.HC)
备注: Accepted to the Posters and Demos Track at the 28th International ACM SIGACCESS Conference on Computers and Accessibility (ASSETS 2026)
Abstract:AI-assisted programming environments have accelerated software development, giving rise to new paradigms like vibe coding. However, their benefits remain largely inaccessible to visually impaired programmers, as existing screen readers and assistive tools offer limited support for these emerging workflows. We introduce LipCoder, a voice-centric programming toolkit designed to deliver editor-level functionality through auditory and speech-based interfaces. LipCoder offers features comprising speech feedback and earcon cues for comprehension and validation, as well as natural language input for navigation and modification. In an exploratory evaluation, 5 visually impaired programmers performed a series of coding tasks comparing LipCoder with a baseline of VSCode, Copilot, and VoiceOver. Quantitative trends and qualitative feedback point to directions for auditory-first design that may broaden accessibility in speech-driven coding environments.
[HC-3] Domain-Grounded Tool Orchestration for LLM -Guided Scientific Analysis
链接: https://arxiv.org/abs/2608.30696
作者: Jeff Lee,Sebastien Jourdain,Cory Quammen,Patrick O’Leary,Berk Geveci
类目: Computational Engineering, Finance, and Science (cs.CE); Human-Computer Interaction (cs.HC)
备注: 27 pages, 12 figures
Abstract:Scientific analysis workflows encode deep domain knowledge through sequences of tightly coupled operations where correctness depends on tool selection, execution order, and parameterization. A CFD engineer investigating flow separation must extract wall shear stress, identify zero-crossings in skin friction, and confirm with boundary-layer profiles: a chain that requires both domain expertise and proficiency with visualization tools. Current approaches to LLM-assisted scientific visualization generate scripts that encode this knowledge implicitly, and often incorrectly, producing code that executes but yields wrong results. We present an architecture that separates intent interpretation (LLM) from execution (deterministic domain tools) from explanation (LLM), connected by the Model Context Protocol (MCP) and grounded by domain ontologies that constrain planning to valid analysis chains. We instantiate the architecture in two domains on the same ParaView server infrastructure: computational fluid dynamics post-processing and topological data analysis via the Topology ToolKit (TTK). Adding the second domain required only an ontology and tool wrappers around existing filters, with no change to the architecture, protocol, or deployment. By construction the design removes whole classes of failure that affect script generation (such as API hallucination and missing pipeline stages) and narrows the strategic errors that remain. An ablation across both domains locates the ontology’s empirical effect: it does not change which tools the planner selects, which is already reliable, but corrects how the model interprets results, raising interpretation accuracy from 0.41 to 0.91, and only when the relevant fact is retrieved in scoped rather than bulk form. ParaView’s client-server model carries analysis to production-scale datasets through a thin browser client.
[HC-4] Inferring Value Criteria from Ordinal Preferences: An Iterative In-Context Learning Framework for Music Generation
链接: https://arxiv.org/abs/2608.30694
作者: Futa Hidaka,Naomi Imasato,Kazuki Miyazawa,Takato Horii
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:Adapting a generative music system to an individual’s taste requires learning what that listener values. Listeners can rank pieces, but their underlying criteria may be tacit and difficult to articulate. We ask whether and under what conditions a large language model (LLM) can adapt symbolic music generation from rankings alone and construct transferable natural-language descriptions of value criteria. In our iterative in-context learning framework, the LLM formulates hypotheses, generates candidate pieces in ABC notation, receives a ranking, and periodically infers and verbalizes value criteria from history to guide later generation. We evaluate the framework against 16 simulated raters in 480 adaptation runs using mixed-effects modeling, an ablation, and transfer tests on unseen music. Overall, the framework did not outperform a feedback-free diverse-generation baseline, but did so for two value functions with targets difficult to reach through simple sampling. How atypical the target was relative to the LLM’s feedback-free generation tendencies predicted adaptation difficulty. Moreover, higher value during adaptation did not imply identification of the criterion as a general rule. On unseen music, acquired descriptions and histories improved generation for more value functions than they improved preference prediction, which remained near chance. Some gains were associated with acoustic proximity to music in the context, but others were not. These findings show that rankings alone can guide generation under limited conditions, while transferable criterion inference remains constrained by the foundation model’s ability to recognize, reason about, and verbalize musical attributes.
[HC-5] User Experience in Human-Machine Interaction: Insights from Field Studies in Autonomous Mobility ECCV2026
链接: https://arxiv.org/abs/2608.30526
作者: Helen Schneider,Svetlana Pavlitska,J. Marius Zöllner
类目: Human-Computer Interaction (cs.HC)
备注: Accepted for publication at ABAW Workshop at ECCV2026
Abstract:Autonomous vehicles (AVs) promise safer, cleaner, and more inclusive mobility, yet large-scale adoption is hindered by user acceptance rather than by technical challenges. Prior studies on acceptance and user experience largely rely on surveys, simulators or Wizard-of-Oz setups, often over-representing technologically enthusiastic participants and focusing on drivers instead of passengers. We address this gap with real-world field studies with AVs in real traffic, totaling 144 participants. Using multi-modal sensing, we evaluated EGG, heartbeat, breathing, camera and voice signals for affect inference in combination with vehicle data. Our results show that breathing, camera and voice measurements are reliable and pratical in naturalistic passenger contexts. We further contribute a validated study protocol, a self-assessment app for real-time assessment during human-machine interaction, and a tailored questionnaire to capture participant attitudes towards AVs. By grounding UX evaluation in real-world contexts, this work lays a foundation for user-centered design of autonomous mobility systems and robotics in general. Our work bridges the gap between affective computing and technical implementation of autonomous vehicles.
[HC-6] SExplorer: An interactive data annotation and exploration tool for time-series data INTERSPEECH2026
链接: https://arxiv.org/abs/2608.30514
作者: Einari Vaaras,Manu Airaksinen,Okko Räsänen
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Software Engineering (cs.SE)
备注: Accepted for publication in Proc. Interspeech 2026 (Sydney, Australia) in the Show Tell track
Abstract:We present TSExplorer, a cross-platform tool for interactive annotation and exploration of time-series data. The tool enables users to inspect high-dimensional datasets through multiple complementary 2D visualizations derived from high-dimensional feature representations. TSExplorer is designed as a general-purpose research tool supporting a wide range of workflows, including exploratory data analysis, annotation of unlabeled or partially-labeled datasets, comparison of feature representations, and post-hoc inspection and refinement of existing labels with interactive visual feedback.
[HC-7] owards Cognitive Process-Aware Proactive Writing Support
链接: https://arxiv.org/abs/2608.30424
作者: Masahiro Yoshida,Atsuya Kobayashi,Kei Tateno,Xiang ‘Anthony’ Chen
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models can support writing, but existing tools require users to explicitly articulate prompts-particularly burdensome in creative writing, where intentions are often ambiguous. Proactive support that infers users’ needs from writing interactions could alleviate this burden, but raises two challenges: determining what support to provide and when to intervene. This work focuses on the former. We hypothesize that Flower and Hayes’ cognitive process theory of writing-which characterizes writing through six cognitive processes-offers an interpretable bridge between observable writing behavior and appropriate support types. Through a formative study and literature review, we identify 14 writing support types associated with these cognitive processes, along with characteristic interaction behaviors linked to each process. We then instantiate this framework in AToM CoWriter, which infers support needs from writing interactions and document context. Two within-subjects studies (N = 21) provide initial evidence that this approach improves expressiveness and idea exploration, and that cognitive process inference increases engagement with proactive suggestions. These findings suggest that cognitive processes can provide a promising basis for support selection in proactive writing systems.
[HC-8] Augmenting Human Performance with an XR Agent Learning from Online Behavior and BCI Evidence
链接: https://arxiv.org/abs/2608.30369
作者: Ziheng Li,Xichen He,Haoyan Chen,Charlie Zou,Sheng Bai,Benjamin Yang,Mengyuan Wu,Jake Ledner,Yi-Jie Cheng,Akito Yamauchi,Dishita G Turakhia,Steven Feiner,Paul Sajda
类目: Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注: To appear in ACM UIST 2026. 30 pages, 23 figures
Abstract:We present OLIVE, a framework for adapting a foundation model to provide real-time assistance in temporally demanding, high-stakes, and dynamic tasks. We show that passive EEG, fused online with behavioral evidence, can meaningfully extend the number of targets users detect and engage beyond their unaided action bandwidth. OLIVE learns from both explicit behavioral signals (the targets the user shoots down in an XR first-person shooter game) and implicit physiological signals (fixation-locked EEG) to provide timely guidance, continuously adapting a frozen vision-language model’s inference on which items are task-relevant by jointly estimating per-source reliability without manual labels or offline training. Through three user studies, including two live deployments of an assistive agent driven by OLIVE in XR, we show that OLIVE Pareto-dominates prior test-time adaptation frameworks, achieving the highest convergence rate at comparable convergence speed. Combining implicit physiological and explicit behavioral signals, the OLIVE agent produces the largest and most reliable within-session improvement to a user’s ability to detect and engage targets, largely independent of the individual’s skill. When the target switches silently, the agent that uses both behavioral and physiological signals reconverges significantly faster than the behavior-only agent (1.27 times faster on average, p = .008), restoring trustworthy guidance at the moment the task changes, precisely when reliable assistance matters most.
[HC-9] Co-Annotator: Expert-Distilled ViT and VLM for Visual and Documentation Guidance in Age-Related Macular Degeneration
链接: https://arxiv.org/abs/2608.30352
作者: Ziheng “Leo” Li,Benjamin Freeman,Akshay Raman,Kavin Aravindhan Rajkumar,Xinxin Fang,Rishabh Srivastava,Steven Feiner,Kaveri A. Thakoor
类目: Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注: 23 pages, 11 figures. To appear in UIST '26: Proceedings of the 39th Annual ACM Symposium on User Interface Software and Technology, November 02-05, 2026, Detroit, MI, USA. DOI: https://doi.org/10.1145/3830398.3830722
Abstract:Clinical AI often optimizes predictive performance without engaging how clinicians decide where to look and what to write. We present Co-Annotator, which distills expert gaze and dictation into two guidance components: a gaze-aligned Vision Transformer producing fixation-aligned areas of interest (AOIs), and an ontology-bounded vision-language model (VLM) that pre-fills editable biomarker summaries for retinal optical coherence tomography (OCT). We first collect expert gaze and dictations (US1) to train the models, significantly improving diagnostic accuracy and biomarker generation. We then deploy the system with ophthalmology residents: a controlled resident study (US2) confirmed each modality is safe and independently beneficial, with AOI guidance producing lasting perceptual efficiency gains through post-guidance carryover and VLM guidance more than doubling biomarker documentation breadth. In a combined deployment across two academic institutions (US3), providing both modalities simultaneously produced efficiency gains that substantially exceeded either modality alone: correct diagnoses per minute increased by 40% and comment editing time fell by 67%, without compromising diagnostic accuracy. Notably, neither modality improved efficiency during guidance in US2, which makes the in-guidance efficiency gain under combined guidance in US3 the more striking result. Expert-distilled multimodal guidance can remove two distinct clinical workflow bottlenecks at once (visual search overhead and documentation burden) without compromising the diagnostic accuracy clinicians already achieve.
[HC-10] DOBI: Dynamic Opportunistic Body Input via Spare Joint Recruitment for Hands-Free XR
链接: https://arxiv.org/abs/2608.30341
作者: Rachel Kim,Xun Qian,Sang Ho Yoon
类目: Human-Computer Interaction (cs.HC)
备注: 14 pages, 12 figures. To appear in the 39th Annual ACM Symposium on User Interface Software and Technology (UIST '26), November 2-5, 2026, Detroit, MI, USA
Abstract:Extended Reality (XR) systems are often most useful when users are engaged in ongoing physical tasks, yet current interaction techniques still largely assume the hands are available. We present opportunistic body input, an interaction paradigm that redirects continuous XR control to whichever available body region remains free in the moment. To investigate how users naturally coordinate these spare-body movements, we conducted an elicitation study across six hand-busy scenarios. We found that while users’ preferred spare body regions shift dynamically based on physical constraints, the resulting spontaneous movements share a consistent, low-dimensional kinematic structure organized around a dominant principal axis. Building on these findings, we present DOBI (Dynamic Opportunistic Body Input), a real-time XR technique that uses gaze to target a UI element, a brief trigger gesture to identify the recruited spare body region, and the region’s subsequent motion to drive continuous 1D control. A 1D Fitts’ law study establishes the baseline motor performance of this paradigm across four distinct body regions, achieving throughputs up to 2.62 bits/s with an overall 5.0% error rate, and a dual-task usability study shows that DOBI supports reliable, low-effort control (SUS = 84.2) while users remain engaged in realistic hand-busy activities.
[HC-11] One AI Signal Many Human Judgments: A Bayesian Cascade Analysis of AI-based Credibility Indicators in Online Information Spread
链接: https://arxiv.org/abs/2608.30311
作者: Zhuoran Lu,Weilong Wang,Yangyang Yu,Xinru Wang,Zhuoyan Li,Zhiwei Liu,Sophia Ananiadou
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI); Social and Information Networks (cs.SI)
备注: 22 pages, 20 figures. Accepted at HCOMP 2026 (2026 ACM Conference on Human-AI Complementarity and Alignment). Supplementary material included as appendices
Abstract:Social media platforms increasingly use AI-based credibility indicators to help users judge misinformation. Unlike individual human-AI decision-making, these indicators are embedded in information spread: users see both an AI prediction and earlier judgments shaped by the same AI, and their own judgments may then enter the public history. Yet how to analytically characterize this process remains under-explored. We therefore introduce a social-learning lens for this setting by extending the classical Bayesian cascade model with the AI indicator as a shared public signal. The resulting Gateway condition compares the evidence from the AI prediction with users’ private impressions. Through this view, we show that AI changes what public history means. Crowd agreement may reflect accumulated independent human evidence, or repeated dependence on the same AI prediction. This creates a preservation-correction trade-off: stronger reliance on AI can preserve correct predictions, but can also lock in incorrect ones by blocking corrective private impressions. We calibrate the model using human-subject data on news veracity judgments. Although the AI outperforms human users, the average user weights it below her own impression but above several peer judgments, while individual users vary from discounting the AI to relying on it enough to cascade. Simulations show that over-reliance on a weak AI is especially harmful, and that diversifying AI signals across users can better keep the crowd informative. We conclude with implications for understanding human-AI interaction in information spread and designing misinformation interventions.
[HC-12] Frontier vision-language models have overtaken young adults at detecting AI-generated portraits – but not their calibration
链接: https://arxiv.org/abs/2608.30210
作者: Sunwhi Kim(1),Sunyul Kim(2),Meounggun Jo(3),Jini Tae(4) ((1) Hwasung Medi-Science University, Dept. of Bio-Healthcare, (2) Yonsei University, Graduate School of Engineering, Dept. of Artificial Intelligence, (3) Hoseo University, (4) Gwangju Institute of Science and Technology, School of Humanities and Social Sciences)
类目: Human-Computer Interaction (cs.HC); Computer Vision and Pattern Recognition (cs.CV)
备注: 35 pages, 8 figures, 5 supplementary figures. Human reference data reused (not newly collected) from arXiv:2603.24048 . Data and code: this https URL
Abstract:AI image generators now create face portraits that are hard to tell from real photographs. Vision-language models (VLMs) are increasingly proposed to flag such images. We benchmarked 19 VLMs on the same 198 face portraits – real photographs and identity-matched ChatGPT-4o and Imagen 3 versions – under the same task as our earlier study of 1,667 adults (85% correct overall; accuracy fell steeply with age). The June-2026 cohort of 14 models only matched adults in their 20s-30s. Four weeks later the ceiling broke. Among five July-2026 releases under the identical protocol, gpt-5.6-sol reached 92.8% balanced accuracy (five-draw mean 92.1%), clearly above adults in their 20s (88.5%), and claude-fable-5 detected every AI image while averaging 91.9%. Model sensitivity now exceeds young adults decisively (d’ up to 3.4 versus ~ 2.4). What has not been overtaken is human calibration. Model criteria spread from c = -1.10 to +1.45 while humans sit near zero at every age; both new leaders are biased (+0.44, -0.97), and only a few mid-ranked models approach the human balance. Changing the labelled examples still flipped about one answer in four. The best machines now out-see young adults here, without matching the human balance between suspicion and trust.
[HC-13] LandmarkLens: Predicting and Presenting Effective Landmarks for Mixed-Reality Urban Exploration
链接: https://arxiv.org/abs/2608.30142
作者: Chu Li,Yotam Sechayk,Jared Hwang,Jon E. Froehlich,Takeo Igarashi
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:People with a poor sense of direction (SOD) struggle to build cognitive maps for effective spatial navigation, and existing navigation tools prioritize efficiency over spatial learning. To understand how navigation strategies differ by ability, we conducted a landmark attention study with 20 participants (ten good SOD, ten poor SOD) who navigated across four Tokyo neighborhoods in virtual reality (VR). We found systematic group differences in both gaze behavior and the types of landmarks they verbally identify as effective. Based on these findings, we built LandmarkLens, a mixed-reality (MR) navigation system that uses a vision-language model (VLM) to identify and highlight navigation-relevant landmarks. A follow-up study with eight poor-SOD participants showed improved performance in scene recognition, suggesting that guided landmark attention can support landmark-level spatial knowledge acquisition for people with poor SOD, a first step toward broader spatial learning.
[HC-14] FocusAdapt: Context-aware Adaptive Focus Assistance in Diminished Reality
链接: https://arxiv.org/abs/2608.30108
作者: Tianyu Zhang,Shutong Wu,Jiankun Yang,Zhen Bai,Yukang Yan
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:Diminished Reality (DR) can reduce visual clutter by removing irrelevant objects. However, removing all task-irrelevant objects may eliminate useful contextual information and reduce situational awareness. We present FocusAdapt, a context-aware DR system that predicts object-level distraction by integrating visual saliency, semantic relevance, and gaze behavior. Based on findings from a formative study, FocusAdapt selectively diminishes highly distracting objects while preserving useful context, enabling adaptive focus assistance during procedural tasks.
[HC-15] Occlusion-induced risk and interventions in pedestrian-autonomous truck interactions on multi-lane roads: A virtual reality study
链接: https://arxiv.org/abs/2608.30066
作者: Yun Ye,Yuan Che,S.C. Wong,Stergios-Aristoteles Mitoulis,Haoyang Liang
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:Autonomous trucks (ATs) may introduce distinct pedestrian-safety risks because of their large physical dimensions, constrained braking capability, limited driver-based communication cues, and potential to occlude surrounding traffic. This study employed a controlled virtual reality experiment with 54 participants to investigate pedestrian-AT interaction risk in an unsignalized multi-lane crossing scenario and to evaluate occlusion-targeted risk mitigation strategies. The experiment examined the effects of near-side vehicle type, weather condition, and far-side vehicle yielding strategy on pedestrian behavior, perceived risk, and objective safety. Based on a representative high-risk scenario, three targeted interventions were designed and tested: an environment-aware external human-machine interface (eHMI), a projected eHMI, and an auditory warning. The results showed that ATs increased perceived risk and encouraged more cautious crossing behavior, suggesting a risk-compensation effect. However, this compensation was weakened under rainy conditions, where braking-related safety margins were reduced. AT-induced occlusion further increased far-side interaction risk by limiting pedestrians’ recognition of hidden vehicles. Among the three interventions, the projected eHMI showed the best overall performance, improving objective safety margins, enhancing risk awareness, and supporting behavioral adjustment. These findings highlight the need for AT-specific interface and warning strategies that address both intention communication and risk localization.
[HC-16] Looking Around by Looking Around: Omnidirectional Gaze-based VR Viewport Control
链接: https://arxiv.org/abs/2608.30014
作者: Hock Siang Lee,Jinghui Hu,Florian Weidner,Haopeng Wang,Hans Gellersen
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:Traditional VR viewport control primarily relies on head and torso movement, which can be effortful and limiting in both constrained and extended-use settings. We introduce Looking Around by Looking Around (LALA), a gaze-based VR pitch-and-yaw viewport control technique designed for natural and effortless omnidirectional exploration via eye movements, without requiring or obstructing movement of the head, hand, or body, offering a low-effort and highly accessible interaction method. Because gaze is primarily used for perception and exhibits oculomotor and perceptual asymmetries, using it directly for control is difficult. To address this, we designed an asymmetric omnidirectional control profile for the eye, then built on it to exploit tendencies for eyes to stay within comfortable regions for viewport control. We evaluated LALA in a user study (N=18) featuring two contrasting tasks: alignment towards known directions and open-ended visual search towards unknown directions. LALA was strongly preferred over the traditional baseline, achieving competitive performance while enabling fully hands-free interaction with minimal physical movement.
[HC-17] Multimodal Takeover Requests for Drivers with Hearing Loss: Implications for AI-Enabled Communication in Automated Vehicles
链接: https://arxiv.org/abs/2608.30013
作者: Aries Chu,Wei-Hsiang Lo,Gaojian Huang
类目: Human-Computer Interaction (cs.HC)
备注: 10 pages, 5 figures. Preprint
Abstract:More than 430 million people worldwide live with disabling hearing loss. Although people with hearing loss are legally permitted to drive and may benefit from conditionally automated vehicles, SAE Level 3 systems still require drivers to respond to takeover requests when automation reaches its limits. Existing takeover requests often rely on auditory information, yet little evidence addresses visual and tactile designs for drivers who cannot rely on sound. This driving-simulator study with 40 participants examined the effects of information type (instructional, informative, and baseline), signal type (visual, tactile, and visual-tactile), and hearing condition (normal hearing and simulated hearing impairment) on takeover performance. Information type significantly affected reaction time, with baseline displays producing the shortest times. Signal type significantly affected reaction and takeover time, with visual-tactile displays producing the shortest times. The interaction between signal type and information type was significant for all three measures. Visual-tactile displays produced the shortest reaction times within every information type. With visual-tactile signaling, simple baseline alerts prompted the fastest reactions and the most abrupt maneuvers, whereas informative content produced the lowest mean maximum resulting acceleration. Hearing condition showed no significant main effect on any measure. These findings suggest that AI-enabled vehicles can support urgent takeover communication through visual-tactile displays and can adapt message content to the time available and the maneuver quality required, with implications for drivers across hearing abilities.
[HC-18] A Cyber-Physical Machine Tool Framework with a Real-Time Machining Process Digital Twin
链接: https://arxiv.org/abs/2608.29955
作者: Khalil Chakal,Tero Kaarlela,Jose Outeiro,Carlos Andrade
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:Digital Twins (DTs) have emerged as a key technology for improving the monitoring, optimization, and automation of manufacturing systems. However, existing Cyber-Physical Machine Tool (CPMT) implementations primarily represent the machine tool, while the machining process remains only partially synchronized with its physical counterpart. This paper extends a previously presented CPMT framework by introducing a hierarchical DT framework that simultaneously maintains DTs of both the machine tool and the machining process. The proposed framework integrates real-time CNC operational data, a voxel-based workpiece representation, synchronized process vibration measurements, and a persistent part DT repository for process replay, traceability, and future synthetic data generation. Experimental evaluation demonstrated real-time operation at a 20 Hz machining-state update rate, interactive visualization exceeding 100 frames per second, and a mean depth reconstruction error of 0.16 mm. The implementation provides a foundation for AI-assisted machining applications while preserving the machine tool monitoring and teleoperation capabilities. Subjects: Human-Computer Interaction (cs.HC) Cite as: arXiv:2608.29955 [cs.HC] (or arXiv:2608.29955v1 [cs.HC] for this version) https://doi.org/10.48550/arXiv.2608.29955 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[HC-19] he Policy Deficit in AI x Social-Emotional Learning Research
链接: https://arxiv.org/abs/2608.29950
作者: Tran Van Cuong,Liu Yihan,Nguyen Van Tuong
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注: 23 pages, 1 figure, journal submission
Abstract:As artificial intelligence (AI) is increasingly integrated into social-emotional learning (SEL) initiatives, the need for evidence-based policy has become paramount. We systematically reviewed 65 peer-reviewed papers that examine the intersection of AI and SEL to investigate how these studies articulate policy implications. Our analysis revealed a substantial “policy deficit” in the current AI x SEL literature: nearly three-quarters of the studies did not mention policy implications at all. Using the “WH-question” framework (Who, What, Why, When/Where, and How), we map the policy implications narratives present in the literature and show that they often lack the specificity and actor-oriented guidance required for effective evidence-informed policymaking. We find a significant association between publication venue and policy engagement, suggesting that current academic incentive structures may prioritize technical innovation and pedagogical feasibility over explicit engagement with governance and regulation. This study identifies a “techno-solutionist” trap, where technical potential is foregrounded while the institutional conditions for responsible implementation remain under-specified. We conclude by proposing a shift from “implication-as-afterthought” to “implication-as-methodology” and offer a set of actionable guidelines for researchers, editors, reviewers, and policymakers to bridge the gap between AI innovation and educational governance. Rather than presenting policy as a generic ethical horizon, we argue that AI-SEL studies should systematically specify Who should act, What actions are recommended, Why these actions are needed, When and Where they apply, and How strongly they are framed, thereby strengthening the translation of AI x SEL innovation into educational policy and practice.
[HC-20] HoopMind: A Real-Time Neural Game-Tree System for Opponent-Aware Possession Planning ICDM2026
链接: https://arxiv.org/abs/2608.29563
作者: Yibo Gong,Cong Guo,Jiacheng Ding
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注: 5 pages, 5 figures, 2 tables, 2 algorithms. Submitted to the IEEE ICDM 2026 Teen Research Symposium. Code and live demo: this https URL
Abstract:School coaches prepare for opponents with game film and intuition. The analytics tools of professional teams stay out of reach. We ask how far public data can close this gap. Professional basketball is our case study, chosen for its data rather than the league. We fuse five public sources into one per-shot dataset of 4.23M shots over 21 seasons. The sources are shot locations, two play-by-play feeds, official matchup tracking, and player biometrics. Alignment across them is 99.5% to 100%. We also report two data pitfalls that are easy to miss. We then model a half-court possession as a sequential game. Shot values come from ShotNet, an embedding multilayer perceptron (MLP). On a held-out season it beats a zone-rate baseline and a logistic baseline, and its probabilities are well calibrated. A depth-limited expectimax search then solves the offensive decision tree, with branch-and-bound pruning to keep it real time. All training runs offline, so the online system stays light. A scouting planner and a playable simulator both run in a single browser page.
[HC-21] owards Effective Generation of Interactive Visualizations with Vibe Coding: An Empirical Study
链接: https://arxiv.org/abs/2608.29550
作者: Yanshan Zeng,Ruixuan Tu,Zuo Xiang,Lijia Feng,Guozheng Li,Chi Harold Liu
类目: Human-Computer Interaction (cs.HC)
备注: 5 pages, 4 figures
Abstract:Constructing interactive visualizations has traditionally required substantial human effort, involving both technical implementation and design decision-making. Recently, vibe coding, a programming paradigm leveraging Large Language Models to generate, interpret, and refactor code from natural language specifications, has emerged as a promising approach to reduce the burden. However, the capabilities and limitations of vibe coding in building interactive visualizations remain unexplored. To address this gap, we conducted a user study with 78 participants that were tasked with constructing interactive visualizations using vibe coding. We further collected users feedback through questionnaires, interviews, and case analyses. Based on this study, we examine (1) the capabilities and (2) user experience of vibe coding in generating interactive visualizations, and (3) the practical human-agent collaboration strategies adopted. Our findings provide the first systematic assessment of vibe coding for interactive visualization construction, revealing both its strengths and limitations, explaining the shift in developer labor and identifying the hybrid collaboration strategies participants adopted. Furthermore, our study offers insights for more intuitive and robust vibe coding practices.
[HC-22] Context or Digits? Balancing Memorability and Efficiency in Virtual Reality Authentication
链接: https://arxiv.org/abs/2608.29531
作者: Yuxuan Huang,Qiao Jin,Tongyu Nie,Victoria Interrante,Evan Suma Rosenberg
类目: Cryptography and Security (cs.CR); Human-Computer Interaction (cs.HC)
备注: This paper has been conditionally accepted to VRST 2026
Abstract:We present Adaptive Direction-Based Authentication (ADBA), a knowledge-based authentication method for Virtual Reality that decouples users’ needs temporally by enforcing password creation based on virtual environment context while supporting both context- and digit-based entries during authentication. This design prioritizes memorability for new passwords and offers both efficient and memorable options to support users’ evolving needs. We conducted a remote longitudinal study with 66 participants comparing ADBA against 6-digit PINs over 2-3 weeks. The results demonstrated that ADBA achieved superior memorability and lower perceived task load. Interestingly, no participant chose to enter via digits in the study, yet they still perceived ADBA to be highly efficiency despite longer objective entry times. ADBA also provided security benefits through randomly-generated digit representations, though some degree of password homogeneity was observed in specific virtual environments. Our findings suggest that ADBA offers solid advantages to the traditional PINs, and successfully addresses the tradeoffs between efficiency, memorability, and security under the usage scenarios considered in the study.
[HC-23] EITWatch: Smartwatch-Integrated Planar Electrical Impedance Tomography for Hand Gesture Recognition ATC
链接: https://arxiv.org/abs/2608.29415
作者: Xuanyou Liu,Novel Alam,Karan Ahuja
类目: Human-Computer Interaction (cs.HC)
备注: 7 pages, 6 figures, 2 tables. Accepted to UIST '26 (39th Annual ACM Symposium on User Interface Software and Technology, Detroit, MI, USA, November 2-5, 2026). Hardware schematics, PCB layouts, and firmware: this https URL
Abstract:Wrist Electrical Impedance Tomography (EIT) senses hand gestures from muscle- and tendon-driven impedance changes, but prior wrist-EIT systems require electrode coverage beyond the watch-back contact patch and separate analog front ends. We present EITWatch, the first wrist-EIT system built around smartwatch case-back geometry, asking whether this contact patch alone can support gesture recognition: eight planar electrodes in a 31 mm ring acquire 35 impedance measurements at 48 Hz. Because a planar array cannot encircle the wrist, EITWatch uses multi-depth scanning to sample multiple source-sink distances and current paths; it beat matched adjacent injection by 15.1/10.4 percentage points (macro/micro) across all 12 participants. In a prompted study, within-session leave-one-round-out accuracy reached 91.4%/92.5% (window/trial) for six macro-gestures, and 90.1%/91.5% (window/segment) for five micro-gestures plus relax; window-level cross-session and leave-one-user-out transfer reached 73.2%/70.4% and 63.1%/55.3% (macro/micro).
[HC-24] Feelium: A Touchable Blimp Body for Aerial Telepresence
链接: https://arxiv.org/abs/2608.29391
作者: George Xi Wang,Henghao Li,Shan Lin,Yunge Wen,Jiaqian Hu,Yuhua Jin
类目: Human-Computer Interaction (cs.HC); Robotics (cs.RO)
备注: 3 pages, 2 figures. To appear in UIST Adjunct '26: The 39th Annual ACM Symposium on User Interface Software and Technology, November 2-5, 2026, Detroit, MI, USA
Abstract:Floating things invite touch. We present Feelium, a blimp-based telepresence platform that enables visual embodiment and touch interaction through its inflatable skin. Through a VR headset, a remote person inhabits the blimp, looking out of it first-person, appearing on its skin as a face or avatar, and steering it through the room. Partners in the room pat it, press a palm against it, draw on it, or lean into it; the skin senses each contact, renders it into the wearer’s view in VR spaces. Touch thus provides a physical interaction channel for remote presence, turning the skin into a shared surface between remote and co-located partners.
[HC-25] Understanding Behavioral Dark Patterns of High BMI Individuals ISWC2026
链接: https://arxiv.org/abs/2608.29328
作者: Manjeet Yadav,Prasenjit Karmakar,Suchetana Chakraborty
类目: Human-Computer Interaction (cs.HC)
备注: 5 pages, 5 figures, accepted to DiversityOne Open Challenge at UbiComp/ISWC 2026
Abstract:Understanding how everyday behaviors influence body weight is essential for designing effective and personalized health interventions. Existing studies largely rely on self-reported questionnaires or limited sensing modalities, making it difficult to capture the temporal dynamics of daily behavior. In this work, we analyze the DiversityOne dataset, comprising four weeks of passive smartphone sensing and ecological momentary assessments collected from 453 university students across eight countries. We extract behavioral features spanning dietary habits, physical activity, screen time, and smartphone usage, and investigate their associations with self-reported Body Mass Index (BMI). Beyond feature-level analysis, we employ Hidden Markov Models (HMMs) to uncover latent behavioral patterns. Our analysis reveals that higher BMI is associated with more frequent consumption of soda, alcohol, and processed meat. We further reveal that overweight and obese individuals spend longer periods in food delivery apps and are more likely to transition back to unhealthy eating and drinking routines after starting to exercise. In contrast, normal-weight individuals lead a more balanced lifestyle. These findings highlight key behavioral patterns that make weight loss particularly challenging.
[HC-26] Measuring the “Interaction Gap” in Drama Therapy with AI
链接: https://arxiv.org/abs/2608.29292
作者: Sora Kang
类目: Human-Computer Interaction (cs.HC)
备注: Presented at workshop ‘Human-Centered AI for Expressive Arts Therapy’ at ACM DIS’ 26(Designing Interactive Systems Conference)
Abstract:Generative AI is increasingly being introduced into expressive arts therapy, where it is often credited with offering a non-judgmental environment that supports psychological safety. Existing HCI work has largely positioned AI as a co-creative material or as a bridge/mediator into human-led care. This paper explores a different position. When a patient performs the same drama therapy task with an AI partner and with a human partner, the resulting self-presentations tend to differ in patterned ways. We propose treating this difference, the Interaction Gap, as a diagnostic lens within drama therapy. Rather than asking which context elicits a truer self, the lens reads the difference between the two performances as information about the social pressures shaping self-expression in each context. We sketch a starting point for task design and measurement signals grounded in drama therapy’s existing use of role and aesthetic distance, and raise provocations for workshop discussion: the observer effect and privacy paradox that measurement introduces, and the question of whose lens the gap is.
[HC-27] oward Cultural Alignment: Human-Centered Evaluation of Multimodal AI Stories Across Five African Communities EMNLP2026
链接: https://arxiv.org/abs/2608.29209
作者: Millicent Ochieng,Felermino D. M. A. Ali,Elizabeth A. Ankrah,Najeeb Gambo Abdulhamid,Migisha Boyd,Stephanie Nyairo,Mercy Muchai,Samuel Chege Maina,Aditya Vashistha,Anja Thieme,Jacki O’Neill
类目: Computation and Language (cs.CL); Human-Computer Interaction (cs.HC)
备注: Accepted to Findings of EMNLP 2026
Abstract:In this paper, we examine how well AI-generated multimodal stories align with the lived practices, relationships, language, values, and visual expectations of the communities they represent. We conduct a community-grounded mixed-methods evaluation with 19 culture representatives across five African communities, combining quantitative annotations with qualitative focus group discussions. We find that cultural alignment depends not simply on recognizable cultural markers, but on how those markers fit social, linguistic, procedural, and visual context. From these evaluations, we develop a taxonomy of cultural alignment comprising five broader cultural marker categories and eight recurring mechanisms of misalignment. We additionally evaluate five multimodal LLM judges to examine whether automated evaluation can approximate community-grounded judgments at scale. Judge reliability and score calibration vary substantially across communities, with no single judge performing consistently across all five settings. These findings motivate community-calibrated evaluation pipelines in which automated judges are validated against community judgments to determine where they can be trusted and where human review remains necessary.
[HC-28] An Eye-Tracking Dataset for Viewing Distance Categories in Real-World Scenarios
链接: https://arxiv.org/abs/2608.29192
作者: Dohwa Kim,Yejin Choi,Seungbok Lee,Chi Yoon Jeong,Eunji Park
类目: Human-Computer Interaction (cs.HC)
备注: 22 pages, 6 figures, Manuscript submitted to Scientific Data
Abstract:Estimating viewing distance from gaze behavior is essential for understanding user intent and enabling distance-aware interactive systems. However, most existing eye-tracking datasets have been collected in constrained settings, such as laboratory environments or static tasks. Consequently, they only partially capture viewing behaviors in real-world situations where viewing distance changes with natural head and body movements. We introduce GazeDepth, an eye-tracking dataset collected from 19 participants using a wearable tracker during tasks reflecting real-world scenarios. GazeDepth includes fixed-distance viewing scenarios with constant observer-target distances at near (33 cm), middle (50 cm), and far (300 cm), as well as variable-distance viewing scenarios in which participants shift gaze among targets at different depths in indoor and outdoor environments. The dataset provides synchronized gaze data, pupil size, 3D eye-vectors, and head-motion signals, along with distance labels. Statistical analyses showed that distance-related gaze features, such as vergence angle and estimated viewing distance, differed consistently across viewing-distance categories. In addition, classification models trained on GazeDepth further demonstrated that the dataset captures gaze characteristics that distinguish the three viewing-distance categories, supporting gaze-based distance inference and distance-aware interaction in realistic scenarios.
[HC-29] Sharing Roughness with Hand-Outline Visualization to Reduce Sensory Asymmetry in VR Collaboration
链接: https://arxiv.org/abs/2608.29040
作者: Minju Baeck,Yoonseok Shin,Hyunjin Lee,Boram Yoon,Sang Ho Yoon,Woontack Woo
类目: Human-Computer Interaction (cs.HC)
备注: 10 pages, 8 figures, 9 tables. Accepted for publication in TVCG Special Issue on the 2026 IEEE Symposium on Mixed and Augmented Reality (IEEE ISMAR)
Abstract:In collaborative VR, asymmetric access to haptic hardware creates a critical information gap: tactile evidence remains private to the haptic user, hindering the shared understanding needed for joint decision-making. While prior work has explored crossmodal sensory cues in virtual environments, it remains unclear how such cues should be designed for asymmetric collaboration, where collaborators receive information through different modalities. In our setting, the haptic user feels roughness through fingertip vibration, whereas the non-haptic user relies on vision alone. To reduce this asymmetry, we propose externalizing an object’s tactile state through a glanceable hand-outline visual proxy. Specifically, we examine whether abstract visual roughness cues based on line shape and motion can encode three discrete roughness levels for both haptic and non-haptic users. Two preliminary studies establish a shared visual semantics by identifying visually distinguishable cues for non-haptic users and validating their visuo-haptic correspondence for haptic users. In a main study of a collaborative sorting task, showing this visualization on both users’ hands significantly reduced completion time relative to a no-visualization baseline. Moreover, NU-side cue visibility was associated with higher confidence and perceived contribution for the non-haptic user. These findings show that hand-anchored abstract visual cues provide a lightweight means of externalizing object tactile state, reducing information asymmetry without compromising social presence.
[HC-30] How Mental Health Self-Disclosure Becomes Visible: Evidence from Eight Conditions on Reddit
链接: https://arxiv.org/abs/2608.29010
作者: Renkai Ma,Lingyao Li,Shanting Chen,Chen Chen,Fan Yang,Yuanyuan Lei
类目: Human-Computer Interaction (cs.HC); Computation and Language (cs.CL); Computers and Society (cs.CY)
备注:
Abstract:People share mental health diagnoses on social media, yet how such language becomes visible around their self-disclosure, and whether community engagement tracks it, remain unexamined across conditions. We analyze 89,605 Reddit posts from 739 users across eight conditions, removing each user’s diagnosis disclosure and aligning their surrounding posts to that anchor. Within the pre-disclosure year, language-visible burden was highest in the month before disclosure for six conditions, earlier for post-traumatic stress disorder and furthest from it for borderline personality disorder, and remained visible afterward rather than resolving. The theme Seeking Clinical Explanations showed the largest early-to-late difference before disclosure in five conditions, yet engagement rarely tracked what users wrote: only 9 of 360 language–engagement correlations survived correction. Disclosure is therefore a waypoint in an unevenly visible process, and we offer implications for community practice and platform design where engagement metrics do not reflect clinical need.
[HC-31] Using LLM s to Mimic the Conversational Dynamics of Reddit Communities
链接: https://arxiv.org/abs/2608.28989
作者: Vedaant Jain,Yoshee Jain,Ishq Gupta,Aditi Shrivastava,Koustuv Saha,Eshwar Chandrasekharan
类目: Human-Computer Interaction (cs.HC)
备注: 5 pages, 2 tables, 1 figure
Abstract:Online communities face a constant battle against toxicity and misinformation. While human moderators struggle to keep pace with the volume of content, LLMs offer a promising solution for automatically generating constructive responses and shaping online interactions. This paper preliminarily investigates if LLMs can mimic the communication styles of Reddit users using their comment history as context. We evaluate two prompting approaches: predicting a target comment and filling in masked comments. We find that LLMs outperform expectations at replicating comment structure and formality, but struggle to accurately capture nuanced emotions, e.g. understating joy and overstating anger. These findings highlight a promising direction for LLMs in guiding online conversations towards prosociality influencing emergent communication patterns and norms within the community. The results of our study inspire future work with more rigorous methods of evaluation to explore the LLMs’ effectiveness across diverse online communities to better understand their broader societal impact.
[HC-32] AREAs-Lab: An Interactive Environment for AI-driven Requirement Elicitation for AI Systems EMNLP2026
链接: https://arxiv.org/abs/2608.28979
作者: Pengshan Cai,Zihao Zhang,Ting Jin,Chenyang Zhu,Kushal Chawla,Sangwoo Cho,Scott Novotney,Yebowen Hu,Fei Liu,Shi-Xiong Zhang,Sambit Sahu
类目: Human-Computer Interaction (cs.HC)
备注: Accepted to Findings of EMNLP 2026. 45 pages, 13 figures
Abstract:Building effective AI systems increasingly depends on writing high-quality task requirements, yet users often struggle to articulate the constraints, preferences, and edge cases that determine success. This problem is especially acute in AI development, where behavior is shaped not only by human expectations but also by data characteristics. We present AREAs-Lab, an interactive environment for AI-driven Requirement Elicitation for AI systems. In AREAs-Lab, an assistant iteratively refines an initially incomplete requirement by analyzing the underlying dataset and asking targeted clarification questions to uncover the user’s latent intent. To study this setting systematically, we construct a synthetic benchmark grounded in 16 public datasets spanning diverse domains and task types. Each benchmark instance includes a user profile, a complete reference requirement, and an intentionally underspecified version that serves as the assistant’s starting point. We further introduce an automated evaluation pipeline based on an AI-simulated user that reveals hidden information only when appropriately prompted, enabling scalable and reproducible assessment of interactive elicitation quality. AREAs-Lab provides a controlled testbed for studying how AI assistants can transform vague user goals into actionable requirements for AI systems.
[HC-33] he Web-CLI: Verifiable Privacy for Tools Models and Inference Engines in the Browser
链接: https://arxiv.org/abs/2608.28950
作者: Tejaswi Gowda
类目: Human-Computer Interaction (cs.HC); Cryptography and Security (cs.CR); Software Engineering (cs.SE)
备注:
Abstract:We introduce the Web-CLI, a novel application architecture deploying powerful computational capabilities (command-line tools compiled to WebAssembly, models run through client-side inference runtimes, and GPU-accelerated engines) as zero-install, offline-capable browser applications that preserve full underlying capability. Unlike web-based alternatives that require server-side processing and expose user data to third parties, Web-CLI applications execute entirely on the client, providing a verifiable privacy guarantee by architecture rather than policy. We define the pattern and its four properties: fidelity, progressive disclosure, offline-first, and zero egress. We present four reference implementations across distinct domains: ffmpeg-webCLI, a browser-based video editor built on FFmpeg; whisper-webCLI, speech transcription via this http URL; chat-webCLI, WebLLM-based language model inference; and 3mf-webCLI, a deterministic tool segmenting 3D models into multi-material files for physical 3D printing. Together they demonstrate that the pattern generalizes across deterministic media processing, neural speech recognition, LLM inference, and geometry processing with a physical output, and we outline how it extends to AI-native interfaces in which a local language model becomes the command surface itself. We further report early, anecdotal signs of independent reuse by third-party tools, suggesting the pattern generalizes beyond its reference implementations. We evaluate the primary implementation against native FFmpeg on performance and feature parity, and argue that progressive disclosure lowers the barrier for non-technical users. We argue that for applications processing sensitive user data (medical, legal, journalistic, or personal), the Web-CLI should be the default architecture, as it makes data locality an independently verifiable technical property rather than a policy promise.
[HC-34] Structured State Reconciliation for Human-AI Task Handover
链接: https://arxiv.org/abs/2608.28907
作者: Kayleigh Bishop,Maria P. Stull,Breanne Crockett,Bradley Hayes
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注: In preparation for conference submission
Abstract:Task handover requires communicating enough current state for a successor to resume work, yet the relevant information is often divided between system records and human observations. System records can be precise and timestamped but only partially observe the task, while human reports capture intent and task knowledge that no log contains but are vulnerable to omission and memory error. We present a provenance-aware pipeline that converts task telemetry and human-authored reports into a shared typed task-state representation, aligns and reconciles their facts, detects conflicts, and generates structured handover reports. We evaluate the approach on 13 paired task states collected in a controlled spatial multitask environment, using task-grounded metrics that estimate the state-reconstruction cost a report would spare a hypothetical recipient and the misinformation burden it would impose. Reconciling both sources preserved greater estimated task-state utility than either the user report or telemetry alone. Relative to a direct end-to-end LLM given the same inputs, structured reconciliation maintained comparable estimated utility while incurring substantially less misinformation, and task-aware rendering retained utility more efficiently (per token) than exhaustive rendering. An exploratory content analysis further shows that human reports contain substantial strategic knowledge that lies outside state-focused metrics. These results support provenance-aware state reconciliation as a design pattern for safer AI-assisted handover.
[HC-35] FlowCheck: Helping End-Users Specify and Verify Intent in Vibe-Coded Web Apps ISSTA2026
链接: https://arxiv.org/abs/2608.28880
作者: Reya Vir,Lydia Chilton,Zhuo Zhang,Eugene Wu
类目: oftware Engineering (cs.SE); Human-Computer Interaction (cs.HC); Programming Languages (cs.PL)
备注: Accepted at the 2nd ACM SIGPLAN International Workshop on Language Models and Programming Languages (LMPL '26), co-located with SPLASH/ISSTA 2026. 14 pages, 6 figures (10 pages main text, plus references and appendix)
Abstract:Vibe-coded applications often contain silent behavioral failures in which the interface appears functional even though user-visible information does not flow to the expected state or output. We introduce FlowCheck, a constraint language to specify these user-visible information flows directly through the application interface, where constraints can also be displayed and inspected without reading code, and are structured enough for reliable LLM generation. FlowCheck translates the constraints into deterministic CodeQL analyses, and we evaluate it across four applications generated via Claude Code, and compare with three coding models as bug-finding baselines. We find that FlowCheck correctly translates and flags all 30 of our injected constraint violations with no false positives. In contrast, frontier models (Claude Opus 4.7, DeepSeek V3, and Gemini Pro) showed significantly lower accuracy when prompted to find bugs in the same code, with none achieving full accuracy. This approach lets vibe coders state intent in terms of the interface they understand, and checks it deterministically against the code they do not.
[HC-36] oward Postural State Classification in Immersive VR with Multimodal Data and Explainability Analysis
链接: https://arxiv.org/abs/2608.28844
作者: Nipa Anjum,Md Irfan Pavel,Robert Gonzalez Jr,Kevin Desai,Alberto Cordova,M. Rasel Mahmud,John Quarles
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Accepted in The 25th IEEE International Symposium on Mixed and Augmented Reality (ISMAR)
Abstract:Ensuring a safe virtual reality (VR) experience requires systems that can predict and respond when users lose their balance. Although prior work has examined fall prediction and motion sickness, many approaches are regression-based and postural state classification remains less explored. This study compares machine learning (ML) and deep learning (DL) models for classifying postural states in VR under visual perturbations. We used a multimodal dataset containing kinematic, electromyographic (EMG), and electrodermal activity (EDA) signals. The data were prepared for a binary task to distinguish balanced from imbalanced postural states, and participant-wise downsampling addressed class imbalance. All models were evaluated with Leave-One-Participant-Out (LOPO) cross-validation to test generalization to unseen participants. Among the models, the Mamba-inspired CNN (MI-CNN) achieved the highest accuracy of 96.76%. SHapley Additive exPlanations (SHAP) analysis improved interpretability and identified the most influential classification factors. The SHAP results showed that kinematic features were dominant, indicating that body-motion patterns are informative for detecting imbalance in VR. We also evaluated MI-CNN using only the top two-thirds of features ranked by SHAP importance. Despite a 33% reduction in input dimensionality, the model maintained performance, achieving 0.957 accuracy and 0.957 F1-score, with about a 1% decrease compared with the full-feature model. These findings suggest that multimodal sensing, temporal deep learning, and explainable AI can support reliable classification of balance-related instability in VR. Accurate recognition of imbalanced postural states may raise awareness of fall risk and guide safer, adaptive VR systems that respond to instability while improving user safety and experience. Code is available at: this https URL.
[HC-37] Delegating Before Learning: Where Generative AI Sits in Students Professional Communication
链接: https://arxiv.org/abs/2608.28837
作者: Jared Ren,Soobin Cho
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注:
Abstract:We conducted an interview study with twelve students on their use of generative AI in academic communication. Students delegated professional messages to AI most where the pressure to sound professional is highest: email to instructors and administrators. AI involvement ranged from correcting the writer’s own text to working out and writing the message outright, and students checked AI-written text against two criteria: whether it looks like AI and whether it sounds like them. Building on these findings, we model the AI-mediated process of writing a student–instructor email at the highest level of involvement we observed, and compare it with an unaided model of writing the same messages, built from participants’ accounts and a classic model of the writing process. Three differences emerge: the learning loop that builds writing skill is removed, the message is no longer written for its specific recipient, and the confidence a successful exchange returns goes to using the system rather than to the writer’s own ability. From these differences we derive two risks, that individual capacities never form and that authenticity and trust in communication become work. Design can respond to both but is unlikely to be enough, so the risks also need research and policy attention.
[HC-38] Visible but Not Yet Curatable: Characterizing the Curatability of Compact and Derived Open LLM Artifacts
链接: https://arxiv.org/abs/2608.28819
作者: Yiyi Lu,Yilai Qian,Yucheng Jin
类目: Human-Computer Interaction (cs.HC); Digital Libraries (cs.DL)
备注: 11 pages, 5 figures. Accepted at the ACM/IEEE Joint Conference on Digital Libraries (JCDL 2026). Yiyi Lu and Yilai Qian contributed equally; Yucheng Jin is the corresponding author. Code and results: this https URL
Abstract:Open Large Language Model (LLM) research increasingly produces compact and derived artifacts, such as adapters, quantized checkpoints, merged models, and distilled variants, that are distributed across papers, model hubs, model cards, code repositories, and release statements. Although these artifacts are publicly visible, digital libraries often lack sufficient evidence to identify, preserve, and cite them as coherent scholarly objects. We introduce a framework that conceptualizes curatability as a record-level property of distributed scholarly records and operationalizes it through four evidence dimensions: artifact identity, scholarly linkage, upstream evidence, and release assets. Guided by this framework, we conduct the first collection-scale characterization of open LLM curatability using a May 2026 snapshot of 191,375 public Hugging Face repositories and a core corpus of 2,214 scholarly papers. Our results reveal a pronounced visibility-to-curatability funnel. While 90.7% of paper records contain at least one useful curation signal, only 18.1% combine usable upstream evidence with concrete release evidence, and only 6.1% provide sufficiently coordinated evidence to support high-curatability records. Based on these findings, we derive a minimal seven-field curatable record and complementary responsibilities for model hubs, scholarly indexes, and digital libraries, providing practical guidance for improving the preservation and bibliographic control of open LLM artifacts.
[HC-39] Bringing Data to Life: Designing Data Characters for the Emotional Self IEEE-VIS
链接: https://arxiv.org/abs/2608.28780
作者: Diego Abarcar Calugay,Isabella Amador,Keke Wu
类目: Human-Computer Interaction (cs.HC)
备注: IEEE VIS Poster and Summary (2026)
Abstract:Journaling is a common practice for emotional expression, reflection, and processing. However, as entries accumulate, it can become difficult to interpret and compare their affective content, especially since traditional text-based analyses and visualizations often struggle to convey affective nuance. We introduce Data Characters, a visualization approach that represents affective content in journaling through human-like characters. Using a customizable Data Character as a design probe, we investigate the potential of character-based representations for conveying affective experiences and explore what visual encodings emerge through customization. Preliminary walkthroughs with two participants demonstrate the intuitiveness and feasibility of the approach. This work contributes an exploratory approach to studying how affective experiences can be visually represented and encoded through anthropomorphic forms.
[HC-40] A Conceptual Framework for Modeling Team Adaptation in Cooperative Games Through Ludic Knowledge
链接: https://arxiv.org/abs/2608.28729
作者: Caleb Vatral
类目: Human-Computer Interaction (cs.HC); Multimedia (cs.MM)
备注: 10 pages, 4 figures, To be presented at AHFE: Hawaii International Conference 2026
Abstract:With the increasing importance of teamwork skills for modern workplaces, development of teamwork training programs has received substantial attention. Game-based teamwork training is one promising approach that is engaging, cost-effective, and well-suited to increasingly decentralized workplaces. However, design of effective game-based teamwork training requires understanding how a game elicits specific desired teamwork behaviors. Significant progress has been made in characterizing these relationships. However, despite its critical importance, little work has examined how a game’s design influences team adaptability behaviors. This paper presents a preliminary framework for analyzing adaptability in cooperative games by conceptualizing adaptive stimuli as retrieval or disruption of players’ ludic knowledge. We illustrate this framework through a qualitative case study that applies interaction analysis methods to gameplay videos of a Overcooked!, a cooperative cooking game. We examined instances where game events led to players altering their behavior and connected the game’s design features that resulted in each event with three adaptive stimulus cue categories. Although exploratory and limited to a small case study of a single game, the proposed framework is grounded in established theories across teamwork research and game studies, and it offers an initial vocabulary for describing how cooperative games can be designed to create demands for team adaptation. With this continued development, the framework may provide an analytic tool to help inform the design and evaluation of purpose-built game-based teamwork training environments.
[HC-41] Experts Disagree on How to Fight AI Disinformation but Agree That Health and Politics Need Different Solutions
链接: https://arxiv.org/abs/2608.28621
作者: Alexander Loth,Martin Kappes,Marc-Oliver Pahl
类目: Computers and Society (cs.CY); Human-Computer Interaction (cs.HC)
备注: 16 pages, 5 figures. Author-formatted version; identical in content to the version published in Harvard Kennedy School Misinformation Review 7(4), 2026, CC BY 4.0. Replication materials: doi: https://doi.org/10.7910/DVN/BXO2QA
Abstract:When 54 international experts assessed AI-generated disinformation threats, they revealed a surprising pattern: while video deepfakes received the highest average threat ratings in the political domain (M = 6.31/7), the pattern differed in the health domain, where AI-generated text received the highest average rating (M = 5.80). Experts also diverge on what to do: government regulation drew both the most “most effective” (30%) and the most “least effective” (15%) votes, though rating distributions were contested rather than polarized, indicating disagreement over priorities rather than over efficacy. These findings offer an initial expert map of an AI-disinformation landscape that is still rapidly forming.
[HC-42] From GenAI Virtual Patient Dialogue Logs to Teacher-Interpretable Process Evidence: A Learning Analytics Study in Higher Education
链接: https://arxiv.org/abs/2608.28619
作者: Xinyu Li,Zijian Li,Mengyu Xia,Luzhen Tang,Naping Chen,Changmin Lin,Danijela Gasevic,Dragan Gasevic,Yizhou Fan
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注:
Abstract:Medical history taking is a dialogue-based clinical reasoning task in which learners must gather, organise, and integrate patient information while the consultation unfolds. Generative AI-powered virtual patients (GenAI VPs) make repeated history taking practice scalable and preserve full turn by turn dialogue. However, these logs are educationally difficult to use directly. Complete transcripts are too detailed for routine teacher review, whereas final scores obscure whether learners followed up patient cues, checked uncertainty, or used summaries to guide later questioning. This study examined whether coded GenAI VP dialogues can provide teacher-interpretable process evidence of clinical reasoning. We analysed 1,030 GenAI VP dialogues from 210 second-year medical learners across five weeks chest-pain cases. Each consultation was teacher-scored using a rubric assessing the full history taking dialogue, and consultations were classified within each week as high- or low-rated using the weekly median score. To explain how rated performance was reflected in the dialogue process, we applied three analytic layers to the same coded dialogue data: behavioural prevalence, local co-occurrence using Epistemic Network Analysis, and sequential transition using Transition Network Analysis. High-rated consultations involved more history taking activity, but differences were not simply about volume. High rated consultations more often connected information gathering and symptom exploration with communication, checking, organisation, and synthesis. Summarising and organising moves more often led to verification or mechanism-oriented follow-up. These findings show how layered analysis of GenAI VP dialogue logs can reveal process patterns associated with high rated history taking and support process-focused feedback in medical education.
计算机视觉
[CV-0] BRF-GS: Hyperspectral Bidirectional Reflectance Factor Modeling and Image Generation Based on 3D Gaussian Splatting
链接: https://arxiv.org/abs/2608.31159
作者: Yiling Yao,Wenjuan Zhang,Bowen Wang,Bocheng Li,Wentao Song,Bing Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 58 pages, 10 figures, 4 tables
Abstract:The bidirectional reflectance factor (BRF) characterizes the directional radiative properties of terrestrial surfaces. However, existing three-dimensional (3D) radiative transfer models require complex scene construction and computationally intensive radiative transfer solvers, limiting efficient generation of multi-angle hyperspectral reflectance imagery. 3D Gaussian Splatting (3DGS) offers an efficient framework for neural scene representation and novel view synthesis, but its low-order spherical harmonics representation is insufficient for complex directional reflectance, while the high dimensionality and inter-band quality differences of hyperspectral data introduce additional challenges. To address these challenges, we propose BRF-GS, a 3DGS-based framework for BRF modeling and hyperspectral reflectance image generation. BRF-GS introduces a hybrid BRDF-driven kernel to represent complex directional reflectance, selects geometry-reliable spectral bands for robust 3D scene initialization, and adopts a two-stage training strategy that decouples geometry optimization from spectral modeling. We further construct the AIR-BRF dataset, a multi-angle hyperspectral directional reflectance dataset comprising three scenes with diverse natural and artificial targets. Experiments demonstrate that BRF-GS achieves superior spatial and spectral fidelity and accurately reproduces characteristic view-dependent BRF responses. The proposed framework provides an efficient data-driven approach for BRF modeling and multi-angle hyperspectral reflectance image generation in remote sensing scenes.
[CV-1] BLARM: Animating 3D Objects from Video via Blending Latent Rigid Motion Primitives
链接: https://arxiv.org/abs/2608.31113
作者: Pradyumn Goyal,Yizhak Ben-Shabat,Hsueh-Ti Derek Liu,Haomiao Jiang,Snehasish Mukherjee,Kyle Spence,Mark Stauber,Evangelos Kalogerakis,Yunze Zeng
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:We introduce BLARM, a feed-forward method for video-driven 3D mesh animation. Given a monocular video and a static object mesh, BLARM predicts a temporally coherent animated mesh whose motion follows the video. Rather than relying on explicit rigs or directly regressing high-dimensional vertex motion, we represent animation using a compact set of learned, time-varying rigid motion components and time-invariant vertex-to-component skinning weights. This yields a low-dimensional deformation space without requiring skeletons, cages, skinning weights, or rig annotations. Our architecture conditions geometry-derived deformation latents on video features through factorized spatial-temporal attention, then decodes rigid transformations blended by predicted skinning weights. Trained with trajectory reconstruction, entropy regularization, and motion-aware contrastive learning, BLARM produces accurate and temporally stable animations while recovering compact, interpretable motion structure from monocular video.
[CV-2] VeriCam: A Verification Baseline for the Classification of Unknown Data
链接: https://arxiv.org/abs/2608.31107
作者: Lucas Wojcik,Gabriel E. Lima,Sergio M. Silva Jr.,Eduil Nascimento Jr.,David Menotti
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: SIBGRAPI WIP 2026
Abstract:The advent of foundation models have enabled a new era in zero-shot classification. Yet, key challenges persist. Despite their impressive generalization power that leverages the immense pre-training knowledge, both foundation models for image and text as well as vision-text hybrids lack the representational power needed for fine-grained, minutiae-based class separation that some real-world tasks require. To address the current gaps in the literature, we propose VeriCam, a pipeline designed to learn highly specialized features that enable classification of unknown classes in unseen data. VeriCam works by leveraging the representation power of image models trained for the verification task, where the model develops an intricate feature space that incorporates fine-grained details. By training a model to discriminate between pairs of images from the same and different classes, a relational graph is constructed, representing the class relationships between data points. We then present two approaches for graph clustering: a naive algorithm and a specific setup for the Leiden graph clustering algorithm. The pipeline is validated on the LPLCv2 dataset, which comprises real-world traffic surveillance images. We show that the dataset carries an inherent capture device bias that is posed as a generalization challenge for downstream License Plate recognition tasks such as OCR. As such, we dynamically identify capture devices with a label-agnostic approach, enabling the construction of a fair and unbiased benchmark. In the cross-device scenario, our pipeline reaches an F1-Score of 93.45 in the verification baseline and a V-Measure score of 80.13 in the clustering step. All code is publicly available at this https URL
[CV-3] DreamX-Creator: Democratizing Native Audio-Video Generation at 2K Resolution
链接: https://arxiv.org/abs/2608.31106
作者: Jiashu Zhu,Yanhao Zheng,Ruitian Tian,Rujing Dang,Shen Zhang,Bingze Song,Jiachen Lei,Ruimin Lin,Jiahong Wu,Xiangxiang Chu
类目: Computer Vision and Pattern Recognition (cs.CV); Sound (cs.SD)
备注:
Abstract:Recent video generators often omit audio or synthesize it in a separate stage, limiting reciprocal modeling of visual dynamics and acoustic events. We present DreamX-Creator 1.0, a compact native joint audio-video generation system centered on a 7B generator. Conditioned on a first frame and a text prompt, the generator jointly denoises modality-specialized audio and video streams. The streams are processed independently in the first half of the network and coupled in the latter half through Gated Cross-Modal Attention, whose token- and head-wise output gates modulate each active cross-modal attention-head output. A unified Audio-Video Data System constructs and filters temporally coherent clips, produces structured multimodal annotations, and organizes clips into capability-oriented data pools. Progressive Joint Training comprises two audio-video pre-training stages followed by High-Quality Finetuning. Audio-Video Reinforcement Learning further post-trains the generator with Modality-Aware Multimodal Feedback that routes video-, audio-, and cross-modal feedback to the corresponding streams. For high-resolution output, our Autoregressive 1-Step 2K Refinement pipeline adapts a bidirectional multi-step teacher into an autoregressive multi-step refiner and distills it into a student requiring one denoising evaluation per temporal chunk. Overall, DreamX-Creator 1.0 achieves native, synchronized audio-video generation with performance competitive with state-of-the-art open-source systems. By releasing our compact 7B generator and 2K Refiner, we seek to democratize native audio-video generation and provide an accessible foundation for future research in unified audio-video generative modeling.
[CV-4] One Adapter Many Tasks: Task-Conditioned Feature Transformations for Continual Learning
链接: https://arxiv.org/abs/2608.31096
作者: Yunxiang Fu,Meng Lou,Yizhou Yu
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Class-incremental learning (CIL) requires a model to incrementally learn tasks that contain new classes without accessing earlier training data while preserving the ability to recognize all seen classes. Recently, pretrained-model-based approaches have become prevalent by adapting a frozen backbone with additional lightweight trainable modules. Existing methods, however, exhibit limitations: task-specific adapters learn explicit per-task representations but are parameter- and computation-inefficient, while LoRA-based merging methods combine per-task LoRA parameters into a single model whose static aggregated weights cause representation interference during inference. To address these problems, we present \textbfFACET: task-conditioned \textbfFe\textbfAture transformation with \textbfCondition\textbfEd feature consis\textbfTency, achieving excellent parameter efficiency while producing highly discriminative features during inference. When continually trained on a task sequence, FACET learns a single shared adapter that employs a dynamic task-conditioned feature transformation, shaping the overall feature distribution of the adapter into a mixture of overlap-reduced task-specific components. On the other hand, we propose an efficient replay-free task-conditioned feature consistency loss, aiming to mitigate catastrophic forgetting of the learned mixture distribution in the adapter’s feature space. Even when maintaining only a single adapter, FACET demonstrates robust scalability. On both very long task sequences (e.g., 200 tasks) and standard short task sequences (e.g., 20 tasks), our method achieves superior performance while using significantly fewer trainable parameters and GFLOPs. The code will be made open source upon acceptance.
[CV-5] Robust retinal biometrics for patient identity verification and retrieval across age and imaging devices
链接: https://arxiv.org/abs/2608.31094
作者: Jose D. Vargas-Quiros,Dennis Bontempi,Jeroen Vermeulen,Bart Liefers,Sven Bergmann,Caroline C.W. Klaver
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Patient identity errors can compromise longitudinal medical records, research databases, and downstream clinical decisions. We present a retinal biometric system for verifying claimed identities and retrieving the correct identity from color fundus images. We trained a 512-dimensional metric-learning encoder combining a ConvNeXtV2 backbone with ArcFace and triplet losses on 227,004 images from 21,851 patient-eye identities in the Rotterdam Study, spanning multiple imaging devices and up to 32.6 years of follow-up. The system was evaluated on held-out Rotterdam Study data and externally on the UK Biobank and Age-Related Eye Disease Study (AREDS). Before evaluation, we used the model to screen for identity inconsistencies and manually adjudicated flagged images, identifying incorrect assignments in 0.588% of Rotterdam Study images, 0.259% of UK Biobank images, and 0.164% of AREDS images. In retrospective-only verification after removing near-duplicate images, the system achieved AUROCs of 0.9998, 0.9997, and 0.9998 in the Rotterdam Study, UK Biobank, and AREDS, respectively. For identity retrieval using only previously acquired images, Recall@1 was 99.7%, 97.2%, and 97.6%, respectively, from galleries averaging 4436-8510 identities; the correct identity appeared among the top five results in at least 98.6% of cases. Performance remained robust across imaging devices and long follow-up intervals, while lower image quality and inconsistent retinal fields accounted for most failures. These findings establish retinal anatomy as a durable biometric signal, useful for safeguarding the integrity of longitudinal imaging records.
[CV-6] Real-Time Video Anomaly Detection Using YOLO Pose Estimation and CLIP-Based Semantic Scoring
链接: https://arxiv.org/abs/2608.31074
作者: Vanodhya G. Warnasooriya,Amir Hajian,Watchara Ruangsang,Supavadee Aramvith
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Image and Video Processing (eess.IV)
备注: 5 pages, 2 figures, 3 tables. Presented at the 9th IEEE International Conference on Multimedia Information Processing and Retrieval (MIPR 2026); accepted for publication in IEEE Xplore
Abstract:We propose a lightweight two-stage framework for real-time video anomaly detection. The first stage employs YOLO v11n-pose to detect persons and extract seventeen skeletal keypoints in a single forward pass. The second stage encodes each cropped person region through CLIP ViT-B/32 and computes cosine similarity against predefined textual descriptions of anomalous behaviors. This architecture eliminates the need for optical flow, standalone pose estimators, and density-based scoring modules. Experiments on CUHK Avenue, ShanghaiTech Campus, and a custom indoor dataset collected at Chulalongkorn University demonstrate an end-to-end throughput of approximately 51 FPS on an NVIDIA Titan XP GPU, a 3.36x speedup over the multi-feature baseline, while maintaining frame-level AUROC values of 89.26%, 70.26%, and 84.13%, respectively.
[CV-7] LISynSeg: Data-Centric Label-to-Image Synthesis for Cross-Modality Whole-Heart Segmentation
链接: https://arxiv.org/abs/2608.31073
作者: Jiacheng Wang,Ivana Isgum,Ipek Oguz
类目: Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
备注: 12 pages, 4 figures
Abstract:Whole-heart segmentation (WHS) in computed tomography (CT) and magnetic resonance imaging (MRI) is affected by acquisition shifts and heterogeneous cardiac annotations. Existing WHS systems combine architectural design, transfer learning, and generic spatial or intensity augmentation. We investigate whether changes to data augmentation and training supervision can improve cross-modality WHS while the segmentation architecture is held constant. We present LISynSeg, a data-centric approach that augments real-image nnU-Net training with label-to-image synthesis. Synthetic volumes are generated from cardiac label maps using contrast and acquisition perturbations calibrated to the training cohort, then mixed with real images to retain thoracic context absent from the labels (and thus the synthesized images). We model cardiac label variation through controlled changes in myocardial wall thickness and partial supervision of uncertain vessel endpoints. On the CARE Whole-Heart benchmark, synthetic-only training performs worse than the real-image nnU-Net baseline, whereas calibrated real-synthetic training improves cross-modality segmentation without changing the architecture; the improvement is larger for MRI than for CT. The results show that modifying the training data strategy can benefit model development for heterogeneous cardiac data. Code and trained weights will be released at this https URL.
[CV-8] Multimodal Shared Latent Representation of Narration Microscope and iOCT Images for Phase Recognition in Vitreoretinal Surgery
链接: https://arxiv.org/abs/2608.31065
作者: Onur Izmitlioglu,Shervin Dehghani,Tarek Ghannoum,Benedikt Schworm,Nassir Navab
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Surgical phase recognition is key to context-aware computer-assisted feedback in vitreoretinal procedures, yet the scarcity of synchronized multimodal intraoperative data, particularly microscope views and intraoperative OCT, limits approaches that aim to replicate the multimodal integration surgeons perform naturally. Surgical narration, by contrast, is abundantly available online and offers rich semantic supervision. Prior work has mainly explored pairwise contrastive learning (e.g., intraoperative OCT-microscope or microscope-narration), leaving the joint modeling of all three modalities largely unexplored. We introduce a framework that uses microscope views as a shared anchor to bridge surgical narrations and intraoperative OCT (iOCT) without requiring a fully synchronized tri-modal dataset, leveraging real microscope-narration videos and a synthetic dataset of synchronized microscope video and tool-aligned iOCT pairs. Contrastive alignment transfers structural priors from the synthetic domain to real videos lacking iOCT, and a dual-head MS-TCN++ integrates the resulting embeddings for joint macro- and micro-phase prediction. Evaluated on real vitreoretinal surgeries, our framework improves macro-phase recognition over a zero-shot baseline (mean F1 0.38 to 0.53) and provides an exploratory route to estimating fine-grained instrument-tissue measurements that are not directly observable in real microscope video alone; these micro-phase estimates are validated quantitatively on synthetic data and shown only qualitatively on real surgery. To our knowledge, this is the first work to unify microscope view, iOCT B-scans, and surgical narrations in a shared latent space for surgical phase recognition.
[CV-9] Identity-Conditioned Latent Consistency Distillation for Face Synthesis
链接: https://arxiv.org/abs/2608.31053
作者: Tiago Kienen Chaves,Bernardo Biesseck,David Menotti
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted for presentation at the 2026 Conference on Graphics, Patterns and Images (SIBGRAPI)
Abstract:Diffusion models have achieved strong results in high-fidelity image synthesis, but their iterative sampling process makes large-scale generation computationally expensive. This limitation is especially relevant when generating synthetic face datasets for face recognition, where a large number of subjects with many samples in different poses, expressions, ages, etc., are required. In this work, we show that identity-conditioned face synthesis can be performed at a substantially lower computational cost by a latent Consistency Model with few iterations, without compromising image quality. For training, we distill knowledge from the foundation Diffusion Model Arc2Face (teacher) by adapting its original text-to-image pipeline to an embedding-to-face setting, replacing textual prompts with ArcFace identity embeddings. Our distilled model (student) generates identity-conditioned face images with an average inference time of 0.4819 seconds per image, compared with 2.102 seconds for Arc2Face, resulting in a 4.36 \times speed-up. Quantitative results, based on FID scores, show that the distilled model remains competitive with Arc2Face across all evaluation protocols. On 100k generated images, it achieves near-parity on CelebA (13.921 vs. 12.928) and outperforms the teacher on WebFace42M (9.317 vs. 9.802). Further evaluations on Synth-500 and AgeDB show a moderate performance gap for the former but comparable results for the latter. These results indicate that Arc2Face can be accelerated through task-specific latent consistency distillation while preserving high image quality for large-scale synthetic face generation. Our proposal is publicly available at this https URL.
[CV-10] Segmentation of Bovid Dentition Under Imperfect Annotations: A Comparative Study of Convolutional and Attention Models
链接: https://arxiv.org/abs/2608.31052
作者: Keith G. Mills,Evan B. Sanders,Gregory J. Matthews,Juliet K. Brophy
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 13 pages, 2 Tables, 13 Figures
Abstract:Semantic segmentation decomposes an image into distinct mask regions corresponding to different object categories, such as people, cars, signs or buildings. Advances in machine learning (ML) have shifted this task away from traditional rule-based heuristics such as edge detection, towards deep neural networks (DNN) that learn to classify pixels directly. However, semantic segmentation DNNs crucially depend on expertly designed mask targets to learn from, and imperfect or misaligned masks can interfere with a model’s ability to learn effectively. This paper presents a comparative study of segmentation architectures, ranging from convolutional backbones to vision transformers, applied to the B.O.V.I.D. dataset, a corpus of high-resolution bovid dental photographs paired with hand-made segmentation masks not originally designed for ML-based training. We evaluate a range of preprocessing and alignment techniques to mitigate the resulting label imperfections. We find that while these preprocessing choices have limited effect on quantitative metrics such as Dice score and mIoU, their qualitative impact on predicted masks is substantial. Comments: 13 pages, 2 Tables, 13 Figures Subjects: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG) Cite as: arXiv:2608.31052 [cs.CV] (or arXiv:2608.31052v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2608.31052 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-11] FaceSnap: Real-Time Personalized Lightstage Facial Performance Capture
链接: https://arxiv.org/abs/2608.31033
作者: Rukhshanda Hussain,Noé Artru,Emeline Got,Luiz Gustavo Hafemann,Alexandre Messier,Brandon Dearlove,Rafael M. O. Cruz,Abdallah Dib,Eric Granger
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Lightstage facial capture produces production-quality digital humans, but it is resource and labor-intensive. Multi-camera setups, hours of computation, and massive data storage create bottlenecks that hinder iterative workflows. This paper introduces FaceSnap, an end-to-end framework that streamlines capture via a two-stage approach. First, a one-time multi-view optimization from a range-of-motion sequence builds a personalized model encoding both geometry and expression-dependent appearance. This model then enables high-fidelity real-time facial performance capture from a single monocular lightstage camera, with no further multi-view capture required. FaceSnap jointly estimates geometry and dynamic 4K texture at 83 fps. The 4K texture is produced by a novel personalized residual upscaler that recovers subject-specific high-frequency detail, which generic upscalers fail to capture. FaceSnap achieves geometric accuracy competitive with full per-frame multi-view optimization while outperforming feed-forward methods trained on production-quality 3D data, all from a single camera view. Finally, we introduce Multi4D, a public benchmark for evaluating 4D facial reconstruction methods in lightstage environments, enabling topology-invariant geometric comparison across methods.
[CV-12] Driving on Memory
链接: https://arxiv.org/abs/2608.31029
作者: Christian Löwens,Thorben Funke,Alexandru Paul Condurache
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Robotics (cs.RO)
备注:
Abstract:End-to-end autonomous driving models plan future trajectories from raw sensor input. While earlier driving benchmarks often measured deviation from the human trajectory, current benchmarks such as NAVSIM and Bench2Drive evaluate models with richer simulation-based metrics intended to capture safe and compliant driving. A high benchmark score should reflect that a model can understand the scene in front of it and act accordingly. But how much of that score specifically comes from reacting to the dynamic part of that scene? To probe this, we remove a model’s camera input and replace it with memories from prior drives at the same location. The retrieved memories can provide persistent scene information, including road layout and location-conditioned regularities, but not the current traffic state. Surprisingly, memory is nearly sufficient on NAVSIM, reaching or even exceeding the performance of leading end-to-end methods without actually observing the evaluated scene. Our results suggest that a high NAVSIM score does not require a planner to react to the current traffic scene and should be treated with caution. This effect is benchmark-dependent: driving from memory causes substantially larger performance drops on Bench2Drive and RealEngine. We provide our code at this https URL . Subjects: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Robotics (cs.RO) Cite as: arXiv:2608.31029 [cs.CV] (or arXiv:2608.31029v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2608.31029 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-13] Analytic Dynamics: Learning Physics-Grounded Representation for Fast Intrinsic Dynamics Inference from Monocular Videos
链接: https://arxiv.org/abs/2608.31025
作者: Jailing Lin,Jikuan Zhang,Jianhua Sun
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Inferring object dynamics from visual observations is essential for intelligent agents to reason about and interact with the physical world, yet remains challenging due to the fundamental gap between visual evidence and intrinsic dynamics. Existing methods either rely on costly per-scene optimization, limiting efficiency and scalability, or directly map visual evidence to intrinsic dynamics without intermediate physical abstractions, making them prone to appearance and geometry shortcuts. To bridge this gap, we propose Analytic Dynamics, a feed-forward dynamics inference framework that introduces an intermediate physics-grounded dynamics representation between visual observations and intrinsic dynamics. Specifically, we leverage privileged physical states, including position, displacement, and deformation gradient fields, which are available in simulation, to learn a structured dynamics representation that is difficult to discover from visual observations alone. By aligning visual representations with this space, we equip visual models with a physics-grounded inductive bias, guiding them to capture dynamics-relevant patterns for material model classification and parameter regression. To facilitate this research, we develop a dynamics data generation pipeline and benchmark containing paired physical state trajectories, rendered videos, and ground-truth material models and parameters. Extensive experiments demonstrate that Analytic Dynamics achieves efficient, accurate, and generalizable dynamics inference from monocular videos.
[CV-14] SMG: Semantic Motion Graph for Monocular Dynamic Gaussian Splatting ECCV2026
链接: https://arxiv.org/abs/2608.31023
作者: Haozheng Yu,Xinyu Yang,Rundong Luo,Jennifer J. Sun,Bharath Hariharan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: ECCV 2026
Abstract:We study dynamic Gaussian Splatting from monocular videos. While recent advancements in dynamic Gaussian splatting offer a promising foundation for modeling dynamic scenes, they often overfit to the training views and fail under occlusion or complex scene motion due to the lack of reliable regularization signals in under-constrained regions. We propose Semantic Motion Graph (SMG), a novel approach models the Gaussian motion as the low-rank semantic motion. Our key insight is that the real-world scene motion is often structured by semantic coherence: regions that are spatially close and semantically related tend to exhibit consistent dynamics. To leverage this prior, we construct SMG to model structured motion of the scene. The Gaussian motion is driven by the motion of SMG nodes. We further observe that the uncertainty of Gaussian motion arises from both unreliable off-the-shelf priors and weakly constrained regions during optimization. SMG addresses this by using reliable graph nodes to guide the motion of nearby unreliable nodes. To evaluate dynamic Gaussian splatting under challenging real-world scenarios, we introduce a new multiview dataset collected under an ego-exo setup. Extensive experiments demonstrate that SMG achieves state-of-the-art performance on monocular dynamic Gaussian splatting across challenging real-world benchmarks. Project page: this https URL.
[CV-15] MNIST-PRO: MNIST is Back as a Partially Observable World for AI Agents
链接: https://arxiv.org/abs/2608.31022
作者: Vernon Toh,Navonil Majumder,Zhengyuan Liu,Nancy F. Chen,Soujanya Poria
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:AI agents in partially observable environments need to coordinate active sensing with working memory to maintain an evolving perceptual state. However, existing benchmarks struggle to isolate this perceptual-state construction and interpretation capability because they introduce physical and control complexities. We address this with MNIST-PRO, a benchmark that isolates agentic perception by converting MNIST digit recognition into a sequential, glimpse-based search task with lookback constraints. We evaluate ten multimodal models across four memory representations, including raw visual history, textual states, structured metric grid maps, and a consolidated visual canvas. While models excel under full observability, partial observability exposes a clear performance gap. We identify three distinct bottlenecks. First, perceptual-state construction and interpretation present a challenge, as agents struggle to integrate fragmented glimpses. Second, agents often stop exploring before they see the full sequence. Third, models often fail to revise early, incorrect beliefs even when faced with subsequent contradictory evidence. These results show that simply acquiring visual evidence is not enough. Agents must also be able to build and update a reliable perceptual state.
[CV-16] SPFN: A Temporal Tabular Foundation Model for Physiological Time Series Classification MICCAI
链接: https://arxiv.org/abs/2608.31013
作者: Jérémie Stym-Popper,Clément Rambour,Federica Granese,Nicolas Thome,Olivier Bernard
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at STACOM 2026 (MICCAI Workshop). 10 pages, 3 figures
Abstract:Designing models that generalize effectively in low- to medium-data regimes remains a primary challenge in medical machine learning, particularly for physiological time-series classification. While tabular foundation models such as TabPFN offer an attractive alternative to conventional fine-tuning through in-context learning, they are not designed to capture the temporal dependencies inherent to physiological signals. ~In this paper, we introduce TSPFN, a foundation model that redesigns TabPFN’s architecture for time series data. TSPFN integrates structured temporal representations and positional embeddings to capture intra-sample temporal and channel dependencies. To fully leverage its spatio-temporal design, the model is pretrained on 140,000 real-world physiological time series across multiple medical domains. This yields a unified, generalizable framework capable of learning the specificities of medical time series. Experiments across diverse physiological benchmarks demonstrate that TSPFN consistently outperforms standard tabular baselines and TabPFN, and achieves superior cross-domain generalization compared to specialized deep time-series models. All our experiments, ablation studies, and pre-processing scheme are publicly available at this https URL
[CV-17] From Intent to Evidence: Policy-Steered Multi-Strategy Retrieval for Long-Video Agents
链接: https://arxiv.org/abs/2608.31005
作者: Can Zhang,Baofeng Zhang,Xiaotian Han,Junyuan Shang,Yuchen Ding,Shuohuan Wang,Dianhai Yu,Ruirui Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 15 pages, 5 figures, 6 tables (main paper with appendix)
Abstract:Existing long-video agents acquire evidence through one uniform behavior, ignoring whether the required evidence is concentrated, requires broad occurrence coverage, or must discriminate competing hypotheses—which can cause failure before substantive reasoning begins. Prescribing a fine-grained solution procedure for every question is not a satisfactory remedy, as it restricts autonomous exploration. We propose VESTA, a training-free long-video agent organized as a route-conditioned acquire–verify–consolidate loop. Before exploration, an intent router infers an evidence-acquisition policy—focused, recall, or contrastive retrieval over a shared visual–speech scene index—together with an evidence-accounting policy that configures the evidence view maintained during exploration. Policy-steered retrieval yields provisional references that multimodal evidence operations convert into observations, while the Reasoner remains free to verify them, re-query using intermediate findings, or inspect regions outside the retrieved set. A temporal evidence ledger consolidates observations into an adaptive, compressed view of temporal location, provenance, coverage, conflicts, verification outcomes, and hypothesis support, exposing missing and unresolved evidence to guide subsequent acquisition; finalization prioritizes verified observations. On Video-MME-v2, VESTA improves average accuracy by 2.7 points over VideoARM and gains across all six reported metrics. On LongVideoBench, EgoSchema, and LVBench under shared query-time models, it improves by 6.9 points on the LongVideoBench long subset and 1.5 on LVBench, and matches VideoARM on EgoSchema.
[CV-18] DARP: A Calibrated Dual-Arm RGB-D-IR Dataset for Multi-View Robotic Perception
链接: https://arxiv.org/abs/2608.31002
作者: Manish Kansana,Mohammed Yusuf Mujawar,Sudip Mittal,Shahram Rahimi,Noorbakhsh Amiri Golilarz
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Robotic perception from a single viewpoint is often limited by self-occlusion and incomplete surface visibility. This paper presents DARP(Dual-Arm Robotic Perception) this https URL, a calibrated dual-arm RGB-D-IR dataset for object-centered robotic perception using two independently moving eye-in-hand manipulators positioned on opposite sides of a shared tabletop workspace. Each arm carries an Intel RealSense sensor that continuously records RGB, depth, and stereo infrared data while synchronized robot joint states are logged for pose recovery. Objects are placed without fixed poses or marked locations, and the acquisition procedure performs automatic localization, cross-arm confirmation, adaptive viewpoint generation, and continuous multimodal recording. DARP contains ten unique tabletop objects and preserves the original sensor recordings, robot-state logs, object-level metadata, and calibration information required to reconstruct camera trajectories in a shared metric frame. To evaluate the geometric consistency of the acquisition, we implement a deterministic multi-view fusion pipeline that converts calibrated RGB-D observations into complementary partial point clouds and measured surface meshes without using learned or generative completion methods. Evaluation on 224 held-out RGB-D keyframes comprising 1,563,466 three-dimensional query points yields a median point-to-mesh distance of 2.13~mm and an RMSE of 4.04~mm, with 96.56% of points within 10~mm of the measured-surface mesh. DARP is intended as a reusable resource for multi-view reconstruction, collaborative robotic perception, multimodal fusion, active perception, and future learning-based reasoning over partial object observations.
[CV-19] Multi-View Reflective Surface Inspection via Semantic-Saliency Cross-Verification
链接: https://arxiv.org/abs/2608.30997
作者: Van-Giang Nguyen,Thanh-Tuan Tran,Xuan-Hieu Phan,Xiem HoangVan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Submitted to RIVF 2026
Abstract:Reflective smartphone cover glass is challenging to inspect from a single fixed viewpoint because defect visibility varies with viewing geometry and specular reflections. This gives rise to two practical challenges: defects may be weakly observable from certain viewpoints, while the available visual evidence may remain spatially ambiguous. To address these issues, we propose a multi-view inspection framework in which each RGB observation is processed by a shared per-view expert. A vision-language model (VLM) produces class-aware semantic boxes, while a normal-reference reconstruction branch provides class-agnostic saliency. Their spatial agreement is used as supporting evidence to rank semantic proposals without modifying their coordinates or treating saliency as ground truth. The resulting evidence records are combined at product level without cross-view registration. On 282 production-line images, semantic-saliency association improves AP_50 from 52.6% to 62.6% by re-ranking fixed semantic proposals. Across 94 products, cross-view evidence recall R_\rm prod@0.5 increases from 75.5% for the best single view to 88.3% using all three views. These results support the complementary roles of semantic-saliency cross-verification and additional optical observations in reflective-surface inspection.
[CV-20] MR-JEPA: A General Purpose Video Foundation Model for Cardiac MRI MICCAI2026
链接: https://arxiv.org/abs/2608.30975
作者: Athira J. Jacob,Puneet Sharma,Dorin Comaniciu,Daniel Rueckert
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted at STACOM 2026 (MICCAI 2026 peer-reviewed workshop)
Abstract:Cardiac magnetic resonance imaging (CMR) produces rich sequential data such as temporal cine videos and spatial LGE/mapping stacks, yet most deep learning approaches process individual 2D slices, discarding this context. We present MR-JEPA, a self-supervised video foundation model for CMR that extends LeJEPA to 3D spatiotemporal inputs through tubelet tokenization, spatiotemporal masking augmentation, and initialization from a 2D CMR foundation model. Unlike prior CMR video models limited to cine data, MR-JEPA is pretrained on multi-sequence data (cine, LGE, mapping) from 10,505 patients across two centers without annotations. We evaluate the frozen encoder on six downstream tasks using a unified multi-view gated attention architecture: LV ejection fraction, RV ejection fraction, three myocardial strains (GLS, GCS, GRS), and four-class disease detection. MR-JEPA outperforms other compared methods on all five regression tasks, including both a domain-specific CMR model pretrained on more data with text supervision and a natural-video foundation model, achieving an LV EF MAE of 4.79% (r =0.764) and a GLS MAE of 1.87 (r=0.805), with 21-27% MAE reductions over baselines on strain tasks. For disease detection, MR-JEPA achieved a macro AUG of 0.868, remaining competitive with the domain-specific baseline despite using a fully self-supervised pretraining objective. These results demonstrate the potential of a unified video encoder for robust, multi-view utilization of diverse CMR sequences in clinical cardiac quantification and diagnosis.
[CV-21] Vision Models Predict Urban Scene Appraisal with Limited Neural Alignment
链接: https://arxiv.org/abs/2608.30964
作者: Kaizhen Tan,Yuantao Deng
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Pretrained vision embeddings are increasingly used as general-purpose representations for modelling how people appraise urban scenes, and are validated almost entirely by how well they predict human ratings. High predictive accuracy does not establish that these embeddings organise scenes as human perception does. We test the two properties separately against brain data. Using openly released EEG from 63 adults who viewed and rated 56 Berlin street scenes, we estimate the representational geometry of the scenes over time, the proportion of that geometry that is explainable at all, and its correspondence with seventeen feature spaces spanning language-supervised, self-supervised, category-supervised and dense-prediction training, two orders of magnitude of scale, and interpretable controls. Correspondence is low throughout: the best representation, DINOv2 ViT-B, reaches 29.6% of the lower bound of the noise ceiling, the panel spans 11.0% to 29.6%, and a Gabor energy descriptor is indistinguishable from the best model while outperforming every language-supervised model tested. Within a model, deeper layers still match later neural responses, so the hierarchical correspondence found for object recognition survives even at this low overall level. The same embeddings predict held-out appraisal ratings well, up to r = 0.87, and the two measures do not track each other across models; reweighting features towards the neural geometry lowers appraisal prediction for every model tested, against a control of matched dimensionality. Predicting how a street is appraised is therefore weak evidence that a model represents the street as the brain does. The benchmark uses only public data and requires no training, so evaluating a new representation needs only its embeddings for 55 images.
[CV-22] LOCI: A Locator-Critic with Refinement Loop
链接: https://arxiv.org/abs/2608.30959
作者: Walid Bousselham,Mathilde Caron,Arsha Nagrani,Cordelia Schmid
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Vision-Language Models (VLMs) still struggle on tasks requiring complex visual understanding. We argue that the core issue is not high-level reasoning, but instead failing to locate critical details in the image. Due to this shortcoming, VLMs generate often plausible but incorrect reasoning based on flawed perceptual grounding. To address this, we propose Locator-Critic (LOCI), a training-free framework that decouples visual search from evidence verification. LOCI employs a Locator agent to propose candidate visual evidence and a separate Critic agent to evaluate its relevance and sufficiency. These agents engage in an iterative refinement loop, progressively improving the evidence until it is adequate to answer the given question. This decoupled, self-correcting process yields substantial performance gains, achieving state-of-the-art results on multiple complex visual benchmarks. LOCI improves accuracy for both open-weight models like Qwen3-VL (+12.1 on V*, +5.8 on HR-Bench and +11.2 on VisualProbe-Hard) and proprietary models like Gemini 2.5 Pro (+8.9 on V*, +4.3 on HR-Bench, +4.8 on VisualProbe-Hard).
[CV-23] Audio-Driven Adversarial Defense for 3D Talking Face Generation with totally Visual Fidelity Preservation
链接: https://arxiv.org/abs/2608.30951
作者: Rui-Qing Sun,Chen-Hao Cui,Hui-Yang Zhao,Tian Lan,Zhijing Wu,Xian-Ling Mao
类目: Computer Vision and Pattern Recognition (cs.CV); Multimedia (cs.MM)
备注:
Abstract:The rapid development of generative portrait models has raised growing concerns about privacy leakage and identity misuse. In particular, audio-driven 3D talking face generation can reconstruct a reusable 3D portrait of a target person from a monocular video and animate it with arbitrary speech, making realistic identity impersonation alarmingly practical. Existing proactive defenses mainly operate in the visual domain by injecting subtle perturbations into acial regions to disrupt identity acquisition. However, such perturbations often compromise visual quality due to the strong structural priors and social sensitivity of human faces, and are easily weakened by common real-world transformations such as resizing. To overcome these limitations, we propose an imperceptible audio defense for audio-driven 3D talking face generation by shifting protection from the visual modality to the audio modality. Specifically,we exploit psychoacoustic masking to hide protective perturbations within perceptually masked frequency regions of the speech signal, thereby reducing perceptual distortion while suppressing reliable facial animation. Extensive experiments demonstrate that the proposed method effectively degrades 3D talking face generation while preserving favorable perceptual quality. These findings highlight psychoacoustically guided audio perturbations as a practical and promising direction for privacy-preserving portrait protection.
[CV-24] Uncertainty-Aware Trajectory Forecasting from Imperfect Tracking
链接: https://arxiv.org/abs/2608.30899
作者: Stephane Da Silva Martins,Victor Petrovic,Emanuel Aldea,Sylvie Le Hégarat-Mascle
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Most trajectory forecasting models are trained on clean annotated histories, and are often evaluated under the same idealized assumption, although practical deployments rely on trajectories produced by imperfect multi-object trackers. The real-world observations exhibit localization jitter, missed or unstable detections, and data-association ambiguity, which are usually either ignored or removed through denoising. This paper instead treats tracking-derived reliability cues as an informative signal to be propagated to the predictor. We propose a plug-in uncertainty-aware formulation in which each observed state is encoded as an uncertain state representation, modeled by a Gaussian distribution whose covariance combines detection-level localization uncertainty and association-level ambiguity through the law of total variance. Existing backbones are adapted with minimal architectural changes: input trajectories are represented as Gaussian observations, and predicted trajectories are produced as Gaussian forecasts rather than deterministic coordinates. To train predictors that remain robust under structured observation noise, we combine temporally correlated Ornstein-Uhlenbeck perturbations with response-based knowledge distillation from a teacher trained on clean trajectories. Experiments on Oxford Town Centre and VIRAT using real tracker outputs, together with a complementary ETH/UCY pseudo-detection protocol, show that the proposed formulation improves displacement accuracy and the reliability-sharpness trade-off of probabilistic forecasts.
[CV-25] Rad-R: A Raw-ADC Radar Dataset and Capture-Invariant SSM for Hardware-Fault Diagnosis
链接: https://arxiv.org/abs/2608.30896
作者: Mainak Mallick,Junghwan Yim,Seung-Kyum Choi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Automotive mmWave radar can develop vibration, antenna misalignment, radome blockage, and receive-channel degradation that corrupt the signal before perception begins. Data for these faults are scarce because each condition must be induced and measured on physical hardware. We introduce Rad-R, a raw-ADC dataset captured with a 4-chip 77GHz TI MMWCAS-RF-EVM cascade (192 virtual channels). Unlike existing raw-radar datasets, Rad-R pairs each recording with a controlled hardware fault at a calibrated severity, an independent physical severity measurement, and frame-synchronised IMU, temperature, GPS, and camera streams. Rad-R is a single-session dataset, so our generalisation claims are confined to a controlled cross-severity protocol in which train and test use physically distinct captures. A reproducible benchmark evaluates seven representative vision backbones and the proposed raw-IQ Mamba SSM (RadrNet) under within-clip, chirp-wise anytime, few-shot cross-capture, and controlled cross-severity protocols. Within-clip performance is near-saturated ( 0.98 macro-F1), whereas cross-severity generalisation remains difficult: the absolute-phase RadrNet-DS falls to 0.49 macro-F1. RadrNet-DS-CI replaces absolute phase with per-frame-standardised magnitude and relative chirp-to-chirp phase and ranks first on the controlled benchmark ( 0.663 vs. 0.628 for the strongest RD-CNN; three seeds); the RadrNet family also leads on the anytime and few-shot budgets. A descriptive cross-modal analysis further finds that radar micro-Doppler covaries with independently measured IMU vibration energy (pooled Spearman \rho=0.41 across conditions). The complete dataset and code will be released publicly under permissive licences.
[CV-26] A Controlled Evaluation of Model Rankings and Input Reliance in Surface Water Segmentation
链接: https://arxiv.org/abs/2608.30895
作者: Kittipat Phunjanna,Kristóf Karacs,Chayut Ngamkhanong
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Performance evaluation for surface-water segmentation commonly uses an aggregate metric such as global intersection-over-union (IoU) to rank model configurations. However, a configuration ranking does not by itself establish why one system performs better, whether a close ordering is stable, or how strongly predictions rely on individual inputs. We examine these distinctions primarily on Sen1Floods11 through repeated configuration comparisons, paired test-chip analysis, fixed-checkpoint input stress tests, and geographic reweighting, with a targeted secondary evaluation of supervised input configurations on GEOID-Flood. The cross-modal student achieves the highest three-seed mean IoU on Sen1Floods11, but close orderings vary across seeds and geographic weighting, while ancillary-input rankings differ between Swin-UNet and U-Net. The GEOID-Flood evaluation shows substantial agreement in supervised ancillary-input effects, although the exact architecture ordering remains configuration dependent. Fixed-checkpoint tests further establish reliance on terrain and WorldCover without establishing a clean-input performance benefit, while target semantics and the later WorldCover prior restrict the evaluation to retrospective all-water segmentation. These results show that aggregate metrics remain useful for ranking complete configurations, but ranking stability, component attribution, input reliance, and deployment scope require distinct evidence. Performance evaluation should therefore match the evidence reported to the claim being made.
[CV-27] SurgSkill-Bench: A Benchmark for Multimodal Surgical Skill Assessment
链接: https://arxiv.org/abs/2608.30872
作者: Chaohui Dang,Zheheng Jiang,James Glasbey,David Luke,Theodoros Arvanitis,Le Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Objective assessment of surgical technical skill is important for surgical training and structured feedback, but current workflows remain dependent on labor-intensive expert review. Existing automated approaches primarily focus on visual inputs and provide limited support for jointly studying operative performance, structured skill scores, and evaluator feedback. We introduce SurgSkill-Bench, an initial video-score-text benchmark-style dataset containing 214 surgical training simulation videos, six-dimensional OSATS scores, and expert free-text comments. We define two evaluation settings: video-only OSATS prediction for automated assessment and post hoc expert-comment-assisted prediction, where evaluator comments are available as auxiliary information. We provide controlled baseline experiments using representative frozen visual backbones, content-adaptive key-frame sampling, and a simple video-text co-attention fusion module. Under internal video-level validation, content-adaptive sampling improves video-only performance in this dataset, while evaluator comments provide additional score-related signal in the assisted setting. The best mean AUROC reaches 0.88 under dataset-specific median dichotomization. We further discuss evaluation constraints related to dataset scale, metadata completeness, and the interpretation of comment-assisted prediction. Code will be released publicly at a later date.
[CV-28] VCAR: Training-Free 3DGS Segmentation via View Completeness and Axis-Aware Boundary Refinement
链接: https://arxiv.org/abs/2608.30870
作者: Kun Cao,Di Wang,Haibin Zhu,Haozhi Huang,Xu Wang,Zheng Shi,Guanghua Yang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to the 34th ACM International Conference on Multimedia (MM '26). 16 pages, 11 figures, including supplementary material
Abstract:Semantic segmentation in 3D Gaussian Splatting (3DGS) is crucial for advancing 3D scene understanding. Existing methods predominantly rely on feature distillation, which incurs substantial per-scene training overhead and often yields blurred segmentation boundaries. We identify that these boundary artifacts are driven in part by insufficient viewpoint coverage and boundary overflow of anisotropic Gaussian primitives. To address these challenges, we propose VCAR, a training-free coarse-to-fine segmentation strategy based on View Completeness and Axis-aware Boundary Refinement. In the coarse stage, a visibility-based weighted multi-view voting scheme rapidly localizes the target. In the fine stage, an object-centric sphere derived from the coarse result generates supplementary viewpoints via Spherical Spiral Sampling (SSS), allowing multi-view voting on the augmented views to precisely refine object boundaries and suppress irrelevant 3D Gaussians. Moreover, we introduce Axis-aware Boundary Refinement (ABR) to mitigate artifacts from anisotropic primitives. By decomposing the projected 2D covariance into per-axis contributions, ABR identifies the dominant axis responsible for boundary leakage and applies targeted anisotropic compression exclusively along that axis. Extensive experiments on NVOS and LERF demonstrate that VCAR achieves state-of-the-art segmentation accuracy and efficiency without training. Our code is available at this https URL.
[CV-29] GAFT: Geo-Anchored Fine-Tuning for Hazard Identification from Rare Failures
链接: https://arxiv.org/abs/2608.30858
作者: Yanran Xu,Chuanhang Qiu,Yue Wang,Wenbo Wu,Zhaoxing Li
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Off-road navigation can fail when physical structures induce irrecoverable states such as high-centering or entrapment, requiring human interventions. Identifying these structures is crucial, yet challenging. Such failure events are rare and costly to collect, resulting in limited training data. Moreover, the collected data associate frames with outcomes, but do not indicate the visual cues responsible for the failure. Learning directly from these data can therefore exploit scenario-specific visual cues, leading to poor generalization. We propose \textbfGeo-Anchored Fine-Tuning (GAFT), a parameter-efficient method that adapts a vision foundation model with a geometry-derived prior. It guides LoRA adaptation by aligning a spatial attention-rollout map with the geometry prior, while preserving pretrained representations. On an intervention-verified forest hazard benchmark, across ten independently trained adaptations, GAFT consistently outperforms frozen DINOv2 and supervised PEFT baselines, improving the repeated leave-one-scenario-out mean F_2 from 0.0607 to 0.3757 with statistical significance under paired analysis. Within these independently trained models, the best-performing GAFT model achieves a repeated-LOSO F_2 of 0.570. Code and benchmark: this https URL
[CV-30] AMI: Temporally Aligned Missingness-Aware and Interpretable Multimodal Fusion for Mental Health Assessment in Older Adults with Mild Cognitive Impairment
链接: https://arxiv.org/abs/2608.30857
作者: Merna Bibars,Bolaji Omofojoye,Allan I. Levey,Rachel Hershenberg,Gari D. Clifford,Hyeokhyen Kwon
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 19 pages, submitted to IEEE Transactions on Affective Computing
Abstract:Depression and anxiety in older adults with Mild Cognitive Impairment (MCI) are frequently underdiagnosed due to limited access to care. Multimodal analysis of remote clinical interviews is a scalable screening approach, but existing methods have three limitations. First, they do not correct temporal misalignment across multimodal features extracted at different resolutions, inducing spurious cross-modal associations. Second, remote recordings exhibit uneven modality dropout, but missing values are often zero-filled, making them indistinguishable from valid near-zero measurements. Finally, they do not jointly attribute predictions to modalities, questions, and interview moments, limiting fine-grained clinical interpretation. We propose a Temporally-Aligned, Missingness-Aware, Interpretable (TAMI) multimodal fusion framework. TAMI aligns speech, language, facial, and physiological features within question-answer segments on a shared timeline, encodes modality-level missingness over time, and conditions fusion on question context. In interviews with 49 older adults with MCI, TAMI achieved area under the receiver operating characteristic curve (AUROC) scores of 0.68 (depression) and 0.69 (anxiety). Fine-grained temporal alignment of multimodal features produced the largest performance gain ( \Delta\geq0.1 ). Multi-level interpretability analysis revealed that depression classification relied on eyegaze and open-ended questions, while anxiety classification depended on eyegaze and head pose, with attribution uniformly distributed across questions. Using only responses to the open-ended questions (5.1min), the depression model achieved an AUROC score of 0.67, which was not significantly different from using the full interview (19min) ( p0.05 ). Our findings support designing interview protocols centered on open-ended questions for depression screening in older adults with MCI.
[CV-31] Pretrained Curriculum-Tuned and Ensembled: A Tracer-Aware Interactive Segmentation Pipeline for AutoPET V
链接: https://arxiv.org/abs/2608.30844
作者: Xinglong Liang,Chunyao Lu,Tianyu Zhang,Jiaju Huang,Tao Tan,Yunchao Yin,Lishan Cai
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Interactive lesion segmentation in whole-body PET/CT requires a model to provide a strong initial prediction while also responding efficiently to sparse corrective scribbles during inference. This setting is particularly challenging because tracer distributions, physiological uptake patterns, lesion appearance, and acquisition characteristics differ substantially between FDG and PSMA studies. We present TRIAGE, Tracer-aware Refinement via Interactive Anatomy-Guided sEgmentation. The core backbone is a 3D STU-Net initialized through masked autoencoding pre-training with an asynchronous masking strategy, aiming to learn transferable anatomical and cross-modal representations before task-specific fine-tuning. In parallel, we train an auxiliary organ segmentation model whose predictions provide explicit anatomical context and help distinguish physiological uptake from malignant lesions. A dedicated tracer classifier first routes each study to an FDG- or PSMA-specific branch. Within each branch, a first-stage segmentation model consumes CT, PET, and organ context to generate an initial lesion mask. The initial prediction is then combined with cumulative foreground/background scribbles and refined by a second interactive segmentation network. The FDG and PSMA branches share the same overall processing pipeline but are trained independently to account for tracer-specific appearance and error modes. We additionally employ curriculum-style training and model ensembling to improve robustness across interaction steps and heterogeneous cohorts. Experiments are conducted using the official AutoPET V data and ten-fold split; quantitative results, ablations, and final test-set performance are left as placeholders to be completed after the challenge evaluation. Code: this https URL.
[CV-32] Physical Adversarial Examples for Person Detectors in Thermal Images Based on 3D Modeling
链接: https://arxiv.org/abs/2608.30839
作者: Xiaopei Zhu,Siyuan Huang,Zhanhao Hu,Jianmin Li,Jun Zhu,Xiaolin Hu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by TPAMI 2025
Abstract:Thermal Infrared detection is widely used in autonomous driving, medical AI, etc., but its security has only attracted attention recently. We propose infrared adversarial clothing designed to evade thermal person detectors in real-world scenarios. The design of the adversarial clothing is based on 3D modeling, which makes it easier to simulate multiangle scenes near the real world compared to 2D modeling. We optimized the black patch layout pattern of 3D clothing based on the adversarial example technique and made physical adversarial clothing using the aerogel. The idea is to paste a set of square aerogel patches, which display black squares in thermal images, in the inner side of clothing at specific locations with specific orientations. To enhance realism, we propose a method to build infrared 3D models with real infrared photos and develop texture maps for 3D models to simulate varied infrared characteristics over time and location. In physical attacks, we achieved an attack success rate of 80.11% indoors and 76.85% outdoors against YOLOv9. In contrast, randomly placed patches yielded much lower success rates (26.53% indoors and 23.03% outdoors). The adversarial clothing also showed good transferability to unknown detectors with an ensemble attack method, demonstrating the effectiveness of our approach.
[CV-33] Reliable Benchmarking of Artifact Detection in Computational Pathology: A Reproducibility and Uncertainty Analysis
链接: https://arxiv.org/abs/2608.30835
作者: Konstantinos Moutselos,Ilias Maglogiannis
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 5 figures
Abstract:Background and Objective: Quality control is a prerequisite for whole-slide image analysis, yet the benchmarks on which quality-control methods are compared share four properties that make their reported differences hard to interpret: few independent slides, annotation concentrated in a minority of them, pooled ratio metrics with no closed-form standard error, and a single inherited train/test partition. We propose a reliability protocol for such benchmarks. Methods: The protocol quantifies four sources of variability - test-set sampling, training stochasticity, partition composition, and undocumented preprocessing - a claim is reportable only if it survives all four; three of the four cost minutes of compute. We apply it to an independent reconstruction of a published diffusion-based artifact detector, evaluated on the original 24-slide partition and against a supervised baseline. Results: The method’s central mechanism reproduces: the auxiliary contrastive term improves pooled F1 from 0.673 to 0.688 and replicates under a second seed (+0.0156, p = 0.031; +0.0190, p = 0.005), although it acts on pen marking rather than the artifact types cited to motivate it. Its comparative claims do not: differences between design variants, and against the supervised baseline, fall inside the uncertainty of the evaluation. Four of 24 slides carry 70% of scored annotated pixels, giving an effective sample size of 6.2, and the inherited partition sits at the 7th percentile. An unreported tissue-restriction step excludes 41.4% of out-of-focus annotation against 2.6% of air bubble; such a gate is confounded with blur by construction. Conclusions: Small-cohort benchmarks support far weaker conclusions than current reporting implies. The four checks are cheap enough to accompany any evaluation on such a resource and separate reproducible effects from differences the evaluation cannot resolve.
[CV-34] Whole-Body MRI Classification via Prompt-Based Clinical Conditioning
链接: https://arxiv.org/abs/2608.30824
作者: Laura Daza,Marta Hasny,Cristina González,Julia A. Schnabel
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Combining whole-body magnetic resonance imaging (WB-MRI) with clinical variables has the potential to improve systemic disease diagnosis by leveraging complementary sources of patient information. However, structured clinical variables are often incomplete or missing, limiting the applicability of conventional multimodal fusion methods that assume fixed inputs. In this work, we propose TACTIC (Tabular-Attribute Conditioned Transformer for Image Classification), a prompt-based multimodal framework that integrates WB-MRI and structured clinical data through conditional visual feature learning. By encoding clinical attributes as prompts, TACTIC supports an arbitrary number of tabular inputs and naturally handles missing data without requiring imputation or fixed input structures. We evaluate TACTIC on five WB-MRI classification tasks spanning systemic and oncologic applications, including diabetes, chronic obstructive pulmonary disease (COPD), breast cancer, prostate cancer, and metastasis diagnosis. Across all tasks, TACTIC consistently improves performance over image-only baselines when clinical information is available while maintaining strong predictive capability under incomplete tabular inputs. Our results demonstrate the effectiveness of prompt-based models as a flexible approach for improving WB-MRI analysis using clinical context. The model weights and code are available at this https URL
[CV-35] Lucida: Parse Generate and Place for Composable Real-to-Sim Scene Modeling
链接: https://arxiv.org/abs/2608.30821
作者: Minghan Qin,Yuang Wang,Xiuyu Yang,Yushi Long,Yujian Zhang,Ruihuan Wang,Kai Ye,Yangang Zhang,Hang Li
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Project Page: this https URL
Abstract:Composable scene modeling aims to recover a real indoor scene as complete, editable object assets arranged as observed, giving robot simulation and embodied AI a simulation-ready replica of the real environment whose objects can be manipulated individually. Existing pipelines decompose the task into three steps—parse the observations into instances, generate an asset for each, and place each asset back—but every step presumes an input that a cluttered capture rarely provides: accurate instance geometry, unoccluded views, and assets that accurately match the observations. We propose Lucida, which keeps this order but redistributes the requirements, so each step consumes only what a real capture reliably provides and precision is reached at the end of the pipeline rather than demanded at its start. Lucida parses the video into a scene graph whose nodes carry per-instance multi-view evidence, generates a complete asset for each instance from its evidence, and places assets with GizmoAct, a VLM policy that casts placement as multi-turn GUI interaction, manipulating the object’s gizmo in a closed loop and deciding itself when alignment is reached. Across scene-level 3D object detection, object pose estimation, and scene reconstruction, Lucida improves mAP over Boxer by 69% on R2S-Scene, raises ADD-SB@0.05 from 57.8% to 83.4% on CA-1M, and increases scene F-Score from 0.794 for SAM3D to 0.924.
[CV-36] RealOOB: A Definition-Consistent Real-World Oriented Occlusion Boundary Benchmark
链接: https://arxiv.org/abs/2608.30820
作者: Lintao Xu,Yinghao Wang,Chenchu Rong,Xuchong Qiu,Chaohui Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 8 pages, 6 figures, 5 tables
Abstract:Occlusion boundaries (OBs) are pixel-level image boundaries corresponding to surface visibility discontinuities caused by occlusion. Through precise boundary localisation and occlusion orientation, OBs encode local surface layout and depth ordering, providing geometry-driven mid-level cues for scene understanding. However, progress in pixel-level OB estimation has been limited by fragmented supervision: Existing benchmarks often suffer from limited coverage, category-specific designs, missing self-occlusion annotations, or inconsistent annotation definitions. Meanwhile, modern edge detectors and monocular depth estimators have become strong boundary and geometry predictors, yet their relationship to definition-consistent OBs remains underexplored. We introduce RealOOB, a carefully annotated real-world benchmark with 4.26M definition-consistent, geometry-grounded OB labels covering both inter-object and self-occlusion boundaries, together with validity-aware occlusion-orientation maps that restrict supervision to pixels whose cross-boundary depth ordering is reliably measurable. Based on RealOOB, we evaluate forty OB estimators and edge detectors alongside six monocular depth estimators. Our evaluation reveals a clear gap in occlusion reasoning: modern edge detectors perform competitively with OB methods in localisation, whereas orientation prediction remains challenging for all evaluated methods. Meanwhile, even strong depth estimators often fail to exhibit measurable geometry at true OBs. We believe RealOOB provides a strong reference benchmark for the OB estimation community and a real-world testbed for assessing depth discontinuities and geometry fidelity in broader low-level vision tasks. Dataset and code will be released.
[CV-37] A Composition-Aware Pretraining Framework for Geospatial Foundation Models
链接: https://arxiv.org/abs/2608.30817
作者: Aryan Kashyap Naveen,Abhishek Srinivas,Pranav Moothedath,Shrutilipi Bhattacharjee
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Geospatial foundation models have emerged as state-of-the-art methods for downstream Earth observation tasks. However, existing pretraining methodologies process imagery through a single-concept lens, failing to capture the highly compositional nature of complex satellite scenes. We propose a composition-aware pretraining framework that explicitly encodes fractional land-cover mixtures. Each satellite image cell is mapped to a histogram representing its fractional land-cover distribution, which we term the “composition target”. These targets serve as the primary prediction objective and are distilled into the backbone using Earth Mover’s Distance. Experimental evaluation shows that composition-aware pretraining yields substantial gains on region-level understanding tasks requiring semantic similarity judgment, including zero-shot image retrieval and scene classification, while remaining competitive on tasks requiring fine-grained spatial precision, such as segmentation and object detection. With a 36.8M-parameter backbone, our framework outperforms SatMAE and Prithvi-EO-2.0, which contain 303M and 600M parameters, respectively, in most retrieval and scene classification settings. On the fine-grained ForestNet-12 dataset, a rigorous testbed for compositional discrimination, our method boosts baseline mAP@10 from 0.279 to 0.434, a 55.6% relative improvement, providing direct evidence for the effectiveness of explicit composition modeling. The code implementation can be found at this https URL
[CV-38] Camera trap classification with deep learning under ground truth uncertainty
链接: https://arxiv.org/abs/2608.30789
作者: Leonard Hockerts,Peter S. Stewart,Sarthak Arora,Tiffany J. Vlaar
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 28 pages, 5 figures
Abstract:Supervised deep learning methods enable the rapid processing of ecological image data, but depend on a costly annotation process. Consequently, training labels are commonly derived from volunteer citizen science projects. However, disagreement among volunteers introduces uncertainty in the “ground truth” data that are assumed to be correct for model training and validation. Using two datasets containing camera trap images with associated volunteer and expert classifications, we investigated the effects of training under higher ground truth uncertainty. We observed improved overall test accuracy, particularly for images that were more difficult for volunteers. Species-level accuracy also generally improved, but generalisation to a different dataset did not. The benefits of ground truth uncertainty were enhanced by pre-training on ImageNet. Pre-training also reduced the number of training epochs required; further reductions in computational cost, but not gains in accuracy, resulted from additional pre-training on other camera trap images. With unbalanced training data, we still observed a clear benefit of increased ground truth uncertainty for overall accuracy, especially on difficult images. Class imbalance improved accuracy for common species, reduced rare species accuracy, and changed patterns of misclassification to more closely resemble mistakes made by volunteers. Our findings have implications for applying deep learning across ecological image types with multiple labels. Practitioners can improve accuracy, especially on difficult examples, by including moderate levels of label disagreement during training and using models pre-trained on general image data. In addition to improving the use of citizen science-derived labels in model training, our study suggests avenues for more effectively integrating human and deep learning classifications in combined workflows. (abridged)
[CV-39] PixelIR: Fidelity-Perception Decoupling via Pixel-Space Image-Residual Flow Matching for Efficient One-Step Real-World Super-Resolution
链接: https://arxiv.org/abs/2608.30782
作者: Bingtian Qiao,Yue Shi,Yong Guo,Wenjun Zhang,Jiezhang Cao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Real-world image super-resolution (Real-ISR) aims to preserve structures supported by the degraded observation while reconstructing perceptually realistic details. However, existing Real-ISR methods largely optimize fidelity and perceptual quality within a shared network, causing the two objectives to interfere throughout training and making their balance difficult to control. Recent one-step methods reduce sampling steps, yet often inherit both this coupled optimization behavior and the expensive high-resolution backbone of their multi-step predecessors. We argue that efficient Real-ISR requires not only a shorter sampling trajectory, but also specialized modeling of faithful reconstruction and perceptual detail synthesis. Based on this insight, we propose PixelIR, a fidelity-perception decoupling framework built upon pixel-space image-residual flow matching. PixelIR first learns an image flow that maps the degraded observation to a faithful reconstruction. Then, a residual flow synthesizes the missing perceptual details from noise without repeatedly relearning or overwriting the complete restoration solution. We further distill the teacher into a deployment-oriented one-step student within a coarse-to-fine pyramid architecture. Extensive experiments show that PixelIR achieves leading PSNR, SSIM, and LPIPS on both RealSR and DRealSR. The final model completes pixel-space restoration in a single evaluation with only 32.9M parameters, 89.7G MACs, and 8.5ms latency, demonstrating a strong practical fidelity-perception-efficiency balance.
[CV-40] CORAL: A Benchmark for Structure-aware and Brain-wide Neuron Reconstruction in Light Microscopy
链接: https://arxiv.org/abs/2608.30768
作者: Zekang Yang,Jiamin Li,Zhenghua Li,Jiaqi Fan,Zengcai Guo,Xiaolin Hu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Automatic neuron reconstruction from light microscopy images is a central problem in computational neuroanatomy. While recent methods have achieved encouraging results on local image blocks, it remains unclear whether such progress translates to reconstruction that is both structurally accurate and scalable to the whole-brain scale. We present CORAL, the first benchmark for structure-aware evaluation of automatic neuron reconstruction from light microscopy images at both local and whole-brain scales. Built on a high-quality whole-brain fMOST dataset with carefully curated annotations, CORAL establishes two progressive tasks: block-level reconstruction, which evaluates reconstruction methods under limited spatial context, and brain-wide reconstruction, which assesses complete neuron reconstruction at the whole-brain scale. To account for topological correctness beyond geometric distance similarity, we introduce a structure-aware metric based on fiber prediction. To further achieve complete neuron reconstruction across the entire brain, we develop a brain-wide neuron tracing framework that extends arbitrary local reconstruction methods to the whole-brain scale through an iterative local-to-global process. Using this benchmark, we provide the first structure-aware comparison of mainstream methods for local neuron reconstruction and further evaluate their performance in brain-wide reconstruction. Our results underscore the importance of structure-aware evaluation and the need for more robust methods for complete neuron reconstruction.
[CV-41] ChessQueries: Toward Better Chess Board Recognition
链接: https://arxiv.org/abs/2608.30762
作者: Joël Seytre
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Chess board recognition is the task of mapping the image of a chess board to the information of which piece is on which square. So far this task has two established benchmarks: ChessCog is synthetic, and ChessReD comes from smartphone pictures of a single chess board setup. We introduce ChessQueries, a new method combining a ViT encoder with a DETR-style decoder, which outperforms existing methods. On the ChessReD benchmark, we improve the state of the art from 15.3% to 99.2%, and demonstrate strong capabilities on out-of-distribution datasets. Our method saturates the task on the two datasets, with an average 0.01 wrong squares per board (vs. SotA: 3.4 / 0.15 respectively). We also share a new, harder public dataset, parsed from broadcasted top-level chess tournaments. Code, model weights and the SLCC data will be released.
[CV-42] CheXGround: Anatomical Region Tokens for Grounded Longitudinal Chest X-ray Interpretation BMVC2026
链接: https://arxiv.org/abs/2608.30758
作者: Adonay Demewez Gebremedhin,Wessam Shehieb,Sara Alansari,Mohamad Alansari,Muzammal Naseer,Sajid Javed,Naoufel Werghi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted for publication at the 37th British Machine Vision Conference (BMVC2026)
Abstract:Recent radiology multi-modal language models have made substantial progress in chest X-ray report generation, visual question answering, and temporal reasoning. While longitudinal chest X-ray interpretation compares sequential examinations to describe change, visual grounding aims to connect clinical language with localized image evidence. Although longitudinal modeling and visual grounding have each advanced radiology language models, how localized visual evidence can support longitudinal interpretation remains under-explored. We introduce CheXGround, a region-grounded longitudinal chest X-ray language model that represents paired studies through corresponding anatomical regions. CheXGround extracts anatomical regions from current and prior radiographs, encodes them as temporally enhanced Region-of-Interest (ROI) tokens, and combines them with global temporal image context during generation. To connect these region tokens with clinical text, we propose Temporal Region–Phrase Alignment, a pretraining objective that aligns temporal anatomical representations with localized report phrases. We evaluate CheXGround on single-study and longitudinal Visual Question Answering (VQA), longitudinal findings generation, temporal grounded VQA, and anatomical grounding. Across these tasks, CheXGround improves clinical language quality, temporal reasoning, and localization accuracy over recent baselines. Our results suggest that organizing longitudinal evidence at the anatomical level is a strong representation for grounded radiology language modeling. Project page: this https URL
[CV-43] Autoregressive Mosaics: Probing 2D Spatial Reasoning in Text-Only Language Models WACV2027
链接: https://arxiv.org/abs/2608.30751
作者: Ashwin Nedungadi,Stefan Oehmcke,Stefan Lüdtke
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注: WACV 2027 Submission Pre-Print
Abstract:Large language models (LLMs) trained only on text and code can sometimes generate programs that draw recognizable images. However, it is unclear whether this reflects an internal representation of 2D spatial layout or simply the ability to translate spatial descriptions into code. We introduce Autoregressive Mosaics (AM-Bench), a benchmark that separates these factors: First, a translation task gives a model a fully specified geometry of a picture in words as a prompt and asks for the code that produces it. Second, a layout task requires the model to compose an image from an underspecified prompt. Across eight open-weight text-and-code-only models, all models reliably translate specified geometry into code, but their open-ended layout performance differs substantially, indicating that these differences are not explained by code-generation ability alone. An output-medium ablation further shows that the interface or medium of expression that the model uses matters: replacing procedural code with raw SVG improves layout scores across all models. Finally, probing model activations shows that a coarse layout plan is present before generation, but reflects only the layout implied by the prompt. During generation, models track the evolving geometric state instead of executing an initially fixed plan. Overall, these results show that 2D spatial performance in text-only LLMs depends on both the model and the output medium, and is not explained by code-generation ability alone.
[CV-44] RailGen: Improving Railway Intrusion Detection via Agent -Guided Small-Scale Foreign Object Generation
链接: https://arxiv.org/abs/2608.30727
作者: Quan Hao,Ziyang Tao,Chenxi Zhang,Yudong Wang,Rui Shi,Liguo Zhang
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Small-object detection under long-tailed data distributions is a fundamental yet challenging problem in multimedia. Railway Foreign Object Detection (RFOD) epitomizes this challenge with easily confused small intrusions and scarce samples. To address these issues, we propose a generative-augmented detection paradigm that leverages multimodal image generation to enrich the feature space of rare and small objects. We first construct RailGen, a multimodal image generation agent based on large models. Under semantic constraints, RailGen automatically invokes tools to generate railway scenes, calibrate intrusion positions, extract foreign objects, and fuse them into realistic intrusion effects. This process produces high-quality synthetic samples that effectively densify the feature representations of tail classes and complete the small-object feature space. Within this paradigm, we further propose FocalDEIM, a detection framework designed to enhance training with generated data. FocalDEIM improves dense matching with Focal Modulation for better small-object discrimination and adopts Focal Loss to emphasize hard samples, thereby alleviating blurred inter-class boundaries in complex railway scenes. Experimental results demonstrate that RailGen can generate high-quality small-scale foreign objects, reducing the object pixel area by up to 58x and 13.85x on average. Equipped with these challenging samples, our paradigm surpasses the baseline DEIM by 5.6% and 7.5% in mAP@50 and mAP@(50-95), respectively, and outperforms existing state-of-the-art methods. Ablation studies verify RailGen’s feature-space enrichment and FocalDEIM’s boundary discrimination. The paradigm provides an effective multimodal generative solution for long-tailed small-object detection in safety-critical applications.
[CV-45] SegWave: Wavelet-Driven Segmentation of Tampered Regions ATC ECCV2026
链接: https://arxiv.org/abs/2608.30714
作者: Siddhi Pravin Lipare,Vishesh Kumar,Akshay Agarwal
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at the PFATCV Workshop, ECCV 2026
Abstract:Verifying image authenticity is increasingly difficult, posing serious risks across journalism, law enforcement, and political domains. Most existing forensic methods rely on high-level visual artifacts and treat frame detection as a simple binary task. To address this, we propose SegWave, a hybrid framework that jointly leverages spatial and frequency-domain cues for image tampering detection. SegWave integrates a transformer-based architecture with the Discrete Wavelet Transform (DWT) to capture localized, multi-scale frequency inconsistencies indicative of manipulation. To further improve localization effectiveness, we introduce an Adaptive Sub-band Attention module (ASA) that dynamically highlights the informative high-frequency wavelet components. Extensive experiments on multiple benchmark datasets demonstrate that SegWave consistently outperforms state-of-the-art tampering detection methods in challenging evaluation settings.
[CV-46] RailSyn: Diagnosis-Guided Image Generation for Traceable Data Completion in Railway Foreign Object Detection
链接: https://arxiv.org/abs/2608.30709
作者: Quan Hao,Chenxi Zhang,Ziyang Tao,Yuyuan Zhou,Yudong Wang,Rui Shi,Lechuan Xu,Changhao Liu,Liguo Zhang
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Railway foreign object detection (RFOD) is critical to safe railway operation, yet scarce real positive samples incompletely represent task-relevant variations in object scale, intrusion relation, railway scene, illumination, and adverse weather. Existing synthetic augmentation can improve RFOD detection, but its gains lack an explicit account of the task-relevant deficiencies complemented by the generated data. We therefore introduce RailSyn, a diagnosis-guided framework comprising a real-referenced Inspector and a requirement-aligned Generator. The Inspector constructs a variable-radius empirical cover from finite real observations to localize candidate completion regions and profile synthetic pools. The resulting audit identifies railway-context, intrusion-semantic, and visual-consistency requirements; the Generator addresses them through domain adaptation, agent-planned placement and physical contact relations, and plan-consistent conditional refinement. Using the Inspector, we further trace representation-space changes across generation variants; the complete system attains a local-shell occupation of C_gap to 13.64%, which measures generated coverage of real-derived completion regions. Extensive experiments show AP50–95 gains of up to 4.9 points and consistent improvements across nine mainstream detectors, demonstrating broad cross-architecture utility.
[CV-47] VisLens: Single-Pass Interpretable Visual Search for Multimodal LLM s
链接: https://arxiv.org/abs/2608.30705
作者: Jingyi He,Sanghwan Kim,Zeynep Akata
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Multimodal large language models (MLLMs) struggle with fine-grained Visual Search, the task of locating small or rare objects in high-resolution images. Existing remedies fall into two families: (1) Training-free methods based on attention or confidence scores are accurate but slow, since they require multiple MLLM queries per example. (2) Reinforcement Learning (RL) trained tool-use models are faster at inference but opaque, since their tool calls remain uncontrollable and hard to interpret. To overcome this, we propose \emphVisLens (Visual Focus via Logit Lens), a Visual Search method built on the logit lens, which decodes the semantics held in a hidden state by projecting it through the LLM head. VisLens further uses a lightweight tuned-lens that maps early hidden states into the final hidden state space, so visual tokens can be read out from early layers. These tokens are matched to target words in the query to generate a crop of the relevant region, which is fed back in alongside the original image to produce the final answer. The whole process, from decoding to the final answer, completes in a single forward pass without repeated queries. VisLens matches or exceeds prior baselines while delivering a substantial latency advantage, running 8.5 – 9.9\times faster than Thyme and up to 22.2\times faster than training-free multi-pass search methods.
[CV-48] UE-Detector: A Tool-Using Expert MLLM -Based Detector for AI-Generated Videos
链接: https://arxiv.org/abs/2608.30704
作者: Yichen Wu,Haoxuan Qu,Yongxing Dai,Yan Bai,Yihang Lou,Yuqi Lin,Hossein Rahmani,Jun Liu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 32 pages, 7 figures, 28 tables; includes supplementary material. Code: this https URL
Abstract:AI-generated video detection, which aims to distinguish AI-generated videos from real ones, has recently received increasing research attention. To perform this task reliably, a key challenge lies in accurately identifying subtle-yet-measurable unnatural artifacts. In this work, we address this challenge from a novel perspective of tool-mediated evidence discovery and propose Tool-Using Expert MLLM-based AI-generated Video Detector (TUE-Detector), a novel framework for AI-generated video detection. TUE-Detector trains a general MLLM into a task-tailored tool-using expert detector that learns to invoke suitable tools, collect concrete evidence of unnaturalness, and reason over the evidence for reliable detection. Meanwhile, TUE-Detector further introduces novel designs to equip the expert detector with high-quality and suitable tools. Extensive experiments demonstrate the effectiveness of our framework.
[CV-49] Can Video World Models Track Unobserved World States?
链接: https://arxiv.org/abs/2608.30692
作者: Joonghyuk Shin,Yicong Hong,Jaesik Park,Xun Huang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Project webpage: this https URL
Abstract:Video world models are increasingly used as simulators, yet visual fidelity alone does not show that a model maintains the hidden state of the world. We examine this gap with an action-conditioned video Shell Game, a visual analog of S_5 state tracking that decouples visual rendering from compositing the hidden state underneath. Bidirectional and autoregressive Transformers, Mamba, and linear attention restricted to nonnegative transition eigenvalues all fit the training horizon of 5 swaps and then fall toward chance on longer swap chains (extrapolation) while still rendering plausible video with additional denoising steps providing no benefit. The pixel-based diffusion target never supervises the unseen hidden state, so the generated frames cannot carry it and the state has to live inside the architecture rather than in the tokens. For a Transformer, that architectural state is only an append-only KV cache, so the model has to re-derive the hidden arrangement from the whole history at every chunk. We find two mechanisms that do extrapolate, and both carry a state across chunks and revise it in place. Linear attention succeeds once its transition eigenvalues may be negative, and TTT with a nonlinear fast weight succeeds by updating the feature map through which it reads its own state. We further examine harder cases in dynamic world exploration tasks, and discuss the broader implications for building stateful video world models.
[CV-50] Failure or Drift? Evaluating Monocular SLAM under Synthetic and Real-World Corruptions ECCV2026
链接: https://arxiv.org/abs/2608.30690
作者: Abhay Skaria Thomas,Shashank Agnihotri,Margret Keuper
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注: Accepted at the 3rd NeuSLAM workshop at ECCV 2026
Abstract:Visual SLAM is commonly evaluated on clean trajectories, although deployment failures are often caused by adverse weather, illumination, blur, and sensor artifacts. Controlled corruptions are attractive because they isolate such factors, but a synthetic stress test is useful only when it leads to the same engineering conclusion as the condition it is intended to approximate. This work examines that question for monocular SLAM. We evaluate a classical feature-based system and two learned trackers under image-space, geometry-aware, and compound corruptions, and compare their behavior with adverse conditions from 4Seasons. Rather than reducing robustness to a single trajectory error, the evaluation separates explicit tracking failure from drift accumulated by methods that remain active. The results show that learned trackers largely replace catastrophic loss with sustained, and sometimes severe, drift. More importantly, the apparent ordering of the learned systems changes with the physical fidelity of the corruption: structured rain and fog proxies preserve the real-world ordering, whereas a simple illumination proxy does not. Code is available at: this https URL.
[CV-51] CANVAS: Consistency-Aware Navigation via Visual Adaptive Sampling for Long-Context Text-to-SVG Generation
链接: https://arxiv.org/abs/2608.30689
作者: Yichen Wu,Haoxuan Qu,Yihang Lou,Hossein Rahmani,Jun Liu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 30 pages (including supplementary material), 9 figures, 10 tables. Code: this https URL
Abstract:Autoregressive large models have recently advanced Text-to-SVG generation from simple icons to complex, long-context graphics, yet standard autoregressive decoding often fails to maintain global consistency across geometry, layout, occlusion, and composition. We introduce CANVAS (Consistency-Aware Navigation via Visual Adaptive Sampling), a training-free, render-aware inference framework that combines power-sharpened trajectory likelihood with visual feedback from rendered futures and derives a stroke-wise navigation rule. It effectively estimates each candidate stroke’s future value under a limited generation and rendering budget and adaptively allocates samples according to candidate uncertainty, decision influence, and rollout cost. Experiments across multiple autoregressive SVG backbones and complementary benchmarks demonstrate improvements in global consistency, which includes sound geometric relationships, spatial layouts, occlusion ordering, and overall composition, without additional training, demonstrating the effectiveness and generalization ability of our framework.
[CV-52] UFPR-PEs: A Brazilian Face Recognition Benchmark with Self-Declared Race/Color Labels
链接: https://arxiv.org/abs/2608.30688
作者: Alexandre Diano,Bernardo Biesseck,Gabriel Polo,Vinicius Gregorio,Laura Lopes,Diego Addan,David Menotti
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted for presentation at the 2026 Conference on Graphics, Patterns and Images (SIBGRAPI)
Abstract:While face recognition systems are widely deployed, ensuring their demographic reliability and robustness under uncontrolled visual conditions remains a critical challenge. To bridge this gap, we present UFPR-PEs, a benchmark for face recognition bias evaluation using public videos of elected Brazilian politicians annotated with official self-declared race/color categories. The dataset adopts the Brazilian census taxonomy, including the parda category, which has no direct equivalent in the U.S.- or Europe-centric schemas commonly used in prior benchmarks. Our benchmark is built from compressed public video and preserves difficult samples so that performance can be analyzed under realistic conditions. We describe the construction pipeline, report dataset statistics, and evaluate face recognition performance across verification and (closed- and open-set) identification settings, including subgroup analysis by race/color and difficulty level. The results show that recognition performance varies substantially with image quality, and that subgroup gaps must be interpreted jointly with visual difficulty rather than in isolation. Overall, UFPR-PEs provides a reproducible and demographically grounded setting for studying face recognition bias under challenging public video conditions.
[CV-53] InfraOcc: An Infrastructure Occupancy Benchmark with Static-to-Dynamic Reasoning
链接: https://arxiv.org/abs/2608.30657
作者: Lei Yang,Xiaokai Bai,Boqi Li,Chunmian Lin,Li Wang,Ziying Song,Jiahuan Zhang,Enhui Ma,Haibao Yu,Jiaqi Ma,Kaicheng Yu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 17 pages, 12 figures
Abstract:Fixed-viewpoint infrastructure sensors repeatedly observe the same traffic space, making roadside 3D occupancy structurally different from ego-vehicle perception: a near-persistent static scaffold is overlaid with sparse, short-lived dynamic events. Existing occupancy benchmarks and methods, however, are built around moving ego vehicles and neither measure nor exploit this structure, instead treating occupancy as flat one-shot voxel classification. We address this gap from both data and model perspectives. We build InfraOcc, to our knowledge, the first real-world infrastructure-side semantic occupancy benchmark, with dense voxel annotations for 290 multi-modal sequences in a fixed roadside frame, a static-dynamic decoupled annotation pipeline, unified camera-only, LiDAR-only, and multi-modal evaluation, and diagnostics for static and dynamic occupancy. InfraOcc shows that static infrastructure fills 97.3% of occupied voxels and persists across frames, whereas dynamic participants have a median occupied-frame ratio of only 1.8% per location, revealing a structural static-dynamic asymmetry beyond semantic long-tailedness. We further propose ProSD-Occ, which reformulates occupancy as progressive static-to-dynamic evidence reasoning: it explains persistent layout, exposes residual dynamic evidence under static-confidence guidance, and recomposes static, dynamic, and free-space evidence into a unified field. ProSD-Occ ranks first in overall, dynamic, static, and geometric occupancy on every track, e.g., a 23.5% relative camera-only dynamic-mIoU gain over the strongest baseline and 65.87 multi-modal overall mIoU, establishing fixed-viewpoint roadside occupancy as a distinct problem with its own reasoning paradigm. The benchmark and code will be publicly available at this https URL
[CV-54] APT: Anchor-aligned Perturbations for Tamper Localization in Fully Regenerated Images ECCV2026
链接: https://arxiv.org/abs/2608.30656
作者: Suhyeon Ha,Woo Jae Kim,Joonsung Jeon,Sooel Son,Sung-eui Yoon
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to ECCV 2026
Abstract:Proactive tamper localization embeds an imperceptible signal into an image prior to distribution, enabling pixel-level manipulation detection. Existing methods assume a spliced (SP) setting, where synthesized regions are composited onto the original background, leaving embedded signals intact. However, real-world diffusion-based inpainting operates in a fully regenerated (FR) setting, where the entire image undergoes denoising, disrupting background signals and rendering existing frameworks ineffective. We propose APT, a semi-fragile latent-space perturbation that embeds a dense, vector-wise localization signal. By aligning each spatial feature vector toward a fixed anchor direction, APT localizes tampering via the alignment disparity between synthesized foreground and anchor-aligned background features after inpainting. The proposed hard negative mining loss and noisy perturbation branch further enforce uniform alignment. Experiments on COCO demonstrate that APT achieves an FR IoU of 0.92, outperforming the strongest baseline (WAM, 0.84), while existing methods collapse to near-random performance (AUC 0.5), establishing APT as a practical forensic framework generalizable across tampering types unknown at test time.
[CV-55] Fine-Grained Multi Image Object Hallucination Benchmark CVPR2026
链接: https://arxiv.org/abs/2608.30653
作者: Joonki Min,Chaeyun Kim,Hyungwook Choi,Yejin Kim,Kihyun Kim,Yohan Jo,Joonseok Lee
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Accepted at CVPR 2026
Abstract:Multimodal Large Language Models (MLLMs) are increasingly deployed in multi-image scenarios requiring complex reasoning across visual contexts. However, current MLLMs remain fundamentally limited by object hallucination-generating plausible yet factually inconsistent descriptions about objects. Existing benchmarks, designed primarily for single-image settings or providing only high-level multi-image assessments, cannot systematically diagnose how visual complexity and reasoning demands trigger hallucination. To address this gap, we introduce MIOH, a fine-grained multi-image object hallucination benchmark that systematically evaluates object hallucination across four foundational tasks (existence, counting, attribute, position) through three multi-image reasoning patterns (comprehensive, comparative, selective) under three controlled adversarial pressures (visual context scale, perceptual difficulty, contextual bias). Through evaluation of 29 models, we reveal that even state-of-the-art systems like GPT-5 and Gemini-2.5-Pro exhibit distinct failure patterns across different reasoning patterns and tasks. Our evaluation reveals that hallucination stems not merely from perceptual failures but from integration-stage limitations when maintaining object representations across multiple images. MIOH provides a controlled framework for analyzing multi-image object hallucination and serves as a critical evaluation tool for developing more reliable multimodal AI systems.
[CV-56] Quantum-Grassmann-Plucker Token Mixing for Deep Learning-Based Post-Disaster Damage Assessment
链接: https://arxiv.org/abs/2608.30633
作者: Kooroush Farahkhah,Umut Lagap,Taha Rezaei,Saman Ghaffarian
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Timely post-disaster building damage assessment from satellite imagery is a critical engineering decision support task, yet it remains constrained by class imbalance, ambiguous intermediate damage states, and limited cross-event transferability. This study presents, to our knowledge, the first application of Grassmann-Plucker (GP) token mixing to computer vision and introduces two extensions for image classification: the Quantum-inspired Grassmann-Plucker (QGP) head and the Hybrid Quantum Machine Learning Grassmann-Plucker (HQML-GP) head. The GP head represents multiscale relationships among image patch tokens by encoding subspaces formed by token pairs with Plucker coordinates; QGP enriches these coordinates with amplitude-derived probability features, whereas HQML-GP incorporates expectation values generated by a simulated quantum circuit into the geometric token representation. Paired pre- and post-event image patches from the xBD tornado dataset were processed using a frozen six-channel Vision Transformer base encoder with 16 x 16-pixel patches. The three GP-based heads were compared with multilayer perceptron and Transformer baselines under identical training, checkpoint selection, and evaluation protocols. Joplin and Moore tornado samples were used for model development and seen-event testing, while Tuscaloosa was reserved for unseen-event evaluation. QGP led both test sets in accuracy and macro-F1: 83.46% and 64.50% for the seen events, and 66.45% and 52.70% for the unseen event. Although HQML-GP obtained the highest validation macro-F1 of 65.63%, it did not surpass QGP on either test set and required substantially more training time per epoch. These results establish GP token mixing as a competitive attention-free alternative to conventional Transformer-based token mixing for paired satellite image damage classification.
[CV-57] Cost-efficient Active Learning for Referring Image Segmentation and Grounding EMNLP2026
链接: https://arxiv.org/abs/2608.30621
作者: Junbeom Hong,Seonghoon Yu,Hyung Rok Jung,Sundong Kim,Jeany Son
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted to EMNLP 2026 Findings
Abstract:Collecting natural-language referring expressions along with region annotations, such as masks or boxes, is a major bottleneck in visual grounding (VG), as annotators must write descriptions that distinguish target regions from visually similar ones. We tackle this by formulating active learning (AL) for VG under the realistic setting where only raw images are available without accompanying text. Since ground-truth text is unavailable, sample selection must estimate which images contain ambiguous regions that would require discriminative referring expressions. To address this, we generate auxiliary region-text pairs using foundation models, and introduce Referred Region Ambiguity, a new acquisition function that measures whether the model’s confidence collapses onto a single region or disperses across multiple candidates. It allows our method to prioritize images with strong cross-region competition, which are more informative due to their visual ambiguity. We also design a referring-expression annotation interface that helps annotators quickly focus on writing discriminative language with a few clicks. Experiments on RIS and REC benchmarks show that our AL framework consistently outperforms several AL baselines, while a user study shows up to 1.6X faster description labeling of ours.
[CV-58] AQ3D: Adaptive Query Transformer for 3D Instance Segmentation
链接: https://arxiv.org/abs/2608.30618
作者: Keno Moenck,Thorsten Schüppstuhl
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Transformer-based decoders for 3D instance segmentation typically commit to a fixed number of queries and positional modeling calibrated on the training distribution rather than on the scene at hand. Indoor scans vary widely in spatial extent and object count, so a fixed query set over-initializes small scenes and under-initializes large ones, while learned absolute and relative encodings are bound to the training scenes’ extents and can saturate. We present AQ3D, which is designed to handle scenes of various sizes during training and inference. Queries are instantiated at a fixed ratio of the scene’s superpoints, forming an overcomplete set whose background rejection is entirely left to the decoder. Positional information is encoded using 3D RoPE over quantized metric coordinates, replacing learned bounded lookup tables of prior decoders. Further, we improve the decoder itself by using attribution-based superpoint pooling, a mask refinement branch, and a cosine classifier for background rejection. Experiments show our method sets a new state-of-the-art on validation and hidden test splits across the datasets ScanNetV2, ScanNet200, and ScanNet++V2 among decoder methods trained without additional data augmentation. Code is available at \hrefthis https URLthis http URL.
[CV-59] RealCAD: Towards Real-World Image-to-CAD Reconstruction under Domain Shift and Parameter Bias WWW
链接: https://arxiv.org/abs/2608.30617
作者: Yihe Sun,Ziyu Lu,Kaihua Tang,Xian-Sheng Hua
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: The code and dataset are publicly available. Code: this https URL . Dataset: this https URL
Abstract:Reconstructing editable Computer-Aided Design (CAD) models from images is essential for downstream modification, manufacturing, and design reuse. However, existing image-to-CAD methods are developed predominantly on synthetic renderings and face two coupled obstacles: a substantial appearance domain gap between synthetic and real images, and a previously overlooked parameter bias in widely used CAD data. We show that the local normalization adopted by DeepCAD concentrates several geometric parameters around a few discrete values while encoding substantial information in a single scale factor. Consequently, a model can achieve deceptively high parameter accuracy by exploiting these frequent values rather than inferring geometry from the input image. In this paper, we propose RealCAD, a unified framework that addresses these limitations at the representation, image, and feature levels. At the representation level, we redistribute scale information to the corresponding geometric parameters, producing less concentrated parameter distributions in a shared scale space. At the image level, geometry-constrained translation converts synthetic renderings toward the real-image domain while conditioning on object contours. At the feature level, a multi-positive contrastive objective aligns representations of the same CAD model across viewpoints and image domains, enabling CAD sequence prediction from each individual view. We further introduce OpenRealCAD, comprising four-view photographs of 392 3D-printed objects paired with ground-truth command sequences. Experiments show that the revised representation substantially reduces the accuracy attainable from parameter-frequency priors, making parameter accuracy a more reliable measure of image-conditioned geometric inference. RealCAD further improves real-domain command and parameter accuracy, while retaining competitive synthetic-domain performance.
[CV-60] OCR-Based Field Extraction for Archaeological Pottery Metadata: The CENTURIA Dataset
链接: https://arxiv.org/abs/2608.30616
作者: Gissu Valentina Naghavi,Dominik Hagmann,Martin Kampel,Irene Ballester
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Pottery is a primary source for reconstructing the chronological and economic dimensions of past societies. Archaeologists often document ceramic finds through technical drawings and handwritten metadata. This metadata is critical for dating, provenance attribution, and cross-site comparison, but remains inaccessible to computational analysis, requiring manual transcription of every record. We investigate whether state-of-the-art document analysis models can address this task, and introduce CENTURIA, a dataset of 507 pottery records from the Roman site of Carnuntum, providing transcriptions, bounding boxes, and structured field-level labels across seven metadata categories. Benchmarking five OCR models reveals a substantial domain gap: zero-shot transcription error reaches 15-32% SpACER-M, far exceeding rates on printed archival documents, with domain-specific fields recovered in fewer than 3% of cases. LoRA fine-tuning on just 57 samples, reflecting a realistic archival annotation budget, closes this gap, reducing transcription error to below 1.5% and recovering overall field-level accuracy above 87%. Our results show that a small expert-validated fine-tuning set suffices to convert handwritten pottery documentation into structured, searchable metadata ready for archaeological databases.
[CV-61] DiffSAC: Diffusion-guided Sampling for Consensus-based Robust Estimation
链接: https://arxiv.org/abs/2608.30603
作者: Chang Nie,Guangming Wang,Zhe Liu,Hesheng Wang
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Robust estimation is a core computer vision task frequently tackled using sample consensus. However, traditional methods suffer from inefficient sampling as they struggle to identify effective minimum sets before hypothesis evaluation. To address these challenges, we propose a novel Diffusion-guided Sampling for Consensus-based Robust Estimation (DiffSAC) framework. DiffSAC introduces a diffusion model to learn the distribution of effective minimum sets. It refines the confidence for each data point, indicating whether it belongs to a good minimum set, rather than ranking the data points as in previous work. This significantly reduces the need to process numerous bad sets. To constrain the refinement direction, geometric features are incorporated as conditions within our diffusion model. Consequently, DiffSAC outputs a small number of high-quality minimum sets, enabling identification of the best hypothesis via consensus evaluation. Notably, compared to previous works requiring evaluating over ten thousand hypotheses, DiffSAC achieves state-of-the-art performance with only dozens, significantly boosting efficiency. Extensive experiments across five classic computer vision tasks demonstrate the superiority of DiffSAC. The diffusion model’s sampling accelerators enable real-time operation, and DiffSAC can be used as a plug-and-play module to improve existing sample consensus methods.
[CV-62] Learning Compositional Spatio-Temporal Video Grounding with Synthetic Curriculum
链接: https://arxiv.org/abs/2608.30584
作者: Xingjian Wang,Shijian Wang,Yibo Wang,Zihao Yu,Runhao Fu,Xuelian Cheng,Zongyuan Ge
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 20 pages
Abstract:Despite the impressive progress of recent MLLMs on spatio-temporal video grounding (STVG), existing evaluations and training data focus primarily on simple queries. They largely overlook the compositional queries prevalent in real-world scenarios, where a target must be disambiguated by jointly reasoning about its attributes and relations to other entities. To bridge this gap, we propose Compositional Spatio-Temporal Video Grounding (CompSTVG), a task that requires models to process complex textual queries where every intertwined attribute and relational cue is essential for disambiguation. To facilitate this task at scale, we build a synthetic data engine that leverages a spatio-temporal scene graph as a difficulty measure and casts difficulty-controlled query synthesis as a constraint programming problem, producing difficulty-graded data for both evaluation and training. Built on this engine, we introduce STVG-CompBench, a benchmark stratified by explicit difficulty levels that jointly capture temporal complexity and spatial interference. Evaluating 11 representative STVG models on STVG-CompBench reveals that current models perform poorly on compositional queries, exhibiting a sharp performance drop that is typically obscured by overall dataset-level averages. We further construct synthetic training data and propose CurrSTVG, a curriculum reinforcement learning framework that delivers consistent gains, with the largest improvements observed on the most challenging compositional queries.
[CV-63] Modality Disentangled Learning for Incomplete Multimodal Emotion Recognition: A Primitive Memory Distillation Perspective EMNLP2026
链接: https://arxiv.org/abs/2608.30563
作者: Jiaqi Zhang,Zheng Pang,Mengting Li,Yiqi Wang,Guangyuan Dong,Chao Xue,Yusen Wu,Zihao Li,Huy Phan,Sicheng Zhao,Björn W. Schuller,Jiachen Luo
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 19 Pages, 8 Figures, 13 Tables. Accepted to EMNLP 2026 Findings
Abstract:Multimodal Emotion Recognition (MER) systems often suffer from missing modalities in real-world scenarios. Existing methods usually generate, align, or distill missing modalities as a whole, overlooking the heterogeneous nature of the information carried by each modality. Such holistic treatment mixes inferable shared semantics with uncertain modality-specific details, yielding unstable representations and degrading robustness. To address this issue, we propose the Primitive Memory Distillation (PriMD) framework. Unlike existing methods, PriMD takes an intra-modal perspective and focuses on how different types of information within a modality differ in recoverability within each modality. PriMD first disentangles cross-modal shared semantics from modality-specific representations, and then discretizes the latter into learnable semantic primitives to construct modality-specific memory banks. When modalities are missing, PriMD is a teacher-student framework that the student model uses the shared semantics of available modalities as queries to dynamically retrieve primitives. It compensates for missing modality-specific information within a constrained memory space and aligns with the teacher model. Extensive experiments on IEMOCAP, CMU-MOSI, and CMU-MOSEI demonstrate that PriMD achieves state-of-the-art performance and consistently stronger robustness across a wide range of missing-modality settings, while mitigating the instability caused by holistic feature inference. Our code and project website are available at this https URL and this https URL, respectively.
[CV-64] GarmentWeaver: Schema-Aware Structured Synthesis for Multimodal Sewing Patterns
链接: https://arxiv.org/abs/2608.30550
作者: Yinwen Lu,Weihao Luo,Yueqi Zhong
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注: 10 pages, 7 figures
Abstract:Multimodal Sewing pattern generation aims to infer executable sewing patterns from design cues such as sketches and textual descriptions. As an interpretable and simulation-compatible representation, sewing patterns are particularly valuable for digital garment creation. However, existing methods often model garment specifications as flat long sequences, which entangles garment structure with detailed parameters and leads to redundant components, inaccurate local details, and poor simulation compatibility. In this paper, we present GarmentWeaver, a schema-aware framework for multimodal Sewing pattern generation. GarmentWeaver constructs compact hierarchical targets by activating garment-relevant structural branches and predicts executable Sewing patterns in a structured manner. Specifically, we introduce a schema-aware target construction strategy, build the generator on top of a pretrained vision-language model for multimodal garment understanding, and impose feasibility-aware regularization to encourage structurally valid and simulation-compatible outputs. Extensive experiments show that GarmentWeaver produces more accurate and more executable sewing patterns than strong baselines, while also yielding better simulation results. These findings demonstrate the effectiveness of schema-aware structured generation for reliable multimodal Sewing pattern prediction.
[CV-65] Minerals in the Wild: A Hyperspectral-XRF Dataset for Elemental Composition Estimation
链接: https://arxiv.org/abs/2608.30537
作者: Eleftheria Tetoula-Tsonga(1),George Arvanitakis(2),Theodoros Giannakas(1) ((1) Institute of Communication and Computer Systems, Athens, Greece, (2) Geonova, Athens, Greece)
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Rapid mineral characterization is essential for applications ranging from mineral exploration to industrial ore processing. To this end, Hyperspectral Imaging (HSI) has emerged as a promising sensing modality thanks to its fine spectral resolution, enabling mineral discrimination in both close-range and remote sensing settings. However, the scarcity of publicly available datasets with reliable ground-truth labels hinders the development and evaluation of HSI-based mineral identification methods. We release Minerals in the Wild, a multi-purpose dataset comprising 1,132 rock specimens collected across Europe. For each specimen, we provide an HSI acquisition together with an elemental characterization obtained via an XRF sensor. We define the task of elemental characterization on our dataset and propose a pruning mechanism that removes distant signatures from the USGS dictionary prior to a convex optimization approach for matching HSI pixels with USGS spectral signatures. Finally, we empirically show that our approach outperforms simpler baselines.
[CV-66] MEOM: Multi-View Expected-OKS Maximization for Human Pose Triangulation
链接: https://arxiv.org/abs/2608.30521
作者: Ziliang Xiong,Henglin Shi,Per-Erik Forssen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Conventional algebraic triangulation solves 3D human pose estimation (HPE) from multi-view 2D keypoints. The typical approach, decoding 2D keypoints from predicted heatmaps, is unreliable as heatmaps can be multimodal under occlusion, and collapsing them into single peaks discards their spatial distribution. We seek to use the entire heatmap to estimate 3D poses more accurately, which requires solving two problems: how to robustly fuse heatmaps across views, and how to assess the reliability of heatmaps. For the former, we introduce a novel objective, Multi-viewExpected-OKS Maximization (MEOM), that locates a 3D joint where the views agree in probability mass. For the latter, we adopt highest-density-region (HDR) calibration as a diagnostic of that mass, independently of distance-based metrics. The proposed framework covers two settings, with and without 3D supervision. Without 3D supervision, we optimize 3D poses from pretrained heatmap predictors by maximizing MEOM, achieving comparable performance with state-of-the-art methods that rely on larger backbones, temporal fusion, and simulated 3D data. On ambiguous Human3.6M (H36MA) and occluded CMU Panoptic frames, the advantage is substantial. When 3D labels are available, we train the model end-to-end with a combined MEOM and MSE loss, achieving 19.11 mm absolute MPJPE on Human3.6M outperforming the state-of-the-art volumetric approach on absolute MPJPE at half the inference cost.
[CV-67] VisER: Visual Evidence and Reliance for Object Hallucination Detection in LVLMs EMNLP2026
链接: https://arxiv.org/abs/2608.30480
作者: Afsaneh Hasanebrahimi,Hanxun Huang,Christopher Leckie,Sarah Erfani
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: Accepted at EMNLP 2026 Main Conference
Abstract:Object hallucination remains a persistent reliability issue in large vision-language models, where generated object mentions may sound plausible but lack visual grounding. Recent training-free detectors use internal signals such as token likelihood, attention, visual confidence, or image-text similarity to identify hallucinated objects. These signals are useful, but they are often source-confounded. They measure how strongly an object is supported inside the model without distinguishing whether that support comes from object-specific visual evidence or the generated text prefix. In difficult cases, a hallucinated object can still receive high internal support because it fits the scene, is associated with nearby visual cues, or follows naturally from the generated text prefix. We propose VisER, a training-free two-sided metric for object-level hallucination detection. VisER evaluates each generated object mention from two complementary views. Visual Evidence measures whether object-context compatibility is backed by object-specific evidence from image tokens. Visual Reliance measures whether the object is supported more by the image than by the generated prefix. Combining these views gives a more source-aware grounding score, while avoiding additional object-level verification generations. Across multiple LVLMs and benchmarks, VisER improves AUROC and AUPR over a range of baselines.
[CV-68] HorizonNet for visual terrain navigation
链接: https://arxiv.org/abs/2608.30471
作者: Bertil Grelsson,Andreas Robinson,Michael Felsberg,Fahad Shahbaz Khan
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注: 7 pages, 7 figures, 1 table. Published at IEEE IPAS 2018. An extended version appeared in Journal of Field Robotics 37(6):951-971, 2020, doi: https://doi.org/10.1002/rob.21929
Abstract:This paper investigates the problem of position estimation of unmanned surface vessels (USVs) operating in coastal areas or in the archipelago. We propose a position estimation method where the horizon line is extracted in a 360 degree panoramic image around the USV. We design a CNN architecture to determine an approximate horizon line in the image and implicitly determine the camera orientation (the pitch and roll angles). The panoramic image is warped to compensate for the camera orientation and to generate an image from an approximately level camera. A second CNN architecture is designed to extract the pixelwise horizon line in the warped image. The extracted horizon line is correlated with digital elevation model (DEM) data in the Fourier domain using a MOSSE correlation filter. Finally, we determine the location of the maximum correlation score over the search area to estimate the position of the USV. Comprehensive experiments are performed in a field trial in the archipelago. Our approach provides promising results by achieving position estimates with GPS-level accuracy.
[CV-69] Beyond Accuracy: Quantifying Pulmonary Attribution in Anatomy-Guided Chest X-Ray Classification Under Domain Shift
链接: https://arxiv.org/abs/2608.30467
作者: Abdullah Al Mamun,Md. Nasif Osman Khansur,Md Ashraful Hossen Akash,Md. Kishor Morol,Tze Hui Liew
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 34 pages, 8 figures, 7 tables. Code available at this https URL
Abstract:Deep-learning models can achieve strong chest X-ray (CXR) classification performance without establishing whether their predictions predominantly rely on pulmonary image content. This study evaluates pulmonary attribution containment as an anatomy-related reliability property distinct from diagnostic performance. We propose DBCA-SegNet-MGAP, a multi-task anatomy-guided CNN-Transformer framework that combines complementary feature representations through bidirectional cross-backbone attention, predicts a soft lung mask, and incorporates this anatomical prior directly into classification through Mask-Guided Adaptive Global Average Pooling (MGAP). Pulmonary attribution containment is quantified using the Anatomical Local Energy Ratio (ALR) and high-intensity cumulative ALR (cALR@0.9). Experiments were repeated across three training seeds using the COVID-19 Radiography Database for four-class internal testing and a locked Shenzhen-to-Montgomery protocol for zero-shot external tuberculosis testing. On COVID-19, the proposed model achieved a weighted F1 of 0.9615 \pm 0.0015 and macro ROC-AUC of 0.9906 \pm 0.0007 . In an architecture-matched dual-bridge comparison, replacing conventional GAP with MGAP increased ALR from 0.3878 \pm 0.0098 to 0.7086 \pm 0.0104 and cALR@0.9 from 0.5265 \pm 0.0101 to 0.9905 \pm 0.0018 , while weighted F1 remained essentially unchanged ( 0.9618 \pm 0.0015 vs. 0.9615 \pm 0.0015 ). Under locked external transfer to Montgomery, ROC-AUC remained 0.9080 \pm 0.0043 and pulmonary ALR remained 0.6466 \pm 0.0081 , whereas weighted F1 decreased to 0.7528 \pm 0.0080 and ECE increased to 0.1683 \pm 0.0055 . These findings show that diagnostic discrimination, calibration, and pulmonary attribution containment are distinct model properties and support their joint evaluation under internal testing and external domain shift.
[CV-70] Federated Multi-Task Learning for Bladder Tumor Segmentation and MIBC Classification Using a Hybrid CNN-Transformer Architecture
链接: https://arxiv.org/abs/2608.30458
作者: Malhar Udmale,Divyanshu Dwivedi,Aarohi Dhand,Sachin Dudda Nagaraju,Mayank Rai,Bagesh Kumar
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Accurate bladder tumor segmentation and assessment of mus- cle invasion from T2-weighted MRI are important for treatment plan- ning, but developing robust models across institutions is challenging be- cause patient data cannot be centrally pooled and imaging characteristics vary across scanners and acquisition protocols. We propose a federated multi-task learning framework for joint bladder tumor segmentation and MIBC/NMIBC classification across four clinical centers. The proposed Swin Hybrid model combines a ResNet-34 branch for local texture and boundary information with a Swin-Tiny Transformer for global anatomi- cal context. A segmentation-guided classification mechanism further uses tumor localization information to support MIBC prediction. We also investigate several augmentation strategies under both centralized and federated training to improve robustness to multi-center variability. Ex- periments on the FedBCa dataset show that the Swin Hybrid provides the best overall balance between segmentation and classification among the evaluated architectures. Under federated training, Geo+Elastic aug- mentation achieved a DSC of 0.8100 and a patient-level AUC of 0.8931, yielding the highest combined score of 0.8474. These results demonstrate that joint segmentation and classification can be effectively performed across multiple institutions using federated training without centralizing patient data.
[CV-71] SeqAlign3DVG: A Sequence-Aligned Benchmark and Voxel Reasoning Framework for 3D Visual Grounding
链接: https://arxiv.org/abs/2608.30451
作者: Yi Zhang,Yi Wang,Yueting Wu,Kaiyue Yang,Yuejiao Su,Lap-Pui Chau
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ACM Multimedia 2026 (MM '26)
Abstract:Image-based 3D visual grounding is critical for embodied agents, yet existing benchmarks suffer from loose text-observation alignment and neglect temporal ordering. We introduce SeqAlign3DVG, a novel benchmark dedicated to temporally ordered and strictly observation-aligned image-based 3D visual grounding. Unlike prior works using order-agnostic views or global point clouds, SeqAlign3DVG ensures all expressions are human-verified and strictly grounded in the provided RGB observations (single frames or ordered observation sequences). It comprises 9,622 single-view and 14,493 sequence samples featuring rich descriptions, complex relations, and multi-instance ambiguities. To tackle this benchmark, we propose a unified voxel-based pipeline featuring Relevance-Ordered Voxel Memory (ROVM) and Progressive Language-Voxel Fusion (PLVF). ROVM dynamically ranks and aggregates multi-view evidence via a conservative memory to mitigate noisy observations, while PLVF performs coarse-to-fine spatial-linguistic reasoning for precise disambiguation. Our approach achieves state-of-the-art performance under the depth-free protocol, significantly improving localization for targets defined by complex relations and appearance cues.
[CV-72] FlowVVTON: Flow-Guided Mask-Free Video Virtual Try-On
链接: https://arxiv.org/abs/2608.30450
作者: Shengyao Chen,Xianbing Sun,Liqing Zhang,Jianfu Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Video virtual try-on aims to transfer a target garment onto a moving person across video frames. Current methods rely on human parsing masks or pose keypoints that frequently fail under large motions and occlusions, causing boundary artifacts and temporal inconsistency. A further limitation is that most approaches rely solely on attention mechanisms for temporal modeling, providing no explicit motion supervision. We propose FlowVVTON, a mask-free framework that eliminates parsing mask dependency entirely. Optical flow is used solely as a training-time supervision signal: a flow-warped latent loss, applied across all layers of the generation model, enforces multi-scale temporal consistency by aligning adjacent-frame features under explicit physical motion constraints. A two-stage training strategy establishes mask-free spatial alignment before introducing flow-guided temporal supervision. Experiments on TikTokDress show that FlowVVTON outperforms baselines by substantial margins, particularly in temporal consistency (5.7 \times VFID-R improvement over SwiftTry), while requiring no segmentation masks, pose keypoints, or region annotations at any stage.
[CV-73] ObjectSplat: Improving Mesh Fidelity and Interactivity for 3D Scenes via Object-Level Mesh Splatting
链接: https://arxiv.org/abs/2608.30423
作者: Minhas Kamal,Hiranya Garbha Kumar,Mahedi Kamal,Balakrishnan Prabhakaran
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Splatting-based algorithms reconstruct photorealistic, real-time-renderable, and mesh-exportable 3D scenes from regular images, but they represent a scene as a single monolithic field. Therefore, the reconstruction has no object-level structure, leaving it infeasible for downstream editing or interaction. Moreover, regions that are never directly observed in the input scans are contaminated by the surrounding texture and left uncorrected, capping both mesh fidelity and novel-view synthesis. We propose a decompose-before-reconstruct approach: we segment the instances out of every frame, consider the remaining as background and inpaint it, reconstruct each instance and the background independently with mesh splatting, and compose them into a single scene. Our method significantly improves mesh fidelity (over a 5% gain in F-score) and novel-view synthesis, while supporting object-wise modifiability and interactivity. The code will be made publicly available.
[CV-74] Whole-Slide Image Analysis under Realistic Few-Shot Annotation Protocols
链接: https://arxiv.org/abs/2608.30420
作者: Tiffanie Godelaine,Maxime Zanella,Karim El Khoury,Benoit Macq,Christophe De Vleeschouwer
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Preprint under review
Abstract:Automating the analysis of whole-slide images has high clinical value, since characterizing cancers requires examining them in detail. Such analysis increasingly relies on vision-language models that provide patch-level zero-shot predictions. However, these predictions remain noisy and must be refined with a few annotations. A promising paradigm for this refinement is few-shot transduction. Rather than treating each patch independently, these methods leverage the relations between patches, together with a few annotations, to refine all predictions jointly. However, current transductive methods are evaluated under conditions that overlook key properties of whole-slide images: (i) datasets consist of independent patches extracted from multiple slides, ignoring the complex tissue organization; (ii) datasets are mostly balanced, whereas a single whole-slide image exhibits severe class imbalance, with several classes absent; and (iii) annotations are sampled at random, without reflecting how a pathologist annotates a limited number of regions. To align the transduction paradigm to realistic whole-slide settings, we introduce the following contributions. First, we propose SlideCRF, which adapts conditional random fields for whole-slide images by combining spatial and biological cues while accounting for classes that may be absent from a given slide. Second, we provide a set of realistic annotation protocols, based on spatially localized clicks and scribbles, modeling different pathologist interactions, such as the iterative correction of model errors. Across four datasets, we show that SlideCRF outperforms current transductive methods in macro F1, improving over the zero-shot predictions by +24.2% and +37.5% with one and 16 clicks per present class, respectively.
[CV-75] SePArate: Segmenting Patterns from Defects in Wafer Manufacturing Using Weak Supervision
链接: https://arxiv.org/abs/2608.30410
作者: Dain Kwon,Changmin Shin,Sunjong Park,Kanghyun Choi,Hyeyoon Lee,Jaewon Jang,Minseok Choi,Jinho Lee
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 7 pages, 8 figures. Accepted at the 63rd ACM/IEEE Design Automation Conference (DAC 2026)
Abstract:In semiconductor manufacturing, defect analysis is essential, but manual inspection cannot scale. However, existing automated inspection methods remain insufficient for root-cause analysis and process optimization. To this end, we present SePArate, a weakly supervised wafer defect segmentation method. SePArate enables pixel-level separation of patterns by leveraging only image-level annotations. It consists of a three-phase training: encoder pretraining, knowledge transfer to learn spatial cues, and training on synthetic mixed-defect data for accurate segmentation. Experiments demonstrate that SePArate outperforms the baselines.
[CV-76] ImageCAS-X: a dataset and benchmark for coronary artery segmentation and centerline extraction in coronary CT angiography
链接: https://arxiv.org/abs/2608.30404
作者: Kit M. Bransby,Esther Øksnebjerg,Kristoffer Kjær,Jacob Kirkeby,Yasmin El Youssef,Aïda Jiménez,Philip R. Pedersson,Martina C. de Knegt,Klaus F. Kofoed,Rasmus R. Paulsen
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Pre-print (under review)
Abstract:Accurate segmentation of the coronary vessel lumen is a prerequisite for quantitative assessment of atherosclerotic plaque and perivascular adipose tissue in coronary computed tomography angiography (CCTA). Cardiologists rely on semi-automated methods for this task because manual vessel tracing and segmentation are labour-intensive. Although many automated methods have been proposed, their validation remains limited by the lack of large, high-quality publicly available datasets. We provide a new dataset of voxel-wise annotations of the vessel lumen and coronary segments, alongside centerlines, and mesh surfaces for 800 scans from the publicly available ImageCAS dataset. Using this dataset, we benchmark established lumen segmentation methods against inter-observer variability, stratifying performance by disease, image quality, coronary dominance, coronary segment, vessel diameter, and lumen attenuation. These labels allow segmentation accuracy to be described in anatomical and clinical context rather than reported as a single aggregate score. The dataset supports the development and validation of methods for lumen segmentation, plaque and perivascular quantification, and haemodynamic modelling.
[CV-77] Real-Time Scene-Adaptive Tone Mapping for High-Dynamic Range Object Detection NEURIPS2025
链接: https://arxiv.org/abs/2608.30400
作者: Gongzhe Li,Linwei Qiu,Peibei Cao,Fengying Xie,Xiangyang Ji,Qilin Sun
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by NeurIPS 2025
Abstract:High-dynamic-range (HDR) images, with their rich tone and detail reproduction, hold significant potential to enhance computer vision systems, particularly in autonomous driving. However, most neural networks for embedded systems are trained on low-dynamic-range (LDR) inputs and suffer substantial performance degradation when handling high-bit-depth HDR images due to the challenges posed by extreme dynamic ranges. In this paper, we propose a novel tone mapping method that not only bridges the gap between HDR RAW inputs and the LDR sRGB requirements of detection networks but also achieves end-to-end optimization with downstream tasks. Instead of relying on the traditional image signal processing (ISP) pipeline, we introduce neural photometric calibration to regularize dynamic ranges and a scaling-invariant local tone mapping model to preserve image details. In addition, our architecture also supports performance transfer finetuning, enabling efficient adaptation from the LDR sRGB images to the HDR RAW images with minimal cost. The proposed method outperforms traditional tone mapping algorithms and advanced AI-ISP methods in challenging automotive HDR scenes. Moreover, our pipeline achieves real-time processing of 4K high-bit-depth HDR inputs on NVIDIA Jetson platforms.
[CV-78] PRISM: Predictive Recomposition via Semantic Latent Decomposition for View-invariant Video Representation Learning
链接: https://arxiv.org/abs/2608.30388
作者: Youngchae Chee,Hosu Lee,Sungjune Park,Junho Kim,Yong Man Ro
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Cross-view video representation learning aims to capture viewpoint-invariant action semantics despite substantial appearance changes across egocentric and exocentric videos. However, existing methods encode each video as a unified embedding, where view-invariant and view-variant semantics inevitably entangle under co-occurrences - a failure mode we show persists even in cross-view methods explicitly trained for view-invariance. Our key insight is that a view-invariant feature is truly disentangled when it can be sufficiently recomposed with an arbitrary view-variant feature while preserving their independent semantics. Building on this, we propose PRISM, that decomposes video into view-invariant and view-variant latents and recompose them under language supervision encouraging clean decomposition of the two streams. PRISM achieves state-of-the-art results on EgoExo4D, EgoExoLearn, AE2, even surpassing in-domain models under zero-shot setting. Code is available at this https URL.
[CV-79] MCSeg: Pre-training and Fine-tuning Volumetric Pyramid Transformer for Multi-modal Cardiac Image Segmentation
链接: https://arxiv.org/abs/2608.30371
作者: Zhiyu Ye,Hairong Zheng,Tong Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Automatic cardiac image segmentation is pivotal for diagnosing and treating cardiac diseases. In this work, we introduce MCSeg, a volumetric transformer-based network tailored for multi-modal cardiac segmentation. To overcome the architectural mismatch inherent in existing hybrid networks, we propose a novel Scaling Feature Pyramid (SFP). Unlike conventional skip connections, the SFP effectively bridges the single-scale 3D Vision Transformer (ViT) encoder and the multi-scale CNN decoder by transforming the ViT’s output into a hierarchical feature pyramid, ensuring that global contextual information is effectively leveraged. For the training paradigm, the ViT encoder first undergoes self-supervised pre-training via masked image modeling. Subsequently, the network is fine-tuned on downstream tasks, during which a regional mutual information (RMI) loss is integrated to improve boundary segmentation accuracy. In experiments, MCSeg consistently outperforms eleven SOTA methods on CT dataset ImageCHD, multi-modal dataset MM-WHS, MRI dataset HVSMR-2.0 and MSD Heart, highlighting the effectiveness of our MCSeg for multi-modal cardiac segmentation tasks. Furthermore, MCSeg’s superior performance in few-shot experiment showcases its significant potential in adapting to limited data scenarios. Codes and pre-trained ViT-B weights are open-sourced at this https URL
[CV-80] Seeing the Unseen: Camouflaged Object Detection Beyond the Visible Spectrum
链接: https://arxiv.org/abs/2608.30355
作者: Avi Gupta,Trasha Gupta
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Recent advances in camouflaged object detection (COD) have led to substantial progress in challenging low-visibility scenarios, with pioneering studies demonstrating notable success in localizing objects in camouflaged scenes. Despite these achievements, existing approaches predominantly rely on conventional three-channel RGB imagery, thereby constraining the available visual information to a limited spectral range. Multispectral images offer a wide range of information about a scene by capturing fine-grained spectral signatures. Hence, by leveraging multispectral images for COD, we introduce a novel approach to detect camouflaged objects from the corresponding multispectral inputs. In particular, we propose an end-to-end framework, \textbf\textitMSFormer, that takes a multispectral camouflaged image as input and predicts a binary mask for it. Additionally, we also provide empirical justification for integrating multispectral bands for this complex low-vision task. Our extensive experiments demonstrate the effectiveness of our method, which outperforms existing methods.
[CV-81] Proximity3D: Shape from Capacitive Proximity on Sensing Manifold
链接: https://arxiv.org/abs/2608.30344
作者: Hao Chen,Chenming Wu,Chun Ping Lam,Xiangjia Chen,Guoxin Fang,Charlie C. L. Wang,Yeung Yam,Juncong Lin,Chengkai Dai
类目: Computer Vision and Pattern Recognition (cs.CV); Computational Geometry (cs.CG); Graphics (cs.GR); Robotics (cs.RO)
备注:
Abstract:Most shape reconstruction methods assume measurements defined over planar sensing domains, such as RGB images or depth maps. In this paper, we use a curved capacitive textile as a shape sensor, treating its surface as a non-planar sensing manifold. Each scan is represented as a capacitive proximity field on this manifold, induced by the interaction between the curved electrode layout and nearby object geometry. We introduce a multi-view feedforward reconstruction model that aggregates these fields across known sensor views and recovers the observed object shape. Simulated and physical experiments demonstrate robust reconstruction from capacitive proximity signals acquired on curved sensing surfaces, pointing toward a new route to robotic near-field geometric awareness via embodied sensing.
[CV-82] CapFrame: Text-Instructed Viewpoint Grounding in 3D Gaussian Scenes via Geometric Pseudo Labels ECCV2026
链接: https://arxiv.org/abs/2608.30342
作者: Jirong Li,Satoshi Ikehata,Shuhei Kurita,Ikuro Sato
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at ECCV 2026
Abstract:3D Gaussian Splatting (3DGS) enables photorealistic real-time novel view synthesis, yet placing a virtual camera to capture a desired frame remains largely manual. Existing language-guided approaches in 3D scenes mainly focus on object-centric grounding, determining what to observe but rarely controlling how it should appear in a single frame, such as subject orientation or frame layout. To address this limitation, we introduce a new task, Text-Instructed Viewpoint Grounding (TIVG), which aims to identify a 6-DoF camera pose in a 3D Gaussian scene whose rendered frame aligns with a text instruction. To solve this task, we propose CapFrame, a partially differentiable framework that converts language into geometric pseudo labels for camera pose optimization. CapFrame follows a Retrieve-Translate-Refine pipeline: it retrieves relevant views and ranks them through a Question-Evaluation process with MLLMs, translates the instruction into orientation and layout pseudo labels, and refines the camera pose via differentiable optimization with layout and orientation losses in 3DGS. Experiments on 38 real-world scenes with 135 instructions indicate that CapFrame produces viewpoints better aligned with texts than heuristic viewpoint search and adapted trajectory generation baselines, validated by VLM metrics, MLLM judges, and user studies. Code is available at: this https URL
[CV-83] Knowing Beyond the Known: Reinforced Knowledge Specification for Multi-Label Class-Incremental Learning
链接: https://arxiv.org/abs/2608.30316
作者: Aoting Zhang,Dongbao Yang,Chang Liu,Xiaopeng Hong,Can Ma,Yu Zhou
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Existing class-incremental learning methods struggle in multi-label scenarios (MLCIL) due to the inherent contradiction of learning objectives arising from co-occurring and incomplete labels. We argue that the core obstacle is the model’s ambiguous boundary between known and unknown knowledge, which undermines historical knowledge retention, complicates current task learning, and limits adaptability to future concepts. To address this, we propose KBK (Knowing Beyond the Known), a reinforced knowledge specification framework that explicitly models what is known or not to unify historical, current, and prospective learning. Specifically, to clarify known knowledge, we develop a hierarchical feature purification module that disentangles fine-grained class-specific features from global features, where high-level semantic abstraction is reinforced with low-level visual features. Additionally, an uncertainty-aware recall enhancement strategy suppresses unreliable predictions based on distribution priors, improving the quality of historical recall. For probing the unknown, KBK leverages semantic correlations to synthesize informative unknown features under co-occurring, preserving embedding space for future learning. Furthermore, to mitigate heterogeneous forgetting, we design a category-balanced gradient compensation loss that dynamically reweights gradient backpropagation according to forgetting speeds. Experiments on multiple benchmarks validate the effectiveness and robustness of KBK, which surpasses prior best methods by 2.7% in Avg. Acc on MS-COCO B0-C10 setting even without any replay buffers.
[CV-84] ScenePilot: Grow-and-Repair Policy for Text-Driven 3D Indoor Scene Generation
链接: https://arxiv.org/abs/2608.30307
作者: Jiawei Zhang,Hongsong Wang,Pan Zhou
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Text-driven 3D indoor scene generation has advanced from dataset-bound layout modeling to open-vocabulary synthesis with large language and vision-language models. Yet existing methods remain limited: one-pass generators often yield geometrically invalid layouts, heavy post-hoc optimization is costly and unstable, and prompt-only planners lack reusable layout priors for functional grouping and object relations. We propose \textbfScenePilot, a retrieval-augmented \textbfGrow-and-Repair framework that formulates scene generation as prior-guided incremental growth with learned rectification. Given a prompt, the Hierarchical Retrieval-Augmented Planning (HRAP) module retrieves room-, group-, and anchor-level layout priors to support functional group planning. A text-driven base generator then inserts object groups sequentially, while the Reinforcement Multimodal Repair (RMR) module performs lightweight local correction after each insertion and a final global repair after completion. To train this policy, we construct \textbfSceneReverse-17k, a repair-trajectory dataset built by perturbing high-quality 3D scenes in position, rotation, and scale, then using inverse operations as executable rectification targets. The policy predicts structured \emphmove–rotate–scale actions from rendered views, scene state, retrieved priors, and edit history. By combining HRAP with RMR, ScenePilot offers an efficient alternative to one-shot generation and heavy full-scene optimization, improving physical plausibility, functional coherence, and controllability while preserving diversity.
[CV-85] Learning to Restore More: Continual Capability Expansion for Pretrained Image Restoration Models
链接: https://arxiv.org/abs/2608.30305
作者: Hu Gao,Yulong Chen,Lizhuang Ma
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Image restoration models are typically trained with a fixed set of capabilities. When new restoration requirements emerge, existing solutions usually train additional models or jointly retrain the original model with both new and historical data. Instead of designing another restoration backbone, we investigate how a trained restorer can continually acquire new capabilities without forgetting those learned previously. We propose RestoreMore, a continual capability-expansion framework that preserves the pretrained restoration model as a frozen capability anchor and learns residual expansion modules for newly arriving degradations. RestoreMore introduces a capability-oriented bi-level routing mechanism at multiple feature stages. The first routing level identifies restoration capabilities relevant to the current input, while the second selects and combines a sparse set of complementary degradation experts. This design enables newly introduced tasks to selectively reuse historical restoration knowledge and progressively enriches the expert bank available for subsequent restoration tasks. Extensive experiments on a wide range of restoration benchmarks demonstrate that RestoreMore consistently acquires new restoration abilities while preserving and improving previously learned capabilities.
[CV-86] Dynamic Hub-and-Spoke Memory for Streaming Video Understanding EMNLP2026
链接: https://arxiv.org/abs/2608.30294
作者: Xinru Jiang,Lin Zhao,Xi Xiao,Yunbei Zhang,Janet Wang,Chenrui Ma,Haolin Li,Yanzhi Wang,Yifan Gong,Octavia Camps
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to Findings of EMNLP 2026
Abstract:Streaming video understanding requires answering questions at arbitrary times over a continuously growing visual stream. The central challenge is to compactly remember long-range history while effectively retrieving question-relevant evidence. We propose Dynamic Hub-and-Spoke Memory (D-HSM), a training-free framework that represents distant history as structured textual memory while preserving the recent frames as visual tokens for fine-grained perception. Specifically, D-HSM turns selected historical video chunks into typed textual observations and stores them in an entity-centered hub-and-spoke memory, with entities as hubs and related evidence as spokes. When answering a question, D-HSM dynamically retrieves a compact question-aware memory subset, expands it through hub-and-spoke links, and combines it with the recent visual window for frozen-VLM answer prediction. Extensive experiments on both streaming and long video benchmarks show that D-HSM consistently and substantially improves VLM backbones and outperforms other state-of-the-art online and offline video understanding baselines.
[CV-87] SELECT: SELEctive Context Transfer for Class-Incremental Semantic Segmentation BMVC2026
链接: https://arxiv.org/abs/2608.30281
作者: Avi Gupta,Saurabh Yadav,Koteswar Rao Jerripothula,Tammam Tillo
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to BMVC 2026
Abstract:Class-Incremental Semantic Segmentation (CISS) is fundamentally challenged by catastrophic forgetting and background shift, where learning new concepts degrades performance on previously seen classes. While existing methods attempt to balance stability (retaining old knowledge) and plasticity (learning new knowledge), they often fail to leverage prior knowledge effectively. These approaches typically rely on indiscriminate knowledge transfer or ambiguous initializations, which can dilute crucial semantic information. To overcome this limitation, we propose SELECT, a novel approach for Selective Context Transfer, which instead grounds each new class in a small set of semantically similar past classes. Its core is a Context Transfer Attention mechanism that aggregates the learned tokens from similar classes into a structured initialization for the new class. To ensure this transfer does not corrupt the borrowed representations, we add a controlled noise perturbation and a margin-based context-transfer loss that enforces separation between the new class token and its source tokens. Extensive experiments on Pascal VOC and ADE20K show that SELECT consistently outperforms prior work, achieving mIoU of 2.2% on VOC and 2.8% on ADE, providing an effective handle on the stability-plasticity dilemma. Code is available at this https URL.
[CV-88] Motion-Saliency Complementary Masked Modeling for Point Cloud Video Understanding
链接: https://arxiv.org/abs/2608.30279
作者: Wei Wang,Yiding Sun,Yuyan Wang,Zhuoyue Zhang,Zhengqiao Li,Dongfu Yin,Chen Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Point cloud video representation learning is crucial for 3D dynamic scene understanding. In this paper, we propose MoSaiC, a novel Motion-Saliency Complementary masked modeling framework for self-supervised point cloud video representation learning. MoSaiC couples three components: Curriculum Motion-Saliency Masking (CMSM), which guides the masking process toward motion-salient tokens under a curriculum schedule; Normal-Flow Motion (NFM) modeling, which supervises the local rigid rotation of each token in the Lie algebra so(3) as an explicit geometric motion target; and Cross-view Token Consistency Prediction (CTCP), which enforces consistency between two complementary masked views at the token level. Together, these components allow MoSaiC to effectively capture both appearance and motion dynamics. Extensive experiments on multiple downstream tasks, including action recognition, temporal action segmentation, and point-level semantic segmentation, demonstrate the effectiveness of our approach.
[CV-89] Centering before Pruning: Lightweight Geometry Correction for Diversity-Based Visual Token Pruning in LVLMs
链接: https://arxiv.org/abs/2608.30263
作者: Shunjie Wen,Jaeyeon Lee,Dong-Wan Choi
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Large vision-language models (LVLMs) incur substantial inference costs due to their long and highly redundant visual-token sequences. Diversity-based pruning mitigates this cost by selecting token subsets based on pairwise cosine similarity. We find, however, that similarities between raw visual tokens are strongly concentrated in the positive range, limiting their ability to distinguish non-redundant tokens. A natural way to improve this resolution is to center token features before computing cosine similarity. Centering indeed reveals a substantially richer pairwise structure, yet unexpectedly degrades pruning performance when used alone. We show that this apparent contradiction arises because the raw geometry does more than represent pairwise diversity: it also implicitly favors globally distinctive tokens, which tend to contain semantically informative content. Centering better resolves subset diversity but loses this useful token-wise preference, revealing that diversity and distinctiveness are entangled in the raw geometry. Based on this analysis, we propose the \textbfCentered Geometry \textbfPruner (Cen-Prune), which measures subset diversity using centered cosine similarity while retaining raw-space distinctiveness as a complementary token-wise preference. This lightweight, plug-and-play correction leaves the underlying selection mechanism unchanged and incurs negligible computational overhead. Extensive experiments across multiple image- and video-understanding benchmarks and LVLM architectures demonstrate that Cen-Prune provides robust improvements in overall performance across existing diversity-based pruners.
[CV-90] OPUS: A Simple yet Effective Unified Framework for Open-Vocabulary Detection
链接: https://arxiv.org/abs/2608.30247
作者: Xiaoyan Wei,Zhimin Yao,Ruilin Yang,Wei Zhang,Yong Dai,Yi Zhang,Wei Ge
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Recent unified open-vocabulary detection (OVD) supports heterogeneous prompts, including text queries, visual exemplars, and their combinations, but often rely on increasingly complex designs such as heavy cross-modal fusion, staged training, and iterative annotation pipelines. We revisit whether such complexity is necessary in the era of stronger foundation models. Our finding is that unified OVD can be made substantially simpler with semantic-rich visual representations and scalable grounding supervision. We present OPUS (\textbfOpen-vocabulary, \textbfPrompt-\textbfUnified, \textbfSimple), a unified detector supporting text, interactive visual, generic visual, and mixed prompting within one framework. OPUS adopts a simple three-part design. Its model architecture combines a semantic-rich visual encoder, built on a DINOv3-ConvNeXt-B backbone with efficient hybrid encoding, with a prompt-aware decoder that avoids prompt-specific branches for unified prompt reasoning. OPUS is trained with a one-stage text-visual training strategy with Instance-level Contrastive Alignment (ICA), and is supported by a SAM3-based single-pass data engine for heterogeneous grounding supervision. Experiments on COCO, LVIS-minival, and ODinW35 show that OPUS achieves state-of-the-art Visual-I performance, reaching 68.1/69.2/54.7 AP, while maintaining balanced Text and Visual-G accuracy. OPUS also turns mixed prompting from interference into complementarity, improving over text or visual prompt alone. These results show that simplicity and strong unified prompting capability can be achieved together.
[CV-91] Motus2: A Self-Evolving General World Model for Dexterous Manipulation
链接: https://arxiv.org/abs/2608.30237
作者: Hongzhe Bi,Zihao Zhou,Yihang Tang,Jingrui Pang,Shuhe Huang,Haitian Liu,Runqing Wang,Shuai Huang,Yichen Wang,Yiming Cheng,Ruowen Zhao,Zhenghua Li,Hengkai Tan,Xiaolong Liu,Jinhui Wan,Jiabao Liu,Min Zhao,Fan Bao,Jun Zhu
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:General embodied agents should perceive, predict, act, evaluate, and improve within a unified system. World models have shown great promise in building such agents, yet existing models typically append an action output head to a world simulator, without coupling them into a closed decision-and-learning loop for policy improvement. We present Motus2, a self-evolving general world model for dexterous manipulation. Motus2 advances world modeling through model scaling and data scaling. For model scaling, a single model with shared weights exposes three control interfaces: a policy (world-action model), a simulator (action-conditioned world model), and an evaluator (value model). The policy proposes candidate action chunks, the simulator predicts their visual consequences, and the evaluator assesses the predicted outcomes. Their coupling forms a closed decision-and-learning loop for policy improvement. This formulation uses curated expert demonstrations for action learning, while failed and suboptimal interactions provide valuable evidence for dynamics modeling and value learning. For data scaling, Motus2 progresses from large-scale monocular egocentric data to synchronized stereo egocentric data, followed by robot-domain adaptation with robot trajectories and supplementary human-robot alignment data. Motus2 further studies global-autoregressive and hybrid-memory extensions of its sliding-window context, adds tactile feedback for contact-aware control, and is instantiated on a fully biomimetic platform with stereo vision, dual arms, dual dexterous hands, and tactile sensing. Together, egocentric data scaling and closed-loop general world model scaling provide a general path toward self-evolving dexterous manipulation.
[CV-92] Semantic-Spatial Discriminability Enhancement for Generalized Visual Grounding ACM-MM2026
链接: https://arxiv.org/abs/2608.30233
作者: Kaiyan Lei,Xu-Yao Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ACM MM 2026
Abstract:Generalized Visual Grounding (GVG) task aims to localize targets in an image based on referring expressions, extends the classical visual grounding paradigm by integrating multi-target and non-target scenarios. Previous methods typically rely on global semantic matching or coarse-grained region interactions for localization, where the discriminative cues are primarily derived from sentence-level semantics or regional context. In complex multi-target scenarios, such approaches tend to confuse visually similar targets, making it difficult to establish stable instance-level decision boundaries. To address these limitations, this paper proposes a novel Semantic-Spatial Discriminability Enhancement (SSDE) framework for generalized visual grounding, which aims to enhance the discriminative ability on fine-grained semantics and spatial localization, improving both cross-modal understanding and instance-level grounding. Specifically, to enhance the semantic discriminability of query representations at the fine-grained level, we propose a Semantic Discriminability Enhancement (SeDE) module, which leverages spatially guided cross-attention to disentangle fine-grained target-relevant visual attributes and integrates them with the textual subject semantics. Furthermore, to strengthen the spatial discriminability of the referred targets, we introduce a Spatial Discriminability Enhancement (SpDE) module, which models an instance center density map to characterize the spatial distribution of targets, and explicitly constructs instance separation structures in the spatial domain by employing them as an auxiliary supervision signal. Extensive experiments show that SSDE achieves superior performance on ten datasets across both classic and generalized visual grounding tasks.
[CV-93] Amortized Anchor Refinement for Deployable Continuous-Time 4D Gaussian Reconstruction
链接: https://arxiv.org/abs/2608.30218
作者: Jingong Chen,Qingwen Zhang,Sanghyeon Jun,Chulwoo Pack,Kyle Gao,Kwanghee Won
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 18 pages, 13 figures
Abstract:Continuous-time 4D reconstruction remains impractical on standalone XR headsets. Per-scene optimization demands deployment-infeasible compute, and lower budgets cause collapse rather than degrade gradually. Feed-forward prediction is fast, but struggle to recover scene-specific detail. We present Amortized Anchor Refinement, which uses a frozen backbone to predict an initial Gaussian representation and a short optimization to specialize it under a fixed compute budget, with a capacity floor preserving representational density. A training-free stage then applies a persistent-homology constraint to prune unstable Gaussians while preserving topologically persistent structures, and streams the resulting trajectories directly as scene flow. On the Stage-Capture benchmark, Amortized Anchor Refinement achieves 24.31 \pm 2.22dB, while our deployment experiments demonstrate reconstruction within the target budget on a single consumer GPU and playback on a standalone XR headset.
[CV-94] DICS: Exploring Data Intrinsic Consistency for Visual Instruction Selection EMNLP2026
链接: https://arxiv.org/abs/2608.30209
作者: Yuyang Hong,Jinhui Guo,Jiaqi Gu,Lubin Fan,Ruixiang Wang,Kun Ding,Yue Wu,Shiming Xiang,Jieping Ye
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by EMNLP2026
Abstract:Visual instruction tuning is crucial for advancing the vision-language alignment and instruction-following capabilities of Vision-Language Models (VLMs). However, identifying optimal subsets under a fixed ratio constraint from rapidly expanding datasets remains a significant bottleneck. While existing methods largely depend on distribution diversity or heuristic filtering, they often overlook the internal coherence within individual samples. To bridge this gap, we propose Data Intrinsic Consistency (DIC), a self-scoring metric designed to quantify the sample-level inter-component consistency. DIC consists of two modules: Visual Information Consistency (VIC), evaluating the alignment between visual content and instructions, and Response Information Consistency (RIC), assessing response coherence relative to the instruction. Building upon DIC, we introduce Data Intrinsic Consistency Selection (DICS), an adaptive data selection method that optimizes the trade-off between high intra-sample consistency and global distributional diversity under varying data budgets. Extensive experiments demonstrate that DICS consistently outperforms state-of-the-art methods across diverse dataset scales and model architectures, surpassing full-dataset fine-tuning while using only 25% of the LLaVA-1.5-665K data. We further curate DICS-6M, a 6M-sample multi-modal instruction corpus that enables the largest-scale visual instruction selection study to date; remarkably, DICS reaches 94.52% of the official InternVL3-8B-Instruct performance using less than 25% of its reported training data. Code can be seen at this https URL
[CV-95] NoisEasier: Test-Time Noise Optimization for Text-to-Video Generation ECCV2026
链接: https://arxiv.org/abs/2608.30194
作者: Yujiang Pu,Yu Kong
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to ECCV 2026
Abstract:Diffusion models have recently advanced text-to-video (T2V) generation, yet they still struggle with fine-grained compositional alignment, such as attribute binding, spatial relations, and object interactions. While reward-based fine-tuning improves alignment, it is susceptible to reward hacking and adapts poorly to new prompt distributions. In this work, we propose NoisEasier, a test-time scaling framework that improves T2V generation through differentiable reward-guided noise optimization without modifying the underlying model. By combining efficient short-step generators with a multi-objective reward formulation, NoisEasier enables stable and practical test-time optimization under realistic inference budgets. Our key insight is that jointly optimizing the entire stochastic trajectory accelerates reward convergence and improves compositional alignment over optimizing only the initial latent, with negligible additional computational and time cost. Experiments on VBench and T2V-CompBench demonstrate consistent improvements across multiple backbones, achieving over 10% average gains on challenging dimensions such as attribute binding, object interaction, and numeracy. Overall, NoisEasier serves as both a flexible alternative and a complementary enhancement to reward-based fine-tuning, establishing test-time scaling as an effective paradigm for controllable text-to-video generation.
[CV-96] ATGS: Anchored Temporal Gaussian Splatting for Long Volumetric Video Representation SIGGRAPH’2026
链接: https://arxiv.org/abs/2608.30184
作者: Jiahao Wu,Jie Liang,Die Hu,Jiayu Yang,Kaiqiang Xiong,Xiang Li,Xiaoyun Zheng,Chao Wang,Ronggang Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: ACM ToG(SIGGRAPH’2026)
Abstract:Volumetric video enables immersive free viewpoint rendering of dynamic real world scenes, yet existing methods struggle with long sequences and complex motions, often leading to temporal instability and visual artifacts. To address these challenges, we propose \ourname, a Gaussian splatting based framework for volumetric video reconstruction. Our key insight is that explicitly tracking long term complex motion with individual Gaussian primitives is inherently unstable. Instead, we organize Gaussians around time conditioned anchors that localize their spatial and temporal support, thereby reducing long range motion complexity. We further introduce a temporal windowing strategy to activate only anchors relevant to the queried time, which improves scalability and temporal coherence. In addition, to ensure spatial and temporal stability, we design a compact set of multi level anchor features that encode global features, local spatial features, and local temporal features, jointly constraining Gaussian generation. Extensive experiments demonstrate that \ourname \ consistently outperforms prior methods on long sequence volumetric videos with complex motions. Project page: this https URL.
[CV-97] AI-enabled Low-Cost 3D Maize Ear Morphometry Platform at Breeding Scale
链接: https://arxiv.org/abs/2608.30161
作者: Therin Young,Elijah Rodriguez,Lisa Coffey,Talukder Zaki Jubery,Adarsh Krishnamurthy,Patrick Schnable,Baskar Ganapathysubramanian
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Maize ear geometry (length, width, curvature, and volume) is closely tied to yield and grain-filling outcomes, but existing high-throughput phenotyping pipelines remain constrained by the cost, labor, and specialized hardware they require. We developed and validated a low-cost pipeline that reconstructs a watertight 3-D mesh of a maize ear from a single 20-second video captured with a consumer-grade DSLR on a motorized turntable under uniform LED illumination. Camera poses from a multi-seed COLMAP procedure initialize a Neural Radiance Field (NeRF), and a cylindrical holder of known diameter, visible in every frame, provides automatic metric scaling with downstream geometric quality control. Applied to 300 ears spanning a diverse maize inbred panel, 250 (83.3%) passed automated processing and quality control. Skeleton length agreed with manual caliper measurements across all 250 ears (R^2 = 0.964, RMSE = 4.68 mm), and convex-hull volume agreed with water-displacement volume on a 15-ear subset spanning the full size range (R^2 = 0.982, RMSE = 5.26 mL). Residual length error grew with ear curvature, whereas bounding-box height, which records the same straight-line chord as calipers, showed no such trend; the discrepancy therefore originates in the measurement definition, since calipers record the chord while skeleton length traces the geodesic arc. The capture hardware costs approximately 607 USD, and operator involvement fell from roughly five minutes to one minute per ear, with all downstream processing running unattended. The platform provides a foundation for breeding-scale 3-D ear phenotyping.
[CV-98] CedarCypress3D: an annotated UAV-LiDAR dataset of individual trees in planted cedar and cypress forests
链接: https://arxiv.org/abs/2608.30149
作者: Katsuto Shimizu(1),Fumiaki Kitahara(2),Tomohiro Nishizono(2),Hideki Saito(3),Masayoshi Takahashi(2),Shingo Obata(4),Shunsuke Tei(4),Naoyuki Furuya(2),Tomoya Goto(5),Eiji Kodani(2),Yusuke Yamada(6) ((1) Shikoku Research Center, Forestry and Forest Products Research Institute, (2) Department of Forest Management, Forestry and Forest Products Research Institute, (3) Forestry and Forest Products Research Institute, (4) Hokkaido Research Center, Forestry and Forest Products Research Institute, (5) Green Kogyo, (6) Graduate School of Bioagricultural Sciences, Nagoya University)
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 15 pages, 8 figures
Abstract:Individual tree measurements derived from Light Detection and Ranging (LiDAR) mounted on Unmanned Aerial Vehicles (UAV) provide valuable information for forest inventory, ecosystem monitoring, and sustainable forest management. Recent advancements in machine learning have increased the demand for annotated datasets to develop and evaluate point cloud-based approaches, especially for individual tree segmentation. However, publicly available annotated UAV-LiDAR datasets in temperate forests are limited. In this article, we present CedarCypress3D, a manually annotated UAV-LiDAR dataset collected in Japanese cedar (Cryptomeria japonica) and Japanese cypress (Chamaecyparis obtusa) plantations in Japan. The dataset consists of UAV-LiDAR point clouds and field survey measurements from 34 circular plots across two sites with different topographic characteristics, along with terrestrial LiDAR point clouds available for a subset of 22 plots. A total of 1,627 trees were measured in the census field survey and manually annotated to match the corresponding trees in the UAV-LiDAR point clouds. For the subset of plots with terrestrial LiDAR data, semantic labels (i.e., stem and non-stem) were additionally assigned to tree points in the UAV-LiDAR data. CedarCypress3D provides high-quality annotated UAV-LiDAR data for developing and evaluating individual tree instance segmentation and semantic segmentation methods in temperate planted forests. The dataset can also support research on tree attribute prediction and multi-platform LiDAR analysis. The dataset is publicly available at this https URL.
[CV-99] Repeatability Characterisation and Error Budget of a Consumer Structured-Light Scanner for 3-D Wound Geometry:A Rigid-Phantom Study
链接: https://arxiv.org/abs/2608.30143
作者: Pushkal Kumar,Aadit Aggarwal,Karlen Aleksanyan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 8 pages, 2 figures
Abstract:We characterise the measurement error of a free- hand consumer structured-light scanner used to derive three- dimensional geometric wound descriptors. Rigid wound-care phantoms cannot change, so every difference between repeat scans of one site is measurement error; all repeats come from a single scanner unit, nine sites and 23 scans, so this charac- terises one instrument. The 95 percent repeatability limit for reconstructed surface area is a factor of 4.8, with a confidence interval from 3.0 to 7.0, so rescanning an unchanged site can shift the reading from a 79 percent decrease to a 377 percent increase. Forty-four of 45 descriptors fall below an intraclass correlation of 0.50, and none of 248 descriptor and pipeline combinations reaches 0.75. An error budget formed by holding the analy- sis region fixed, leaving sensor and reconstruction untouched, bounds the share of variance attributable to how much surface the operator captured at 75 percent for surface area, 91 percent for hull area and 95 percent for bounding-box diagonal; only hull volume is majority instrumental, at 48 percent, so a better sensor would buy little. No wound is delineated anywhere in the chain, so the comparison against the four-week area reduction used clinically to predict healing, a ratio near 2.1, is a lower bound on the noise an unsegmented pipeline must overcome, not a measurement of wound-area reproducibility. Standardising the analysis region cuts the limit to 2.13, meeting that ratio rather than clearing it. Statistical outlier removal imposes a measured systematic area deficit near 11 percent.
[CV-100] Efficient and High-Quality Depth Estimation via Pixel-Space Diffusion with Linear Attention ECCV2026
链接: https://arxiv.org/abs/2608.30129
作者: Bingde Liu,Wu Ran,Jinglei Zhang,Huanhuan Yuan,Chao Ma
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ECCV 2026
Abstract:This work presents \textbfLapis , a \textbfl inear- \textbfa ttention-based \textbfpi xel- \textbfs pace generative framework that achieves efficient and high-fidelity depth estimation with one-step diffusion. While generative frameworks have significantly advanced monocular depth estimation with superior detail fidelity, the \mathcalO(N^2) complexity of standard attention and the multi-step denoising process introduce prohibitive computational costs when scaling them to high-resolution image applications. Although linear attention and one-step prediction are intuitively viable, directly applying them leads to poor structural consistency, detail loss, and noise. Lapis rectifies these limitations through a coarse-to-fine hierarchy. Specifically, a Patch-level Consistency Module restores structural coherence by integrating semantic and spatial priors. Subsequently, a Pixel-level Refinement Module recovers sharp geometric boundaries via skip-connection-based pixel correspondence. Furthermore, to mitigate sampling noise inherent in one-step diffusion, we leverage the manifold assumption and adopt a direct \mathbfx -prediction strategy to target the clean data manifold. Extensive evaluations on multiple benchmarks demonstrate that Lapis consistently achieves state-of-the-art (SOTA) accuracy and boundary sharpness across various resolutions, reducing inference latency by up to 7.6 \times at 1080P and 10.9 \times at 1440P resolution compared to previous SOTA generative models.
[CV-101] Aligning Multi-Trajectory Supervision with Policy Optimization for VLA Driving
链接: https://arxiv.org/abs/2608.30122
作者: Tian Zhang,Zhuo Huang,Hongrui Ye,Yu Wu,Zengmao Wang,Kaixuan Zhou
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Vision-language-action (VLA) driving methods increasingly combine multi-trajectory imitation learning with group-relative policy optimization (GRPO), making trajectory selection critical to final performance. However, some high-scoring trajectories that improve imitation can degrade subsequent GRPO by inducing advantage estimates misaligned with the current policy’s feasible behavior distribution, driving updates away from safe and compliant behaviors. To address this, we propose a novel framework that aligns multi-trajectory supervision with policy optimization. To address the policy gradient bias induced by infeasible noisy trajectories outside the feasible region, augmented trajectories are constrained to a neighboring manifold of the ground-truth feasible region, and a Pareto-optimality criterion is adopted in place of the conventional aggregate score, retaining only non-dominated candidates and thereby filtering out conflicting samples at the source. To ensure that expanded trajectory supervision is effectively absorbed during policy optimization, we introduce two complementary mechanisms: feasibility-first advantage assignment and dynamic distillation. The former adapts Pareto credit to the feasibility composition of each rollout group and guides fully infeasible groups toward safe references. The latter updates teacher trajectories across refinement rounds to continually transfer useful supervision. Together, they progressively translate the benefits of expanded supervision into policy improvement. On NAVSIM v1 and v2, our method achieves 91.4 PDMS and 89.1 EPDMS, respectively, under single-trajectory inference, and recovers 440 of 658 initially failed scenes, 11.1% higher than the original GRPO baseline.
[CV-102] A Hybrid State-Space Approach for Census-Tract Population Estimation
链接: https://arxiv.org/abs/2608.30094
作者: Jackson R. Ye,Alexandre V. Morozov
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Sequence models—the architecture family behind large language models and, increasingly, state-of-the-art image recognition—have redefined how machines learn from high-dimensional data. Yet population estimation from satellite imagery, a task that underpins infrastructure planning, public health, and disaster response, has scarcely benefited: leading systems still bind population to a uniform raster, disaggregating census counts onto grid cells through weighting surfaces built from ancillary data (e.g., in WorldPop and LandScan), which can introduce systematic spatial bias, and predicting population per grid cell with convolutional neural networks. In this approach, the administrative-unit structure in which the census was actually collected is discarded. We close this gap with MambaPop, which renders each administrative unit as a single polygon-masked satellite image and treats tract-level population estimation as a sequence-modeling problem over its image patches, pairing each tract image directly with its population label and eliminating the disaggregation step entirely. Built on the hybrid state-space–attention MambaVision backbone, MambaPop is, to our knowledge, the first method to learn population directly from an administrative unit’s own image as well as the first to apply a state-space based (Mamba) hybrid architecture to the population estimation task. Across all \sim 84,000 contiguous-US census tracts of the 2020 census, MambaPop attains a mean absolute error (MAE) of 1,141 persons per tract, matching the strongest convolutional baseline (YOLOv11, MAE 1,122 ).
[CV-103] he Nearest Target Is the Wrong One: Target Separation in Arc2Face Identity Unlearning ECCV2026
链接: https://arxiv.org/abs/2608.30087
作者: Zeynel Tok
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 16 pages, accepted at the ECCV 2026 Workshop on Unlearning and Model Editing (UME)
Abstract:Unlearning an identity from a face-conditioned generator by redirecting its conditioning embedding can silently fail if the redirected output is still verified as the original person. We show that this failure depends on a controllable choice of how far the redirection target lies from the forget identity in recognition space, and that the most intuitive target, the nearest neighbour, is the one most likely to cause it. We audit Arc2Face with a locked ArcFace protocol and a projection adapter that redirects identity conditioning before generation. On a hard-neighbour stress test built from the hardest 0.5% of eligible identities, four target-selection policies show a monotonic response: clean forgetting rises from 9/30 groups under the nearest hard target to 30/30 under the least similar one. Mean forget-identity re-identification falls from 51.9 to 0.0 while mean retention stays flat. This reflects successful redirection rather than outputs becoming unverifiable: 710 of 720 least-sim-hard generations arrive at the chosen target, with no leakage to unrelated identities. Re-verifying identical images with an independent recogniser (AdaFace) preserves that trend, correlating at r=0.94, arguing against a verifier artefact. Target separation is thus a first-order, reportable design variable for identity unlearning.
[CV-104] racing Generated Samples to Training-Data Clusters in Flow-Matching Models
链接: https://arxiv.org/abs/2608.30081
作者: Rania Briq,Ohad Fried,Michael Kamp,Stefan Kesselheim
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Understanding which training samples influence a generated image is an important problem in generative modeling. In flow matching, training samples influence the generated image through the velocity field along the generation trajectory. Removing samples to examine their counterfactual influence changes the velocity field, and the resulting effect on the final image depends on how the change propagates through the trajectory. Consequently, local changes in the velocity field do not necessarily predict the final counterfactual effect. This work investigates attribution in flow-matching models through a hybrid analytical–learned approach, and uses it to derive trajectory-based attribution scores at the cluster level. We evaluate these attribution scores using independently retrained leave-one-cluster-out (LOO) models, and compare with several attribution baselines using two different flow-matching latent spaces. Our experiments show that semantic similarity constitutes a strong baseline, while the closed-form trajectory-based attribution is competitive in some metrics without requiring counterfactual retraining or model gradients. Our results show that attribution in flow matching depends not only on semantic similarity to training samples, but also on the latent representation, trajectory dynamics, and how influence is propagated to the final output. Subjects: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2608.30081 [cs.LG] (or arXiv:2608.30081v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2608.30081 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-105] AKE 85: Testing Audiovisual filmmaKers intEnt across 85 Hours of Film
链接: https://arxiv.org/abs/2608.30068
作者: Kaishuu Shinozaki-Conefrey,Olivier Pascaud,Robin Courant,Xi Wang,Dimitris Samaras,Vicky Kalogeiton
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Films communicate through deliberate creative choices, including lighting, color, composition, editing, dialogue, music, and sound. Humans naturally interpret these signals as directorial intent, yet current multimodal large language models (MLLMs) are evaluated almost exclusively on understanding what happens rather than why it is presented that way. We introduce TAKE 85, the first benchmark for directorial-intent understanding, comprising 398 short films (85 hours) with expert-verified question-answer pairs spanning global and fine-grained visual and audio intent. Through controlled modality ablations, TAKE 85 enables systematic evaluation of multimodal reasoning. Experiments on state-of-the-art MLLMs reveal a substantial gap between perceptual recognition and intentional understanding: while models accurately describe events and narratives, they consistently fail to infer the communicative role of filmmaking decisions. Our results establish directorial intent as a previously overlooked dimension of multimodal understanding: even the strongest model reaches only 58 out of 100, and our ablations show that no input modality is sufficient on its own. All code, QAs, and models are publicly available from this https URL
[CV-106] When 3D Gaussian Splatting Recovers Real Surfaces ECCV2026
链接: https://arxiv.org/abs/2608.30054
作者: Songhe Wang,David Johnathan Miller
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at ECCV 2026. 30 pages, 3 figures, 1 table; includes supplementary material
Abstract:When does 3D Gaussian Splatting (3DGS) recover the true scene surface rather than just overfitting view-dependent appearance? We answer this by developing a mathematical framework based on a first-hit rendering abstraction that cleanly isolates geometry from appearance. We prove that geometric misalignment forcefully converts spatial textures into high-frequency angular signals via parallax. This establishes a strict identifiability window: if angular capacity is bounded, surface-consistent solutions are mathematically preferred; if unrestricted, the same images can be perfectly explained by an incorrect, opaque billboard geometry. Experiments on synthetic stress tests confirm this prediction, showing billboard failures emerge precisely at high angular capacities. Conversely, in the real-world datasets we evaluate under standard capture protocols, reconstructions remain surface-consistent even at high SH degrees, which is consistent with the prediction that rich spatial texture can push billboard solutions outside the tested angular-capacity range.
[CV-107] Input-Adaptive Gating of a Dehazing Front-End for On-Device Perception in Smoke-Obscured Environments NEURIPS
链接: https://arxiv.org/abs/2608.30034
作者: Seongjun Kang,Ishaan Garg,Vishnu Bharadwaj
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Submitted to NeurIPS
Abstract:Two-stage vision pipelines often place an enhancement network before a task network, on the assumption that a cleaner input produces a better output. We evaluate this in a firefighter assistance pipeline, where a dehazer precedes an edge detector that renders smoke-filled rooms as structural outlines. Both were designed for a Raspberry Pi 4, at 355K and 23K parameters, and quantized to UINT8 via TensorFlow Lite. The float dehazer reaches 18.60 dB peak signal-to-noise ratio (PSNR) on held-out real smoke against 13.60 dB unprocessed and 17.08 dB for an AOD-Net trained on the same data, and the edge detector reaches an F-measure at optimal dataset scale (ODS) of 0.738, outperforming an optimized Canny’s result of 0.692. Dehazing improves edge extraction under dense smoke but degrades it on clear and lightly hazed frames, where the dehazer discards more detail than the haze obscures. We therefore run the dehazer only when a dark channel haze estimate exceeds a threshold, a 10.1 ms test that lets the pipeline save 469.6 ms on the dehazing stage. Averaged over four haze levels, gating is more accurate than either fixed decision, at 0.675 mean ODS against 0.664 for always dehazing and 0.630 for never dehazing. It reduces the mean per-frame time on the Raspberry Pi from 569 ms to 321 ms, and on clear frames increases the frame rate fivefold, from 1.8 to 9 frames per second.
[CV-108] OPAL: Orthonormal Prototype Alignment Learning for Interpretable Image Classification ECCV2026
链接: https://arxiv.org/abs/2608.30003
作者: Ilán Carretero,Gustavo Jesús Angulo,Rocío del Amor,Valery Naranjo
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at ECCV 2026
Abstract:Prototypical part-based models provide explainable predictions by comparing input regions to learned prototypes. However, current approaches are burdened by complex, multi-stage training pipelines and heavily rely on auxiliary regularization to prevent prototype collapse. To overcome these limitations, we introduce Orthonormal Prototype Alignment Learning (OPAL), a single-stage, end-to-end framework that simplifies interpretable classification. Our approach anchors the latent space using predefined orthonormal bases, embedding each class within a dedicated subspace spanned by fixed part-prototypes. To achieve precise part localization, OPAL enforces spatial competition across feature maps. This mechanism isolates sparse, discriminative regions, directing each prototype to consistently attend to the same semantic concept across different images. By framing classification as a direct representation alignment task, our method eliminates the need for auxiliary losses. Extensive experiments on fine-grained benchmarks demonstrate that OPAL outperforms both its non-interpretable counterparts and state-of-the-art part-prototype methods, delivering granular visual explanations by explicitly revealing the specific image regions driving every prediction. Code is available at this https URL.
[CV-109] Discrete Diffusion Bridges for Spatiotemporally Aligned Image Translation and Generation ECCV2026
链接: https://arxiv.org/abs/2608.29997
作者: Xing Xie,Jiawei Liu,Shijun Zhou,Huijie Fan,Zhi Han,Yandong Tang,Liangqiong Qu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to ECCV 2026
Abstract:We propose Discrete Diffusion Bridges (DDB), a novel framework designed to resolve the fundamental spatiotemporal misalignment of standard discrete diffusion in image translation and generation. By corrupting data into a pure mask state via a random schedule, the conventional forward process induces a twofold misalignment: spatially, this pure-mask destination entirely discards the rich structural priors of the source image; temporally, the random masking order inherently contradicts the ``easy-first, hard-last’’ decoding mechanism used during inference. To address this, DDB constructs a direct and efficient trajectory between domains. Spatially, we introduce a hybrid absorption mechanism that redefines the absorbing state to a stochastic mixture of mask and source tokens, effectively injecting source prior as spatial anchors into the latent space. Temporally, we design an information-guided noise schedule that quantifies semantic variation to prioritize the corruption of high-information regions at earlier timesteps. This ensures the model learns to resolve difficult semantic changes using robust context from invariant regions. Extensive experiments validate the versatility and robustness of our framework across diverse generative paradigms. DDB effectively balances edit alignment with structural fidelity across both text-guided semantic manipulation and pure structural image translation, while inherently complementing text-to-image generation and guaranteeing robust high-quality decoding under extremely low sampling steps. Code and models are available at \hrefthis https URLthis https URL.
[CV-110] Partition-Aware Unlearning for Removing Spurious Correlations in Large Vision-Language Models
链接: https://arxiv.org/abs/2608.29996
作者: Aditi Sarker,Nazreen Shah,Rafi Ibn Sultan,Rhongho Jang,Dongxiao Zhu,Prashant Khanduri
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 35 pages, 11 figures
Abstract:Large Vision-Language Models (LVLMs) achieve strong performance across many multimodal tasks; however, they often exploit spurious object-background correlations, resulting in predictions driven by contextual shortcuts rather than object-relevant visual evidence. Despite growing interest in hallucination and robustness evaluation, existing benchmarks provide limited control over whether model predictions are grounded in the target object or induced by correlated background cues. In this work, we introduce PURGE (\underlinePartition-aware \underlineUnlearning for \underlineRemoving spurious-correlation \underlineGenerated \underlineErrors), a framework for constructing, benchmarking, and mitigating spurious-correlation-induced failures in LVLMs. The framework consists of: – (1) Structured dataset construction wherein we develop three complementary structured data construction strategies that partition examples by object-relevant evidence and spurious background cues, enabling controlled diagnosis of shortcut reliance; and – (2) Partition-aware unlearning, which uses these partitions to selectively remove spurious object-background associations while preserving object-based reasoning. We evaluate the \algo~framework across multiple LVLMs, including LLaVA-1.6-7B, Qwen3-VL-8B-Instruct, and Qwen3.5-9B, together with CLIP as a vision-language encoder, on a diverse suite of benchmarks, including CHAIR, POPE, Causal-HalBench, MM-SpuBench, AMBER, MMHal, and Waterbirds. Our results show that PURGE consistently reduces hallucinations and spurious-correlation-driven errors while maintaining or improving overall performance in most evaluated settings, providing both a reusable evaluation protocol and an effective mitigation framework for more reliable LVLMs.
[CV-111] SVI2LoD3: Agent -Driven Reconstruction of LoD3 Facade Openings in Semantic 3D City Models from Volunteered Street View Imagery using Large Language and Visual Models
链接: https://arxiv.org/abs/2608.29992
作者: Elmehdi Kanna,Lukas Arzoumanidis,Huynh Duc An Son Nguyen,Youness Dehbi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted for publication in the ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Abstract:This paper presents an end-to-end, agent-driven pipeline for the LoD3 reconstruction of facade openings in 3D city models, producing directly usable CityGML-conform outputs. In contrast to existing approaches that rely on supervised semantic segmentation and therefore require large amounts of manually annotated training data, the proposed method employs a zero-shot segmentation strategy. This substantially reduces the annotation effort while still achieving strong performance in our benchmark on the eTRIMS dataset. A further key contribution is the enforcement of correct partonomic hierarchies, thereby producing CityGML-conform LoD3 building models. Beyond the reconstruction pipeline itself, this work also introduces a novel evaluation metric for facade reconstruction, termed Facade Feature Distance (FFD). Unlike conventional metrics such as mIoU or FRDS, which assess similarity primarily through pixel-wise overlap, FFD measures distance in a high-level feature space derived from a vision transformer. In doing so, it captures both semantic correctness and architectural layout, providing a more suitable assessment of facade reconstruction quality. The proposed pipeline and evaluation strategy together offer a practical and scalable contribution toward the automated generation and analysis of semantically enriched 3D city models. The developed code is published at: this https URL.
[CV-112] FIS-OT: Feature-Induced Optimal Transport for Unsupervised Action Segmentation ICME2026
链接: https://arxiv.org/abs/2608.29980
作者: Linxiang Peng,Xinyao Qin,Jinhan Li,Di Yang,Jiangtao Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ICME2026
Abstract:Unsupervised action segmentation is a challenging task. It involves finding action categories and boundaries in videos without labels. Existing Optimal Transport (OT) methods use global constraints. This causes them to overlook the use of local information. Furthermore, existing Optimal transport architectures are prone to confirmation bias because they overly trust the pseudo-labels they generate. This causes models to learn from noise in the early training stages. To address these issues, we propose FIS-OT. It is a novel Feature-Induced Structured Optimal Transport framework. First, we introduce a Feature Enhanced Generator (FEG) module. It serves as an internal regularizer. By using triplet loss, FEG captures local consistency. It provides robust supervision that is independent of noisy pseudo-labels. Second, we propose a Feature-Induced Residual Structural Prior. This combines a fixed temporal backbone with dynamic feature similarities. This design ensures temporal continuity. It also allows the solver to adapt to complex action structures. Finally, we establish a cyclic optimization loop. This aligns local feature learning with global structural alignment. Extensive experiments on the three datasets show the effectiveness of our method.
[CV-113] Confidence-Aware Ensemble and Long-Word Refinement for Artistic Text Recognition
链接: https://arxiv.org/abs/2608.29970
作者: Lucas A. Dias,Henrique A. Schulz,Rafaela de Miranda,Guilherme L. Peres,Pedro L. Bittencourt,Rayson Laroca
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted for presentation at the 2026 Conference on Graphics, Patterns and Images (SIBGRAPI)
Abstract:Artistic Text Recognition (ATR) remains challenging because word images often combine decorative fonts, curved layouts, object-like characters, clutter, and severe distortions. This paper studies WordArt-V1.5 as a standardized benchmark for this setting and evaluates recent scene and artistic text recognizers under a common protocol. We propose a confidence-aware ensemble that combines SVTRv2, PARSeq, and MAERec after fine-tuning on the official training split. The ensemble selects predictions using the minimum confidence over disagreement positions, emphasizing characters that separate competing hypotheses. For long words, where a single character error can invalidate the whole prediction, we add a targeted refinement stage based on Needleman-Wunsch alignment and lexicon-guided correction. On the WordArt-V1.5 Test B split, the proposed system reaches 89.90% Word Recognition Accuracy, improving the best individual fine-tuned model by 1.77 percentage points. The long-word refinement produces a modest global gain, but improves the targeted long-word subset by 2.72 percentage points. Finally, an error analysis of all remaining mistakes shows that 48.8% are associated with labeling issues, visual ambiguity, or illegible samples, highlighting the value of diagnostic reporting for future ATR benchmarks and models. Our source code is available at this https URL.
[CV-114] RIDGE: Region-Informed Derivative-Guided Evidence Selection for Long Video Understanding EMNLP2026
链接: https://arxiv.org/abs/2608.29958
作者: Shanqing Xu,Meng Luo,Mengchen Qian,Yuhui Gao,Siyue Peng,Xiaohan Zhong,Xiaojin Zhang,Zhongyu Wei,Wei Chen,Xiang Bai
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: EMNLP 2026
Abstract:Long videos contain far more visual content than Large Vision-Language Models (LVLMs) can process under a fixed visual-token budget, making frame selection essential. Existing query-aware selectors usually estimate frame-query relevance and build a compact subset from high-scoring frames. Although their mechanisms differ, the similarity sequence is still often treated primarily as values to rank or sample from, rather than as an ordered signal whose shape reflects how query-relevant evidence emerges, peaks, and fades over time. This can obscure frames that explain, contextualize, or follow an event, because such evidence may lie on the rising or falling sides of a nearby relevance peak and receive lower absolute scores. We propose RIDGE, a frame selection framework that reads the frame-query similarity curve as a temporal signal. By using local changes and curvature, RIDGE partitions the timeline into structural regions and applies region-specific selection to preserve event cores, transitions, buildup, aftermath, and contextual frames under a fixed budget. It is a lightweight post-processing step on precomputed frame-query scores and requires neither training nor iterative LVLM calls. Across four long-video benchmarks and three backbones, RIDGE achieves the best performance in most settings and remains competitive in the others.
[CV-115] On the Role of MRI Sequences in Cross-Dataset Generalization for Brain Tumor Segmentation
链接: https://arxiv.org/abs/2608.29944
作者: Henrique Zan Grande,João G. Pitol,Lucas B. Schuck,Rafael V. Serenato,Rayson Laroca,Andre Gustavo Hochuli
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted for presentation at the 2026 Conference on Graphics, Patterns and Images (SIBGRAPI)
Abstract:Brain tumor segmentation in magnetic resonance imaging (MRI) is a critical task for diagnosis and treatment planning. Despite the success of deep learning architectures such as U-Net and its variants, performance degradation across datasets remains a major challenge, particularly under domain shift and limited annotated data. To address this issue, this study systematically evaluates how individual MRI sequences influence model robustness across two well-known datasets. A ResUNet-based framework is employed, where each modality is trained independently to isolate its effect under a controlled cross-dataset evaluation protocol with tumor size stratification, without target-domain training, or with limited domain adaptation. Results show that the T2f/FLAIR sequence achieves the best cross-dataset performance, with Dice scores exceeding 75%. It consistently outperforms other modalities across most tumor size ranges, while multi-sequence training further improves performance. Additionally, even limited target-domain adaptation yields rapid initial gains, reducing the need for extensive annotations and costly retraining. Our source code is publicly available at this https URL.
[CV-116] Evaluating 2D and 3D-Aware Vision Foundation Models for Vehicle Attribute Recognition
链接: https://arxiv.org/abs/2608.29929
作者: Alexandre V. Delazeri,Gabriel E. Lima,Eduil Nascimento Jr,Rayson Laroca,David Menotti
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted for presentation at the 2026 Conference on Graphics, Patterns and Images (SIBGRAPI)
Abstract:Vehicle attribute recognition is an important task in intelligent transportation systems, particularly when Automatic License Plate Recognition (ALPR) is unavailable or unreliable. Although vision foundation models have shown strong transferability across domains, their effectiveness for fine-grained vehicle classification remains underexplored. Moreover, given the inherently three-dimensional structure of vehicles, it is unclear whether emerging 3D-aware foundation models offer advantages over standard 2D architectures. This paper presents an empirical benchmark of 14 state-of-the-art 2D and 3D-aware vision foundation models. Using the challenging real-world UFPR-VeSV dataset, we evaluate these models as frozen feature extractors via linear probing for vehicle type, make, and model recognition. We further stress-test the best-performing models under few-shot learning and Out-of-Distribution (OOD) domain shifts. Our results show that standard 2D self-supervised models, particularly DINOv3, substantially outperform 3D-aware models in fine-grained tasks, achieving over 93% Macro-Accuracy for make and model recognition. However, the 3D-aware Depth Anything v2 exhibits stronger invariance to viewing angles in vehicle type classification. These findings motivate hybrid approaches that combine 2D and 3D priors for robust vehicle recognition. Our code is publicly available at this https URL.
[CV-117] Biomechanical 3D Body: Self-Supervised Distillation of Biomechanical Pose from a 3D Body Foundation Model ECCV2026
链接: https://arxiv.org/abs/2608.29928
作者: R. James Cotton,J.D. Peiffer,Lucinda Williamson,John Leske,Georgios Pavlakos
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to the ECCV 2026 Workshop MoCha
Abstract:State-of-the-art monocular body recovery methods predict mesh vertices and angles on the corresponding kinematic tree, but their outputs lack biomechanically defined joint angles that downstream applications like clinical and biomechanical analyses require. We extend an existing foundation model, SAM-3D-Body, with an additional biomechanical prediction head that, from a single RGB image, regresses the joint angles and scales of a biomechanical model. Training this model presents a challenge, as there are limited datasets of paired images and biomechanical fits. To overcome this, we supervise biomechanical outputs with in-loop optimized targets from a Levenberg-Marquardt solver performing inverse kinematics fits against markers from the mesh predictions. This allows distilling the biomechanical head from the mesh head, even from unlabeled images. To make this work with GPU-optimized biomechanical models in MuJoCo, the entire model was implemented in JAX using Equinox. We trained this distilled output head on the publicly released SAM-3D-Body dataset. We then validated this model on biomechanical fits to two publicly available marker-based datasets, MoVi and BioCV, as well as movements from a clinical cohort captured with multiview markerless motion capture. The resulting model outperforms existing models for direct regression of biomechanics from images while only slightly underperforming the state-of-the-art monocular biomechanics method that performs more costly inference-time optimization of entire trajectories.
[CV-118] Everybody Tracking Every Body
链接: https://arxiv.org/abs/2608.29927
作者: Daeyun Shin,Yunhan Zhao,Shu Kong,Alexander C. Berg,Charless Fowlkes
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:We address the problem of 3D body pose estimation of multiple interacting people from their egocentric views with centralized coordination. Each individual wears a camera recording egocentric video and IMU data. Processing this video with VIO SLAM provides high-quality tracking of each egocentric camera through space. The first-person view from one individual provides third-person observations of other people, although these exocentric observations are sparse, intermittent, and of highly variable reliability as both cameras and subjects move. To integrate these synchronized data streams, we propose a diffusion-based approach that fuses estimates of pose based on head motion derived from egocentric camera motion with exocentric pose observations, conditioning on both observation content and reliability. Our model is trained on a mixture of single-person motion-capture data and multi-person video in order to learn rich priors for body motion trajectories and video observation reliability. Evaluation on challenging multi-person datasets suggests our fusion approach improves over motion-only and vision-only baselines in terms of both absolute and relative pose accuracy.
[CV-119] Dior: Drawing the Light of Image via Material-Decoupled Illumination Representation
链接: https://arxiv.org/abs/2608.29925
作者: Xuanpu Zhang,Xuesong Niu,Haoxiang Cao,Ruidong Chen,Jianhao Zeng,Changqian Yu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Controllable image relighting is an important problem in image editing, and hand-drawn scribbles provide an intuitive interface for specifying the desired illumination. However, existing methods do not establish a consistent and effective mapping between scribble inputs and relighting results, limiting their ability to control illumination intensity, chromaticity, and complex spatial distributions. We address this limitation by introducing a material-decoupled illumination representation, termed the Lumi Map, which establishes an explicit mapping between user scribbles and the resulting illumination, thereby improving both relighting accuracy and controllability. Specifically, we use a renderer to synthesize source image-Lumi Map-relit image triplets and train the model to predict the target relighting result conditioned on the Lumi Map. To mitigate the domain gap introduced by synthetic data, we further perform reconstruction training on real relighting pairs, improving the model’s generalization to real-world images. Finally, we present Dior-Light, an image relighting method controlled by hand-drawn strokes. Extensive experiments demonstrate that our method outperforms existing approaches in relighting accuracy and enables effective control over illumination intensity and chromaticity on in-the-wild images.
[CV-120] Hallucination Mitigation for Large Vision-Language Models via Implicit Feature Stabilization
链接: https://arxiv.org/abs/2608.29924
作者: Aditi Sarker,Rafi Ibn Sultan,Hui Zhu,Dongxiao Zhu,Prashant Khanduri
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 28 Pages, 12 Figures
Abstract:Large Vision-Language Models (LVLMs) are prone to hallucinations: they fluently describe objects, attributes, and scenes that are not in the image. We connect part of this failure to a measurable property of their representations, feature instability, where mild semantics-preserving perturbations of the input cause large changes in the learned embeddings; hallucination rates rise together with this variability. Existing stability-motivated remedies are explicit, in the sense that they intervene at inference time through latent steering or constrained decoding, and pay for it on every query. We propose implicit stabilization instead: perturbation-invariance is built into the model weights during fine-tuning, and nothing extra runs at deployment. Our framework, INFUSE, first stabilizes visual and textual representations around perturbation-averaged and ground-truth anchors, then aligns the stabilized representations across modalities with bidirectional contrastive objectives. We prove that the anchor’s root-mean-square deviation from the perturbation-mean representation shrinks at rate 1/\sqrtK in the number of views, and that under a Lipschitz decoder, this bounds how much any perturbation can change the model’s hallucination behavior. On LLaVA-1.5, LLaVA-1.6, and Qwen3-VL-8B-Instruct, INFUSE reduces AMBER CHAIR by 46-63% relative to each base model, improves ObjHal, MMHal, HallusionBench, and POPE, and preserves VQA-v2 and TextVQA, all with no inference-time overhead.
[CV-121] owards Continual Test-Time Adaptation of Vision-Language Models in Open-Vocabulary Semantic Segmentation
链接: https://arxiv.org/abs/2608.29923
作者: Chandler Timm C. Doloriel,Yunbei Zhang,Sarthak Kumar Maharana,Muhammad Salman Siddiqui,Tor Kristian Stevik,Fadi Al Machot,Kristian Hovde Liland,Habib Ullah
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: under review. code available at this https URL
Abstract:Open-vocabulary semantic segmentation (OVSS) relies on vision-language alignment to recognize arbitrary text-defined categories, yet this alignment is fragile under continual test-time distribution shift. Our diagnostic analysis reveals that entropy minimization drives patch-level class collapse, continual updates erode vision-language alignment, and redundant gradients from low-shift samples waste computation. We propose Diversify, Anchor, and Filter (DAF), a stabilization framework that augments entropy-based adaptation with a marginal diversity loss that resists collapse, a cross-modal anchor consistency loss that constrains feature drift relative to a frozen source model, and feature salience filtering that skips low-value backward passes to offset part of the source-anchor overhead. We evaluate on five datasets spanning natural scenes, autonomous driving, underwater imagery, and remote sensing with their corrupted variants. Across the evaluated continual shifts, DAF remains stable where entropy minimization collapses, improving mIoU by over 8 points on Pascal VOC20-C, over 9 points on LoveDA, and over 3 points on Foggy Cityscapes compared to the source model, and is robust to aggressive adaptation and learning rate choices.
[CV-122] Continual Test-Time Adaptation via Entropy Sensitivity-Guidance in Strict Online Setting
链接: https://arxiv.org/abs/2608.29920
作者: Chandler Timm C. Doloriel,Yunbei Zhang,Muhammad Salman Siddiqui,Tor Kristian Stevik,Fadi Al Machot,Kristian Hovde Liland,Habib Ullah
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: under review. code available at this https URL
Abstract:Test-time adaptation (TTA) promises robustness under distribution shift by updating a pretrained model on unlabeled test data, but strict online TTA with batch size one and no access to source data is especially prone to drift or collapse. We introduce Sensitivity-Guided Erasing Adaptation (SEGA), a method for strict online continual TTA (CTTA) on corruption-style streams. SEGA uses a small number of structured erasures to probe how predictive entropy changes as information is removed, and uses the resulting per-sample sensitivity trajectories to coordinate recovery and sample selection rather than relying on raw entropy or batch statistics. This yields a practical feedback signal for long-horizon batch-size-one adaptation without periodic resets or model reservoirs. In experiments on ImageNet-C, CIFAR10/100-C, and corruption-generated aquaculture streams treated as controlled corruption-style proxies, SEGA yields consistent robustness and stability gains over strong CTTA baselines while reducing backward passes through sensitivity-based gating.
[CV-123] FoundYou: A Unified Model for Personalized Segmentation and Retrieval ECCV2026
链接: https://arxiv.org/abs/2608.29917
作者: Gabriele Trivigno,Marcos Alfaro,Claudia Cuttano,Gabriele Berton,Luis Payá,Carlo Masone
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: ECCV 2026. Project page: this https URL
Abstract:Personalized segmentation and personalized retrieval both aim to identify the same physical object across different images. While the former localizes the object within a target image, the latter retrieves images where it appears. Despite this shared instance-level objective, the two tasks have largely evolved separately and are addressed with distinct solutions. In this work, we introduce FoundYou, a unified framework built on the observation that Segment Anything 2 (SAM 2), trained to preserve object identity across video frames, inherently captures instance-level cues. We leverage this property to match objects across independent images, enabling segmentation and retrieval to emerge as two outcomes of the same instance alignment process. This unified view unlocks new capabilities beyond traditional benchmarks, including few-shot personalized retrieval and promptable personalized segmentation with flexible prompts. Extensive experiments show consistent gains over unified and task-specific methods, including +18.4 mIoU on PerMIS and +17.8 mAP on ILIAS. Performance scales with additional references and remains robust to weaker prompts. Beyond personalization, FoundYou achieves state-of-the-art results on category-level retrieval benchmarks. Notably, our approach keeps the SAM 2-small model entirely frozen and adds only 5.9 M trainable parameters, yielding a 52 M-parameter model that is over 75x faster and 20x smaller than the only prior unified solution. Code is available at this https URL .
[CV-124] Matrix-Game 3.5: Enhancing Real-Time Streaming Interactive World Models with Patch Memory
链接: https://arxiv.org/abs/2608.29910
作者: Runjia Qian,Zile Wang,Jihai Zhang,Kai Zou,Wei Yu,Jiaxing Li,Zexiang Liu,Yaokun Li,Fei Kang,Kaichen Huang,Mengyin An,Haobo Zhang,Biao Jiang,Jiahua Wang,Haofeng Sun,Yang Liu,Yangguang Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: this https URL
Abstract:Interactive world models extend video generation from offline clip synthesis toward persistent simulation of interactive virtual worlds, enabling applications in games, robotics, embodied agents, and XR. Achieving stable long-horizon interactive generation, however, remains challenging, as the model must simultaneously preserve scene geometry, dynamic consistency, and camera control while supporting real-time autoregressive generation. Building upon Matrix-Game 3.0, we present Matrix-Game 3.5, as shown in Figure 1, which advances real-time interactive world generation toward geometry-aware and long-horizon consistent simulation through three key improvements. First, we propose a unified geometry-aware memory framework, whose patch-memory and tiled-PRoPE components introduce no additional learnable parameters, combining explicit 3D patch retrieval with projective camera conditioning to enable geometry-consistent camera control and faithful long-horizon scene recall. Second, we introduce a static-dynamic disentangled world representation that separately models static scene geometry and dynamic subjects, preserving both geometric consistency and subject identity throughout long-horizon generation. Third, we develop a two-stage progressive real-time distillation framework that converts a bidirectional diffusion model into a few-step causal generator through Perceptual Flow Matching and curriculum based Self-Rollout DMD, enabling minute-long real-time interactive generation. Extensive experiments demonstrate that, with a unified training corpus spanning Unreal simulation environments, open-world games, and internet videos, MatrixGame 3.5 achieves strong performance in long-horizon scene recall, precise camera control, subject consistency, prompt-driven world generation, and stable real-time open-world interaction.
[CV-125] OrnaStyler: Ornament-Aware Latent Editing for Content-Preserving 3D Stylization
链接: https://arxiv.org/abs/2608.29905
作者: Tomohiro Aizawa,Shigeru Kuriyama,Chunzhi Gu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Text-guided style editing of 3D assets is essential for adapting existing objects to diverse visual aesthetics in digital content creation. Despite rapid progress in 3D shape modeling, faithfully stylizing an existing asset remains challenging when the desired stylization involves fine-grained structural ornamentation, which requires the model to preserve the source geometry and object identity, while coherently integrating new style-specific details. We propose \textbfOrnaStyler, a zero-shot framework for text-guided ornament-aware 3D stylization. Built upon rectified flow-based generative modeling, OrnaStyler introduces an inversion-guided editing strategy that recovers content-aware latent representations at both geometry and appearance levels in a staged manner to facilitate faithful editing. Our core idea is to explicitly model the spatial configuration of stylistic elements, thereby mitigating the fundamental tension between content preservation and style expression in the voxel space. Specifically, at the geometry level, we manipulate voxel representations through flow inversion to synthesize ornament-enhanced structures while preserving the spatial identity of the source asset. Then, at the appearance level, we introduce an adjacency-aware feature inpainting mechanism to harmonize newly generated ornaments with the original content, yielding coherent geometry-appearance integration. Our approach operates solely in the inference phase and enables selective editing over geometric augmentation or appearance stylization. Extensive experiments on both generated and real-world 3D assets against prior methods demonstrate that OrnaStyler achieves state-of-the-art editing performance in terms of content preservation, style fidelity, and overall visual realism. Code is available at: this https URL
[CV-126] Off-Manifold Refinement: Guiding Video Generators with a Frozen World Model BMVC2026
链接: https://arxiv.org/abs/2608.29904
作者: Hai Nguyen-Truong,Tuan-Anh Vu,Dang Huynh
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at BMVC 2026. Project page: this https URL
Abstract:Modern video generators routinely fail at physical dynamics: objects float, trajectories violate gravity, contacts vanish. Standard denoising and flow-matching objectives fit visual data distributions but do not explicitly penalize such physical violations. Existing remedies can improve physical consistency, but typically add substantial inference or training cost. Candidate-selection methods generate and score multiple videos, while gradient-based world-model guidance repeatedly decodes and re-encodes intermediate estimates. Generator-internal refinement adds perturbation and re-denoising loops, whereas post-training requires curated data and additional optimization. We propose Off-Manifold Refinement (OMR), an inference-time method that instead injects world-model feedback directly into a single sampling trajectory. During scheduled middle ODE steps, we augment the generator velocity with the gradient of an adapter-space V-JEPA 2.1 surprise energy. This external correction can move the latent away from the uncorrected sampling trajectory and toward regions ranked as more physically plausible by the frozen predictor, after which the generator continues rendering from the corrected state. A small trained latent-to-embedding adapter keeps the gradient tractable at inference, and both the video generator and the world model remain frozen. On our fixed 400-prompt VideoPhy-2 detailed subset, OMR lifts the joint Semantic-Adherence-and-Physical-Commonsense metric from 47.0% to 52.0% (+5.0pp absolute, +10.6% relative) over the base Wan2.2-T2V-A14B sampler. On a separate fixed 50-prompt efficiency subset, it requires 1.71 \times the base runtime rather than the multiplicative cost of reward/search alternatives. Project page: this https URL.
[CV-127] MASQ: Mask-Aware Spatiotemporal Quantization for Unsupervised Skeleton Action Segmentation
链接: https://arxiv.org/abs/2608.29891
作者: Xinyao Qin,Linxiang Peng,Youbao Ye,Di Yang,Jiangtao Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Unsupervised skeleton-based temporal action segmentation is a crucial task for understanding human behavior in long untrimmed sequences. Recent approaches often rely on discrete quantization to discover action boundaries from motion representations. However, when spatial masking is introduced for representation learning, it can introduce representation ambiguity, while discrete quantization further amplifies small fluctuations in the latent space. The interaction between these two factors often leads to unstable code switching and severe temporal jitter near action this http URL address these limitations, we propose a novel Mask-aware Action Spatiotemporal Quantization (MASQ) framework. Our framework decouples the conflicting tasks of spatial feature inference and temporal this http URL the spatial dimension, we introduce a Joint-Level Structured Dropout (JLSD) mechanism that masks the entire temporal trajectory of selected joints, to encourage the model to learn discriminative inter-joint coordination patterns. In the temporal dimension, we design a mask-aware velocity loss that enforces motion consistency only on visible joints, that prevents gradient conflicts caused by masked signals and stabilizing temporal predictions. Extensive experiments on three widely used skeleton datasets, including HuGaDB, LARa, and BABEL, demonstrate that the proposed MASQ framework significantly outperforms existing state-of-the-art unsupervised methods. In particular, our model establishes a comprehensive and substantial leading advantage in the Mean over Frames accuracy.
[CV-128] OptiGeo: Efficient Monocular Geometry for Embodied Perception in Optically Challenging Scenes
链接: https://arxiv.org/abs/2608.29881
作者: Muxin Liu,Tianbo Liu,Jing Xia,Xiaoyang Lyu,Xiaoshan Wu,Bo Wang,Peng Dai,Zhongrui Wang,Shaoshuai Shi,Xiaojuan Qi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Project Page: this https URL
Abstract:Monocular depth estimation has achieved strong open-domain generalization, yet reliable robotic deployment remains difficult in transparent, reflective, and specular environments, where depth sensors often produce missing or biased depth. Existing methods often handle such optical failures with scene-specific preprocessing, auxiliary modules, or post-hoc fine-tuning. While effective in constrained settings, these designs increase architectural redundancy and can over-specialize general geometry models to narrow optical scenarios. We revisit this problem as a localized failure mode within base-model training and identify sensor-induced supervision bias as a key bottleneck: models inherit sensor failure patterns from biased real-depth supervision in optically challenging regions. We then introduce OptiGeo, a bias-aware training framework that rehabilitates biased real supervision using a clean-geometry teacher and residual-trimmed alignment. We redefine transparency-targeted rendering as a compact source of clean optical geometry, rather than a large domain-specific fine-tuning set. With only a small targeted rendering set, OptiGeo learns the geometric structure of transparent objects and regions, correcting local geometry distortions that real sensors cannot reliably supervise. Despite only 30M parameters, OptiGeo outperforms substantially larger 300M-scale monocular models and billion-scale multi-view baselines on transparent-scene benchmarks, while remaining competitive on general zero-shot depth and boundary sharpness. Real-world navigation cases further validate its practicality as an efficient perception module in optically challenging scenes.
[CV-129] MariSat: A Maritime Dataset for Instance Segmentation of Objects in Satellite and Aerial Images
链接: https://arxiv.org/abs/2608.29852
作者: Amir Abbes,Ines Harrabi,Lucas Justin Yirepoa Kinda,Rim Trabelsi,Adnane Cabani,Fatma Abdelkefi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Automated maritime surveillance from satellite and aerial imagery requires large, precisely annotated datasets, which remain scarce for the instance-segmentation task, particularly for small vessels in cluttered port environments. We present MariSat, a new benchmark dataset of 1260 aerial and satellite images covering diverse port and coastal scenes, annotated at the pixel level for eight maritime object classes (sailboat, yacht, jet-ski, fishing boat, cruise ship, military vessel, tugboat and cargo ship). The dataset was produced through a semi-automatic annotation pipeline combining the textpromptable segmentation model SAM 3 with a cascade of geometric and colorimetric post-processing filters, followed by a manual correction and quality-control pass performed with the CVAT annotation platform. We describe the image-collection methodology, the annotation and correction process, and the resulting data organization. We also report class-wise statistics for the training, validation, and test splits. MariSat has already been used to fine-tune and benchmark segmentation and detection models (SAM 3 and YOLO11) for real-time maritime monitoring. We report detailed quantitative and per-class results for both tasks. The MariSat dataset is publicly available on GitHub : this https URL
[CV-130] Null-Space Diffusion Restoration with Adaptive Uncertainty-Guided Fusion for Ultrasound Speckle Reduction
链接: https://arxiv.org/abs/2608.29820
作者: Juneyong Lee,Jaeyoung Choi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 14 pages, 4 figures, 5 tables. Accepted for publication in IEEE Access
Abstract:Ultrasound B-mode imaging commonly suffers from speckle noise and artifacts, requiring a delicate balance between contrast, resolution, and preservation of anatomical structures. Although recently developed despeckling methods have achieved some progress, supervised learning approaches remain fundamentally limited by the ground truth paradox, which arises from the absence of noise-free, ground truth reference images in in vivo scenarios. Existing unsupervised diffusion-based methods typically enforce data consistency directly in the nonlinear log-compressed domain, which can disproportionately amplify background artifacts when mapped back to the envelope domain. To overcome these limitations, we propose an uncertainty-guided null-space diffusion (UGNS) framework, a novel label-free solution that enforces consistency correction on a stabilized positive-envelope proxy obtained via inverse log compression. The proposed UGNS introduces several technical novelties: (a) extraction of a structural prior in the stabilized envelope domain to produce a robust signal envelope that preserves anatomical structure, (b) development of an adaptive range-null reconstruction mechanism that uses an adaptive weight mask to preserve tissue regions via range-space projection, and © introduction of uncertainty-guided fusion in an adaptive way to mitigate sampling variability. Extensive and comparative experiments were conducted using the PICMUS benchmark and in vivo datasets. The results demonstrate that UGNS achieves competitive generalized contrast-to-noise ratio (gCNR) values across diverse datasets. In addition, it is successfully validated that UGNS effectively suppresses speckle noise while preserving fine spatial resolution. Code is available at this https URL.
[CV-131] PhasorNet: Learning Structure from Frequency for Real-Time Stereo Matching
链接: https://arxiv.org/abs/2608.29819
作者: Md Raqib Khan,Santosh Kumar Vipparthi,Subrahmanyam Murala
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Accurate stereo matching remains challenging in ill-posed regions such as fine structures, reflective, or transparent objects, where appearance cues are often ambiguous or unreliable. To tackle this, we propose PhasorNet, a lightweight yet powerful framework that boosts geometric discrimination via frequency-domain cues. At its core, the Phase-Augmented Transformer (PAT) injects Fourier-derived phase information into the attention mechanism, yielding photometrically robust, structure-preserving features that prioritize structural consistency in difficult areas. Additionally, we develop a Geometry-Context Fusion Refinement Module (GCFRM) that combines a full-resolution convolutional stream with a lightweight attention-based stream (leveraging WQA and CDGA blocks) to efficiently preserve fine details and object boundaries without excessive overhead. Training is further enhanced by a multi-scale Edge-guided High-Error Region (EHR) loss that adaptively focuses optimization on high-error and edge regions, guiding hierarchical cost volume refinement. With only 5.3M parameters, PhasorNet achieves state-of-the-art performance on the challenging ETH3D benchmark while exhibiting excellent cross-domain generalization on KITTI, delivering an efficient and practical solution for accurate real-time stereo matching.
[CV-132] RegionCache: Semantic-Aware Region Reuse for Efficient Multi-Turn Image Generation IJCAI2026
链接: https://arxiv.org/abs/2608.29809
作者: Peizheng Li,Xin Ai,Hanyuan Liu,Qiange Wang,Yanfeng Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at IJCAI 2026
Abstract:Real-world image generation often involves multi-turn editing, where users iteratively modify small regions while most image content remains unchanged. However, existing diffusion transformer (DiT)-based editing pipelines recompute the entire image at every turn, causing substantial redundant computation. Existing DiT acceleration methods further ignore semantic correspondence across prompts, leading to unnecessary recomputation or unsafe reuse that harms editing quality. To address this, we propose RegionCache, a semantic-aware reuse framework for multi-turn image editing that selectively reuses diffusion states from unchanged regions. RegionCache detects reusable regions through semantic overlap between consecutive prompts and cross-attention localization, and adopts an adaptive reuse schedule based on prompt similarity and contextual consistency. Experiments on PixArt-alpha demonstrate that RegionCache achieves 1.43x–2.55x end-to-end speedup while maintaining comparable image quality. Code is available at this https URL.
[CV-133] Beyond Global Realism: Virtual Try-On Evaluation and Optimization with Dimension-wise Garment Fidelity Assessment
链接: https://arxiv.org/abs/2608.29804
作者: Kaidong Zhang,Yukang Ding,Xiaoyu Liu,Ying Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 12 pages, 6 figures
Abstract:Virtual try-on (VTON) requires not only realistic generation but also faithful preservation of garment characteristics. However, existing evaluation metrics such as PSNR, SSIM, KID and FID struggle to measure the consistency between the generated and reference garments, particularly in capturing the multi-dimensional characteristics of garment fidelity. To address this, we propose DAT: a Dimension-wise Assessment framework for virtual Try-on, which decomposes garment consistency into seven interpretable dimensions: silhouette, color, neckline and sleeve shape, major decoration and structure, material texture, fine-detail fidelity, and logo preservation, each formulated as a discrete attribute-level prediction task. To train this specialized assessment model, we adopt a two-stage learning paradigm comprising large-scale weak supervision on 50K samples, followed by refinement on 10K higher-quality annotations obtained via multi-model voting. Furthermore, we employ weighted cross-entropy loss to mitigate the severe label imbalance inherent across evaluation dimensions. Beyond its role as an evaluation framework, the assessment model can be integrated into reinforcement learning optimization of Qwen-Image-Edit for VTON, where dimension-wise rewards are adaptively aggregated to emphasize under-optimized aspects during training. Experimental results show that our method (8B parameters) achieves state-of-the-art performance in terms of balanced accuracy, SROCC, and PLCC, outperforming strong proprietary models such as Gemini-3.1, Qwen3.7-plus, and GPT-5.5, while also serving as an effective optimization signal for reward-guided VTON generation
[CV-134] Foundation and Multimodal Large Language Models for Face Presentation and Morph Attack Detection
链接: https://arxiv.org/abs/2608.29802
作者: Hatef Otroshi Shahreza,Asif Hussain Khan,Peter Lorenz,Alain Komaty,Sébastien Marcel
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Face recognition systems are increasingly deployed in security-critical applications, yet they remain vulnerable to presentation and morph attacks. Presentation attack detection (PAD) and morphing attack detection (MAD) are therefore essential components of trustworthy face biometrics. Despite advancements in PAD and MAD methods, existing detectors suffer from limited generalization and degrade in cross-dataset evaluation. In this paper, we systematically investigate whether general-purpose foundation models (FMs) and multimodal large language models (MLLMs) encode PAD-relevant and MAD-relevant information, and how such models can best be deployed for both tasks. We study five approaches with increasing access to the internal information of the model: (i) zero-shot prompting of off-the-shelf MLLMs; (ii) training a shallow model on the next-token logit probabilities at the output of the MLLM; (iii) parameter-efficient fine-tuning on task-specific question-answer data, yielding two specialized MLLMs, called PADLLM and MADLLM, which additionally provide textual reasoning for their decisions; (iv) linear probing of frozen vision encoders; and (v) fine-tuning of vision encoders of FMs and MLLMs. We benchmark 16 open-weight MLLMs and 30 vision encoder backbones on four PAD datasets (MSU-MFSD, CASIA-FASD, Replay-Attack, and OULU-NPU) and four MAD datasets (FFHQ, FRGC, FRLL, and FERET). Our experiments show that FMs and MLLMs can achieve significant performance for PAD and MAD. In addition, the fine-tuned models achieve state-of-the-art detection performance in cross-dataset evaluation, indicating that general-purpose pretrained representations carry substantial attack-relevant information. Source code of all our experiments will be publicly released.
[CV-135] Source-Dependent Deference in Medical Imaging Agents Under Falsified Findings: A Pilot Audit
链接: https://arxiv.org/abs/2608.29800
作者: Ridam Roy,Md Shahriar Rashid,Md. Rajib Mia
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Tool-using agents are being proposed for medical imaging, and their behaviour when a tool returns a false finding is largely unmeasured. We audit whether a ReAct-style tool-calling agent abandons an answer it has already given correctly once a falsified finding arrives, and whether that depends on how the finding is presented. On 20 VQA-RAD closed questions across four vendor-designated model tiers, the agent commits to an answer from the image alone; a negated finding is then delivered either as JSON from an analyze_image tool the agent invokes itself, or as quoted prose attributed to a radiologist. Our outcome is the commission-error rate over cases answered correctly without any tool. Deference is much higher under the prose-attributed claim: at the strongest tier the agent revised its correct answer in 10 of 13 cases against 1 of 13 under the tool (exact McNemar p=0.0039, Holm-adjusted 0.012). We do not claim this isolates the source label. Attribution travels with the delivery channel in our design, and exposure differs because the tool claim reaches the agent only when it calls the tool. The finding is a joint source-and-delivery asymmetry from a small-scale pilot whose pre-specified stopping rule was not met.
[CV-136] GridFlow: Structured Latent Flow for Seamless City-Scale 3D Point Cloud Generation
链接: https://arxiv.org/abs/2608.29793
作者: Xinyu Wang,Muhammad Ibrahim,Atif Mansoor,Ajmal Mian
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Generating realistic 3D city environments from remote sensing data is important for simulation, urban planning, and mixed reality, yet existing point cloud generation methods are limited to single objects or bounded indoor scenes and cannot handle the scale, seamless tiling, and partial observability challenges of city-scale generation. We present \ours, a multi-stage framework that generates dense, colored point clouds ( 10^5 points per 150\textm\times150\textm tile) at city scale, conditioned on satellite imagery, semantic segmentation maps, and digital surface models (DSM). A \emphGrid-Aligned VAE encodes each tile into a topology-preserving latent grid where tokens correspond to fixed spatial regions, enabling spatially coherent multi-modal conditioning and compact latent-space edge consistency that implicitly aligns thousands of boundary points for seamless cross-tile generation. A conditional rectified flow model synthesizes geometry latents from the fused multi-modal conditions, and an orientation-aware diffusion colorizer separately handles satellite-visible horizontal surfaces and occluded vertical façades. To support standardized evaluation, we build on public 3D data sources to introduce \emphCity3D-MultiGen, a benchmark of 163 K densely annotated tiles from Melbourne and London with aligned point clouds, satellite images, semantic maps, and elevation data. Experiments show that \ours outperforms adapted point cloud generation baselines across all geometry metrics and produces visually coherent colored point clouds with seamless boundaries over arbitrarily large urban extents. Our benchmark details are available at this https URL
[CV-137] InspectorGPT : A Comparative Reasoning Enhanced VLM for Comprehensive Industrial Anomaly Detection
链接: https://arxiv.org/abs/2608.29783
作者: Weifei Chen,Honghao Zhang,Zhiyuan You,Xinyi Le
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Industrial anomaly detection is a critical component of modern manufacturing. Most traditional unsupervised methods rely on modelling normal feature distributions, inherently limiting generalization to unknown categories. To improve generalizability, some recent methods incorporate vision-language models (VLMs) for zero-shot detection via text prompts. However, we observe that reasoning-oriented post-training can cause anomaly discrimination to collapse, with some fine-tuned models performing worse than their base VLMs. Existing methods also provide only textual decisions or coarse boxes, without pixel-level segmentation. A more explicit detection principle comes from human inspection: anomalies are identified by comparing a query image with a defect-free reference. Inspired by this, we propose InspectorGPT, a VLM framework centered on comparative reasoning. Given a normal reference and a query image, InspectorGPT compares them to identify discrepancies and perform multiple inspection tasks with detailed reasoning. We internalize this capability through Chain-of-Thought (CoT) fine-tuning and Group Relative Policy Optimization (GRPO) with tailored, verifiable rewards. We further introduce InspectorGPT-Seg for pixel-level anomaly masks. Segmentation supervision improves anomaly discrimination but weakens semantic reasoning, while joint training fails to balance them. We therefore train the two branches separately and combine them through task-vector fusion. Extensive experiments demonstrate superior multi-dimensional performance and generalization to unseen benchmarks, validating comparative reasoning for comprehensive industrial inspection.
[CV-138] SynCrash: A Multi-Stage Pipeline for Zero-Shot Accident Detection and Localization in Traffic Surveillance Video CVPR2026
链接: https://arxiv.org/abs/2608.29759
作者: Arkya Jyoti Bagchi,Ritul Jangir,Varun Raskar
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted at the CVPR 2026 AUTOPILOT Workshop (non-archival)
Abstract:We present SynCrash, a multi-stage pipeline for zero-shot accident detection, spatial localization, and collision-type classification in fixed-view CCTV surveillance video. Our approach addresses the ACCIDENT at CVPR 2026 Challenge, which requires predicting when an accident occurs, where in the frame the impact happens, and what type of collision it is, all without access to labeled real-world training data. The pipeline operates in three decoupled stages: (1) Temporal localization via a VideoMAEv2-giant backbone fine-tuned on CARLA-based synthetic clips with metadata-aware embeddings and dense sliding-window inference; (2) Spatial localization using YOLO for object detection combined with a physics-informed hybrid heuristic that leverages bounding-box overlap and trajectory-based reasoning to predict the impact point; and (3) Collision-type classification using a lightweight rule-based strategy derived from the number and configuration of detected vehicles. The key insight is that temporal understanding benefits from supervised fine-tuning on synthetic data, whereas spatial understanding is better served by pretrained object detectors and physics priors that transfer naturally across domains.
[CV-139] Drift Calibration in Geometric Eye Tracking Systems
链接: https://arxiv.org/abs/2608.29739
作者: Jiaqi Liu,Zixuan Wang,Yuhong Zhang,Dingkang Liang,Jane Hanqi Li,Tzyy-Ping Jung,Gert Cauwenberghs
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 11 pages, 6 figures, 3 tables
Abstract:Geometric eye trackers can provide the spatial accuracy required for gaze-based interaction and multimodal studies, but their measurements remain sensitive to residual session-specific calibration error. Research on correcting this error is difficult to compare because methods are typically evaluated with different devices, target layouts, and error definitions. We present a calibration-focused dataset containing 163 trials from 12 participants, with separate 18-point fitting and 32-point test grids, and use it to evaluate global, local, and composite correction functions under a common spatial-extrapolation protocol. We further introduce a lightweight neural refiner that combines ranked predictions from complementary calibrators. On this controlled dataset, post-vendor correction reduces the mean angular error from 1.53^\circ to 1.03^\circ with the strongest classical composite and to 0.96^\circ with the refiner. In a closed-loop gaze task, lower residual error is associated with higher performance across four online correction conditions. These results provide a reproducible data-quality benchmark for using gaze as a behavioral signal in interactive modeling.
[CV-140] XDG: Accelerated Visual Disambiguation
链接: https://arxiv.org/abs/2608.29733
作者: Gonglin Chen,Ben Southall,Hanyuan Xiao,Wenbin Teng,Haolin Xiong,Tianwen Fu,Junyi Ouyang,Kshitij Singh Minhas,Supun Samarasekera,Rakesh Kumar,Yajie Zhao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Visual aliasing, also known as the doppelganger problem, remains a key challenge for structure-from-motion (SfM): visually similar but physically distinct surfaces can produce incorrect image matches and degrade reconstruction quality. Previous work mitigates this issue with geometry-aware foundation-model features, but places a heavy transformer classifier on top of the backbone, making large-scale disambiguation expensive. We introduce XDG, an efficient visual disambiguation model designed for scalable SfM. Our key observation is that a 3D foundation model already performs the cross-view geometric reasoning necessary for visual disambiguation, so doppelganger classification should adapt the backbone representation directly rather than relearn pair reasoning in a separate heavy decoder. XDG fine-tunes Depth Anything 3 with lightweight LoRA adapters and repurposes its camera tokens as compact pair-level classification tokens. A compact MLP head predicts whether a candidate image pair observes the same 3D surface. Extensive experiments show that XDG provides a favorable accuracy-efficiency tradeoff: it remains competitive with the state-of-the-art disambiguation method across pairwise and reconstruction benchmarks and delivers more than a 3x inference speedup. On individual LaMAR scenes containing thousands of images, XDG saves more than 10 hours of visual disambiguation processing. Code is available at this https URL.
[CV-141] A Calibration Audit of Confidence in Feed-Forward 3D Reconstruction
链接: https://arxiv.org/abs/2608.29705
作者: Nanxing Nick Deng,Qing Cheng,Niclas Zeller,Daniel Cremers
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Feed-forward 3D reconstruction models emit a per-pixel confidence that downstream systems read as a reliability signal. It is trained as a loss weight, not as an uncertainty magnitude, and whether it can be used as an error prediction has not been measured. We audit seven released backbones on thirteen datasets and score the confidence on four properties, how well it ranks error, whether its level is right on average, whether it holds across the confidence range, and whether its intervals cover the truth. The confidence ranks error well, but the predicted uncertainty is too low when it is read under conditions that are not exactly those of training. The median case is off by 2.4x across all seven models, and the error prediction is further off the more confident the model is. We show that this phenomenon can appear even though the loss’s optimum is reached. A released model resumed under its own loss reaches that optimum on its training data within a few hundred updates and stays overconfident on unseen frames. A power law with two constants per backbone and dataset corrects the overall magnitude of the predicted uncertainty and leaves the ranking untouched. What no rescaling reaches is the scene, which we attribute to the model’s missing knowledge of scale across predictions. Every correction we tried is close to right on average and still leaves two thirds of held-out scenes outside a five-point band, because what a scene is missing is a shape rather than a shift. We release the audit protocol, its results, and the fitted constants per model and dataset. Fitted with the target dataset held out, the constants bring the median case from 2.4x off to 1.35x, and a refit on a few labelled scenes of that dataset reaches 1.12x.
[CV-142] GeoRay: Gauge-Aware Feed-Forward Satellite 3D Reconstruction in the Geodetic Frame
链接: https://arxiv.org/abs/2608.29680
作者: Zhe Dong,Wanqing Wu,Yuzhe Sun,Haochen Jiang,Yuchen Ma,Lecheng Ren,Tianzhu Liu,Yanfeng Gu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Feed-forward 3D foundation models reconstruct perspective scenes in one pass. Satellite photogrammetry needs a different product, one that domain adaptation alone does not deliver: dense surface height in an absolute geodetic frame under non-central rational polynomial cameras (RPCs). Perspective-pretrained features are not reliably observable along RPC height rays, absolute elevation carries a low-order height–datum gauge exchangeable with sensor bias to first order, and monocular and multi-view cues fail in different regions. \method treats all three. Lightweight ray-consistent adapters make a frozen backbone matchable along native RPC rays. An explicit datum mechanism separates relief from absolute level and is equivariant to the vertical origin by construction, so one trained model serves zero-, one-, and sparse-control inference. Calibrated inverse-variance fusion combines the two relief streams. \bench, our absolute-frame benchmark of eighteen systems across in-domain, cross-dataset, and cross-city tiers, scores absolute placement without registration or test-reference leakage. On 26 held-out US3D tiles, \method attains 2.99 ,m absolute MAE at 91.9% coverage, improves completeness-aware accuracy by 46.4 points over the strongest compliant feed-forward baseline, remains the most accurate such system under both transfer shifts, and runs in 24 ,s model-forward time per tile. Code and models will be released at this https URL
[CV-143] MedSegBenchmarker: A Raw-Count-First Framework for Controlled 2D Medical Image Segmentation Benchmarks
链接: https://arxiv.org/abs/2608.29677
作者: Vanessa Borst,Lukas Horn,Daniel Grillmeyer,Thomas Prantl,Samuel Kounev
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 10 pages, 4 figures;
Abstract:Despite rapid advances in MIS, fair and reproducible comparisons of segmentation models remain challenging due to heterogeneous datasets, inconsistent evaluation protocols, and rapidly evolving architectures. In particular, comparisons often implicitly assume that model rankings are invariant to data partitioning, preprocessing, metric aggregation, uncertainty estimation, and computational constraints. The lack of extensible and unified evaluation frameworks further limits systematic investigation of new models, datasets, and training paradigms. We present MEDSEGBENCHMARKER (MSB), a configuration-driven framework for controlled benchmarking of 2D MIS. It integrates duplicate and near-duplicate image detection, group-aware data splitting, YAML study specifications, resumable training, hyperparameter optimization, cross-validation, and checkpoint-based evaluation. Rather than retaining only aggregate performance measures, MSB exports sample- and class-level pixel counts and predictions together with the evaluation context. These elementary artifacts enable post-hoc analyses without repeated inference. We demonstrate MSB in a case study involving three heterogeneous 2D datasets and multiple MIS and general-purpose vision models evaluated at 256- and 512-pixel input resolutions. Reaggregation of identical predictions changes the top-ranked architecture in three of six dataset-resolution settings, despite high rank correlations between aggregation strategies. Increasing input resolution produces model- and dataset-dependent performance gains and losses that must be considered alongside empirically measured inference complexity. These results show that seemingly minor choices in evaluation and experimental setup can affect benchmark conclusions. MSB, available at GitHub, provides a practical and extensible basis for making benchmark conditions and evaluation choices explicit and reproducible.
[CV-144] PhysVR: Vision-Language Model Guided Interference-aware Temporal Feature Refinement for Remote Physiological Measurement
链接: https://arxiv.org/abs/2608.29663
作者: Zixu Li,Jianjun Qian,Hang Shao,Daoheng Li,Lei Luo,Jian Yang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Remote photoplethysmography (rPPG) enables contactless physiological measurement from facial videos, yet its subtle pulse-related variations are easily affected by illumination variation, head motion, facial blur, and region-of-interest instability. Existing methods mainly suppress interference during feature learning, while whether the learned temporal features remain affected by interference and how to further suppress such interference before rPPG estimation are rarely examined. To address this limitation, we propose PhysVR, a vision-language model guided interference-aware temporal feature refinement framework for rPPG estimation. Specifically, a physiological backbone produces global temporal features and a coarse rPPG prediction, from which signal-derived physiological reliability evidence is constructed from local temporal characteristics. In parallel, a frozen vision-language model processes sampled facial frames under an interference-oriented prompt, and an evidence head extracts visual interference evidence from the VLM output. Temporal cross-attention integrates the physiological and visual evidence with the global temporal features to construct interference-aware temporal context. Guided by this context, a shared temporal correction unit performs general refinement, while four interference-specific experts selectively suppress different interference through adaptive routing. The refined temporal features are then used for final rPPG estimation. Extensive experiments on five public benchmarks demonstrate that PhysVR consistently outperforms representative methods under both intra-dataset and cross-dataset evaluation protocols.
[CV-145] SPLG-Mamba: Structure-Preserving Local-Global Mamba Network for Salient Object Detection in Optical Remote Sensing Images
链接: https://arxiv.org/abs/2608.29626
作者: Yi Xu,Ruichao Hou,Tongwei Ren,Gangshan Wu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Salient object detection in optical remote sensing images (ORSI-SOD) requires dense predictions that preserve object completeness and structural continuity under complex backgrounds, scale variation, and irregular object shapes. Existing methods often localize salient regions, but their predictions may still suffer from structural degradation, including fragmented, incomplete, or locally missing foreground responses. This degradation is closely related to hierarchical feature propagation, where shallow details can introduce texture-induced background responses, deep semantics may over-smooth weak structures, and uncontrolled cross-scale fusion can disturb coherent regions. To address this issue, we propose a novel Structure-Preserving Local-Global Mamba Network, SPLG-Mamba, for ORSI-SOD. Specifically, SPLG-Mamba integrates Smooth-Detail Recalibration (SDR), hierarchy-aware Local-Global Mamba, and Gated Cross-Scale Fusion (GCSF). SDR recalibrates smoothed responses and detail residuals before state-space modeling, Local-Global Mamba assigns local modeling to shallow feature levels and global modeling to deep feature levels, and GCSF controls cross-scale detail injection during decoding. Experiments on ORSSD, EORSSD, and ORSI-4199 demonstrate state-of-the-art results and improved structural completeness and continuity. The code is available at this https URL
[CV-146] CineForge: Self-Improving Agents for Long-Horizon Video Generation
链接: https://arxiv.org/abs/2608.29621
作者: Junxiang Liu,Lin Wang,Haiyu Shi,Hongxu Ma,Xiaoyu Yang,Chunjie Chen,Xiaoxiao Xu,Kaiqiao Zhan,Boao Wang,Shuizhou Shi,Tianyun Zhu,Jie Li,Jiangtong Li
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Long-horizon story-driven video generation requires a production agent to coordinate narrative decomposition, state tracking, shot design, prompt construction, rendering, and revision across interdependent scenes. Existing adaptive video systems primarily refine requests or reusable skills, leaving recurring production failures disconnected from persistent, stage-targeted improvements across stories. We introduce CineForge, a self-evolving video-production agent framework that couples CineForge-Produce for video generation with CineForge-Evolve for cross-story policy evolution. CineForge-Produce organizes each source story into typed narrative, character, spatial, and cinematic states, uses them to coordinate asset and clip generation, and records the process as a canonical production trajectory. CineForge-Evolve applies Case-to-Pattern-to-Policy Evolution (CPPE) to review trajectory evidence, consolidate recurrent findings into bounded stage-local patches, and deploy validated updates through structural replay and confidence-controlled paired evaluation. To measure complete story realization, we introduce CineScope, which combines a 100-script CineScope-Data suite with a human-aligned, multiscale CineScope-Metric spanning causal state, directorial orchestration, pacing and resource allocation, and character arc. Across CineScope-Data and two public benchmarks, the evolved CineForge policy improves CineScope-Metric from 4.024 to 4.380, outperforms three long-video baselines with consistent gains under ScriptAgent, and reduces review LLM calls by 37.0% on new stories. These results establish production trajectories as actionable experience for video agents that improve cumulatively across long-form storytelling tasks.
[CV-147] See the Change Keep the Flow: Unsupervised Action Segmentation via Spectral-Temporal Representation Learning
链接: https://arxiv.org/abs/2608.29611
作者: Yun Li,Jun Xiao,Cong Zhang,Kin-Man Lam
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Unsupervised action segmentation aims to discover latent action categories and their temporal organization without action annotations. Optimal transport-based methods provide structured frame-to-action assignments, however, their pseudo-label quality is fundamentally conditioned on the representation space used to construct the transport cost. We argue that reliable OT pseudo-labeling requires a representation geometry that is simultaneously sensitive to discriminative action changes and coherent along local temporal progressions. Based on this insight, we propose SpecT-OT, a spectral-temporal representation learning framework built upon an unbalanced optimal transport pseudo-labeling concept. SpecT-OT introduces a Spectral Reparameterization Projector (SRP), which parameterizes projector weights with fixed Fourier bases and learnable coefficients to improve the modeling of rapidly varying discriminative features, and Temporal Affinity Regularization (TAR), which imposes distance-aware, label-free constraints on pairwise frame affinities to stabilize local temporal structure. The two components jointly produce more discriminative and temporally stable transport costs, yielding more reliable pseudo-labels for iterative representation learning. Experiments on four benchmarks demonstrate strong performance compared with state-of-the-art methods. SpecT-OT achieves the best results on 13 of 15 metrics, including 4.1-point MoF and 7.4-point F1 gains over the baseline on Breakfast and Desktop Assembly, respectively.
[CV-148] nnMNet: Baseline for Martian Terrain Semantic Segmentation
链接: https://arxiv.org/abs/2608.29609
作者: Ming-Han Lee,Chi-Yeh Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Semantic segmentation is a crucial task for understanding Mars, the most Earth-like planet in our solar system. However, it is challenging because the Martian surface is highly unstructured and complex, making accurate pixel-level prediction and fine-grained annotation difficult. Recent advancements in deep learning have introduced numerous methods and datasets to address these challenges. Nevertheless, the field lacks a robust, publicly available, and reproducible baseline, as well as a unified benchmark to facilitate fair evaluations. In this work, we present nnMNet, a new baseline model designed for Martian terrain semantic segmentation. Building upon nnWNet, we integrate linear attention to better capture global context and employ lightweight convolutions to reduce computational overhead. To bridge the gap between local and global representations, we introduce the Spatially-Aware Fusion Block (SAFB), which augments and combines features with diverse characteristics. Furthermore, we establish a new benchmark by curating and standardizing three high-quality datasets for thorough evaluation. nnMNet achieves new state-of-the-art 86.61%, 83.25%, and 88.24% mIoU on SynMars-TW, SynMars-Air, and MarsScapes, respectively. Our code, models, and datasets are publicly available at this https URL.
[CV-149] mathcalN_0-Foundation: Towards the Age of Tactile Intelligence
链接: https://arxiv.org/abs/2608.29601
作者: NeoteAI Team,Fudan TEAI Team
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 13 figures, 5 tables
Abstract:We present \mathcalN_0 -Foundation, a paradigm for tactile-enabled embodied manipulation, which integrates tactile sensing hardware, large-scale multimodal data, tactile representation learning, and standardized evaluation. First, we engineer the infrastructure for scalable data collection, including a vision-based tactile sensor, a tactile Universal Manipulation Interface (UMI), and a synchronized visuo-tactile data collection system supporting both robot embodiments and UMI-based demonstrations. Leveraging this infrastructure, we construct NeoData, which contains more than 30000 hours of synchronized visual and tactile demonstrations, spanning six embodiments, 450 tasks, and billions of paired RGB and tactile frames collected through a mixture of real-robot teleoperation and UMI-based demonstrations. To facilitate open research, we further release OpenNeoData, a 5000-hour open-source subset of NeoData. The dataset addresses a central limitation of existing manipulation corpora, critical for deformable-object manipulation, precise assembly, delicate force control, and sustained surface interaction. Capitalizing on the large-scale, heterogeneous tactile measurements, we propose NeoForce, a visuo-tactile representation model that learn transferable tactile representations across different sensor designs. To enable systematic evaluation of tactile embodied models built upon our infrastructure, datasets and tactile representations, we further propose a comprehensive benchmark, which combines the real-world NeoReal suite and the simulated NeoSim suite for standardized evaluation. Experiments across both suites show that policies benefit from the physical contact state rather than from the device-specific appearance of the tactile signal. We release the dataset, the representation, and the benchmark, aiming at supporting future work on tactile-enabled embodied manipulation.
[CV-150] Guardrail-Agnostic Societal Bias Evaluation in Large Vision-Language Models ECCV2026
链接: https://arxiv.org/abs/2608.29590
作者: Yusuke Hirota,Michael Ross Boone,Arun George Zachariah,Jibin Rajan Varghese,Yu-Chiang Frank Wang,Boyi Li,Ryo Hachiuma
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at ECCV 2026
Abstract:We propose a societal bias evaluation method for large vision-language models (LVLMs) in the era of strong safety guardrails. Existing benchmarks rely on prompts that ask models to infer attributes of people in images (e.g., “Is this person a CEO or a secretary?”). However, we find that LVLMs with strong guardrails, such as GPT and Claude, often refuse these prompts, making evaluations unreliable. To address this, we change the prior evaluation paradigm by decoupling the task from the depicted person: instead of inferring person’s attributes, we use prompts that do not ask about the person (e.g., “Write a fictional story about an imaginary person.”) and attach the image as provisional user information to implicitly provide demographic cues, then compare outputs across user demographics. Instantiated across three tasks — story generation, term explanation, and exam-style QA — our method avoids refusals even in guardrailed LVLMs, enabling reliable bias measurement. Applying it to 20 recent LVLMs, both open-source and proprietary, we find that all models undesirably use user demographic information in person-irrelevant tasks; for instance, characters in stories are often portrayed as mechanic for male users and nurse for female users. Although still biased, proprietary models like GPT-5 show lower bias than open-source ones. We analyze potential factors behind this gap, discussing continuous model monitoring and improvement as a possible contributor for reducing bias.
[CV-151] RINITY: A Multi-Perspective Benchmark for Personal-Style Video Highlight Detection ECCV2026
链接: https://arxiv.org/abs/2608.29577
作者: Qianqian Chen,Hyun Bin Kim,Denzel Elden Wijaya,Yang Yi,Bo Liu,Yangkai Ding
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 32 pages, 9 figures. Accepted to ECCV 2026
Abstract:Traditional video highlight detection relies on a narrow, event-centric definition of saliency, which often fails to generalize to unconstrained personal videos where highlights are heterogeneous and perspective-dependent. To address this, we introduce TRINITY, a multi-perspective benchmark that decomposes highlight saliency into three complementary dimensions, Event, Emotion, and Nature, within a unified temporal framework. Leveraging this multi-faceted view, we propose a shared-backbone multi-branch architecture designed for parallel multi-perspective prediction via view-specific experts. Comprehensive experiments demonstrate that our method significantly outperforms state-of-the-art baselines, achieving gains of +7.15/+3.62 mAP (rho=15%/50%) on Mr. HiSum and +10.82 mAP on YouTube Highlights. These results validate that multi-perspective modeling provides a more robust and comprehensive formulation of video saliency, especially for complex real-world scenarios. The benchmark and relevant codes will be released upon acceptance. The benchmark is available at this https URL and the code is available at this https URL.
[CV-152] MotionSync: Non-Causal Refinement of Causal Tracker for Label-Efficient 3D Perception
链接: https://arxiv.org/abs/2608.29567
作者: Rahul Ahuja,Bala Murali Manoghar Sai Sudhakar,Shashwata Gupta,Venkatraman Narayanan,Varun Ravi Kumar,Senthil Yogamani
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Three-dimensional box-and-track annotation is the cost bottleneck in autonomous-driving data engines, and the offline systems built to relieve it replace the online perception stack outright, so a team needing both regimes maintains and reconciles two. MotionSync makes the causal/non-causal boundary an explicit architectural seam instead. A strictly causal tracker, built on a strong published baseline and extended with innovation-driven uncertainty calibration, frame-rate-invariant kinematic association gates, and multi-hypothesis motion with learned mode selection, emits a valid online result. A non-causal pass then revises the buffered trajectories with Rauch–Tung–Striebel smoothing applied separately to pose, extent and yaw, physics-validated gap completion, and semantic pruning of ghost tracks against LiDAR point labels. The refiner never writes back, so one system serves both regimes and refinement’s effect is a delta over an unaltered causal estimate. Used as an auto-labeller, a fixed 3D detector trained on 25% human labels plus MotionSync pseudo-labels reaches 96.9% of its full-supervision mean average precision (mAP) on Waymo, and at a 10% budget the non-causal pass accounts for +3.3 mAP/L2 over pseudo-labels from the same tracker’s causal stage. Re-fitting the online tracker on its own refined output recovers 73% of the benefit of human supervision, while its causal output is worse supervision than no re-fitting at all. As a tracker MotionSync is at parity with the leading published offline entries on the headline metric and ahead of them on error composition, which is where a refinement pass can act at all: it reduces misses and fragmentations together, the signature of gap completion rather than of a tuned detector.
[CV-153] NepScript Genesis: Neural Architecture Search for Handwritten Devanagari Digit Synthesis
链接: https://arxiv.org/abs/2608.29540
作者: Mausam Gurung,Prabin Neupane,Sajjan Acharya
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 11 pages, 4 figures, 3 tables
Abstract:This paper introduces NepScript Genesis, a Neural Architecture Search (NAS) framework for automated Generative Adversarial Network (GAN) discovery, applied to conditional Devanagari handwritten digit synthesis. We compare five NAS strategies against a carefully constructed Deep Convolutional GAN (DCGAN) baseline (FID=332.28). Architecture selection utilizes a two-stage pipeline guided by a novel domain-aware evaluation metric (Enhanced Score). Results demonstrate that Adaptive Exploration achieves the optimal quality-efficiency trade-off, attaining an FID of 79.12 – a 76.19% improvement over the baseline – and the highest mode coverage among the NAS strategies (Recall=0.531) in under one GPU-hour. Furthermore, we demonstrate that incorporating script-specific structural heuristics into the search phase prevents early-stage mode collapse. In a downstream low-resource evaluation, augmenting 250 real training samples per class with GAN-generated digits from the best NAS model improves CNN classification accuracy from 91.0% to 96.5% (+5.5 percentage points), demonstrating that NAS-optimized synthesis produces digits of sufficient quality to benefit practical recognition pipelines when real data is scarce.
[CV-154] FuncRoom-Agent : Sequential Feed-Forward 3D Functional Indoor Scene Generation
链接: https://arxiv.org/abs/2608.29519
作者: Hao Feng,Zhi Zuo,MingJian Liang,Jingyu Hu,Xiaowei Hu,Liupengfei Wu,Dian Zhang,Guoxin Fang,Zhengzhe Liu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:We introduce Function-Room Generation, a new indoor 3D scene generation setting that creates rooms supporting explicit functional goals rather than merely visually plausible layouts. Existing agentic and executable methods improve controllability, but often depend on costly test-time generate–evaluate–revise loops, making functional room generation slow and computationally expensive. We address this challenge with three technical contributions. First, we design a recursive domain-specific language to effectively organize the hierarchical object compositions required by functional rooms, from room structure and major furniture to dense support-surface and nested small objects. It represents rooms as staged executable programs with explicit geometric and functional relations. Second, we propose a sequential feed-forward scene construction framework that distills recursive construction traces into a scene construction expert. At inference time, the expert writes executable DSL code stage by stage, and a deterministic executor directly instantiates each stage without teacher agents, online critics, or iterative repair. Third, we introduce ScenePRM, an execution-grounded process reward framework that improves the expert through reinforcement learning with functional, geometric, relational, and future-constructability feedback. We further establish a function-oriented benchmark and show state-of-the-art performance on both general indoor scene generation and function-room generation, achieving stronger functional completeness, relation correctness, geometric executability, and generation efficiency.
[CV-155] ARMOR: Manifold-Oriented Training for Adversarially Robust Aerial Object Detection under Data Scarcity
链接: https://arxiv.org/abs/2608.29510
作者: Haoran Wang,Matthew Lau,Alec Helbling,Matthew Hull,ShengYun Peng,Mansi Phute,Martin Andreoni,Willian T. Lunardi,Duen Horng Chau,Wenke Lee
类目: Computer Vision and Pattern Recognition (cs.CV); Cryptography and Security (cs.CR); Machine Learning (cs.LG)
备注: 10 pages, 5 figures
Abstract:Aerial object detection is increasingly deployed in real-world applications, but models remain vulnerable to physical, universal adversarial patches that cause them to miss objects. Furthermore, defenders face the practical constraint of training data scarcity: aerial imagery is costly to collect and label, so a deployment site typically yields hundreds of images rather than the tens of thousands that adversarial robustness benchmarks assume. To tackle model vulnerability and training data scarcity, we propose Adversarial Robustness with Manifold-Oriented Training (ARMOR), a novel defense that realizes the core insights of on-manifold adversarial training (OMAT) in low-data regimes. ARMOR builds on the insight of OMAT to model the data manifold - the compact structure capturing the data’s relevant features - to learn and robustify these features during training. While OMAT relies on the data-intensive operations of training large generative models and adversarial training to achieve this, ARMOR adopts a data-efficient approach that reuses labels the detection task already supplies: ARMOR (i) masks image backgrounds to retain object-relevant features, and (ii) injects randomized patches on objects to improve feature robustness. Our low-data experiments with physically-realizable adversarial patches evaluate both query-free transfer attacks and defense-aware attacks. ARMOR maintains strong clean performance of over 0.90 model confidence, while improving adversarial robustness by up to 0.32 in model confidence over state-of-the-art defenses. Physical experiments with printed patches confirm that these gains survive deployment. Overall, ARMOR translates insights from manifold-based training to defend object detectors amidst training data scarcity.
[CV-156] SpatialTrust: A Benchmark for Environmental Risk Recognition in Secure Authentication
链接: https://arxiv.org/abs/2608.29489
作者: Junbin Lu,Hsiang-Wei Huang,Saesha Wadhwa,Yu Ting Hsu,Jenq-Neng Hwang
类目: Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV)
备注: 9 pages, 4 figures
Abstract:Visual environmental risk recognition plays an important role in secure authentication, where a user’s surroundings may reveal sensitive information or introduce potential security risks. However, existing evaluations of multimodal large language models (MLLMs) rarely examine whether models can reliably recognize, localize, and explain such risks in spatially grounded authentication scenarios. We present SpatialTrust, a question-answering benchmark for evaluating environmental risk recognition in secure authentication. SpatialTrust assesses five complementary abilities: sensitive factor detection, direct factor identification, indirect factor identification, direct factor explanation, and indirect factor explanation. We evaluate both proprietary and open-source MLLMs and find that current models show limited performance, especially in understanding and explaining indirect risks, indicating that spatial risk awareness remains a challenging capability for MLLMs. In addition, we introduce SpatialTrustGuard, a structured QA-and-audit pipeline that improves Qwen3-VL-30B-A3B-Instruct from 36.78% to 41.12% overall. Our findings highlight the need for dedicated benchmarks and structured inference methods to improve the trustworthiness of MLLMs in secure authentication.
[CV-157] Seeing Through Extreme Visual Sparsity: Surface Understanding from a Single Random Visual Patch
链接: https://arxiv.org/abs/2608.29475
作者: Sindhuja Penchala,Sudip Mittal,Noorbakhsh Amiri Golilarz
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Surface material recognition from incomplete visual observations remains a challenging problem in robotic perception and environmental understanding. This paper discusses Sparse Surface Understanding Framework (SSUF), a unified dual-task learning framework that adapts four pretrained architectures-Convolutional Autoencoder (ConvAE), Vision Transformer (ViT), Swin Transformer, and Masked Autoencoder (MAE) for si-multaneous surface reconstruction and material classification. Experiments were conducted on the Touch-and-Go dataset using a sparse observation protocol in which only 10% of the original image remained visible while the remaining regions were masked. To enable a fair comparison, reconstruction-oriented models were extended with classification heads, whereas classification- oriented models were augmented with reconstruction decoders. The resulting architectures were assessed using reconstruction quality, classification performance, model complexity, and in-ference efficiency metrics. Experimental results revealed distinct strengths across the models. Swin Transformer achieved the best classification performance with an accuracy of 89.21%, an F1-score of 0.8922, and a ROC-AUC of 0.9813. In contrast, MAE produced the highest reconstruction scores among evaluated models, with a PSNR of 16.06 dB and an SSIM of 0.4501, while ViT provided the best overall balance between reconstruction and classification performance. Furthermore, all models achieved real-time inference, requiring less than 5 ms per image. Over-all, the results show that pretrained architectures can support material recognition under severe visual sparsity, while accurate image reconstruction remains challenging.
[CV-158] Co-Evolutionary Prompt Optimization with Cross-Category Transfer for Zero-Shot Anomaly Detection EMNLP2026
链接: https://arxiv.org/abs/2608.29467
作者: Sisi Zhu,Changwei Yu,Renshuai Tao,Zhenliang Ni
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 25 pages, 25 figures. Camera-ready version. Accepted to EMNLP 2026 Main Conference
Abstract:Zero-shot anomaly detection (ZSAD) has gained significant attention for its practical value in industrial inspection. Recently, CLIP-based approaches have been widely adopted in ZSAD due to their strong vision-language generalization capabilities. However, existing methods commonly employ continuous prompt embeddings for prompt optimization and encode semantics in latent vectors, which lack interpretability and scalability. To this end, we propose CoEvoAD, a co-evolutionary framework for discrete prompt selection. CoEvoAD performs prompt search in the discrete natural-language space using an evolutionary algorithm. Candidate prompts are iteratively generated, evaluated, and selected throughout population evolution, thus preserving the interpretability and composability of natural language. Furthermore, we introduce a Cross-Category Transfer Objective (CCTO), which treats held-out source categories as proxies for unseen categories and scores prompt rules based on their estimated cross-category transferability, effectively improving cross-category generalization. Extensive experiments are conducted to validate the effectiveness of CoEvoAD, and the results show that it achieves state-of-the-art performance across multiple anomaly detection datasets. The code is available at this https URL.
[CV-159] xt-Guided Diffusion-Based Adversarial Attacks on Chest X-Ray Images
链接: https://arxiv.org/abs/2608.29456
作者: Basudha Pal,Arjun Narayanan,Neha Ajith,Vikas R Bhat,Muhammad Umair
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:As artificial intelligence is increasingly integrated into chest X-ray (CXR) interpretation, triage, and clinical decision support, understanding its vulnerability to adversarial manipulation is critical for safe deployment. Existing robustness evaluations, however, predominantly rely on pixel-space attacks that introduce numerically constrained perturbations but may not represent plausible radiographic variation. This limitation is particularly important in multi-disease CXR classification, where models simultaneously evaluate multiple overlapping pathologies and adversarial failures may alter several diagnostic predictions. We propose a text-guided diffusion-based adversarial framework that optimizes learnable text conditioning while keeping the diffusion generator and target classifier frozen, enabling adversarial generation through a learned image prior rather than direct pixel manipulation. We evaluate the framework across multiple classifier architectures in both binary atelectasis and multi-disease CXR classification and compare it with FGSM, PGD, and Carlini-Wagner attacks. Our approach consistently produced the greatest degradation in classifier performance, reducing AUROC to 0.3885-0.5646 in binary classification and 0.4441-0.4878 in the multi-disease setting, while achieving superior image fidelity (SSIM 0.9080, LPIPS 0.1670, FID 51.23). Importantly, clinician interpretation remained unchanged for 95.9% of binary and 73.8% of multi-disease adversarial images despite substantial changes in model predictions. These findings reveal a clinically important discrepancy between human and machine interpretation and demonstrate the need to extend medical AI robustness evaluation beyond conventional pixel-space attacks toward generative threat models that can expose failures under visually and clinically plausible image variations.
[CV-160] Does Latent Planning Survive Point Clouds? Action-Conditioned JEPA World Models for Geometric Observations
链接: https://arxiv.org/abs/2608.29434
作者: Fabio F. Oberweger,Michael Schwingshackl
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:JEPA world models make latent-space planning a practical route to control, but they are built almost exclusively on images. Whether latent prediction survives geometric observations is unclear: point clouds are sparse, unordered, and self-occluded, and with 0.3-15% of scene points moving, the slow-feature optimum of latent prediction compounds with the geometric shortcut of 3D self-supervision. We lift three canonical JEPA designs to point clouds, frozen-encoder, distribution-prior, and action-sensitive, and re-sense the stable-worldmodel benchmark so that only the observation differs from the image baselines. All three plan without collapse: the distribution-prior model is statistically equivalent to its re-evaluated image counterpart on every benchmark, and the action-sensitive model attains the strongest result in our controlled comparison where the most geometry moves. Probing explains why: object positions are almost perfectly linearly decodable and attention falls on the few moving points. Planning withstands heavy dropout never seen in training, though range noise defeats the thinnest scene. Geometry finally makes a commanded 3D target a natural goal interface: we construct the goal latent from the target and the current latent, at no cost in success rate, without a goal observation.
[CV-161] Calibration and Comparative Analysis of Forward-Looking Sonar and 3D Sonar for Enhanced Underwater Object Recognition
链接: https://arxiv.org/abs/2608.29433
作者: Aditya Penumarti,Khanh Dong,Zi-Hao Zhang,Yongkyoon Park,Zhenqi Wu,Trung Dong,Shahriar Negahdaripour,Xiaomin Lin,Jane Shin
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注: 6 pages, Accepted to IEEE OCEANS 2026
Abstract:Sonars generate a significant amount of noise. With the advent of new technology capable of producing full 3D point clouds, the noise is amplified in sparse point clouds, making it challenging to recognize features for navigation, recognition, or reconstruction. To address this challenge, we propose using two different sonar modalities: one that produces a 2D intensity image and another that generates a 3D point cloud. By implementing auto-calibration, we can filter out noisy features between the modalities to enhance feature extraction. Experiments demonstrate that auto-calibration improves performance over manual calibration by 5% and that filtering enhances feature extraction by more than 40% relative to the raw point cloud. Code and datasets are given at this https URL
[CV-162] Polis: 3D Self-Supervision at City Scale
链接: https://arxiv.org/abs/2608.29426
作者: Alexander Rusnak,Sophia Kovalenko,Jingru Wang,Ismail Moudden,Xiru Wang,Frédéric Kaplan
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Reliable semantic representations derived from city-scale 3D models are increasingly important for urban analysis, infrastructure monitoring, autonomous systems, and heritage conservation. However, urban scenes of large spatial extent captured through aerial surveying differ substantially from the indoor, object-level, and self-driving LiDAR data used to pretrain most 3D self-supervised models. We introduce Polis, to our knowledge the first application of Sketched Isotropic Gaussian Regularization (SIGReg) as an objective for a native point cloud encoder, and evaluate it through a frozen-feature benchmark spanning fourteen city- and building-scale corpora. Polis combines geometrically matched cosine invariance, SIGReg, and VICReg-style anti-collapse terms with a 12.8k-scene outdoor pretraining mixture and gravity-preserving spatial view sampling. Controlled ablations show that this objective outperforms student–teacher architecture alternatives, as well as Polis versions without anti-collapse terms, on the same representative outdoor corpus. On three pretraining-disjoint city datasets, Polis reaches 23.8% mean mIoU versus 16.3% for the next-best encoder under high-capacity frozen probing, and 17.3% versus 16.1% at a matched point and voxel budget. The same city-scale lead holds on datasets whose training sets were seen in pretraining. On localized terrestrial captures with fine-grained facade and streetscape labels, the ranking reverses. Our results show that distributionally-regularized joint embedding architectures can be successful on challenging city-scale 3D scenes, and that transfer improves when self-supervision is designed for the capture geometry and spatial context of this domain while also revealing the limits of this specialization.
[CV-163] A Visual Question Answering Model to Automate Nondestructive Evaluation Image Analysis
链接: https://arxiv.org/abs/2608.29408
作者: Mehrdad Shafiei Dizaji,Hoda Azari
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 7, 6
Abstract:This study introduces a Visual Question Answering model designed specifically for nondestructive evaluation applications. VQA models allow inspectors to interactively query NDE images, asking targeted questions like, Is there a crack or Where is the defect located and receive precise answers from the model. Leveraging deep learning and natural language processing, the developed system integrates image feature extraction (via a ResNet-50 model) and language generation capabilities (via GPT-2) to provide accurate, informative feedback. By enabling direct question-and-answer interactions, this VQA model significantly improves inspection efficiency, reduces potential errors, and enhances usability in practical field scenarios.
[CV-164] GATE: Reliability-Gated Gaussian Evidence Fusion for Training-Free Test-Time Adaptation of Vision-Language Models BMVC2026
链接: https://arxiv.org/abs/2608.29395
作者: Pedram MohajerAnsari,Amir Salarpour,Run Wang,Mert D. Pesé
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to the British Machine Vision Conference (BMVC 2026)
Abstract:Vision-language models such as CLIP and SigLIP provide strong zero-shot recognition, but their predictions can degrade when deployed on target data that differ from the pretraining distribution. Test-time adaptation offers a practical way to improve robustness without source data or target labels, yet existing methods often rely on either prompt-side adaptation or image-side target evidence alone. In this work, we introduce GATE, a training-free two-pass transductive test-time adaptation framework that uses the unlabeled target set while keeping the image encoder, text encoder, and prompt parameters fully frozen. Instead of representing each class with a single prototype, GATE builds two complementary Gaussian sources of evidence in the shared vision-language feature space: a text Gaussian estimated from multiple language descriptions and an image Gaussian estimated from reliable unlabeled target samples. A class-wise reliability gate controls the influence of image-derived pseudo-evidence, and a score-level generalized Product-of-Experts fusion produces a normalized residual correction to the original zero-shot logits. Across fine-grained recognition datasets, ImageNet-family distribution shifts, multiple CLIP backbones, and SigLIP-B/16, GATE achieves the best average accuracy in every benchmark/backbone group. It improves zero-shot performance by an average of 5.41 points and outperforms the strongest non-GATE baseline by 1.94 points, demonstrating the benefit of reliability-gated distributional evidence for frozen VLM adaptation.
[CV-165] FiLM-GPNet: Geometry-Aware Pseudo-Supervised Phase Restoration with Zero-Shot Generalization for Large Temporal InSAR Stacks WWW
链接: https://arxiv.org/abs/2608.29384
作者: Getnet Demil,Muhammad Farhan Humayun,Tomi Westerlund,Jukka Heikkonen,Mourad Oussalah
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: Its IEEE Geoscience And Remote Sensing Symposium (IGARSS) 2026 Data Fusion Contest Winner Paper. reference this https URL
Abstract:The growing availability of dense commercial Synthetic Aperture Radar (SAR) time series enables temporal Interferometric SAR (InSAR) analysis, but fixed classical filters fail under heterogeneous acquisition geometries, degrading phase quality and temporal consistency. We propose FiLM-GPNet, a geometry-conditioned network for wrapped-phase restoration that explicitly adapts to acquisition differences using Feature-wise Linear Modulation (FiLM) and a 7D per-pair geometry descriptor. The model is trained with pseudo-supervision from Goldstein-filtered interferograms and regularized by interferometric physics via triplet-closure consistency, while also estimating per-pixel aleatoric uncertainty. Experiments on three Capella Spotlight stacks from the IEEE GRSS 2026 Data Fusion Contest show that FiLM-GPNet reduces temporal residual by 68% (Hawaii) and 66% (Western Australia) relative to the Goldstein baseline, alongside closure error reductions of 10% and 13%, respectively. In Western Australia, it further improves unwrapping success rate by 7.7 percentage points and Digital Elevation Model (DEM) Normalized Median Absolute Deviation (NMAD) by 31%. The model also shows strong zero-shot generalization to a geographically and geometrically distinct third stack (Los Angeles) without retraining, supporting geometry-conditioned restoration as an effective alternative to fixed classical filtering across heterogeneous stacks.
[CV-166] hink Look and Revise: Inconsistency-Aware Visual Self-Correction in MLLM s EMNLP2026
链接: https://arxiv.org/abs/2608.29374
作者: Yu Cheng,Arushi Goel,Hakan Bilen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: accepted by EMNLP2026
Abstract:Tool-augmented multimodal reasoning integrates external tools (e.g., object detection, depth estimation) into multimodal large language models (MLLMs) to address perceptual bottlenecks in complex visual tasks. However, existing approaches rarely verify tool outputs, limiting their ability to detect and recover from tool failures. We propose ReVISE, a framework that equips MLLMs with verification and dynamic error recovery for tool-augmented reasoning. ReVISE introduces (1) a curated training dataset that supervises reflective behaviors, enabling models to validate tool-derived evidence, reformulate queries when visual mismatches arise, and fall back to intrinsic grounding when external tools are unreliable; and (2) a reinforcement learning based targeted rewards that encourage internal reflection and penalize spatial misalignment. Experiments on several benchmarks demonstrate consistent improvements over existing methods, highlighting the importance of error detection and correction in tool-augmented multimodal reasoning.
[CV-167] Sketch2Inspire: Structure-Sensitive Evaluation for Product Retrieval PRICAI2026
链接: https://arxiv.org/abs/2608.29364
作者: Ge Kong
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 15 pages, 2 figures. PRICAI 2026 version. The paper presents Sketch2Inspire, a structure-sensitive evaluation resource for multimodal product retrieval
Abstract:Early-stage product design retrieval often requires more than category recognition: designers may need reference examples that match both a short semantic intent and a rough structural cue. Existing product-image resources and generic image–text retrieval benchmarks rarely separate category retrieval from within-category structural fit. We present Sketch2Inspire, built from a curated subset of Amazon Berkeley Objects with aligned text queries, edge-based sketch-proxy queries, and fused text–sketch queries. The resource separates broad category-level retrieval from structure-sensitive within-category retrieval and includes a human-graded reference protocol for calibration. We evaluate a lightweight reference system based on pretrained CLIP-family encoders, comparing text-only retrieval, sketch-only retrieval, weighted late fusion, and text-first reranking without updating model weights. Under broad relevance, late fusion obtains the highest score (nDCG = 0.9962). Under automatic structure-sensitive relevance, late fusion again obtains the highest score (nDCG = 0.7015), exceeding text-only retrieval (nDCG = 0.5912). In the human-graded results, late fusion obtains the highest nDCG@10 (0.9133), while text-only retrieval ranks second (0.9030). These results show that the retrieval gain from multimodal input depends on how relevance is defined. Sketch2Inspire therefore provides a diagnostic resource for evaluating modality contribution and supports the development of structure-aware product-retrieval protocols with independent human annotation.
[CV-168] GSPotential: Camera Potential Field for Sparse-View 3D Gaussian Splatting
链接: https://arxiv.org/abs/2608.29346
作者: Zeyuan An,Yanghang Xiao,Zhiying Leng,Yijun Feng,Xiaohui Liang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 10 pages, 8 figures, 5 tables. Accepted to Pacific Graphics 2026 Conference Track
Abstract:3D Gaussian Splatting has achieved remarkable success in photorealistic rendering, yet it suffers from severe overfitting and geometric artifacts in sparse-view scenarios due to the inherent deficiency of photometric supervision. Recent advances have attempted to regularize optimization by incorporating external priors, such as depth, point clouds, or diffusion models. However, these methods typically overlook the non-uniform distribution of supervision across the viewing space, resulting in limited specificity in prior use and primitive control. In this paper, we propose GSPotential, a framework that quantifies view-space supervision imbalance using a Camera Potential Field. Our key insight is to identify supervision valleys where photometric constraints are most deficient, and use the potential field to guide reconstruction from two complementary aspects. First, we devise a probabilistic spherical sampling strategy that places informative virtual cameras in low-potential regions. Point-cloud renderings from these views then provide targeted geometric guidance. Second, the same field provides a directional coverage cue for conservative Gaussian updates in weakly covered spatial sectors. Extensive experiments demonstrate that GSPotential achieves high reconstruction fidelity while maintaining competitive training efficiency.
[CV-169] GenFirst: Generation Before Reconstruction for Stable End-to-End Latent Generative Modeling
链接: https://arxiv.org/abs/2608.29335
作者: Guangting Zheng,Yiyuan Zhang,Tao Yang,Yunpeng Chen,Rui Zhu,Jiajun Deng,Yanyong Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Latent generative models typically follow a two-stage pipeline, training a variational autoencoder for reconstruction and then a generative model on the frozen latent space. Since reconstruction-optimized latents are not necessarily generation-friendly, jointly training both models is an appealing alternative. However, direct end-to-end training remains challenging, as it is prone to latent collapse and faces a generation-reconstruction conflict. We revisit this problem by analyzing how different objectives shape the latent space and identify two key insights. First, the entropy term in the Kullback-Leibler divergence objective is essential for preventing collapse: reconstruction and prior fitting tend to shrink the posterior, while entropy preserves non-degenerate latent uncertainty. Second, reconstruction and generation exhibit asymmetric learning dynamics: reconstruction is fast and strongly supervised, whereas generation is slower and harder to optimize. Based on these insights, we achieve the first direct end-to-end training without latent collapse and propose GenFirst, a simple generation-before-reconstruction strategy. The generative objective first shapes the latent space under weak reconstruction pressure, after which reconstruction is progressively strengthened to recover visual details. We validate GenFirst with continuous autoregressive priors with exact likelihoods and SiT priors with implicit likelihoods. With our end-to-end objective and GenFirst, SiT achieves a gFID of 0.97 with CFG and 1.45 without CFG on ImageNet-256, while MMDiT reaches a GenEval score of 0.90 on text-to-image generation. Beyond image generation, we extend the framework to shared visual latents for generation and representation learning, and to continuous unified text-image generation. These results demonstrate the generality of stable end-to-end latent learning across generative priors and modalities.
[CV-170] Neural video codecs quality assessment dataset and benchmark
链接: https://arxiv.org/abs/2608.29331
作者: Nikolay Safonov,Nikita Gornostaev,Alexandra Dubonos,Dmitriy Vatolin
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Video traffic constitutes a significant share of global web traffic. To reduce its volume, video codecs have been developed and continuously improved. While the industry has achieved substantial progress in traditional video coding, neural video codecs (NVCs) have recently emerged as a new approach that applies deep learning to video compression. This creates new challenges for compression quality assessment, which is essential for the further development and improvement of such codecs. In particular, it is important to evaluate the novel temporal compression paradigms introduced by NVCs. In this work, we present a large-scale subjective dataset of videos compressed with both neural and traditional video codecs. The subjective scores were collected through crowd-sourced pairwise comparisons. The proposed dataset provides a valuable resource for the development and benchmarking of video quality metrics tailored to neural video codecs. The dataset is available at the following link: this https URL
[CV-171] st-Time Scaling for Video Diffusion Models via Diagnosis-Guided Candidate Recycling SIGGRAPH
链接: https://arxiv.org/abs/2608.29322
作者: Hangzhou He,Lunhao Duan,Shanshan Zhao,Kaiwen Li,Qing-Guo Chen,Weihua Luo,Yanye Lu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ACM TOG (SIGGRAPH Asia 2026)
Abstract:Recent video diffusion models have achieved remarkable generation quality, but high-fidelity results still largely depend on closed-source systems or costly large-scale infrastructure. Test-time scaling (TTS) offers a training-free way to improve lightweight generators by spending additional inference compute, yet existing methods mostly remain within a noise-search paradigm: they sample, select, or perturb denoising trajectories and discard low-scoring candidates after expensive generation. This generate-and-discard process wastes not only computation but also the partial motion, layout, or appearance structure already encoded in recoverable samples. We present \textbfGEARS (\textbfGuided \textbfEditing for \textbfAdaptive \textbfRecycling \textbfSearch), a training-free framework that introduces diagnosis-guided candidate recycling into video TTS by turning such candidates into editable priors through a generation-evaluation-editing loop. GEARS consists of two collaborative components. The \textbfStage-Aware Scheduler determines what to repair, when to repair it, and which candidates should be preserved, recycled, or discarded. The \textbfCandidate Recycler diagnoses recoverable failures from keyframes and multi-dimensional reward feedback, derives candidate-specific repair prompts, and repairs the corresponding candidates through manifold-aware latent SDEdit. The repaired candidates are recycled into the search pool, creating refinement paths beyond standard noise perturbation while preserving useful structure. Under matched NFE budgets, GEARS consistently outperforms existing video TTS methods on VBench, bringing a 1.3B model to a total score comparable to a 14B counterpart, and ablations verify the necessity of adaptive scheduling, diagnosis-conditioned editing, and manifold-aware re-denoising. Code is available on GitHub.
[CV-172] Hyper3-CLIP: Hierarchy-Conditioned Hyperbolic Vision-Language Training ECCV2026
链接: https://arxiv.org/abs/2608.29313
作者: Matin Mahmood,Antonio Rueda-Toicen,Mohamed ElBassat,Seifeldin Elkerdany,Weixing Wang,Gerard de Melo
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 16 pages, 2 figures, ECCV 2026 Beyond Euclidean Workshop
Abstract:CLIP-like vision-language models (VLMs) trained with contrastive objectives learn strong global image-text representations, but their Euclidean embeddings and global pooling fail to encode relational structure such as part-whole and parent-child relations. Hyperbolic VLMs address this gap with entailment-based objectives, and text-conditioned variants improve fine-grained alignment through sentence- and phrase-level queries. However, these two lines of work remain separate: hyperbolic VLMs use static image and region features, while query-conditioned methods lack hierarchical geometric structure. We present Hyper3-CLIP, a hierarchy-conditioned hyperbolic VLM that combines global, local, and global-local contrastive learning with query-conditioned visual pooling. To train the model, we construct lightweight query hierarchies from text, comprising full captions, sentence fragments, localized part descriptions, and extracted phrases. Each query conditions the pooling of visual patches, and the resulting representations support image-text, whole-part, and parent-child entailment losses. Query-conditioned pooling is active only during training. Hyper3-CLIP improves R@5 and R@10 retrieval on COCO and Flickr, as well as multi-label classification on VOC and COCO, while remaining competitive on hierarchy metrics. We also audit zero-shot prompt sensitivity under fixed prompt regimes and study the effect of the localized GRIT part budget used during training. Code is available at this https URL.
[CV-173] AOI-Net: Structural Face AOI-Guided Eye-Gaze Track Representation Learning for Autism Spectrum Disorder Detection
链接: https://arxiv.org/abs/2608.29289
作者: Zhanpei Huang,Binbin Sun,Jialiang Chen,Yiou Wang,Taochen Chen,Yuzhu Ji,Yiqun Zhang,Yiu-Ming Cheung
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Emerging Technologies (cs.ET); Multimedia (cs.MM)
备注: 14 pages, 8 figures
Abstract:Eye-movement tracking has emerged as a promising non-invasive approach to Autism Spectrum Disorder (ASD) screening, with systematic differences in attentional allocation and revisit behaviors observed during socially interactive tasks. Existing computational methods typically characterize eye-movements using discrete gaze trajectories and fixation events, yielding representations dominated by short-range temporal dynamics and limiting models that primarily emphasize long-range dependencies. Meanwhile, gaze behavior is naturally organized across semantically meaningful Areas of Interest (AOIs), whose attention allocation and transitions provide important structural cues, yet their relationships are rarely modeled explicitly. To address these limitations, we propose a structural face AOI-guided Eye-Gaze Track Network (AOI-Net) that jointly models short-term temporal dynamics and AOI-level structural organization. A network gating mechanism adaptively integrates the complementary temporal and structural representations according to their contributions to gaze-behavior characterization. To mitigate the pronounced class imbalance commonly encountered between individuals with ASD and Typically Developing (TD) participants in clinical datasets, class-distribution-aware learning is further employed to facilitate discriminative embedding learning under skewed class distributions. Experiments on a unique and large-scale clinical eye-tracking database comprising eight stimulus subsets and more than 1,300 participants show that AOI-Net consistently outperforms state-of-the-art methods. The proposed framework also enables interpretable gaze-behavior modeling and provides a practical basis for scalable AI-driven ASD screening in real-world healthcare. The code is available at this https URL
[CV-174] PERSIST: Persistent-State Discrimination for Shot Boundary Detection BMVC2026
链接: https://arxiv.org/abs/2608.29287
作者: Tingyu Lin,Christian Stippel,Armin Dadras,Jakob Zenzmaier,Florian Kleber,Wolfgang Aigner,Robert Sablatnig
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at BMVC 2026. Main paper: 18 pages, 5 figures. Supplementary material: 16 pages, 4 figures
Abstract:Shot boundary detection (SBD) is widely treated as the localisation of local visual discontinuities, yet many false positives such as hand-held shake, illumination flicker, motion blur, occlusion, and damaged archival material produce equally sharp local change without introducing a new shot. We reformulate SBD as boundary semantic discrimination: a frame is favoured as a boundary only when its local change evidence is accompanied by a persistent update of the video’s latent temporal state, rather than a transient excursion that returns to the surrounding trend. This persistence test is operationalised with a continuous latent state from a FiLM-conditioned sinusoidal representation network and a structured discriminator that combines three semantic cues, local change, transient impulse, and return-to-trend, into a single interpretable per-frame signal over a dual-rate temporal backbone. The resulting framework, PERSIST, turns every decision into an inspectable one: the persistence criterion is trained into the classifier, its per-frame effect stays readable from the gate triple, and its learned latent state is measurably boundary-discriminative. On a 2,727-video per-subtype diagnostic it removes 33-80% of flash, text-overlay, and archival false positives relative to an identically trained cue detector, and at matched true-transition recall it roughly halves TransNetV2’s pseudo-event false positives on that diagnostic and cuts its false positives on ClipShots footage by about a quarter, while preserving recall. It does so while reaching parity with the strongest public detector across online, broadcast, short-form, and historical-archive transfer evaluations, under markedly stricter training: it learns from ClipShots real transitions only, whereas the anchor draws on additional corpora whose transitions are 85% synthetic. Code is available at this https URL.
[CV-175] 3D-MRL: Nested Multimodal 3D Representations via Matryoshka Representation Learning BMVC
链接: https://arxiv.org/abs/2608.29285
作者: Márcus Lobo,Vitor Matias,Jeová Farias,Moacir Ponti
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at the British Machine Vision Conference (BMVC) 2026. Project page: this https URL
Abstract:Vision-Language Models align point clouds with image and text embeddings, enabling zero-shot recognition, retrieval, and open-vocabulary understanding of 3D shapes. Existing multimodal 3D pre-training methods produce fixed-dimensional embeddings, requiring separate models for different computational budgets. We propose 3D Matryoshka Representation Learning (3D-MRL), a multimodal 3D pre-training framework based on Matryoshka Representation Learning. 3D-MRL learns nested 3D representations by aligning point clouds with frozen CLIP image and text embeddings while applying contrastive supervision across multiple embedding dimensions. The Matryoshka objective is applied only to the 3D encoder, allowing a single model to produce representations at different dimensionalities without retraining. Experiments on the Objaverse-LVIS, ModelNet40, and ScanNet datasets show that 3D-MRL achieves competitive performance on zero-shot and few-shot 3D recognition tasks. In addition, the learned representations support retrieval across different embedding dimensions within a single model. On Objaverse-LVIS, 3D-MRL improves Top-1 accuracy from 46.8% to 50.9%. Retrieval experiments further show that different embedding dimensions yield varying levels of semantic and geometric specificity.
[CV-176] Elastic Token Compression for Pixel-Space Diffusion Transformers
链接: https://arxiv.org/abs/2608.29281
作者: Eduard Zamfir,Christian Reisswig,Zongwei Wu,Yongqin Xian,Radu Timofte
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Natural images concentrate their detail in a small fraction of the frame, yet diffusion models spend a full token on every patch, in every layer and at every timestep. The waste is largest in pixel-space models, with no autoencoder to absorb low-level redundancy first. Probing a pretrained pixel text-to-image transformer, we find its middle-block tokens redundant wherever the image is flat. The redundancy occupies connected, content-shaped regions, and exploiting it requires tokens with the same geometry. Cutting a Hilbert ordering of the patches provides them. Consecutive positions are always image neighbours, so any contiguous run is a connected region whose size and shape follow the content, and grouping in two dimensions becomes a cut in one. Existing reductions each lose part of this. Similarity merging scatters its groups, latent bottlenecks discard position, and skipping deletes what it should summarize. We cut where the model’s features change most and pool each run into one region token. Our Region Token Interface (\method) adapts a diffusion model to these tokens, with the region count drawn at random during fine-tuning so one checkpoint serves every budget. \method leads prior reduction methods at matched budgets, matches dense quality at 2.0\times the speed, and stays close at 2.6\times . The code and models are open-sourced at this https URL
[CV-177] RAG Diffusion: From Macro-Retrieval to Micro-Fidelity Alignment for Garment Generation
链接: https://arxiv.org/abs/2608.29280
作者: Yuhan Li,Xianfeng Tan,Fangao Zeng,Wenxiang Shang,Pipei Huang,Hao Zhou,Zhiyu Jin,Wenjun Zhang,Bingbing Ni
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Standard clothing asset generation—restoring forward-facing flat-lay garment images from diverse real-world contexts—holds immense commercial value yet demands both macroscopic topological accuracy and microscopic physical fidelity. Although our previous work RAGDiffusion effectively eradicated large-scale structural hallucinations via retrieval-augmented macro-constraints, achieving industrial-grade micro-texture realism remains an unsolved bottleneck. We formally identify this limitation as High-Frequency Trajectory Collapse: supervised fine-tuning (SFT) converges to the conditional mean of the training distribution, which is dominated by smooth, low-frequency textures, causing high-frequency patterns (e.g., fabric weaves, intricate logos) to become nearly un-sampleable. Naively applying Reinforcement Learning (RL) post-training further triggers Artifact Hacking, where models exploit semantic biases in generic reward models by generating deceptive checkerboard noise. Our key insight is that RL can fundamentally reshape the sampling distribution of flow models—elevating the probability of high-fidelity trajectories under accurate reward guidance—while adversarial regularization prevents exploitation of reward blind spots. Realizing this principle requires three prerequisites: (i)inherent capacity, established through a 27,725-pair high-complexity garment dataset (STGarment-Plus) and a Dual-Image-Stream FLUX architecture upgrade; (ii)perceptive reward, provided by a novel attribute-aware reward model (Garment-RM) trained on 500K images via fine-grained contrastive learning, achieving 84.67% human preference accuracy; and (iii)hacking prevention, enforced by our Adversarial-Regularized GRPO (AR-GRPO) strategy that integrates a dynamic discriminator into the RL sampling trajectory to penalize artifacts while enriching authentic high-frequency details.
[CV-178] LightFuse: Relightable Interactive Gaussian Scene Reconstruction via Multi-Scan Fusion and 2D Gaussian Ray Tracing
链接: https://arxiv.org/abs/2608.29269
作者: Haonan Zhou,Gaoxiang Linghu,Youlin Jia,Hongyu Cui,Kewei Wei,Kaiyue Zhou,Bruce X.B. Yu,Gaoang Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Relightable interactive scene reconstruction aims to build an editable 3D model from scans of different object arrangements and render new layouts under novel illumination. Existing methods either bake lighting into appearance or recover material and illumination only for fixed scenes, leaving edited layouts with inconsistent shadows and indirect lighting. We present LightFuse, a 2D Gaussian framework that extends interactive scene reconstruction with explicit material-illumination decomposition and physically based relighting. LightFuse first fuses observations across states to reconstruct a shared background and movable objects. It then conducts ray-tracing-oriented geometry refinement to produce more complete and consistent surfaces. On the refined geometry, staged training with differentiable one-bounce ray tracing separates shared metallic–roughness material from state-specific environment lighting. The resulting scene supports object rearrangement, material editing, and relighting, while ray tracing recomputes appearance after each interaction. Experiments across synthetic scenes demonstrate state-of-the-art relighting quality, outperforming the strongest baseline by +9.74,dB PSNR and +0.121 SSIM on average. Project page: this https URL
[CV-179] Learning to Ground Before Reading: Unified PCB Engineering Drawing Parsing with Compact Vision-Language Models
链接: https://arxiv.org/abs/2608.29268
作者: Jinghao Liu,Xingrun Liu,Gengchen Sun,Han Xiao,Xingyu Chen,Yuhui Deng
类目: Computer Vision and Pattern Recognition (cs.CV); Multimedia (cs.MM)
备注: 14 pages, 5 figures, and 6 tables
Abstract:PCB engineering drawings mix sparse graphics, dense tables, and text whose meaning depends on page position. Localizing the regions and sending crops to specialized recognizers are determined as the methods for most parsers, so missed regions cannot be recovered downstream. We train a compact VLM to read the full page and get a sequence of region classes, normalized boxes, and text or HTML content. Bounding boxes are converted to coordinate tokens for supervision. Inference uses no detector or crop parser. The joint target is difficult to optimize because class and box tokens are sparse relative to the much longer content sequences. Our localization-first curriculum learns the class-box format before adding content targets with content-aware resampling. On the fixed validation split of the Engineering Drawing Dataset (ED dataset), Localization-First improves strict localization F1 by 0.0955 over joint training (paired image-bootstrap 95% interval: [0.0350, 0.1572]). G-Unified has the lowest NED, highest cell F1, and only nonzero exact-match score. It provides a detector-free baseline for full-page PCB drawing parsing.
[CV-180] QCell: Recombining and Aligning Cell Queries for Overlapping Instance Segmentation BMVC
链接: https://arxiv.org/abs/2608.29253
作者: Yaroslav Prytula,Anton Popov,Dmytro Fishman
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Accepted at the British Machine Vision Conference (BMVC) 2026. Project page/code/models/dataset: this https URL
Abstract:Instance segmentation of overlapping cells in microscopy remains challenging due to semi-transparent structures that produce weak boundaries and mixed visual evidence in overlap regions. Existing methods address this through local regions of interest or shape priors but lack global reasoning across overlapping objects. We present QCell, a novel query-based model that de-overlaps cell instances in microscopy scenes. Our approach combines (i) an instance recombination module that decomposes and recombines query representations in latent space, enabling the model to reason about complete object structure under overlap, and (ii) a contrastive query alignment objective that combines distinctive instance feature learning and separation of overlapping cell queries. We additionally introduce a new Organoid dataset benchmark for overlapping cell segmentation. We show that QCell outperforms state-of-the-art methods across multiple benchmarks, achieving +2.2 AP and +2.7 AJI on ISBI2014. Code is available at this https URL
[CV-181] DARD: Zero-Shot Degradation-Aware Retinex-Guided Diffusion for Low-Light Image Enhancement
链接: https://arxiv.org/abs/2608.29243
作者: Wenjie Cai,Yuezhe Yang,Jianyang Xia,Xingbo Dong,Zhe Jin
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Existing diffusion-based enhancement methods provide strong generative capability for low-light image enhancement (LLIE), yet they either rely on paired supervision or lack reliable scene constraints in zero-shot settings, often leading to structural inconsistency and color drift. Motivated by conventional Retinex models, which offer physically interpretable priors that can serve as reliable scene constraints yet struggle with mixed degradations in real-world scenarios, we propose DARD, a zero-shot Degradation-Aware Retinex-guided Diffusion framework for LLIE. DARD first extracts image-specific physical priors from the degraded input through a test-time degradation-aware Retinex decomposition, thereby providing reliable structural guidance for zero-shot restoration. It then injects these priors into reverse diffusion through a timestep-adaptive frequency fusion strategy to balance structural anchoring and detail generation. Finally, a guided reverse refinement process with physical consistency and Contrastive Language-Image Pre-training (CLIP)-based semantic guidance is introduced to suppress structural artifacts and semantic drift during sampling. Extensive experiments show that DARD achieves strong distortion and perceptual performance and consistently outperforms existing zero-shot baselines across multiple real-world low-light benchmarks. To further validate the practical utility of our method for downstream applications, we evaluated its impact on semantic segmentation. Experiments demonstrate that images enhanced by DARD achieve a 28.10% relative improvement in mIoU over AGLLDiff.
[CV-182] AGRICAM: A Track-Mounted Crop Pollination Monitoring Robot
链接: https://arxiv.org/abs/2608.29237
作者: Malika Nisal Ratnayake,Adel N. Toosi,James Cook,Romina Rader,Alan Dorin
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Robotics (cs.RO); Systems and Control (eess.SY); Quantitative Methods (q-bio.QM)
备注: 20 pages, 8 figures
Abstract:Insect pollination is critical for global food production, yet monitoring pollinators at commercial farm scale remains a challenge. Recent advances in computer vision and deep learning have enabled detailed analysis of pollinator behaviour, but monitoring must trade-off detail against spatial coverage and human or technological resources. This paper presents the Automated Guided Robot for Insect and Crop Activity Monitoring (AGRICAM), a purpose-built robotic system designed to meet the requirements of large-scale pollination monitoring in protected cropping systems. AGRICAM operates autonomously on low-cost, easily installed track for movement along crop rows, without disrupting farm operations or insect behaviour. The platform integrates two RGB cameras, microclimate sensors, GPS and RFID modules, motion sensors, and 4G cellular network connectivity for data transmission. A web interface enables remote device configuration and scheduling. The system autonomously captures video and image data of insects’ locations and local environmental conditions. These are transferred to the cloud and analysed using computer vision models to quantify pollinator visitation and spatio-temporal activity variation. We deployed the system on a commercial blueberry farm to demonstrate and test its capability. It successfully mapped insect pollination patterns across 80 m long industrial polytunnels over 30 hours. This data enabled spatial analyses of insect activity we used to confirm a uniform pollinator distribution within polytunnels, as desired by the farm management team. The data also highlighted variation of insect activity associated with time of day and microclimate. AGRICAM therefore has been shown to be a scalable, automated crop pollination monitor that can support data-driven decisions to enhance pollination management, thereby improving crop productivity and food security.
[CV-183] Uncertainty-Aware Multimodal Anti-UAV Detection via Evidential Fusion and Conflict-Discounted Belief Aggregation
链接: https://arxiv.org/abs/2608.29235
作者: Sharanda Suttorp,Seyed Sahand Mohammadi Ziabari,Ali Mohammed Mansour Alsahag
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Anti-UAV perception systems must remain reliable when sensor streams degrade under occlusion, fast motion, or modality-specific failure. Existing multimodal anti-UAV systems fuse RGB and thermal streams deterministically, without modeling predictive uncertainty, and cannot express doubt when streams disagree. Evidential Deep Learning (EDL) produces calibrated per-class uncertainty in a single forward pass. EDTC already exploits this for thermal-only perception, yet cross-modal evidential fusion remains unaddressed. This paper extends EDTC to multimodal RGB-Thermal perception via Discounted Belief Fusion (DBF), which converts inter-modal conflict into uncertainty mass before aggregating stream opinions. Bounding boxes are resolved by selecting the lower-uncertainty modality. On the Anti-UAV benchmark, multimodal fusion consistently outperforms either single stream (test Acc 0.670 vs. 0.604 IR, 0.598 RGB) at real-time speed (at least 38 FPS). However, DBF is empirically indistinguishable from undiscounted averaging: near-zero inter-modal conflict on this presence-dominated benchmark leaves the discounting step inert. The fused uncertainty is well-calibrated (ECE 0.057) yet expectedly a weaker localization failure detector than spatial variance (AUROC 0.626 vs. 0.739). The null result is structural: the benchmark’s near-universal presence and vacuous miss-encoding jointly suppress inter-modal conflict, a diagnosis that delimits where conflict-aware fusion provides measurable benefit. Subjects: Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2608.29235 [cs.CV] (or arXiv:2608.29235v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2608.29235 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-184] Generalization over Memorization: Generalization-Aware Diffusion Adaptation for Single-Image Multi-View Synthesis
链接: https://arxiv.org/abs/2608.29233
作者: Jie Li,Xingchen Zou,Yuxuan Liang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: ACM Multimedia 2026 Grand Challenge Track winning paper; presents the 1st-place solution among 293 registered teams. 7 pages, 2 figures
Abstract:We present the winning solution to the ACM Multimedia 2026 Grand Challenge on Single-Image Guided Multi-Angle Image Synthesis. It ranks first among 293 registered teams; 56 teams obtained at least one scored submission on the public Phase-A leaderboard. With only 40 training scenes, the challenge requires 26 target views from one RGB model and one forward pass per view; it prohibits explicit geometry, external rendering, chained generation, candidate selection, and post-processing. We identify a critical model-selection failure: shared training and validation scenes make memorization appear as transferable view control. We therefore introduce GoM. Short for Generalization over Memorization, the framework combines scene-disjoint validation, exposure-matched selection, and targeted diffusion adaptation. Its synthesis model adapts a 4B rectified-flow DiT using rank-32 LoRA, optimizer restarts, late-checkpoint averaging, and VAE decoder tuning. More than 300 offline experiments and 24 online submissions show that validation design and training-trajectory control can matter as much as architecture scale in small-data generative modeling.
[CV-185] Background-Free Objectness Learning for Class-Agnostic Detection BMVC
链接: https://arxiv.org/abs/2608.29232
作者: Dania Batool,Liliana Lo Presti,Marco La Cascia,Filippo Vella
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Robotics (cs.RO)
备注: Accepted at the British Machine Vision Conference (BMVC) 2026. This arXiv version includes supplementary material
Abstract:Object detectors are typically trained under closed-set supervision, where unlabeled regions are implicitly treated as background. Under incomplete annotations, this assumption introduces objectness bias: visually valid but unlabeled objects are used as negatives, tying objectness to the annotated taxonomy rather than generic object structure. This limitation is particularly problematic for class-agnostic and open-world detection. This paper proposes Background-Free Objectness Learning (B-FOR), a dense class-agnostic detection framework that learns objectness without explicit background supervision on unlabeled regions. B-FOR formulates detection as the prediction of dense multi-scale object-center and scale fields, from which object hypotheses emerge as local spatial structures. Supervision is confined to reliable annotated regions through spatially structured soft targets, avoiding foreground-background discrimination. To support decoding from emergent local maxima, the paper further introduces displacement-aware scale fields that model object extent as a spatially varying property of the learned objectness field. Experiments on PASCAL VOC, MS-COCO, and Open Images demonstrate strong generalization to unseen categories and cross-dataset object distributions. B-FOR improves recall by more than +10 AR points over prior class-agnostic baselines. Ablation studies show that both localized objectness supervision and displacement-aware scale fields are critical for class-agnostic localization under incomplete annotations. Code available at: this https URL.
[CV-186] Compact Snapshot Spectral Imaging with Calibration-Free Aperture Diffraction
链接: https://arxiv.org/abs/2608.29230
作者: Tao Lv,Quan Yuan,Shiqiao Li,Chenglong Huang,Linsen Chen,Chongde Zi,Shuming Wang,Xun Cao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Submitted to IEEE TPAMI. Under Review
Abstract:Snapshot Spectral Imaging (SSI) provides high-dimensional temporal-spatial-spectral observation to uncover intrinsic physical characteristics. However, its complex system and repetitive calibration requirements hinder edge applications. Here, we propose a compact, cost-effective, calibration-free SSI method, Aperture Diffraction Imaging Spectrometer (ADIS), which consists only of a diffractive lens with a binary mask and a Bayer-filtered sensor, requiring no additional physical footprint compared to standard RGB cameras. ADIS disperses and multiplexes wavelengths, mapping energy to distinct sensor locations, enabling full-resolution recovery from superpixel-level encodings. ADIS directly leverages theoretically computed PSFs to enable calibration-free spectral reconstruction, while tolerating lens-dependent variations across different optical configurations and bridging the gap between simulation and reality. To achieve SSI by solving a sparsely-constrained inverse problem, we introduce the Orthogonal Diffraction-Aware Unfolding Framework (ODAUF) with Voxel Shift Transformer (VST) for improved orthogonal diffraction perception. Integrating VST into ODAUF forms the efficient Orthogonal Diffraction-Aware Unfolding Voxel Shift Transformer (ODAUVST), delivering excellent recovery and reduced parameters. By elaborating on theory, systematic and comprehensive comparing, and demonstrating real SSI results, we validate the superiority of ADIS, achieving calibration-free full-resolution SSI within a commercial camera footprint.
[CV-187] Using Channel Representations in Regularization Terms: A Case Study on Image Diffusion
链接: https://arxiv.org/abs/2608.29227
作者: Christian Heinemann,Freddie Åström,George Baravdish,Kai Krajsek,Michael Felsberg,Hanno Scharr
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:In this work we propose a novel non-linear diffusion filtering approach for images based on their channel representation. To derive the diffusion update scheme we formulate a novel energy functional using a soft-histogram representation of image pixel neighborhoods obtained from the channel encoding. The resulting Euler-Lagrange equation yields a non-linear robust diffusion scheme with additional weighting terms stemming from the channel representation which steer the diffusion process. We apply this novel energy formulation to image reconstruction problems, showing good performance in the presence of mixtures of Gaussian and impulse-like noise, e.g. missing data. In denoising experiments of common scalar-valued images our approach performs competitive compared to other diffusion schemes as well as state-of-the-art denoising methods for the considered noise types.
[CV-188] Computational Depth Measurement in Thermographic Video: Overcoming Spatial Overfitting via Spatio-Temporal Decoupling
链接: https://arxiv.org/abs/2608.29223
作者: Zain Ul Abidin,Habeeban Memon,Junaid Ahmed
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Accurate through-thickness measurement of subsurface delamination depth in Carbon Fiber Reinforced Polymer (CFRP) is important for structural assessment because defect location determines affected load-bearing layers. Optical pulsed thermography (OPT) provides a two-dimensional thermal video rather than volumetric measurements, so depth must be inferred from temporal heat-diffusion responses. A challenge is spatial dataset bias: when calibration defects follow regular grids, regression models may memorize their geometry instead of learning physical relationship between thermal decay and depth. This work introduces a spatio-temporal decoupling architecture that separates spatial defect localization from temporal depth measurement. Defect regions are first localized using segmentation methods, after which thermal responses are spatially averaged and converted into sixteen physics-informed temporal, energy, statistical, and geometric features. These features expose the one-dimensional heat-conduction relationship while withholding pixel coordinates from the depth model. Four regression models are evaluated using specimen-level cross-validation: Random Forest (RF), Gradient Boosting Machine (GBM), Advanced Multi-Layer Perceptron (Adv-MLP), and XGBoost. Unregularized trees and over-parameterized Adv-MLP exhibit calibration collapse under geometric shifts, with errors exceeding 0.5 mm. In contrast, regularized XGBoost with L1/L2 penalties and column sampling maintains cross-specimen calibration, achieving a mean absolute error (MAE) of 0.056 mm and root mean square error (RMSE) of 0.085 mm. Predicted depths are merged with masks to generate Delaunay-triangulated three-dimensional defect models in three to five seconds per specimen. Results show that mathematical regularization and spatio-temporal decoupling reduce spatial memorization in thermal-video depth regression.
[CV-189] SGPDFuse: Semantically-Guided Physics-Disentanglement General Multi-Modal Image Fusion
链接: https://arxiv.org/abs/2608.29220
作者: Haozhen Wei,Chengjun Jiang,Yutong Guo,Xinrui Ju,Xingyuan Li,Xiang Chen,Jinyuan Liu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ACM Multimedia 2026
Abstract:Multimodal image fusion (MMIF) aims to integrate complementary sensor data into a single representation that preserves intrinsic scene reality while eliminating environmental interferences. Most existing approaches rely on blind feature aggregation, which excels at signal accumulation but fails to distinguish essential content from physical degradations. We propose SGPDFuse, which bridges this gap by mapping inputs into a physics-disentangled structural representation via a Semantic-Physical Parametric Bridge (SPPB) built on pretrained vision foundation models, utilizing the Intrinsic-Variation principle to decouple invariant scene attributes from transient environmental factors. To guide this decomposition, we introduce a Semantic Alignment mechanism: we explicitly anchor the fused representation to salient semantic features in the same foundation model feature space via cosine similarity to preserve critical targets, while enforcing physical texture fidelity through Gram-matrix regularization to strictly eliminate unnatural artifacts. Extensive experiments demonstrate that SGPDFuse achieves state-of-the-art performance across infrared-visible, multi-focus, and multi-exposure benchmarks using a single architecture.
[CV-190] Asymmetric Phase Coding Video Watermarking
链接: https://arxiv.org/abs/2608.29212
作者: Guang Yang,Fengchen Liu
类目: Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR)
备注: 10 pages, 7 figures, 2 tables
Abstract:Existing video watermarking systems are symmetric: the party that can verify a mark holds the extractor weights or generator secret and can therefore also embed one. Benchmarks confirm the consequence, reporting that white-box forgery defeats all evaluated methods. We present a training-free video watermark that removes the shared secret. The signer embeds a complete Ed25519 signature into the phase spectrum of the chroma plane; any party holding the 32-byte public key and public per-video metadata verifies offline, with no model, no registry, and no network. The payload, 1024 bits of signed message with error correction, is an order of magnitude above common learned payloads and is carried by three design elements: a run-length temporal layout whose decoder identifies payload groups by correlation and never reads a frame index, a payload-free search that recovers scale, rotation, and translation from the carrier itself, and a closed-loop signing procedure that selects each video’s embedding strength by self-verification through the unchanged public verifier. On 1000 uncurated real-world clips the system ships a verifying signature for 99.3% of the corpus and accepts a wrong public key zero times in 1000 attempts. An attack-aware acceptance gate yields embeddings that survive H.264 re-encoding at 100% and 50% rescaling at 97.4% on gated clips. The signature also verifies through a real display and capture loop, an axis absent from published evaluations.
[CV-191] Ground-to-Satellite Localization in Unconstrained Image Collections for 3D Scene Reconstruction
链接: https://arxiv.org/abs/2608.29211
作者: Angel Daruna,Ben Southall,Niluthpol Chowdhury Mithun,Kshitij Minhas,Nicholas Meegan,Qiao Wang,Bogdan Matei,Supun Samarasekera,Rakesh Kumar
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: IEEE International Geoscience and Remote Sensing Symposium (IGARSS) 2026
Abstract:Ground image localization with respect to satellite imagery is a key enabler for metrically-accurate, geo-localized 3D scene reconstruction from unconstrained image collections. Existing cross-view localization methods have strict requirements such as panoramic imagery or known initial locations, limiting their applicability for in-the-wild reconstruction settings. We propose a robust hierarchical cross-view localization framework that leverages geometric constraints from Structure-from-Motion (SfM) models derived from unconstrained ground image collections. Our method generates coarse-to-fine pose hypotheses through a cross-view matching approach and aggregates noisy predictions across SfM model(s) using Kernel Density Estimation to recover consensus alignments while filtering outliers. Experiments demonstrate reliable localization performance from challenging image collections. Empirically we found satellite-referenced alignment enables accurate metric scale estimation, doppelgänger detection, and merging of disjoint SfM reconstructions, resulting in more complete, geo-localized site models than are possible with SfM alone.
[CV-192] Bayesian-Optimized Superpixel-GrabCut for Traceable Optic Disc Segmentation
链接: https://arxiv.org/abs/2608.29196
作者: Shraddha Changune,Vivek Noel Soren,Gautam Das,Tapan Kumar Gandhi
类目: Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
备注: Accepted at 15th International Conference on Image Processing Theory, Tools and Applications 2026
Abstract:Optic disc (OD) segmentation is essential for diagnosing ophthalmic pathologies from retinal fundus images. However, prevailing deep learning approaches operate as opaque black boxes, lacking the inference-stage mathematical traceability–a critical requirement for algorithmic auditing and failure analysis in clinical workflows. This paper presents a fully algorithmically traceable and trainable segmentation pipeline that jointly combines superpixel decomposition, hybrid brightness-proximity superpixel scoring, morphological regularization, iterative GrabCut refinement, and elliptical shape fitting. The hyperparameter optimization is formulated as an objective function and solved via Bayesian optimization to eliminate manual parameter tuning. A quantitative evaluation on the Drishti-GS dataset demonstrates that our method achieves a Dice coefficient of 0.9536, matching state-of-the-art performance. By maintaining explicit mathematical transparency across all processing stages, our framework offers a deterministic, traceable alternative to black-box architectures for medical review and debugging.
[CV-193] RLG-TPV: Radar- and LiDAR-Guided Tri-Perspective View Fusion for Camera-Radar 3D Object Detection
链接: https://arxiv.org/abs/2608.29194
作者: Ahmet Mete Dokgoz,A. Enes Doruk,Hasan F. Ates
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Tri-Perspective View (TPV) representations describe 3D scene structure through top, side, and front feature planes, but existing TPV lifting is primarily camera-based, leaving the depth of sampled image evidence ambiguous along projected camera rays. We propose RLG-TPV, a multimodal TPV framework for camera-radar 3D object detection in which radar and training-time LiDAR provide complementary geometric guidance during representation construction. A ray-guided deformable-attention lift weights sampled image features using LiDAR-supervised camera depth probabilities and radar frustum occupancy, while radar additionally refines the depth distribution before lifting. Because conventional radar provides limited elevation information, LiDAR-derived class-occupancy targets supervise the side and front planes during training; the corresponding heads are removed at inference, so deployment requires only cameras and radar. For temporal aggregation, Doppler-guided temporal fusion aligns past features using a motion field anchored by measured radar radial velocity, with gating that limits warping in regions without supported motion. An RCS-aware radar scatter further allows radar evidence to spread over spatial neighborhoods conditioned on radar cross section. On the nuScenes validation set, RLG-TPV achieves 0.4981 mAP and 0.5959 NDS, reducing orientation and velocity error by 31.9% and 30.7% relative to the published CRN baseline. Ablation studies show that ray-level geometric guidance is a major contributor to the final performance.
[CV-194] OPUS-V2: Bridging the Gap between Sparse Points and Dense Voxels
链接: https://arxiv.org/abs/2608.29187
作者: Jiabao Wang,Qiang Meng,Liujiang Yan,Ke Wang,Qibin Hou,Ming-Ming Cheng
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:The point-based occupancy prediction paradigm has achieved an attractive trade-off between accuracy and efficiency by modeling 3D space sparsely. However, its predictions inherently mismatch the dense voxel-based occupancy required by self-driving systems, necessitating hand-crafted heuristics during training and inference that limit final performance. To overcome these limitations, we propose OPUS-V2, a novel framework built upon the pioneering OPUS (occupancy prediction using a sparse set) point-based approach. OPUS-V2 incorporates a lightweight point-voxel transformation (PVT) module behind the decoder to adaptively map sparse predictions into the dense voxel space, eliminating the need for suboptimal operations and improving model accuracy. Furthermore, our architecture decouples feature and occupancy generation processes, allowing OPUS-V2 to adapt to arbitrary occupancy resolutions. OPUS-V2 achieves a state-of-the-art rayIoU of 44.0 on the Occ3D dataset. On the more challenging OpenOccupancy dataset, it attains a competitive 16.4 mIoU while running in real time at 20.6 FPS.
[CV-195] Multi-Scale Temporal Domain Alignment for Federated Video Domain Adaptation
链接: https://arxiv.org/abs/2608.29186
作者: Lee En-Yi Hannah,Haozhi Cao,Yuecong Xu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 10 pages, 1 figure, 5 tables. Open-source will be available
Abstract:Federated Video Domain Adaptation (FVDA) enables collaborative learning across distributed and non-IID video datasets while preserving privacy, but is under-explored due to challenges in aligning temporal information. We propose Multi-scalE Temporal domAin aLignment (METAL), a novel framework that leverages temporal information at multiple resolutions to improve cross-domain video action recognition with only model parameter transfers. METAL trains per-scale transformer encoders on source-clients, then performs independent knowledge voting at each temporal scale to generate robust pseudo-labels on the target-server. A novel L_2 variance penalty enforces cross-scale consistency during scale-based knowledge distillation, preventing a singular dominant scale. The late fusion aggregates features across different scales, where the fusion head is trained via knowledge distillation using confidence-weighted aggregation of scale-wise predictions, enabling the model to effectively exploit complementary temporal information for final predictions. Experiments on Epic-Kitchens-55 and Daily-DA demonstrate state-of-the-art performances, with gains up to 28.47% over current FDA methods. Ablation studies prove that multi-scale distillation and scale coordination are critical for effective temporal knowledge transfer.
[CV-196] Foundational feature fusion for conditional flow matching in 6D pose estimation BMVC2026
链接: https://arxiv.org/abs/2608.29183
作者: Amir Hamza,Davide Boscaini,Fabio Poiesi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at BMVC 2026
Abstract:Conditional flow matching has enabled a step forward in object 6D pose estimation, achieving state-of-the-art performance by progressively denoising and registering object representations to observed scenes. Existing methods require training task-specific encoders supervised on object-scene overlap and rely on trivial feature fusion strategies to resolve pose ambiguities. We present FunFlow6D, a novel flow matching-based formulation that leverages features from geometric and appearance foundation models for pose estimation, eliminating the need for task-specific encoder training. We also introduce a cross attention-based fusion mechanism that dynamically combines geometric and appearance features to provide richer conditioning for the flow matching module. Experiments on four datasets from the BOP benchmark show that FunFlow6D outperforms the previous state of the art while reducing supervision requirements and memory overhead. Extensive ablations validate the contribution of each proposed component. Project website: this https URL.
[CV-197] Dynamic-Robust Photometric-Semantic Reconstruction for Open-Vocabulary 3D Scene Understanding ECCV
链接: https://arxiv.org/abs/2608.29177
作者: Boyu Cai,Li Yang,Yan Xu,Wei Liu,Nian Liu,Sikui Zhang,Yan Wang,Chunfeng Yuan,Weiming Hu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to the European Conference on Computer Vision (ECCV) 2026
Abstract:The integration of novel view synthesis (NVS) and open-vocabulary segmentation (OVS) has recently yielded powerful feed-forward 3D foundation models. However, their inherent reliance on static-scene assumptions leads to severe misalignment of spatial features in unconstrained dynamic environments. To bridge this critical gap, we propose SPAR, a novel joint semantic-geometric encoding architecture that explicitly isolates transient dynamic noise prior to latent space aggregation. Furthermore, we introduce a dynamic-region-aware end-to-end training paradigm that structurally couples motion estimation with multi-view visual and semantic learning. This unified approach enables the network to inherently resolve motion conflicts and distill multi-view consistent, temporally stable scene representations from dynamic inputs. Extensive experiments on the challenging D-RE10K benchmark demonstrate that SPAR achieves state-of-the-art performance. Our end-to-end approach achieves exceptional novel view synthesis quality, yielding a PSNR of 22.15 dB and 23.33 dB given only 3 and 4 input views respectively. Despite being trained in a self-supervised manner, our model achieves an mIoU of 88.5% for motion mask prediction. Furthermore, our analysis reveals a strong inter-task synergy between photometric scene reconstruction and semantic understanding, where semantic synthesis learning consistently enhances photometric fidelity in novel view rendering. Code will be available at this https URL.
[CV-198] A Tensor Variational Formulation of Gradient Energy Total Variation
链接: https://arxiv.org/abs/2608.29172
作者: Freddie Åström,George Baravdish,Michael Felsberg
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:We present a novel variational approach to a tensor-based total variation formulation which is called gradient energy total variation, GETV. We introduce the gradient energy tensor [6] into the GETV and show that the corresponding Euler-Lagrange (E-L) equation is a tensor-based partial differential equation of total variation type. Furthermore, we give a proof which shows that GETV is a convex functional. This approach, in contrast to the commonly used structure tensor, enables a formal derivation of the corresponding E-L equation. Experimental results suggest that GETV compares favourably to other state of the art variational denoising methods such as extended anisotropic diffusion (EAD)[1] and total variation (TV) [18] for gray-scale and colour images.
[CV-199] Feature-Spectral Frag ility in Segmentation: Dataset Dependence Architecture-Specific Localization and Spectral Correlates
链接: https://arxiv.org/abs/2608.29167
作者: Subhash Kashyap
类目: Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
备注: 4 pages, 3 figures
Abstract:Robustness of segmentation models is commonly assessed through input-domain perturbations, while dependence on frequency content within learned feature representations remains less understood. We probe this dependence using targeted post-training low-pass interventions on internal representations of three segmentation architectures, ResNet50-UNet (CNN), VM-UNet (SSM), and Swin-UNETR (Transformer), across CVC-ClinicDB and ISIC2018, with headline evaluations performed on untouched held-out test sets. At cutoff rho=0.25, feature-domain low-pass filtering causes severe degradation on CVC: Dice drops by 100%, 73.2%, and 30.9% for CNN, SSM, and Transformer, respectively, compared with 9.4%, 10.3%, and 0.6% on ISIC. The cross-dataset difference is statistically significant for every architecture. Single-stage interventions further show that sensitivity is localized at architecture-specific depths: the CNN peaks at a mid/late encoder block, whereas the SSM peaks in an early encoder stage on both datasets. Native feature-domain spectral measurements show an inverse association between high-frequency energy and fragility on CVC; the relationship is only partial on ISIC and is therefore treated as a candidate correlate rather than a proven mechanism. Finally, Fourier augmentation improves robustness to input-space low-pass filtering but leaves feature-domain degradation essentially unchanged. These results show that feature-spectral robustness is strongly dataset-dependent, architecture-specific, and distinct from input-domain spectral robustness.
[CV-200] Mapping-Based Image Diffusion
链接: https://arxiv.org/abs/2608.29164
作者: Freddie Åström,Michael Felsberg,George Baravdish
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:In this work, we introduce a novel tensor-based functional for targeted image enhancement and denoising. Via explicit regularization, our formulation incorporates application dependent and contextual information using first principles. Few works in literature treat variational models that describe both application dependent information and contextual knowledge of the denoising problem. We prove the existence of a minimizer and present results on tensor symmetry constraints, convexity, and geometric interpretation of the proposed functional. We show that our framework excels in applications where nonlinear functions are present such as in gamma correction and targeted value range filtering. We also study general denoising performance where we show comparable results to dedicated PDE-based state of the art methods.
[CV-201] Subtraction-Based Tumor Segmentation and Lesion-Centered pCR Prediction for the MAMA-MIA Challenge
链接: https://arxiv.org/abs/2608.29162
作者: Kai Geissler,Raphael Schäfer
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:We describe the submission of team FME to the MAMA-MIA Challenge, which evaluated primary tumor segmentation and prediction of pathological complete response (pCR) from pretreatment dynamic contrast-enhanced breast MRI on an external multi-country cohort. For segmentation, we trained a five-fold residual-encoder nnU-Net ensemble using only the first post-contrast minus pre-contrast image, combined with mirroring test-time augmentation and largest-connected-component filtering. For pCR prediction, we ensembled 25 pretrained 3D video classifiers trained on lesion-centred crops from the pre-contrast and first two post-contrast volumes. FME ranked second in both tasks. The segmentation method achieved a combined performance-fairness score of 0.882, with Dice 0.713 and normalized Hausdorff distance 0.099. The pCR method achieved a combined score of 0.664, balanced accuracy of 0.541, and equalized-odds disparity of 0.212. The results indicate that subtraction-based input and ensembling support robust tumor segmentation under cross-site domain shift, whereas pCR prediction from baseline DCE-MRI alone remains limited. For the submission repository, see this https URL Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG) Cite as: arXiv:2608.29162 [cs.CV] (or arXiv:2608.29162v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2608.29162 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-202] raining-Free Hidden-State Refinement for Flow-Matching Image Generators
链接: https://arxiv.org/abs/2608.29160
作者: Yuanyi Yan,Xinzhe Rao,Canyu Shen,Yang Chen,Yunlu Chen,Meng Tang,Teng Long,Vincent Tao Hu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 7pages,4 figures,5 tables
Abstract:We aim to improve frozen flow-matching image generators by adding inference computation inside the denoiser, without changing model weights or the outer sampler. Existing generators usually spend extra test-time computation by increasing the number of sampling steps, which repeatedly evaluates the entire denoiser and couples quality gains to sampler cost. A key challenge is how to use extra computation inside a frozen transformer denoiser: the method must decide which tokens, layers, and sampling times receive repeated updates while preserving the original generation pipeline. We introduce a training-free looping framework that repeatedly applies selected transformer layers inside each denoising call. Dense and Sparse Token Loop vary the token scope; Sampling-Progress Gating and the loop layer range specify when and where looping is active; loop count and strength control the repeated updates; and Loop Guidance combines ordinary and looped vector-field predictions. Across two Scale-RAE model scales, loop variants improve primary and auxiliary quality metrics with competitive quality–efficiency trade-offs. Loop Guidance further improves both primary metrics across all three tested models; on Scale-RAE DiT2.4B, it raises GenEval from 0.4471 to 0.5691 and DPG-Bench from 0.7656 to 0.8053. Code will be released.
[CV-203] SGRNet: Spatially Guided Radiology Network for Structured Radiological Reporting of Head and Neck Cancer
链接: https://arxiv.org/abs/2608.29153
作者: Ayush Gupta,Vinkle Srivastav,Prateek Upadhya,Amit Gupta,Krithika Rangarajan,Nicolas Padoy
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Automated radiological report generation can alleviate clinical workloads and eliminate observer variability. However, standard free-text generation models pose hallucination risks in dense regions and fail under data scarcity. We address these challenges in Head and Neck Cancer (HNC) from contrast-enhanced CT (CECT) imaging. To enforce factual safety, we reformulate report generation as an anatomically grounded, multi-label, structured reporting task, predicting localized tumor involvement across a hierarchical clinical schema. To bridge the visual gap from missing metabolic imaging (e.g., PET), we introduce SGRNet (Spatially Guided Radiology Network), incorporating two low-cost spatial priors: automated organ segmentations and weakly supervised tumor localization maps modeled via 3D Gaussian heatmaps. These priors are dynamically integrated via spatial feature modulation to guide the network toward subtle tumor-induced structural alterations. Evaluated on a multi-centric dataset of 184 paired HNC CECT volumes and reports, on five clinically salient, densely packed anatomical subsites, SGRNet achieves a mean Average Precision (mAP) of 0.60, an 8.8 percentage-point absolute improvement over strong volume-only 3D baselines.
[CV-204] STARLINC: Satellite Trail Artifact Removal using Inter-Frame Correlation ECCV2026
链接: https://arxiv.org/abs/2608.29145
作者: Shingeon Kim,Hyeyoon Lee,Dain Kwon,Kanghyun Choi,Sunjong Park,Mi-Ryang Kim,Jeong-Eun Lee,Jinho Lee
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted to ECCV 2026
Abstract:The rapid expansion of low Earth orbit satellites such as Starlink is increasingly contaminating astronomical surveys. In practice, contaminated images are often identified through inspection. However, modern surveys generate terabytes of data each night, making manual screening infeasible and necessitating reliable automated methods for satellite trail removal. Unfortunately, existing general-domain line detection methods fail to generalize to astronomical images due to domain mismatch, which are mostly grayscale with sparse bright stars and have a low signal-to-noise ratio. Moreover, training new models from scratch is impractical due to the lack of large-scale annotated astronomical datasets. To address these challenges, we introduce STARLINC, the first ML-based framework for satellite trail removal without requiring tedious pixel-level annotation of astronomical images. STARLINC combines synthetic satellite trail generation for training, inter-frame differential maps from temporally adjacent exposures to highlight transient trails, and heatmaps to provide additional localization cues for pixel-level segmentation. Extensive experiments on real-world data demonstrate substantial improvements over baselines, establishing STARLINC as a scalable solution for next-generation astronomical surveys. Code is available at this https URL.
[CV-205] Chat-Edit-3D: Interactive 3D and 4D Scene Editing via Large Language Models
链接: https://arxiv.org/abs/2608.29137
作者: Shuangkang Fang,Yufeng Wang,Yi-Hsuan Tsai,Wenrui Ding,Yi Yang,Shuchang Zhou,Ming-Hsuan Yang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Project Website: this https URL
Abstract:Recent work on image content manipulation based on vision-language pre-training models has been effectively extended to text-driven 3D scene editing. However, existing schemes for 3D scene editing still have certain shortcomings, hindering their further development as interactive design tools. Such schemes typically adhere to fixed input patterns, limiting flexibility in text input. Furthermore, their editing capabilities are constrained by a single or a few 2D visual models and require intricate pipeline design to integrate these models into 3D reconstruction processes. To address the aforementioned issues, we propose the Hash-Atlas network, which reformulates 3D scene editing as operations on 2D atlas images, thereby achieving a workflow decoupling of the 2D editing and 3D reconstruction processes. Building on this foundation, we introduce a dialogue-based 3D scene editing approach, termed CE3D++, which is centered on a large language model (LLM) that allows arbitrary textual input from users and interprets their intentions, subsequently facilitating the autonomous invocation of the corresponding visual models. Additionally, we extend CE3D++ to monocular 4D scenes by imposing motion constraints on moving objects and further fine-tuning the LLM by creating a trajectory dataset related to editing tasks, which enables the smaller LLM to schedule up to 30 different visual tools accurately. Experimental results demonstrate that CE3D++ effectively integrates multiple visual models to achieve diverse visual editing effects, possessing strong scene comprehension and multi-round dialog capabilities. The source codes and trained models are available at this https URL.
[CV-206] Efficient Language-to-Vision Feature Injection for Referring Single-Object Tracking
链接: https://arxiv.org/abs/2608.29126
作者: Han Wang,Yuxuan Liu,Yuhan Sun,Jian Yang,Xiaotong Xu,Yixuan Lv,Zhuang Zhou,Shengyang Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Referring single-object tracking enables language-grounded target initialization and subsequent tracking by jointly leveraging semantic cues and visual templates. The core difficulty is to use language differently across stages: it is indispensable for grounding but can induce semantic drift during tracking when overemphasized. Meanwhile, current methods often require costly vision-language alignment training. We present LVTrack, a pure transformer framework that introduces a mode-conditioned Gated Feature Injector to adaptively regulate textual guidance and alleviate semantic drift. Together with targeted adaptations, it directly harnesses a frozen vision-language pretrained model, greatly reducing training cost and preserving strong language understanding. To further improve temporal localization, LVTrack integrates hybrid relative-absolute positional encodings with a lightweight memory mechanism and optimizes autoregressive box prediction using a Gaussian-smoothed KL loss. Extensive experiments on standard benchmarks demonstrate that LVTrack achieves strong performance.
[CV-207] Dancing Stick Figures: An Introductory Dataset for Training Video Generation Models
链接: https://arxiv.org/abs/2608.29123
作者: Jin Hyuk Cho
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 9 pages, 4 figures. Dataset, code, reference models, and Colab notebook are linked from the paper
Abstract:Training a video-generation model from scratch is hard for reasons that precede model design. The feedback loop is long: a failure that appears only after a training run can make each attempted fix another run. The data are hard to reach: the corpora and recipes behind strong models are large, heterogeneous, and often unreleased. And scoring is blunt: open-ended generation has no single correct output, and an aggregate score does not by itself establish whether a sample succeeds or which property failed. Dancing Stick Figures is a synthetic video dataset built against these three obstacles. For iteration speed, its 64x64, 64-frame reference task is sized for practical repeated training on a single workstation GPU. For accessibility, the release is a 0.79-GB training tier of 4,020 video clips–1,340 six-second source motions, each rendered from three cameras by a deterministic dataset-generation harness–with checkpoints and a Colab workflow that reruns the reference training pipeline at reduced budget on a 16 GB Tesla T4. For scoring, every frame retains its generating state (ARDY cskel27 joint positions, camera, body parameters, and source motion) and per-pixel depth, surface normals, and part labels. These annotations support dataset-specific metrics for visible topology and part-wise motion; corruptions expose their sensitivities and blind spots.
[CV-208] Acoustically Grounded Cost Learning for Open-Vocabulary Audio-Visual Semantic Segmentation ACM-MM2026
链接: https://arxiv.org/abs/2608.29121
作者: Tianrui Hui,Shaofei Huang,Qisong Han,Yaxiong Wang,Lechao Cheng,Zhedong Zheng,Zhun Zhong,Richang Hong,Meng Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ACM MM 2026
Abstract:Open-Vocabulary Audio-Visual Semantic Segmentation (OV-AVSS) aims to perform pixel-level segmentation of sound-emitting objects from an open set of categories. The previous method relies on a class-agnostic foreground definition, which groups semantically diverse objects into a heterogeneous positive set, causing the model to learn unstable sounding patterns and produce unreliable proposals. To address this, we reformulate the objective to be category-specific and propose a novel Acoustically Grounded Cost Learning (AGCL) framework to transform the static, audio-agnostic visual-text priors into dynamic, audio-grounded cost representations. For intra-category soundingness discovery, we devise Audio-Modulated Cost Generation (AMCG) and Audio-Guided Temporal Aggregation (AGTA) modules to enable both frame-level sounding region highlighting and video-level temporal refinement with a low-intrusive audio injection mechanism. For inter-category distractor discrimination, we introduce a Synergistic Distractor Mining (SDM) strategy, which selectively penalizes acoustically and semantically confusing negative categories to learn more discriminative decision boundaries. Extensive experiments on the AVSBench-OV dataset demonstrate that our method significantly outperforms previous state-of-the-art approaches, particularly on unseen categories. Code is available at this https URL.
[CV-209] GramLoop: Training-Free Gram-Gated Replay for Robust Dense Prediction
链接: https://arxiv.org/abs/2608.29113
作者: Yang Chen,Canyu Shen,Xinzhe Rao,Yuanyi Yan,Yunlu Chen,Meng Tang,Teng Long,Vincent Tao Hu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:We aim to improve frozen DINOv3 dense-prediction models under distribution shift by adding inference computation inside the visual backbone, without changing model weights, task adapters, or prediction heads. The challenge is that repeated transformer-block computation must refine dense features without disrupting the pairwise patch relations that DINOv3 uses to preserve spatial structure. We introduce GramLoop, a training-free framework that replays a short transformer window and controls each replay through final-layer cosine-Gram consistency. Each proposal is propagated through the frozen suffix, measured against the standard DINOv3 trajectory, and accepted through a patchwise gate at the replay-window endpoint. Across object detection and semantic segmentation under corruptions, perturbations, and natural shifts, GramLoop improves all five shifted benchmarks over the paired DINOv3 baseline. On COCO-O, it improves mAP by +0.252 and Effective Robustness by +0.250, while preserving clean ADE20K performance. Code will be released.
[CV-210] NFAD: Nuisance-Filtered Anomaly Detection Under Distribution Shift
链接: https://arxiv.org/abs/2608.29112
作者: Dat Cao,Son Nghiem,Phan Nguyen,Jun Rekimoto,Jhih-Ciang Wu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Recent advances in anomaly detection (AD) for industrial inspection have pushed performance on standard benchmarks toward saturation. However, strong benchmark performance does not necessarily translate to real-world deployment, as these benchmarks are primarily collected under controlled acquisition conditions. Changes in illumination, background, viewpoint, and other environmental factors can shift normal samples away from the learned normal distribution and cause false anomaly responses. We address AD under such distribution shifts by explicitly modeling nuisance variation from changing imaging conditions in feature space. Without anomaly labels or target-domain data, our Nuisance-Filtered Anomaly Detection (NFAD) framework estimates a nuisance subspace from matched feature displacements induced by content-preserving perturbations and suppresses its contribution to anomaly residuals at inference. The same subspace supports two complementary branches: full projection for image-level detection and selective suppression for pixel-level localization, preserving evidence of localized defects. On AeBAD-S, a benchmark specifically designed for AD under acquisition shifts, NFAD achieves 91.0% image-level AUROC, establishing a new state of the art. Notably, this robustness does not come at the expense of conventional AD performance: NFAD remains competitive on standard benchmarks that do not explicitly evaluate distribution shift, including VisA, Real-IAD, and MVTec AD. These results show that explicitly suppressing such nuisance variation improves AD under distribution shift while preserving strong performance in standard settings.
[CV-211] Elastic Triangle Splatting
链接: https://arxiv.org/abs/2608.29106
作者: Tian Shi,Shenhan Qian,Daniel Cremers
类目: Computer Vision and Pattern Recognition (cs.CV); Computational Geometry (cs.CG)
备注:
Abstract:While neural rendering methods such as 3D Gaussian Splatting achieve remarkable visual fidelity, traditional polygonal meshes remain the backbone of established graphics pipelines. Triangle splatting bridges this gap by optimizing triangle primitives as differentiable splats, producing representations that are closer to mesh-based workflows. Central to these methods is the kernel function that softens triangle boundaries to propagate gradients to vertex positions. Existing triangle splatting methods make inconsistent choices of kernel functions, and analysis of these kernels’ optimization behavior has been limited to unstructured triangle soups for novel-view synthesis. In this work, we consider triangle splatting as a generic tool for photometric optimization, comparing kernel properties through two complementary tasks: mesh optimization for shape reconstruction and triangle soup optimization for novel-view synthesis. Along with the analysis, we introduce an elastic kernel function that features bilateral gradient support across the boundary and an adaptive boundary value, which are shown to be essential for robust optimization. Under isolated comparison, our elastic kernel outperforms existing kernels on shape reconstruction and in the majority of novel-view synthesis benchmarks, demonstrating the importance of kernel design in the effectiveness and versatility of triangle splatting.
[CV-212] mperature-Adaptive Transformed Teacher Matching BMVC2026
链接: https://arxiv.org/abs/2608.29099
作者: Hiroaki Aizawa,Yoshikazu Hayashi
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to BMVC2026
Abstract:Temperature scaling is a core component of knowledge distillation, yet its role and effect are still not fully understood. Transformed Teacher Matching (TTM) clarifies the role of temperature scaling by applying it only to the teacher distribution and interpreting the resulting objective as standard distillation with an implicit Rényi entropy regularization on the student. However, TTM still relies on a fixed temperature and does not specify how the teacher-side temperature should be adapted for individual samples. In this paper, we introduce a sample-wise inverse-temperature update for TTM by locally minimizing the Kullback-Leibler divergence between the temperature-scaled teacher distribution and the student’s prediction. We derive closed-form first and second derivatives with respect to the inverse temperature, and show that they can be expressed using variance and covariance statistics of centered teacher and student logits under the transformed teacher weighting. This yields an efficient curvature-aware update that requires one softmax evaluation and a constant number of class-wise weighted sums. Experiments on standard image classification distillation benchmarks show that our temperature adaptation generally improves TTM and WTTM, while remaining competitive with or outperforming prior temperature-adaptive distillation baselines.
[CV-213] SafeAtlas-VL: Beyond Binary Multimodal Safety with Large-Scale Data and Guard Models
链接: https://arxiv.org/abs/2608.29098
作者: Zongrui Wang,Xiangyang Zhu,Sicheng Wang,Han Wang,Dingyi Rong,Zeyu Zhang,Chunyi Li,Yue Shi,Kaiwei Zhang,Zicheng Zhang,Yuan Tian,Qi Jia,Yan Teng,Wei Sun,Ning Liu,Guangtao Zhai
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Multimodal safety moderation requires distinguishing risks arising from visual content, user intent, and assistant behavior. Existing safeguards, however, are typically trained for a single judgment target and reduce safety assessment to a binary decision. Consequently, risk becomes difficult to compare across a multimodal interaction, and ambiguous cases are obscured. We introduce SafeAtlas-VL, a dataset of 1.5M training instances that places image-, request-, and response-level judgments on a five-level ordered scale. We curate a broad collection of safety-relevant data from both real-world and synthetic sources and apply a disagreement-aware annotation procedure. The resulting dataset spans 15 harm categories and 55 fine-grained subcategories, covering a broad range of multimodal safety scenarios. We also construct SafeAtlas-Bench, a held-out set of 5,000 instances for evaluating five-level predictions and continuous risk scores. Upon this dataset, we train the SafeAtlas Guard series of models via target-conditioned tuning for multimodal safety detection. Our models not only perform five-way classification of safety levels but also map safety to continuous scores through a soft cumulative ordinal head. Experimental results demonstrate that guard models trained on our dataset exhibit strong generalization: even without using the training sets of other benchmarks, they achieve competitive performance on the corresponding test sets. Notably, our 8B model attains the overall best performance, outperforming the previous SOTA by approximately 4% in F1 score. Code, data, and models are released to support further research. Warning: this paper contains example data that may be offensive, harmful, graphic, or disturbing.
[CV-214] EviAnchor: Mitigating Hallucinations in Large Vision-Language Models via Regional Visual Evidence Compensation
链接: https://arxiv.org/abs/2608.29092
作者: Sihang Jia,Shuliang Liu,Songbo Yang,Xuming Hu
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Large vision-language models (LVLMs) frequently generate content unsupported by visual inputs. Preliminary experiments show that visual evidence is primarily incorporated into answer-side representations in early-to-middle decoder layers, while its direct influence progressively weakens in later layers. This attenuation suggests that visual evidence acquired earlier may be insufficiently utilized during subsequent generation. Based on this observation, we propose EviAnchor, a training-free and single-branch inference framework that preserves and reactivates visual evidence throughout generation. EviAnchor introduces Regional Evidence Anchor (REA) slots to progressively aggregate dense visual tokens into spatially structured representations. It then strengthens the current decision state’s access to these visual anchors through decision-conditioned evidence routing, mitigating excessive dependence on textual context. Finally, the model resumes its native Transformer computation to integrate the retrieved visual evidence with question semantics and generation history. Experiments across POPE, CHAIR, and MMHal-Bench demonstrate consistent improvements in visual grounding.
[CV-215] AdapToPASS: Ambiguity-aware Adaptive Spherical Transformer for Panoramic Semantic Segmentation
链接: https://arxiv.org/abs/2608.29081
作者: Soumyaratna Debnath,Weiming Zhang,Shriram Damodaran,Dingwen Xiao,Addison Lin Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 25 Pages, 7 Tables, 15 Figures
Abstract:Spherical Transformers have emerged as a promising framework for panoramic semantic segmentation (PASS) by operating directly on spherical geometry and alleviating projection-induced distortions. However, existing architectures often assume canonical spherical structure and stable viewpoints, which are frequently violated in real-world imagery due to unconstrained camera motion, introducing contextual and geometric ambiguity. Consequently, they lack adaptive mechanisms to handle such ambiguity, limiting robustness to unseen spherical transformations. In contrast, biological perception is inherently ambiguity-aware, adapting to fluctuations in cue reliability caused by geometric and contextual variations to maintain stable interpretation under complex transformations. Motivated by this, we first systematically analyze existing PASS architectures under various unseen spherical transformations. We then introduce AdapToPASS, a novel bio-inspired Spherical Transformer that adaptively models contextual and geometric ambiguities for robust PASS. At its core, Adaptive Spherical Attention (AdaSpA) blocks dynamically modulate attention according to local contextual ambiguity, mimicking adaptive, context-driven biological perception. To address geometric ambiguity, AdapToPASS employs Bifocal Spherical Representation to balance field of view and spatial resolution, together with boundary supervision inspired by the boundary-sensitive nature of biological vision. Across indoor and outdoor semantic segmentation, AdapToPASS consistently outperforms prior state-of-the-art methods. Under unseen spherical transformations, it surpasses the next-best method by +13.38% relative mIoU on Stanford2D3D and +18.77% on WildPASS. We further introduce AdapToPASS-Swift, a lightweight variant with fewer than 2M parameters, which surpasses compact baselines while retaining robustness to spherical transformations.
[CV-216] DReSG: Diffusion Residuals for Stylized Gaussian Splatting
链接: https://arxiv.org/abs/2608.29048
作者: Zhongliang Liu,Wenjie Liu,Yang Li
类目: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR)
备注: Accepted to Pacific Graphics 2026 (Conference Track)
Abstract:Reference-guided stylization of scenes represented by 3D Gaussian Splatting (3DGS) is important for efficient and controllable 3D content creation. Existing VGG-feature-based 3D stylization methods provide stable rendered-view optimization, but often under-represent expressive reference style cues; diffusion models offer stronger image priors, yet direct per-view or score-based diffusion guidance can lead to view drift, local artifacts, and hard-to-control appearance updates. We present DReSG, a 3D-grounded residual-feedback framework for stylized Gaussian splatting. DReSG represents attention-guided diffusion proposals as residual targets relative to the current render, and progressively absorbs these residuals into a shared Gaussian scene through multi-view Gaussian feedback. To make this feedback stable and controllable, DReSG modulates residual strength during target construction and combines coverage-aware view selection with conflict-filtered color updates during multi-view fitting. Extensive experiments demonstrate that DReSG achieves competitive reference-guided stylization while better preserving scene structure and cross-view stability. Our project page is available at this https URL.
[CV-217] Di2CycleSB: Towards High-Quality Unsupervised Nighttime Visibility Enhancement via Schrödinger Bridge Transformer
链接: https://arxiv.org/abs/2608.29043
作者: Hanting Li,Xin Sun,Wei Ye,Jungong Han,Liang-jie Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 12 pages, 8 figures, 5 tables
Abstract:Light-effect contamination poses a significant challenge to nighttime visibility enhancement. Most methods suppress light effects by estimating and decomposing them through prior-driven regularization, yet they are often limited by hand-crafted priors and ill-posed nature of decomposition. This work proposes Di ^2 CycleSB, a unsupervised Cycle Schrödinger Bridge Transformer framework guided by dynamic integral image priors, for high-quality unsupervised nighttime visibility enhancement. Specifically, a novel light-effect estimator is introduced to parameterize Gaussian-like adaptive priors by aggregating dynamic integral image representations for non-uniform glow estimation. Then, we propose a prior-informed Generator that exploits light-effect representations to guide long-range dependency modeling within our specific Transformer blocks. We formulate light-effect suppression as a Schrödinger bridge problem and construct forward and backward bridges with cycle consistency constraints to achieve visually pleasing enhancement. Extensive experiments on real-world datasets demonstrate the remarkable effectiveness of our Di ^2 CycleSB in enhancing nighttime visibility. In particular, it achieves effective end-to-end light-effect suppression without any regularization constraints and image decomposition. The code and models are available at this https URL.
[CV-218] sRGB Real Noise Modeling via Noise-Aware Sampling with Normalizing Flows ICLR2024
链接: https://arxiv.org/abs/2608.29038
作者: Dongjin Kim,Donggoo Jung,Sungyong Baik,Tae Hyun Kim
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: ICLR 2024
Abstract:Noise poses a widespread challenge in signal processing, particularly when it comes to denoising images. Although convolutional neural networks (CNNs) have exhibited remarkable success in this field, they are predicated upon the belief that noise follows established distributions, which restricts their practicality when dealing with real-world noise. To overcome this limitation, several efforts have been taken to collect noisy image datasets from the real world. Generative methods, employing techniques such as generative adversarial networks (GANs) and normalizing flows (NFs), have emerged as a solution for generating realistic noisy images. Recent works model noise using camera metadata, however requiring metadata even for sampling phase. In contrast, in this work, we aim to estimate the underlying camera settings, enabling us to improve noise modeling and generate diverse noise distributions. To this end, we introduce a new NF framework that allows us to both classify noise based on camera settings and generate various noisy images. Through experimental results, our model demonstrates exceptional noise quality and leads in denoising performance on benchmark datasets.
[CV-219] DocIntent: Answerability-Guided Agent ic Restoration for Real-World Document Visual Question Answering
链接: https://arxiv.org/abs/2608.29037
作者: Zihan Huang,Shihang Wu,Junle Liu,Peirong Zhang,Yongxin Shi,Xuhan Zheng,Lianwen Jin
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 25 pages, 14 figures
Abstract:Real-world degradations such as blur, shadow, distortion, and moire patterns severely impair the document question-answering capabilities of Multimodal Large Language Models (MLLMs). Applying restoration tools before Visual Question Answering (VQA) is an intuitive solution. However, existing restoration approaches remain limited, as manually designing and executing restoration strategies is labor-intensive and requires domain expertise. Agentic restoration offers new possibilities for automation, yet existing frameworks primarily target natural images and pursue perceptual quality, overlooking that restoration should serve downstream tasks rather than optimize generic image quality metrics. To this end, we explore the value of agentic restoration for real-world degraded document VQA and propose DocIntent, a training-free Answerability-Guided Agentic Restoration framework. DocIntent first assesses question answerability, then identifies task-relevant degradations and selectively invokes restoration tools. A Comparison-Based Rollback mechanism validates each restoration step and reverts it when question-relevant evidence becomes less decipherable. The entire process requires no additional pretrained degradation classifier or image quality assessment model. Extensive experiments on the WildDoc benchmark show that DocIntent consistently improves the average score and consistency of different open- and closed-source MLLMs. The code and experimental data will be publicly available.
[CV-220] Explainable Multi-Loss Distillation Framework for Efficient and Interpretable Shrimp Disease Text Classification
链接: https://arxiv.org/abs/2608.29027
作者: Anh Nguyen Quynh,Khang Nguyen Quoc,Luyl-Da Quach
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at Journal of King Saud University Computer and Information Sciences
Abstract:Shrimp disease classification has become an urgent issue due to its significant impact on the import-export output of producing countries, particularly Vietnam. Most existing studies focus on image-based classification, which typically operates at the late stage of disease manifestation. Therefore, text-based classification has the potential to enable early and timely disease detection. To address this limitation, we introduce the SALT (Shrimp disease text Analysis with multi-Loss disTillation) framework, which incorporates explainability analysis using Local Interpretable Model-agnostic Explanations (LIME) and SHapley Additive exPlanations (SHAP) to evaluate model predictions and interpret the learned linguistic features. Experimental results demonstrate that SALT achieves competitive performance across multiple distillation objectives, outperforming supervised baselines while providing a favorable trade-off between predictive performance and computational efficiency. Moreover, it exhibits strong explainability, accurately identifying key linguistic features and semantic patterns relevant to disease descriptions. These findings highlight the potential of knowledge distillation-based text classification for future applications in early shrimp disease diagnosis and related research directions.
[CV-221] Disentangling Representation using Attributes-based Gaussian Estimation for Medical Sound Diagnosis
链接: https://arxiv.org/abs/2608.29026
作者: Ke Zhao
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Audio and Speech Processing (eess.AS)
备注: 4 figures, 2 tables, code is available at: this https URL
Abstract:Deep learning has a powerful capability of feature extraction. However, the lack of fairness and interpretability in deep neural networks poses limitations to their adoption in the medical domain. This paper proposes a disentangled representation learning (DisenRL) framework, named the Attributes-based Gaussian Estimation for Disentangled Representation (AGEDR), which incorporates Attribute Mapping Embedding (AME) modules designed to map attributes into vectors and align them with a subset of the latent vectors in a Variational AutoEncoder (VAE). This part of the latent vector will be disentangled from the remaining latent vectors by minimizing mutual information. A classifier is then trained using the mean parameters of the latent vectors from the VAE. Extensive experiments demonstrate that AGEDR outperforms both conventional classification models and existing disentangled representation learning methods. The ablation experiments also indicate the disentangling capability and fairness of AGEDR. The source code is publicly available at this https URL.
[CV-222] owards Fully Automated Medical Imaging Code Generation via Validation-based Context Engineering
链接: https://arxiv.org/abs/2608.29016
作者: Zixiao Zhao,Jing Sun,Zhe Hou,Cheng-Hao Cai,Qian Liu,Mengze Li,Zijian Zhang,Jin Song Dong
类目: Computer Vision and Pattern Recognition (cs.CV); Software Engineering (cs.SE)
备注:
Abstract:Large language models (LLMs) have demonstrated considerable promise in program generation for small-scale and conventional application development; however, they remain limited when applied to complex, domain-specific tasks such as medical image processing. General-purpose models lack explicit domain knowledge and robust validation mechanisms to ensure correctness, often requiring substantial human intervention to produce reliable processing pipelines. To address these limitations, we propose AutoMedImg, a multi-agent framework for fully automated medical image processing code generation. AutoMedImg orchestrates specialised agents across two phases: a Planning Phase that performs dataset analysis and architecture design with semantic and formal verification, and a Coding Phase that generates modules in parallel with static checking, execution testing, and assembly validation. This multi-stage validation mitigates error propagation throughout generation, while comprehensive auto-context engineering combining domain-specific knowledge bases, shared memory, and validation feedback automates context construction without manual prompting. A cross-project adaptive pipeline synthesis mechanism further accumulates validated pipelines and retrieves proven components for new tasks based on project similarity, enhancing generation efficiency through cross-project learning. Extensive evaluation across six diverse and well-established medical imaging datasets with five backbone LLMs demonstrates that AutoMedImg achieves zero human intervention, with Dice scores of up to 0.90 for segmentation tasks and 99% accuracy for classification.
[CV-223] RoSe-SLAM: Robust Semantic-Aware Gaussian Splatting SLAM from Dynamic Monocular Videos IROS
链接: https://arxiv.org/abs/2608.29003
作者: Wenting Wang,Jiaxin Guo,Wenzhen Dong,Yun-Hui Liu,Charlie C.L. Wang,Yeung Yam
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted by IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS SYSTEMS (IROS), 2026
Abstract:In dynamic and unstructured environments, conventional SLAM systems generally suffer from significant accuracy degeneration due to their static assumptions. In this work, we propose Robust Semantic-aware Gaussian Splatting SLAM (RoSe-SLAM), to address the dynamic challenge by a holistic semantic scene understanding from uncalibrated monocular inputs, achieving accurate camera tracking and high-quality geometry reconstruction. Unlike conventional semantic SLAM using handcrafted semantic labels, our RoSe-SLAM exploits the semantic feature from 2D foundation model to enhance the dynamic tracking and mapping performance. By distilling the rich semantic features to our Gaussian fields, our method effectively identifies dynamic distractors and achieves semantic-aware multi-view consistency, significantly enhancing the geometric reconstruction and scene inpainting. Specifically, we propose a spatial-temporal motion mask generation module, enabling both long-term motion monitoring and short-term transient dynamics capturing, achieving robust and effective disentanglement of dynamic objects and static backgrounds. During global bundle adjustment, we propose an occlusion-aware keyframe selection mechanism to prioritize the occlusion as metric to pick the keyframes, and a multi-view semantic consistency module to improve the mapping quality in dynamic environments. By combining geometric motion cues with semantic priors, our system dynamically filters unreliable observations and reconstructs accurate static scene geometry. Extensive experiments conducted on benchmark datasets including dynamic TUM, Bonn and Wild-Mocap datasets, demonstrate that our method achieves superior performance in both trajectory estimation and static scene mapping, outperforming existing dynamic RGB SLAM baselines in long-term dynamic indoor environments.
[CV-224] Hydra: A Navigation World Action Model with Discrete Latent Planning and Continuous Flow-Matching Execution
链接: https://arxiv.org/abs/2608.28995
作者: Mohammad Nazeri,Alexandyr Card,Samira Huber,Anuj Pokhrel,Yujun Wang,Ruben Hammele,Daeun Song,Sören Pirk,Xuesu Xiao
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注: 29 pages, 12 figures. this https URL
Abstract:World models let robots imagine possible futures, but exploiting this capability for real-time control is bottlenecked by a representation misalignment: the generative model and the planner operate on decoupled manifolds, so the planner has no shared structure to search over and must instead decode every candidate back into high-dimensional pixel space to evaluate it. This decoding step is a major obstacle to real-time control on physical hardware. In this paper, we present Hydra, a discrete World Action Model that closes this gap by moving the planner, both the sampler and the evaluator, inside the model. Hydra establishes a unified latent manifold over visual states, physical poses, and control actions, then compresses this manifold through modality-specific Vector-Quantized bottlenecks into discrete vocabularies of kinodynamic intents and visual states. Because candidates are now drawn directly from this shared manifold, sampling is informed by the model’s own understanding of the observation rather than proposed blind, and evaluation happens natively within the discrete space: candidates are ranked by a Kinematic-Perceptual Cost, without ever decoding to pixels. We term this Discrete Latent Planning (DLP). Because planning over discrete intents alone cannot supply the smooth, continuous commands physical actuation requires, Hydra pairs DLP with conditional Flow Matching, which maps each selected intent to a continuous trajectory for execution. Evaluated on two physical robotic platforms, Hydra outperforms state-of-the-art world models in goal-directed planning, while matching or exceeding the closed-loop execution capabilities of leading reactive foundation policies.
[CV-225] FractureFields: Contact-Aware Binary Multi-Field Transfer for Fractured 3D Gaussian Simulation
链接: https://arxiv.org/abs/2608.28982
作者: Jianchen Wang,Runyang Qu,Fei Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 10 pages, 4 figures
Abstract:Physics-integrated 3D Gaussian representations make it possible to simulate image-reconstructed assets directly as particles, but current Gaussia-MPM pipelines keep a single Eulerian velocity field even after fracture. When disconnected fragments share interpolation support, they still write to and read from the same grid nodes, producing cross-fragment momentum leakage that appears as residual adhesion and non-physical stretching. We present FractureFields, a topology-adaptive transfer for fractured 3D Gaussian objects. After a structural event assigns persistent fragment identities, FractureFields builds fragment-specific mass and momentum fields in a single P2G pass, advances each field independently, and performs a field-aware G2P update so particles only sample their own fragment’s grid state. To handle re-contact, we add a momentum-conserving contact projection that applies equal and opposite normal impulses only when two fragment fields are approaching, preserving free separation otherwise. Experiments on reconstructed scenes and a controlled re-contact benchmark show that fragment-conditioned routing eliminates realized cross-fragment mixing by construction, while contact projection reduces interpenetration during collision without reintroducing residual coupling. Overall, we argue that post-fracture simulation should treat structural disconnection as a change in local dynamical state, not merely a change in constitutive stress.
[CV-226] NBS: No Bias Stereo
链接: https://arxiv.org/abs/2608.28933
作者: Vage Taamazyan,Zhuowen Shen,Stefan Hinterstoisser,Alberto Dall’Olio,Agastya Kalra,Aarrushi Shandilya,Xin Li,Wenping Wang,Kartik Venkataraman
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Stereo reconstruction is one of the last remaining Computer Vision tasks where all state-of-the-art methods employ a heavy architectural inductive bias. Even though it has been demonstrated that the task can be solved using general-purpose methods, it is widely believed that inductive biases in stereo are strictly necessary for both high-quality results and computational efficiency. We challenge this paradigm. In this paper, we demonstrate that both state-of-the-art accuracy and superior runtime efficiency are achievable with a model completely devoid of architectural inductive biases, relying instead on a simple, end-to-end Vision Transformer. By training on massive synthetic datasets, we show that pure data-driven learning can surpass explicitly engineered geometry. This work proves that explicit inductive biases are no longer a prerequisite for stereo matching, ultimately unlocking true scaling laws for continuous improvement in 3D reconstruction.
[CV-227] Membership is Ownership: A Robust Ownership Verification Framework for Diffusion Models ICDM
链接: https://arxiv.org/abs/2608.28929
作者: Feng Jiang,Zuobin Xiong,An Huang,Zhipeng Cai,Yingshu Li
类目: Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV)
备注: This paper has been accepted to the IEEE International Conference on Data Mining (ICDM) 2026
Abstract:Large-scale diffusion models have fueled numerous profitable downstream applications for AI-related businesses, including visual editing and content creation. Meanwhile, due to the huge amount of resource consumption (e.g., computation and high-quality data) during training, such diffusion models are deemed valuable intellectual property (IP) for tech companies like OpenAI and Google. Yet, the IP assets are vulnerable to various unauthorized uses by adversaries seeking to steal models for customized, usually commercial applications. Some existing approaches have explored IP protection for AI models; however, they mostly face structural limitations in common — using a training-time watermarking by injecting artifacts in the model, which can impose a measurable utility cost and can be weakened by post-hoc fine-tuning. To address these challenges, this work investigates IP protection (i.e., model ownership verification) for diffusion models in a realistic commercial scenario with minimal model utility loss. Specifically, the proposed method builds a framework for model ownership verification, termed ``Membership is Ownership (MiO)‘’, based on a population-level hypothesis test on a private member evidence dataset. MiO verifies ownership using two criteria: model attribution through membership inference and model separation from public references. Both are tested at p10^-6 . We evaluate MiO on DDIM and Stable Diffusion models without modifying the owner model or its sampling pipeline, and report ROC-AUC and true-positive rates at fixed nominal false-positive targets. Furthermore, MiO stays stable under different post-theft fine-tuning and weight perturbation in adversarial scenarios, reflecting better robustness compared to the watermarking methods.
[CV-228] ActiveAugment: Online Active Learning for Augmentation Selection in Deep Learning BMVC2026
链接: https://arxiv.org/abs/2608.28923
作者: Noah Videcrantz,Mostafa Mehdipour Ghazi
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted at BMVC2026
Abstract:Data augmentation is a cornerstone of deep learning pipelines, yet existing strategies treat it as a static, model-agnostic preprocessing step, either relying on expensive dataset-specific policy search or applying transformations uniformly at random, regardless of what the model has already learned. We introduce ActiveAugment, a unified framework that treats augmentation selection as an online active learning problem. For each training minibatch, ActiveAugment generates a pool of candidate augmented views and scores each candidate using a combination of the model’s predictive uncertainty and the feature discrepancy induced by the augmentation. The augmentation under which the current model is most fragile is selected per sample, and the model is then trained with a joint supervised classification and supervised contrastive objective that enforces intra-class invariance to the selected augmentations while maintaining inter-class separation. We evaluate ActiveAugment on eight benchmark datasets spanning natural and medical imaging, using CNN and transformer architectures across three training regimes (training from scratch, full fine-tuning, and linear probing), and comparing eight active selection strategies for augmentation scoring. ActiveAugment outperforms AutoAugment, RandAugment, and TrivialAugment under controlled augmentation shifts across all domains and budgets, with the most pronounced gains at low labelling budgets. On medical imaging datasets, where data is scarce and domain shift relative to natural-image pretrained models is large, ActiveAugment achieves higher test F1 than all baselines, demonstrating strong cross-domain adaptability. Our analysis reveals that the augmentation selection policy evolves meaningfully during training and that strategy choice has a direct impact on generalisation. Code is available at: this https URL.
[CV-229] mmIR: Frequency-Space Inverse Rendering for 3D Millimeter-Wave Radar ADC Synthesis ECCV
链接: https://arxiv.org/abs/2608.28913
作者: Adnan Armouti,Yixuan Gao,Rajalakshmi Nandakumar
类目: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR); Machine Learning (cs.LG); Signal Processing (eess.SP)
备注: Accepted to the European Conference on Computer Vision (ECCV) 2026. Project page: this https URL
Abstract:High-resolution 3D radar data is scarce. Commodity mmWave sensors use small antenna arrays that limit angular resolution to several degrees, and existing datasets provide only 2D range-azimuth maps or sparse point clouds rather than raw analog-to-digital converter (ADC) signals. Hardware scaling is expensive, synthetic-aperture scanning is impractical at fleet scale, and learned synthesis methods are bottlenecked by the very data shortage they aim to address. We present mmIR, an open-source differentiable frequency-modulated continuous-wave (FMCW) radar inverse renderer that fits a physics-based forward model to real captures and re-renders from dense virtual apertures to synthesize high-resolution 3D radar data. Because radar resolution is too coarse to recover geometry directly, mmIR performs LiDAR-assisted inverse rendering: using LiDAR-derived meshes as a geometric scaffold, mmIR optimizes per-vertex International Telecommunication Union (ITU) physics materials, vertex normals, and antenna beam patterns through end-to-end automatic differentiation of a phase-coherent multiple-input multiple-output (MIMO) forward model with multi-bounce propagation, polarization, and free-space diffraction. On seven outdoor and six indoor ColoRadar scenes, mmIR achieves 0.914 mean Pearson correlation on range-azimuth maps versus 0.307 for Sionna-RT. Scenes trained on a cascaded imaging radar transfer to a co-located single-chip radar without re-training (0.554 correlation), and dense virtual arrays (100x100 elements) produce single-frame 3D occupancy validated against LiDAR. Project page: this https URL
[CV-230] Coarse to Fine: Iterative Adversarial Neural Cellular Automata for Medical Image Synthesis
链接: https://arxiv.org/abs/2608.28909
作者: Anh Thi Luu,Nick Lemke,Anirban Mukhopadhyay
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Large-scale, publicly available datasets have driven advances in deep learning, but privacy and legal restrictions often limit data sharing in medical imaging. Synthetic data generation offers a privacy-friendly alternative to enable the training of high-performance models on health data. While most state-of-the-art generative models produce high-quality images, they remain computationally expensive, which limits their applicability on resource-constrained hardware. We propose StyleGANCA, the first lightweight general-purpose NCA-based generative adversarial network. The architecture integrates a StyleGAN-inspired mapping network and adaptive style modulation into a multi-scale NCA synthesis process, enabling latent-controlled image generation through iterative local interactions. We evaluate StyleGANCA on BloodMNIST and PathMNIST against adversarial, variational, diffusion, and NCA-based baselines. Experimental results demonstrate that StyleGANCA achieves competitive image quality with substantially fewer parameters than baseline architectures, achieving the best FID and KID scores on PathMNIST with only 617k parameters. Furthermore, downstream experiments show that the generated images preserve class-specific information and effectively support the training of multi-class classifiers. Our code is publicly available at: this https URL
[CV-231] ReconSplat: Generalizable 3D Scene Reconstruction Beyond Observed Views ECCV2026
链接: https://arxiv.org/abs/2608.28895
作者: Giuseppe Stracquadanio,Kevin Raj,Julia Grabinski,Stefan Roth
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: ECCV 2026. Code and additional visual results are available on our project page: this https URL
Abstract:We introduce ReconSplat, a feed-forward model for 3D scene reconstruction that aims to address the longstanding trade-off between plausible view generation for unobserved regions and geometric consistency, providing both geometrically aligned novel views and sharp depth estimates. Our approach builds on 3D Gaussian splatting (3DGS) as an intermediate differentiable scene representation and integrates it with a multi-view latent diffusion model (MV-LDM) trained to act simultaneously as a refiner and an inpainter for appearance and scene geometry. We enforce geometric consistency by guiding the diffusion process with variational 3D latent features for appearance and geometry, encoded by the feed-forward 3DGS representation and rasterized to 2D latent space. ReconSplat produces both photorealistic novel views and accurate depth maps on real-world benchmarks, RealEstate10K and DL3DV-10K, outperforming existing methods in challenging extrapolation setups. Notably, ReconSplat allows the extrapolation of unseen and challenging viewpoints jointly with coherent and precise scene geometry.
[CV-232] Pixel-wise Geo-registration of Drone and Satellite Images
链接: https://arxiv.org/abs/2608.28891
作者: Qingyang Liu,David G Shatwell,Parth Parag Kulkarni,Mubarak Shah
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Pixel-level cross-view geo-registration aims to align a query image (e.g., drone) to a geo-referenced satellite map so that every query pixel can be mapped to real-world GPS coordinates. Despite strong progress in cross-view geo-localization, existing benchmarks largely provide only GPS labels, limiting evaluation to a single coordinate per image and leaving dense geodetic alignment underexplored. We introduce SkyReg, a dataset and standardized benchmark for pixel-level drone-to-satellite geo-registration, providing dense per-pixel geo-location supervision across diverse settings (orthographic and perspective), scene types (urban, landmark-centric, suburban/rural), and camera configurations. Using SkyReg, we evaluate a broad set of baselines spanning retrieval, feature matching, homography-based alignment, and feed-forward 3D reconstruction. Finally, cross-view pairs from SkyReg, we train a geometry-aware reconstruction pipeline that achieves state-of-the-art results,improving performance by a significant margin.
[CV-233] MWIR-4-Plastic: The Identification of Complex End-of-Life Industrial Plastic using Mid-wave Infrared Hyperspectral Imaging and Machine Learning
链接: https://arxiv.org/abs/2608.28874
作者: Elias Arbash,Andréa de Lima Ribeiro,Filipa Simões,Ahmed Jamal Afifi,Aldino Rizaldy,Yuleika Madriz,Samuel Thiele,Sandra Lorenz,Margret Fuchs,Pedram Ghamisi,Paul Scheunders,Richard Gloaguen
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 15 pages
Abstract:The automated sorting of shredded black plastics from end-of-life (EOF) industrial waste presents a significant challenge in recycling facilities, primarily due to the limitations of current sensing and analytical approaches. Existing studies predominantly rely on single-point contact-based mid-infrared spectroscopy or laboratory hyperspectral imaging (HSI) setups, which fail to provide the spatially resolved analysis necessary for fast, bulk processing. Moreover, available datasets are laboratory-controlled and focus on intact rather than shredded plastics, hindering further recycling refinement. Black industrial plastics, in particular, are underrepresented, while most classification pipelines depend on manual region selection and rule-based spectral matching, neglecting spatial information and modern deep learning (DL) methods. To address these gaps, we introduce the first publicly available HSI dataset of shredded black plastics from EOF vehicle, comprising four industrial polymers across 13 co-registered RGB, VNIR, SWIR, and MWIR scenes and their segmentation pipeline. We developed a multi-modal spectral-spatial framework that integrates foreground isolation, pixel-wise classification, and object-level majority voting. By adapting advanced hyperspectral transformers from earth observation and incorporating chemometric band selection, we achieve accurate classification of complex black plastics. The study establishes the first comprehensive benchmark using nine processing methods, including chemometric, machine learning, and DL architectures. To ensure reproducibility, the complete dataset and methodologies are publicly released, establishing a benchmark for a hyperspectral object-analysis pipeline in industrial inspection.
[CV-234] BlobBoards: Robust Markers for Accurate Pose
链接: https://arxiv.org/abs/2608.28830
作者: James Pritts,Till Sittart,Hendrik Sauer,Silja Janßen,Felix Seegräber,David Nakath,Kevin Köser
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:We propose BlobBoards, a fiducial marker system comprising a dense, multi-scale field of Gaussian blobs and a feature-based pipeline for joint detection, identification, and pose estimation. Each board is registered from hundreds of blob features whose dense spatial coverage constrains pose, while multiple scales preserve detectability across large changes in focal length, distance, and obliquity. Learned local descriptors are matched to the reference pattern and spatially verified, so the correspondences determine pose and certify identity. Against motion-capture ground truth, BlobBoards achieve median translation errors of 3.6-5.0 mm, reducing AprilTag’s median translation error by 89% on small boards and 70% on large ones. They also produce far fewer large-rotation failures than state-of-the-art tag systems. BlobBoards achieve the highest detection rate, 80% versus 74% for AprilTag and 58% for ArUco, with the largest margin on the smallest markers. Under 50% occlusion, they still detect 69% of boards with essentially unchanged median translation error, while AprilTag and ArUco detect none. In experiments BlobBoards give state-of-the-art detection rate, pose accuracy and occlusion robustness.
[CV-235] xt-Driven Artistic Staging: Pose Lighting and Camera References from Paintings
链接: https://arxiv.org/abs/2608.28823
作者: Yunge Wen
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Artists coordinate human pose, illumination, and camera placement to convey narrative and emotion, but existing generative methods typically model these elements independently. We introduce text-to-editable 3D staging, a task that jointly generates human poses, a dominant light, and a camera configuration from an affective description. We construct 11,911 text–staging pairs from 2,328 figurative paintings by reconstructing SMPL bodies, estimating low-frequency illumination, recovering camera parameters, and pairing each scene with ArtEmis descriptions. We train a flow-matching transformer that supports variable numbers of figures and produces multiple staging alternatives for each prompt. On held-out descriptions, the model achieves 32.2% retrieval R@1, compared with 16.6% for CLIP-based nearest-neighbor retrieval, while approximately preserving corpus-level diversity. These results demonstrate the feasibility of generating editable, emotionally conditioned 3D staging references from text.
[CV-236] Explainable Artificial Intelligence (XAI) in Computational Pathology: Definitions Taxonomy and Recommendations MICCAI
链接: https://arxiv.org/abs/2608.28820
作者: Shubham Innani,Suhang You,Adam Shephard,Bhakti Baheti,Francesco Ciompi,Joe Yeong,Nasir Rajpoot,Michael Feldman,Solene Florence Kammerer-Jacquet,Dimitrios Makris,Geert Litjens,Anne L. Martel,Jana Lipkova,April Khademi,Spyridon Bakas, for theMICCAI SIG-CompPath
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注: On behalf of MICCAI SIG-CompPath. More information: this https URL
Abstract:Computational pathology (CompPath) is transforming medicine by leveraging artificial intelligence (AI) algorithms to support diagnosis, prognosis, and treatment prediction from gigapixel whole-slide images. Clinical adoption is progressing, but is constrained by concerns about safety, accountability, and regulatory oversight in high-stakes clinical environments. Explainable AI (XAI) systems hold promise for building trust and enabling verification, yet the literature remains fragmented due to inconsistent terminology, overlapping methodological families, ad hoc validation, and current reviews. This review aims to formalize XAI methods in CompPath through the: i) introduction of a pathology-centric vocabulary comprising seven core terms; ii) development of a taxonomy across methodological families and three orthogonal axes (stage, type, scope); and iii) establishment of a task-driven framework that maps five clinical questions to recommended methods, method evaluation, and deployment context. Five key gaps between current XAI capabilities and clinical deployment are identified, and actionable steps are proposed to advance XAI for CompPath.
[CV-237] FigMirror: Ground It Code It Plot It
链接: https://arxiv.org/abs/2608.28814
作者: Xiaohan Zhao,Jiacheng Liu,Yaxin Luo,Zhiqiang Shen
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Code and data available at this https URL
Abstract:Converting scientific figures into executable code has gained increasing attention, yet existing methods primarily focus on reproducing the reference figure itself. A more practical setting is to plot new data while preserving the visual style of a reference figure (e.g., color scheme and typography). Prior approaches mimic the reference through pixel-level optimization and struggle to carry its style to new data. We show that the key to this task lies in the coordinate grounding and coding capabilities present in modern computer-use models. We propose FigMirror, an agentic framework that unlocks these capabilities through Grounded Measurement, which locates visual elements by coordinates and measures their properties through executable code. We further introduce PlotTwin-Bench, an expert-curated benchmark with fine-grained code and image-level style metrics. Experiments show that FigMirror consistently outperforms existing methods on reference-conditioned style transfer. All plots in this paper are generated by FigMirror, except those produced by other methods for comparison. Our code and data are available at: this https URL.
[CV-238] A Large-scale Evaluation of Text-guided Models for Facial Editing
链接: https://arxiv.org/abs/2608.28802
作者: Rahul Nair,Saurav Pandit,Hannah Kerner
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: ACM Multimedia (ACMMM) 2026 Oral
Abstract:Facial appearance editing powers popular applications like FaceApp and Photoshop. Generative Adversarial Networks (GANs) and 3D Morphable Models (3DMMs) have been widely used for facial editing. GANs can perform varied facial edits (e.g., changing hair color, hairstyle), but often produce unstable edits. 3DMMs produce stable edits, but can only alter pose and facial expression. Recently, text-guided diffusion models like Nano Banana have become popular for image editing. Text-guided models are a compelling alternative to GANs and 3DMMs since they can produce both stable and varied image edits. While text-guided models have been widely tested for whole-scene edits (e.g., ``make the woman play a guitar’'), they have not been comprehensively tested for facial editing. We conducted the first large-scale evaluation ( \sim1 M images evaluated) of six popular text-guided models on a sequential facial editing task. We present Face-Edit-Attributes, the largest collection of 169 facial editing attributes focused on hair, accessories, and pose edits. We compared model performance using two popular celebrity face datasets: CelebA and CelebSET. Our results show that most models performed hair and accessory edits well, but struggled with editing pose. All models over-edit (e.g., changing hair color when asked only to change the hairstyle). We also evaluated demographic biases in each model. Our results show surprising biases in overediting: almost all models created more overedits for dark-skinned male faces and old faces. The code and data for our results (including our repository of \sim 1 M images) can be accessed \hrefthis https URL\textcolorbluehere.
[CV-239] Blind Stereoscopic Omnidirectional Image Quality Assessment Using Predictive Coding Hierarchy
链接: https://arxiv.org/abs/2608.28798
作者: Wei Zhou,André Kaup
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: IEEE TMM
Abstract:Stereoscopic omnidirectional images (SOIs) have provided users with newly immersive quality of experience in virtual reality environments. However, developing efficient and accurate perceptual quality assessment metrics for SOIs remains challenging due to many factors such as freely changeable field of views and binocular vision. In this paper, based on the characteristics of the human visual system (HVS), we propose a Predictive Coding Hierarchy-inspired metric (PCH) for blind/no-reference stereoscopic omnidirectional image quality assessment. Motivated by the viewing process of SOIs, the proposed PCH includes a local cyclopean perception module, a global predictive perception module, and a visual quality regressor. First, observers browse different spherical sceneries from viewports, and aggregate the local visual information to infer the perceptual quality of SOIs. Therefore, we extract various viewports, followed by cyclopean conversion and saliency detection to approach the perception and attention of the human brain. After the local aggregation, viewers then infer the global scene in their minds. Based on the binocular mechanism, we fuse left and right views to perform predictive coding hierarchy modelling. Finally, the visual quality regressor is exploited to obtain the ultimate quality score related to both local and global perceptual cues. Extensive experiments demonstrate that the proposed PCH achieves competitive and consistently improved performance compared with state-of-the-art quality assessment methods.
[CV-240] Breaking Darknet CAPTCHAs with general purpose LLM
链接: https://arxiv.org/abs/2608.28794
作者: Benjamin Fehrensen,Jens Hubler
类目: Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV)
备注: 13 pages, 5 figures, 5 tables
Abstract:Our work evaluates the effectiveness of automated methods for solving CAPTCHA challenges commonly encountered in darknet environments. These CAPTCHAs are typically designed to operate without JavaScript, resulting in distinct characteristics compared to mainstream CAPTCHA systems. Our study considers three representative challenge types: open-circle localization, rotation-based alignment, and object-selection CAPTCHAs. The experiments reveal a systematic limitation of contemporary MLLMs: while they are generally capable of identifying relevant visual structures, they frequently struggle with precise spatial localization and geometric transformations. These deficiencies can be mitigated either through task reformulation or by augmenting the models with specialized image processing tools. These deficiencies can be mitigated by task reformulation or by equipping the model with specialized image-processing tools. We therefore propose a hybrid framework in which an MLLM serves as a high-level reasoning and orchestration layer while delegating geometric computations to deterministic algorithms via the Model Context Protocol (MCP). The resulting system achieves success rates above 90% across all evaluated CAPTCHA types and demonstrates that combining the complementary strengths of MLLMs and classical computer vision yields a more accurate and efficient solver than either approach alone. Comments: 13 pages, 5 figures, 5 tables Subjects: Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2608.28794 [cs.CR] (or arXiv:2608.28794v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2608.28794 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-241] Beyond Representation Learning: A Systematic Study of Joint-Embedding Predictive Generation for 3D Brain MRI
链接: https://arxiv.org/abs/2608.28787
作者: Meng Zhou,Wenhao You,Yuxing Chen,Yueying Tian
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Preprint, code will be released after review
Abstract:Joint-embedding predictive architectures (JEPAs) have primarily been developed for self-supervised representation learning. Denoising JEPA (D-JEPA) recently demonstrated strong generative capabilities on natural images, yet the applicability to 3D medical imaging remains unexplored. Building on the D-JEPA framework, we present Med-D-JEPA, a systematic adaptation and evaluation of joint-embedding predictive generation for 3D brain MRI. Med-D-JEPA operates on continuous latent tokens produced by a 3D KL-regularized adversarial variational autoencoder, and combines masked context prediction, representation-level alignment, per-token diffusion, and iterative next-set-of-token sampling. We evaluate unconditional and class-conditional generation quality on BraTS2019 and OASIS-1 datasets; downstream classification utility; and preliminary whole-tumor segmentation on BraTS2020. Across different generation settings, Med-D-JEPA achieves superior or competitive performance compared to several strong baselines on fidelity and diversity metrics. Compared to training with real samples, Med-D-JEPA-based synthetic pretraining improves classification AUC from 0.63 to 0.85 on BraTS2019 and from 0.78 to 0.87 on OASIS-1. In the segmentation study, pretraining on Med-D-JEPA samples improves Dice from 0.74 to 0.80 and reduces HD95 from 13.40 to 9.56 mm. These findings establish joint-embedding predictive generation as a promising direction for 3D medical image synthesis and encourage further research in this direction.
[CV-242] ClearText-Video: A Large-Scale Text-Centric Video Dataset Bridging Video Restoration and Scene-Text Enhancement
链接: https://arxiv.org/abs/2608.28784
作者: Jinlong Li,Jiaming Ding,Dingfu Lu,Malcolm Hsiu,Chuang Ke,Kangning Yang,Bochen Guan,Lan Fu,Jie Cai,Huiming Sun,Zibo Meng
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: This paper is accepted by 2026 Proceedings of the European Conference on Computer Vision
Abstract:Multimodal Large Language Models (MLLMs) have recently made strong progress in visual–linguistic understanding. However, their performance on text-centric video reasoning remains highly sensitive to input quality. Real-world user-provided videos often contain motion blur, compression artifacts, noise, and low-resolution text, which impair reliable text reading and downstream reasoning. Whether MLLMs can robustly read and reason about real-world scene text under diverse quality conditions remains a fundamental open question. We introduce ClearText-Video (CTVid), a large-scale, scene-text-aware benchmark for studying text-centric video understanding under controlled quality variation. CTVid contains 4,639 real-world text-rich egocentric videos, 550K+ frames, 1.6M human-verified scene-text annotations, and 220K+ spatial/temporal question–answer pairs in Chinese and English. For each high-quality video, CTVid provides content-matched Degraded-Quality and Restored-Quality variants, supporting two task families: Text-Centric Video Restoration and Multi-Quality VideoQA. We evaluate 18 representative restoration methods and 16 state-of-the-art MLLMs on CTVid. The results show that visual enhancement does not guarantee textual fidelity or downstream reasoning gains: blur is more damaging than low resolution, restored videos can alter the textual evidence used by MLLMs, and OCR-only pipelines remain far below direct multimodal reasoning. CTVid exposes the gap between video restoration and text-grounded understanding, providing a rigorous foundation for restoration-aware, quality-robust text-centric video systems.
[CV-243] Adversarial Calibration Attack on Autonomous Vehicles
链接: https://arxiv.org/abs/2608.28778
作者: Liangkai Liu,Qingzhao Zhang,Kang G. Shin
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV); Emerging Technologies (cs.ET); Machine Learning (cs.LG); Systems and Control (eess.SY)
备注: 19 pages, 8 figures
Abstract:Autonomous vehicles (AVs) rely on accurate camera-LiDAR calibration for multimodal sensor fusion. In practice, calibration can drift due to vibration, temperature variation, or minor sensor displacement, motivating online calibration algorithms that detect and correct misalignment at runtime while allowing the vehicle to continue operating without a factory visit. Existing AV attacks largely assume correct calibration. We instead identify online sensor calibration as a new attack plane. A corrupted calibration update can persist across subsequent fusion operations, causing system-wide errors that propagate from perception to planning and control. We present Adversarial Calibration Attack (ACA), the first physical attack against camera-LiDAR online calibration. Using a single adversarial poster, ACA first spoofs the miscalibration detector to trigger the calibration process and then steers the calibration estimator toward an incorrect transformation. A unified optimization jointly designs the poster’s geometry and texture for both objectives. We evaluate ACA across benchmark datasets, simulation, and physical experiments. On benchmark datasets such as KITTI and nuScenes, ACA induces up to 33.9 degrees mean rotational calibration error, thereby severely degrading object detection. In the CARLA simulator, the attack causes a collision when the corrupted calibration is accepted in vulnerable scenarios crafted by the attacker. On a real Husky robot, a printed adversarial poster successfully reproduces the calibration error. These results demonstrate that online calibration is a practical and safety-critical attack surface for AVs.
[CV-244] FairReL: Deepfake Detection using Fairness-Aware Representation Learning BMVC2026
链接: https://arxiv.org/abs/2608.28777
作者: Xiaoman Lu,Jiaqi Li,Shuntian Zheng,Huiping Chen,Yu Guan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 26 pages, 7 figures. Accepted at BMVC 2026. Code: this https URL
Abstract:Although recent deepfake detectors achieve high overall accuracy, their errors remain unevenly distributed across demographic subgroups, with real faces from certain groups more often misclassified as fake. Existing fairness-aware detectors typically regularise the entire feature representation, without identifying or controlling the specific components that drive unfair predictions. Such coarse intervention can over-suppress useful forgery cues while leaving demographic structure in component-specific subspaces. To address this, we identify two subgroup-sensitive components: multi-scale spatial features, which encode local facial and forgery patterns, and fine-tuning-induced residual features, which adapt the backbone to the unfair training distribution. We propose FairReL, a fairness-aware representation-learning framework that targets both components with dedicated demographic supervision. FairReL uses an SVD-decomposed foundation-model backbone to isolate the fine-tuning-induced residual representation, and introduces two complementary losses. Group-Conditional Wavelet Decorrelation (GCWD) suppresses subgroup-imbalanced structure across spatial wavelet sub-bands, while Subspace-Localised Mean Alignment (SLMA) aligns subgroup means within each real/fake class in the residual representation. Experiments on FF++, Celeb-DF, DFD and DFDC show that, against the state-of-the-art fairness-aware detector, FairReL improves unseen-dataset AUC by 3.9% while reducing subgroup FPR disparity by 10.2%. Code is available at this https URL .
[CV-245] Inter-3D VQA: A Roadside Multimodal Benchmark for 3D Spatiotemporally Grounded Visual Question Answering EMNLP2026
链接: https://arxiv.org/abs/2608.28762
作者: Shaozu Ding,Linan Song,Dajiang Suo
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to EMNLP 2026 main conference
Abstract:Recent advances in visual question answering (VQA) and multimodal large language models (MLLMs) have enabled natural-language reasoning over traffic scenes. However, existing benchmarks are largely built from ego-vehicle views or 2D roadside videos, limiting their ability to evaluate 3D-grounded reasoning over real-world distances, trajectories, infrastructure topology, and safety-critical interactions. We introduce Inter-3D VQA, a large-scale roadside multimodal benchmark for 3D spatiotemporally grounded VQA at intersections. Built from synchronized point clouds and multi-view images, Inter-3D VQA contains 407K QA pairs covering lane-level positions, object relationships, motion patterns, and near-miss-oriented interaction reasoning. We further propose Inter-Geo, an MLLM baseline that integrates object- and scene-level aligned LiDAR representations, and Inter-Metrics, a unified evaluation framework for textual consistency, numerical accuracy, and semantic correctness. Experiments show that Inter-Geo outperforms image-based VLMs, especially on grounded spatial and temporal reasoning tasks. Our benchmark and codes are available at this https URL .
[CV-246] MANTLE: A Framework for Adaptive In-Situ Planetary Perception Using a Modular Uplink Principle
链接: https://arxiv.org/abs/2608.28724
作者: Pranav Durai,Gary Doran
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 10 pages
Abstract:Planetary surface exploration missions rely increasingly on autonomous robotic platforms capable of interpreting complex terrain to ensure safe navigation, enable targeted science, and improve operational efficiency, as demonstrated across past Mars missions from Viking through Perseverance. Among the key perception capabilities, landform classification provides contextual information for landing site selection and scientific analysis, while boulder segmentation supports hazard assessment and path planning. This paper presents MANTLE, a multi-task adaptive network for terrain and landform extraction. The model uses a shared DINOv2 backbone for high-level feature extraction with task-specific heads: a classification head for large-scale landform classification, and a segmentation head for pixel-wise boulder localization, each trained on curated datasets built respectively from HiRISE orbital imagery and MSL surface-level imagery. The classification head achieved a test accuracy of 92.56% across seven Martian terrain classes, while the segmentation head achieved a validation IoU of 0.753 and showed strong cross-sol generalization on a held-out test set from previously unseen rover traverses. A key advantage of MANTLE is its modular, extensible design, formalized here as the Modular Uplink Principle: only a shared, frozen backbone needs to remain onboard, while subsequent perception capabilities are trained on Earth as lightweight task-specific heads and uplinked without retraining the full model. This work demonstrates two such high-impact capabilities, terrain classification and boulder segmentation, as an initial realization of a framework built to support many more over a mission’s lifetime. With this foundation, future explorers need not arrive on Mars fully formed, but can continue to learn, adapt, and grow more capable with every uplink.
[CV-247] RoboPhys-3D: A Comprehensive Embodied World Model Evaluation via 3D Reconstruction
链接: https://arxiv.org/abs/2608.28718
作者: Tianyi Wang,Jiazhou Chen,Yiming Xu,Xiangyu Li,Tianyi Zeng,Chih-Hsien Chou,Ning Lu,Liang Peng,Junfeng Jiao,Christian Claudel
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Emerging Technologies (cs.ET); Systems and Control (eess.SY)
备注: 66 pages, 12 figures, 55 tables
Abstract:Video world models increasingly serve as data engines, action planners, and simulators for embodied AI, but conventional embodied world model (EWM) benchmarks lack a unified 3D-grounded protocol for establishing whether generated rollouts preserve the underlying 3D scene state or translate into executable actions. We introduce RoboPhys-3D, a 3D-grounded EWM benchmark built on RoboTwin 2.0, covering 50 manipulation tasks across four regimes, with 5,000 episodes and 25,000 multi-view ground-truth videos. A defining feature of RoboPhys-3D is that generated and ground-truth videos are processed through the same 3D reconstruction pipeline, enabling reconstruction-induced error to be distinguished from generation-induced error. The RoboPhys-3D benchmark organizes 50 complementary metrics into 18 sub-dimensions across four levels: pixel-level fidelity, 3D geometry consistency, state-level understanding, and task-level completeness. We further introduce Average Full Score, a hierarchical score averaging all 50 metrics for comprehensive evaluation, and RoboPhyscore, a compact task-aligned score averaging the metrics most strongly correlated with task success. Among the four representative video world models, Cosmos 3 achieves the highest RoboPhyscore (0.6330, 92.7% of ground truth), while state- and execution-grounded metrics reveal substantial failures that perceptual and vision-language model-based judgments fail to capture. RoboPhyscore further exhibits strong agreement with human evaluation (Pearson r = 0.9761 and Spearman \rho = 0.8962), demonstrating the importance of grounded, execution-aware evaluation for EWM capability.
[CV-248] STEP: A Modular Silent Trial Engine for Operational Evaluation of Digital Pathology AI in Routine Workflow
链接: https://arxiv.org/abs/2608.28708
作者: Gabriele Campanella,Matthew Croken,Olga Lukatskaya,Jane Houldsworth,Ricky Kwan,Peter Schüffler,Chad Vanderbilt
类目: oftware Engineering (cs.SE); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Prospective silent trials provide an important bridge between retrospective validation of artificial intelligence (AI) models and their use in clinical care by evaluating model performance and operational reliability on live clinical data without influencing patient management. In computational pathology, conducting silent trials requires integration across laboratory information systems, digital pathology infrastructure, computational resources, and model inference pipelines, and these workflows are often implemented using application-specific software. We developed the Silent Trial Engine for Pathology (STEP), a reusable software platform for orchestrating prospective silent trials of computational pathology AI models across heterogeneous clinical and computational environments. STEP separates common trial orchestration from institution-specific data access and compute infrastructure through modular adapter interfaces. The platform supports scheduled case discovery, per-slide inference submission, deterministic idempotency, failure recovery, result and ancillary-data ingestion, persistent trial and run state, and audit logging, with compute adapters supporting local execution and high-performance computing environments using LSF and Slurm. STEP was deployed at three institutions to support prospective silent evaluation of EAGLE, an AI model for predicting EGFR mutation status from hematoxylin and eosin-stained whole-slide images. By separating trial-level workflow logic from site-specific integrations, STEP enables a common execution framework to operate across heterogeneous pathology environments while maintaining durable and auditable trial state. This approach may reduce duplicated engineering effort and facilitate systematic real-world evaluation of computational pathology AI before interventional clinical deployment.
[CV-249] Variable-Granularity Tokenization for High-Resolution Object Detection
链接: https://arxiv.org/abs/2608.28706
作者: Khayrul Islam
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:ViT detectors fix a uniform token grid before any learned stage. A native-resolution aerial detector must then choose between resolving few-pixel objects and staying inside compute and memory limits. We introduce VGTok, a training-free tokenizer that sets patch granularity per region from pixels, ahead of the encoder. VGTok scores each region by multi-scale morphological top-hat separability from its surround, then thresholds those scores at a per-image percentile, which fixes the token budget. A structure-tensor gate ( \lambda_\min ) refines only where two-dimensional object structure supports it, leaving one-dimensional clutter coarse. The resulting token set is a strict partition of the image. In a Co-DETR detector with an EVA-02 ViT-L encoder, VGTok clears every published VisDrone-val AP and AP _S at every budget from 40% to 100% of tokens. At 40% it records 44.22 AP with three fifths of the sequence discarded before the first transformer block; dense, it reaches 48.38 AP, 6.08 above the strongest published entry. VGTok transfers to AI-TOD-v2 untouched, same scorer and same rank, and sets a new state of the art at 37.27 AP and 19.51 AP _vt . As a pure drop-in into a frozen checkpoint it reaches 36.29 AP at 78.5% of tokens, above every published entry, where our 376.3M-parameter detector clears a 3.0B multi-expert model. We show that a token budget fixed before the backbone, from local separability and structure geometry alone, holds accuracy on the tiny-object regimes that dominate aerial detection, at 3.1\times less encoder compute and 1.9\times less encoder memory. Code and models are available at \hrefthis https URL\textttthis http URL and \hrefthis https URL\textttthis http URL.
[CV-250] Stochastic Liquid Deformation Fields: An SDE Generalisation of Closed-Form Continuous-Time Cells for Dynamic 3D Gaussian Splatting
链接: https://arxiv.org/abs/2608.28702
作者: Mingzhao Li,Arghya Pal
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: APSIPA ASC 2026 accepted paper
Abstract:Deformable 3D Gaussian Splatting (D-3DGS) reconstructs dynamic scenes by deforming a canonical set of 3D Gaussians through a deformation field of frame time. Replacing its MLP with a stack of Closed-form Continuous-time (CfC) cells-a Liquid Neural Network that solves the Liquid Timeconstant ODE in closed form-gives the field continuous-time behaviour at feed-forward cost. That closed form, however, is only the deterministic limit of a noise-driven system, and drops the stochastic term usually credited for the robustness of liquid networks. We put it back: a small Gaussian perturbation is added to the time gate of every CfC cell, turning the deterministic field into a simple stochastic (SDE) one. The noise is used only during training, needs no solver, and reduces exactly to the CfC when switched off. On the synthetic D-NeRF scenes the stochastic field is on par with the deterministic CfC and beats the MLP baseline on most scenes; on the real-world NeRF-DS scenes the deterministic limit is already best and adding noise does not help. The study thus gives both a clean way to read the CfC field as an SDE and an honest account of when a plain noise term helps and when it does not.
[CV-251] opoAgent : A Structure-Aware Perception-to-Reasoning Framework for Diagram-to-Graph Topology Extraction with Large Vision-Language Models
链接: https://arxiv.org/abs/2608.28701
作者: Bangwei Guo,Xujiang Zhao,Yanchi Liu,Wei Cheng,Shengyu Chen,Dongyue Li,Masaharu Morimoto,Takayuki Kuroda,Dimitris Metaxas,Haifeng Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Diagram-to-graph topology extraction aims to extract a graph of entities and their connections from a structural diagram. This task remains challenging for current vision-language models because it requires both fine-grained perceptual grounding and topology-aware reasoning with global consistency. We present TopoBench-180, a human-verified benchmark for diagram-to-graph topology extraction, and TopoAgent, a structure-aware perception-to-reasoning framework for reliable topology extraction using large vision-language models. TopoBench-180 contains 180 structural diagrams spanning Web-style and Network-style categories, paired with canonical graph annotations. TopoAgent progressively extracts the target graph by combining grounded perception, global structural priors, canonical node inventory construction, node-centric local-to-global relation reasoning, and topological consistency enforcement. Experiments on TopoBench-180 show that TopoAgent outperforms strong vision-language model baselines and recent visual reasoning frameworks, especially on edge extraction. More broadly, this work fills an important gap in multimodal structured understanding by establishing a benchmark and framework for diagram-to-graph topology extraction. The benchmark and associated resources will be publicly released at this https URL.
[CV-252] Beyond Visual Boundaries: Rethinking Scene Segmentation for Movie RAG
链接: https://arxiv.org/abs/2608.28699
作者: Dong-Hee Kim,Seonwoo Choi,Changbeen Kim,Jungmyung Wi,Juyeon Ko,Youngju Choi,Il Hyeon Mun,Hyunwoo J. Kim,Donghyun Kim
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Understanding long-form video remains a fundamental challenge for multimodal large language models (MLLMs). Sparse frame sampling fails to capture fine-grained visual details, while dense sampling quickly exceeds context length limits. Retrieval-augmented generation (RAG) offers a promising middle ground by selectively retrieving relevant video segments for grounded generation, yet its effectiveness critically depends on the quality of the video segments used as retrieval units. In this paper, we investigate RAG for movie understanding, which demands story-level reasoning over characters, events, and narrative arcs spanning hours of content. Scene segmentation, a long-studied problem that partitions movies into semantically coherent units, is a natural candidate for defining such retrieval units. We reexamine whether existing methods actually serve this role through comprehensive evaluation on downstream movie understanding tasks, and find that they consistently fail to outperform naive uniform temporal chunking. Our audit of the most standard scene segmentation benchmarks reveals why: current annotations prioritize visually salient transitions over narrative event structure. Motivated by this mismatch, we introduce NarraScene, a narrative-centric scene segmentation dataset annotated with a three-level cognitive taxonomy spanning physical, character, and narrative change, where every valid boundary requires a narrative-level shift. When used as retrieval units, these narrative-grounded segments outperform uniform chunking on downstream movie understanding tasks, suggesting that the central challenge for scene segmentation in movie RAG is not detecting boundaries, but identifying the narrative event units that matter for movie understanding.
[CV-253] State-Conditioned Visual Evidence Retrieval for Fine-Grained Perception in Document Vision-Language Models
链接: https://arxiv.org/abs/2608.28698
作者: Mingxu Chai,Chenyu Liu,Ziyu Shen,Jiazheng Zhang,Kaidi Zhang,Ruoyu Chen,Jun Long,Jihua Kang,Tao Gui,Qi Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 18 pages, 12 figures
Abstract:Compared with typical vision-language tasks, document parsing places stronger demands on fine-grained visual perception. Existing vision-language model (VLM)-based parsing approaches rely on globally compressed visual tokens, where fine-grained details are entangled within a single representation and repeatedly accessed during decoding. However, we observe that the visual evidence for each prediction is typically localized and conditioned on the current decoding state, whereas such representations must be accessed in full at every decoding step, resulting in inefficient computation. To address this mismatch, we formulate perception as state-conditioned visual evidence retrieval (SCVER) during autoregressive decoding. The model operates on a compact global representation for coarse structure and retrieves a small set of relevant high-resolution regions conditioned on the current token state. This coarse-to-fine design enables on-demand access to fine-grained visual cues, relieving globally shared representations from encoding all fine-grained details. We further find that learning such state-conditioned retrieval in VLMs is challenging and unstable. To stabilize this process, we introduce a Spatially-Guided Learning Objective (SGLO) to guide the retrieval process. Experiments on document parsing benchmarks show that SCVER improves robustness under reduced input resolution and achieves a better accuracy-efficiency trade-off, demonstrating the effectiveness of on-demand visual evidence retrieval for fine-grained perception.
[CV-254] Instruction Distillation: Text Instructions as Visual Examples
链接: https://arxiv.org/abs/2608.28696
作者: Hardik Jindal,Soumyabrata Pal,Sayak Ray Chowdhury
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Visual in-context learning (ICL) with multimodal large language models (MLLMs) is effective for fine-grained visual classification, but each retrieved image example consumes several hundred context tokens, making large- K settings prohibitively expensive at inference scale. We propose Instruction Distillation: an offline procedure in which the MLLM itself generates, for each individual training image, a structured identification instruction encoding general appearance cues, features that differentiate the class from visually similar ones, and a common confusion point. Unlike prior work that produces a single description per class, our instructions are generated per training image, preserving the intra-class visual diversity that per-class descriptions collapse. At inference time, we study five configurations sharing a single CLIP retrieval index: zero-shot, image ICL, instruction-only ICL, and two hybrid variants in which retrieved neighbors are split between images and instructions. Across seven fine-grained benchmarks and two MLLM backbones, instruction based pipelines match, or exceeds image ICL at K=1 and reduces per-query tokens by 2.9\times and inference latency by 3.3\times at K=5 . Hybrid configurations further show that visual and textual ICL signals are complementary, images give visual patterns to learn and see, while instructions give explicit rules and logic. When both of these are provided, the quality of context improves, which is noticeable in the performance.
[CV-255] SNF-Bench: Separating Static Drift from Natural Flow in Long-Horizon Fixed-Camera Video Generation
链接: https://arxiv.org/abs/2608.28694
作者: Matiur Rahman Minar,Seunghun Oh,Ganghyeon Jeong,Unsang Park
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Project page: this https URL
Abstract:Long-horizon video generation is evaluated with whole-frame metrics that reward motion and temporal consistency. For fixed-camera nature scenes this creates an ambiguity: motion of water, fire, smoke, or rain is desirable, whereas motion of the background is an error. A system can therefore score well on motion while its scene drifts, or on consistency while its flow stagnates. We introduce SNF-Bench, an evaluation framework for long-horizon fixed-camera generation that partitions each scene into static support and dynamic flow and reports static fidelity, flow persistence with absolute magnitude, and drift leakage separately, never as one score. Drift leakage is interpretive context rather than a headline measurement. Each factor is validated mechanistically rather than by correlation with preference: we inject global translation, rotation, and scale drift and progressive late freezing at known severity into real generations, and require each factor to respond in its stated direction and to remain selective against corruptions it does not target. Auditing publicly released long-horizon text-conditioned checkpoints under one recorded common inference configuration, plus an image-conditioned track with released-pipeline references and a deployment-sensitivity panel, we find that whole-frame motion and static-region drift induce near-opposite orderings of the same outputs. At maximum controlled translation, fBD and NBF rise to 1.86\times and 1.32\times baseline, but whole-frame Dynamic Degree reaches only 1.07\times —rewarding the corruption. SNF-Bench measures where motion occurs and whether it persists; it does not measure physical realism. Project page: this https URL.
[CV-256] RoboGesture: Real-Time Semantic-aligned Co-Speech Gestures Generation for Humanoid Interaction ECCV2026
链接: https://arxiv.org/abs/2608.28693
作者: Zifan Wang,Ziang Ren,Pengyang Shi,Zirui Wang,Chenghuai Lin,Tianze Wang,Zekun Qi,Liangliang Zhao,He Wang,Li Yi
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at ECCV 2026. Project page: this https URL
Abstract:Enabling humanoid robots to respond to human speech with synchronized and semantically meaningful gestures is fundamental to natural human-robot interaction. However, this task faces three critical barriers: the scarcity of semantically rich datasets, the “modality eclipse” where models ignore audio cues in favor of kinematic inertia, and the sim-to-real gap regarding physical safety. We propose RoboGesture, a robot-centric framework that co-designs data, modeling, and control to power a complete interactive human-humanoid system in which the robot listens, responds, and gestures in real time. We first establish the RoboGesture dataset featuring over 300 gesture categories and develop an automated pipeline to synthesize large-scale collision-free, robot-specific audio-motion pairs. Our architecture features a Hierarchical Semantic-Acoustic Aligner that extracts multi-granular prosodic and semantic cues directly from raw audio tokens. These cues drive a Streaming Conditional Motion Generator based on a diffusion transformer with conditional flow matching. To ensure high responsiveness, we introduce Anti-Inertia CFG Masking, which prevents the model from collapsing into repetitive historical patterns by compelling it to proactively mine control signals from the audio modality. Finally, an MPC-based safety filter ensures real-time, collision-free execution on physical hardware. Experiments on a Unitree G1 humanoid demonstrate that RoboGesture generates safer, more rhythmic, and more semantically appropriate responses compared to state-of-the-art baselines.
[CV-257] Growing a Stand Not a Tree: Joint Canopy Generation Reproduces Crown Shyness
链接: https://arxiv.org/abs/2608.28692
作者: Guang Yang,Fengchen Liu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 8 pages plus references, 8 figures
Abstract:In closed forests, neighboring tree crowns often stop short of touching, leaving a network of narrow gaps known as crown shyness. The pattern belongs to the stand rather than to any single tree, which makes it a natural probe of a question in generative modeling: can a learned model produce a set of objects whose defining structure exists only between them? We formulate stand-level canopy generation as set generation with a flow-matching model, in which attention between trees is the only channel through which coupling can arise. Trained on stands grown by a resource-competition simulation that is provably not reducible to per-tree geometry, the joint model halves the clearance distribution error of an identical-capacity model that generates each tree alone, and the advantage persists at stem densities outside the training range. Against field measurements of a tropical oak forest, a single calibrated scalar yields held-out agreement in gap magnitude and crown asymmetry. The directional statistics of the gaps are controlled by stem placement rather than by the growth rule, and match the field once stem jitter is calibrated. Crown shyness, in both the simulation and the learned model, is a property of the stand and not of the tree.
[CV-258] Defending Wearable VLMs Against Private Attribute Inference
链接: https://arxiv.org/abs/2608.28691
作者: Zhimin Li,Pan Wang,Jingxian Chen,Yuantao Tang,Anthony Chen,Qian Lou,Jingtong Hu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Wearable VLM pipelines promise continuous multimodal assistance from egocentric visual capture: a user asks a task-driven question about the surrounding scene, and the system uses compact visual tokens to support language reasoning. The challenge motivating this work is that the same egocentric evidence needed for useful assistance can also reveal private attributes about the wearer or nearby bystanders. We investigate this as a joint privacy-utility problem for split VLM inference, where visual encoding occurs within a trusted device boundary but intermediate visual tokens may be transmitted to downstream reasoning components. This exposes an understudied leakage surface: even when final textual responses are benign, external attackers or untrusted downstream components can recover private attributes from transmitted visual tokens. To evaluate this tension, we construct a paired privacy-utility benchmark with 3,221 image-question records, each paired with a utility question and privacy labels covering location, income, sex, and interests. We further propose Token-Guided Attribute Privacy (TGAP), a pre-LLM token disentangler that learns a residual transformation of visual tokens before they leave the trusted boundary. TGAP combines utility preservation, identity regularization, semantic privacy suppression, and image-driven representation suppression, avoiding the utility loss caused by coarse hard or attention masking. On the benchmark used for source-model evaluation, TGAP reduces privacy accuracy from 56.7% to 7.4%, a 49.3% absolute drop, while maintaining relaxed utility at 74.4%. These results suggest that securing the compact token interface is a practical path toward privacy-preserving wearable multimodal AI.
[CV-259] Projection-Aware End-to-End Learned Video Compression for 360-Degree Video
链接: https://arxiv.org/abs/2608.28689
作者: Niloofar Maani
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:360-degree video supports immersive applications such as virtual reality, autonomous driving, and education. Because spherical content cannot be processed directly by conventional video codecs, it must first be mapped to a two-dimensional projection. Projection choice affects spatial continuity, sampling uniformity, motion estimation, and compression efficiency. This thesis investigates how projection format influences end-to-end neural compression of 360-degree video. Seven formats supported by JVET 360Lib are evaluated using the scale-space flow model, JVET test sequences, and common test conditions. Each sequence is converted from its source equirectangular projection to a coding projection, compressed at multiple rate points, reconstructed, and converted back. Performance is assessed using PSNR, spherical PSNR, weighted spherical PSNR, and Bjøntegaard delta rate. A differentiable pipeline combining projection conversion, neural compression, and inverse projection is also compared with 360Lib. Results show that equirectangular and padded equirectangular projections provide the highest compression efficiency with the scale-space flow model, while cubemap-based and rhombic dodecahedron projections are less effective. This differs from the conventional HM-16.16 codec, for which cubemap-based formats, particularly equi-angular and adjusted cubemap projections, outperform equirectangular formats. Neural models based on optical flow benefit from the spatial continuity of single-face projections, whereas block-based hybrid codecs better accommodate multi-face layouts. These findings show that projection efficiency is codec-dependent and provide guidance for selecting projections for learning-based 360-degree video compression. Subjects: Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2608.28689 [cs.CV] (or arXiv:2608.28689v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2608.28689 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-260] FLM: Frequency-Aware Language Models for Generative Image Compression
链接: https://arxiv.org/abs/2608.28687
作者: Jiarun Chen,Kejun Wu,Li Li,Chengtao Cai,Zhengguo Li,Chia-Wen Lin
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Generative models have significantly improved the performance ceiling of image lossy compression at low bitrates by exploiting learned priors. However, the generated textures and semantic details may deviate from the source content, thereby affecting the fidelity of image reconstruction. To solve these challenges, we propose FLM, a frequency-aware language model that improves compression efficiency through frequency-domain probabilistic modeling while retaining deterministic reconstruction. At the encoder, the input image is transformed into quantized DCT coefficients, which are organized into discrete sequences using macroblock-based coefficient tokenization. FLM then performs next-coefficient prediction to autoregressively estimate token-wise conditional probability distributions for arithmetic coding, thereby generating a compact bitstream. At the decoder, the LLM and arithmetic decoder jointly recover the frequency-domain data, followed by inverse transformations for image reconstruction. A task-specific frequency-domain dataset and a two-stage fine-tuning strategy are further developed to enable the model to operate across multiple bitrate settings. FLM is a versatile compressor that is compatible with both lossy compression and lossless JPEG recompression frameworks. Experiments show that FLM exceeds conventional and generative lossy compression methods in rate-distortion performance. FLM achieves BD-PSNR gains of 3.30 dB, 3.83 dB, and 3.80 dB than JPEG baseline on Kodak, Tecnick, and CLIC2020, respectively. Better qualitative quality of FLM can be achieved in improving semantically high fidelity and suppressing blocking artifacts. FLM is also validated to be applicable to the lossless recompression task with competitive performance.
[CV-261] Data Diversity Not Frequency Invariance: A Controlled and Self-Audited Study of Compression-Robust Deepfake Detection
链接: https://arxiv.org/abs/2608.28685
作者: Abbas Aliyev,Samir Rustamov
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Multimedia (cs.MM)
备注: 38 pages, 4 figures. Submitted to IEEE Access. Code, per-video predictions and a claims-to-artifacts map: this https URL (v1.0.0); archival deposit: this https URL
Abstract:Frequency features and compression-invariant representation learning are widely assumed to be key to deepfake detection that survives video compression. We test this with CAFRL - block-DCT and FFT-phase streams, compression-level-conditioned band attention, and adversarial (gradient-reversal) compression invariance - and report a controlled negative. Under a pre-registered protocol with capacity- and augmentation-matched controls, a plain EfficientNet-B0 on multi-quality data beat CAFRL as specified at every compression level on the FaceForensics++ test split, by 3.66 AUC points at CRF 40 (paired, single seed). A self-audit of our own negative found four defects biased against the frequency hypothesis, and pre-specified re-tests repairing all four showed the deficit to be a recipe artifact, not an architecture failure: the baseline recipe recovered 3.96 points over the matching shipped-recipe variant. The frequency path made no detectable difference: discriminative alone (standalone validation AUC 0.91-0.98 late in training) but of no marginal value under this fusion, at two feature widths of one 4.0 M trunk, every seed-pooled interval for the intra-dataset compression contrasts including zero; on the single held-out manipulation tested, the fair variants sat below the plain backbone. The adversarial branch, as specified, added nothing and degraded its own conditioning estimator; at the fair recipe it is untested. Robustness under single-pass H.264 re-encoding came instead from data diversity: real constant-rate-factor variants beat synthetic JPEG augmentation by 7.3 points (single runs, non-overlapping intervals). The evidence is FaceForensics+±family, GAN-era and single-codec. Match controls on training recipe as well as capacity, and buy compression robustness with codec diversity before architecture.
[CV-262] Distributed Semantic Segmentation With Improved Rate-Distortion Trade-Off BMVC2026
链接: https://arxiv.org/abs/2608.28684
作者: Danish Nazir,Timo Bartels,Thorsten Bagdonat,Tim Fingscheidt
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: accepted at BMVC 2026 (Oral)
Abstract:Distributed deep neural networks (DNNs) for dense perception tasks such as semantic segmentation execute an encoder DNN on edge devices, and a decoder DNN typically on a large-scale cloud platform with a particular constraint on transmission bitrate. Recent works employ source codecs to enable bitrate-efficient transmission between the edge device and the cloud. However, as these approaches are typically bound to a particular type of source codec and alternative network architectures are often not explored, this results in a suboptimal rate-distortion (RD) trade-off in the low-bitrate regime. In this work, we propose two novel source codecs that \textitenable extremely low bitrates, while improving RD performance. We demonstrate the effectiveness of our proposed source codecs by achieving state-of-the-art performance in distributed semantic segmentation at below 0.2 (0.03) bits per pixel, measured using the mean intersection-over-union metric on ADE20K (Cityscapes).
[CV-263] CARD: Calibration via Agreement in Reverse Diffusion for Out-of-Domain MRI Segmentation
链接: https://arxiv.org/abs/2608.28681
作者: Jiaheng Dai,Weidong Guo,Qingbiao Li,Jie Xu,Yi Guo,Yuanyuan Wang,Zeju Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 10 pages, 5 figures
Abstract:Probability calibration aligns model confidence with predictive accuracy, enabling clinicians to identify unreliable segmentation regions. This alignment breaks down under domain shift, where artifacts and unseen protocols produce confident errors. Existing post-hoc methods adapt the correction at test time, conditioning on predictive entropy, the logit pattern, or augmentation response, but each proxy is read from the terminal prediction, the very quantity that shift corrupts. This motivates reliability evidence beyond the terminal prediction, which categorical diffusion provides in two ways. First, a generative shape prior keeps a capacity-limited reference intact when appearance is corrupted, so its disagreement with the primary segmentor highlights primary-model errors. Second, every reverse step yields a class distribution, separating persistent disagreement from transient discrepancy. Aggregated over the trajectory, this disagreement correlates with Dice at 0.788, against 0.521 for a matched discriminative control. We therefore propose CARD (Calibration via Agreement in Reverse Diffusion), which maps the temporal aggregate of this disagreement to a temperature field applied per pixel across all classes, so that confidence changes while the segmentation does not. Across cardiac, prostate and brain MRI shifts, CARD lowers calibration error in 45 of 49 comparisons against the strongest baseline in each setting.
[CV-264] Multi-Sensor Mapping of Vulnerable Urban Settlements Using SAR Multispectral and Hyperspectral Imagery: A Case Study in Córdoba Argentina
链接: https://arxiv.org/abs/2608.28680
作者: Luigi Russo,Anabella Ferral,Silvia Liberata Ullo,Paolo Gamba
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Submitted to IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
Abstract:Informal settlements represent a major urban challenge in rapidly expanding cities, yet their identification from Earth Observation (EO) data remains difficult because of their heterogeneous appearance and incomplete official inventories. This work presents a multi-sensor deep learning (DL) framework for slum-likelihood mapping in Córdoba, Argentina, integrating high-resolution PlanetScope multispectral (MS) imagery, COSMO-SkyMed (CSK) Synthetic Aperture Radar (SAR) data, and medium-resolution PRISMA hyperspectral (HS) observations. The problem is formulated as a patch-level classification task using the official Registro Nacional de Barrios Populares (ReNaBaP) inventory as reference, and the models are evaluated through four geographically partitioned folds. SAR-only and MS-only baselines, their configurations with PRISMA HS support, and early fusion (EF), middle fusion (MF), and late fusion (LF) strategies are systematically compared. Results show that LF+HS provides the best overall balance between classification performance and spatial selectivity, while PRISMA contributes complementary spectral information alongside the higher-resolution MS and SAR representations. Beyond the standard evaluation against ReNaBaP, an external municipal vulnerability layer is used to interpret detections outside the official polygons, showing that several apparent false positives overlap broader vulnerable urban areas. Thermal analysis further shows that ReNaBaP settlements exhibit significantly higher surface temperatures than their immediate surroundings during a heatwave event, indicating localised surface-heat amplification. Taken together, these results suggest that multi-sensor EO fusion can support both the mapping of ReNaBaP settlements and the interpretation of broader urban vulnerability patterns.
[CV-265] Evaluating Constrained Iterative Refinement for Scalable Vector Graphics Generation with Off-the-Shelf VLMs
链接: https://arxiv.org/abs/2608.28678
作者: Matthew Perlman,James Beetham,Niels Da Vitoria Lobo,Amrit Singh Bedi,Mubarak Shah
类目: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR)
备注: Accepted to Pacific Graphics 2026 Poster Track. 2 Pages. 2 Figures. 1 Table
Abstract:Scalable Vector Graphics (SVGs) power much of the modern visual ecosystem, yet state-of-the-art generative models focus almost entirely on rasterized images. We explore whether inference-time methods can unlock SVG generation capabilities in off-the-shelf vision-language models (VLMs). We systematically evaluate a constrained iterative refinement harness that combines visual feedback, structured editing, and constrained decoding to characterize the capabilities and limitations of current VLMs for SVG generation. Across multiple VLMs and generation settings, we find that constrained decoding improves compilation success rates, while iterative refinement reveals a deficit in visual reasoning and self-correction. Our results highlight both the promise and current limitations of using inference-time methods to adapt general-purpose VLMs for SVG generation.
[CV-266] Automated pipeline for herbarium label digitization
链接: https://arxiv.org/abs/2608.28676
作者: Hiba Abbad,Hanane Ariouat,Eva Perez Pimpare,Nicolas Turenne,Eric Chenin,Abderrazak Sebaa,Edi Prifti,Jean-Daniel Zucker,Youcef Sklab
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Digitized herbarium collections, now comprising over 100 million freely accessible specimen images, have become a critical resource for addressing fundamental questions in ecology and evolutionary biology. Yet the rich metadata encoded in herbarium labels (collector identities, geographic localities, collection dates, and ecological observations) remains largely inaccessible at scale, constraining both biodiversity informatics and the construction of specimen-specific image-text corpora for multimodal AI. We present HERBIOME, a modular end-to-end pipeline for automated herbarium label digitization, integrating YOLOv8-based component detection, CRAFT Hezar word-level text localization, fine-tuned TrOCR for recognition of mixed handwritten and printed text, and GPT-4o Mini for semantic metadata structuring into standardized fields. TrOCR was trained on a multi-source dataset combining general transcription corpora (CREMMA-AN, PictoCatalogs) with herbarium-specific data (RéColNat), achieving a Character Error Rate of 4.05-4.10%. End-to-end evaluation on 450 French herbarium specimens, using a dual-metric framework of Maximum Window Similarity (MWS: 0.614-0.618) and Semantic Metadata Accuracy (SMA: 0.440-0.445), reveals that hybrid training strategies improve semantic fidelity while random sampling maximizes surface similarity, with taxonomic fields remaining the principal bottleneck. By automating the extraction of structured metadata from complex, heterogeneous labels, HERBIOME reduces transcription burden, enables the construction of paired image-text datasets that faithfully capture specimen individuality, which is a prerequisite for next-generation multimodal biodiversity AI systems.
[CV-267] Multi-exposure HDR Imaging: A Review of Pixel-level and Feature-level Reconstruction Methods
链接: https://arxiv.org/abs/2608.28674
作者: Qian Tao,Wei Wang,Chaobing Zheng,Zhengguo Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Published in Sensors, 2026, 26(14), 4649
Abstract:Multi-exposure is an efficient way to capture real-world high-dynamic-range (HDR) scenes. However, HDR imaging suffers from severe ghosting artifacts in dynamic scenes due to the temporal gap between sequential exposures. In this article, we categorize the literature on two important topics on HDR imaging: multi-exposure fusion (MEF) and ghost removal. Conventional filter-based and data-driven methods are studied in pixel space and feature space. For popular deep learning-based approaches, we provide a granular taxonomy based on their alignment and fusion domains: pixel-space methods, which typically employ explicit motion compensation such as optical flow or spatial transformers, and feature-space methods, which leverage implicit alignment through deformable convolutions, attention mechanisms, or latent representation merging. Representative works are compared across different supervision settings, and key design principles are summarized. In addition, this survey summarizes commonly used datasets and evaluation metrics, discussing their applicability under diverse output forms. Finally, major bottlenecks and promising directions for future research are outlined.
[CV-268] AdaptAV: Continuous Adaption of Vision Models for Autonomous Vehicles Using Cloud-based Oracle
链接: https://arxiv.org/abs/2608.28673
作者: Yuheng Zhu,Dhruva Ungrupulithaya,Boluo Ge,Man-Ki Yoon
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 7 pages, 11 figures
Abstract:Deploying vision perception models in autonomous vehicles requires that we prioritize inference speeds, resulting in a model with shallower architectures and lesser model parameters (i.e., more pruned). Such small models do not generalize well, which could result in poor performance when encountered with novel scenarios. We propose a system that overcomes this by continuously retraining the vision models on the cloud with data uploaded by vehicles. We leverage the abundant compute resources, including machine learning accelerators, of the cloud to run a highly-accurate oracle model that will guide the retraining process of the on-vehicle model. This newly trained model is transmitted to the vehicle over the network and is utilized by the vehicle for perceptions, leading to improved inference accuracy over time.
[CV-269] FrameScope: Temporal Data Valuation for Stream Active Learning in Autonomous Vehicle Systems
链接: https://arxiv.org/abs/2608.28672
作者: Yuheng Zhu,Man-Ki Yoon
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 15 pages
Abstract:Autonomous vehicles operate in dynamic, ever-changing environments where new scenarios and edge cases constantly emerge. As a result, static learning models are inadequate for ensuring safe and reliable operation. Continuous learning is essential for adapting to these evolving conditions and maintaining robust performance across diverse real-world settings. However, autonomous vehicles generate massive streams of visual data during operation, and existing continuous learning approaches typically rely on heuristic sampling methods that fail to capture temporal dynamics, often overlooking critical learning opportunities or selecting redundant frames. In this paper, we introduce FrameScope, a temporal data valuation framework for continuous learning in autonomous vehicles. FrameScope extends neural tangent kernel theory to temporal domains, enabling principled valuation of streaming visual data. Unlike cloud-centric methods that transmit all video data for processing, our approach performs principled, local frame selection on the vehicle and queries a cloud-based oracle model only for labels of those high-value frames. Extensive experiments across multiple domain shifts show that FrameScope consistently outperforms existing methods, achieving higher sample efficiency and significantly reducing catastrophic forgetting in autonomous vehicle perception. By valuing data on the vehicle and querying only labels for selected frames, FrameScope reduces bandwidth requirements, enabling scalable operation with a lightweight cloud labeling service.
[CV-270] Measuring Similarity between Artistic and AI Generated Images using Siamese Neural Networks NEURIPS2025
链接: https://arxiv.org/abs/2608.28671
作者: Diego Castro Elvira,Navil Pineda Rugerio,Jesús García-Ramírez,Cecilia Reyes-Peña,Ricardo Ramos-Aguilar
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted to LatinX in AI Research Workshop at Neurips 2025
Abstract:AI-generated art has sparked debates around potential plagiarism, as these images may closely resemble existing artworks. This research quantifies the similarity between original pieces and AI-generated counterparts, particularly those produced by the Stable Diffusion XL Refiner 1.0. We use Siamese Networks with frozen CLIP encoders and cosine similarity optimized through triplet loss. A dataset of paired original and generated images was built using image-to-image generation and custom prompts, enriched with semantic descriptors and BLIP-2 captions. Prior studies report up to 81% style replication and 90% visual similarity. Our results show high discriminative performance: training accuracy reached 99.9%, and the best model configuration achieved 99.4% test accuracy with strong inter-class separation ( \delta \mu = 0.677), demonstrating the effectiveness of our semantic-visual embeddings.
[CV-271] Memory-Efficient Training-Free Acceleration of Diffusion Transformers with BaryCache
链接: https://arxiv.org/abs/2608.28670
作者: Chengjie Lu,Tianchi Deng,Zhengqi He,Zhijian Gao,Huisi Wu,Xueliang Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ICITES 2026
Abstract:Diffusion Transformers achieve high-fidelity image and video generation, but their iterative sampling remains expensive, for each denoising step requires large matrix operations. Existing cache-based acceleration reduces redundant computation yet increases the VRAM footprint by storing intermediate states, which can directly constrain inference batch size. In this work, we propose a training-free acceleration method that performs stepwise forecasting for DiT sampling using a Barycentric Extrapolator. By leveraging barycentric extrapolation, our predictor is numerically stable and alleviates oscillatory artifacts analogous to the Runge phenomenon during forward forecasting. Across extensive experiments on both image and video generation, our approach provides a favorable trade-off between memory usage and perceptual quality, while delivering up to 3.30x end-to-end sampling speedup compared with baseline DiT inference.
[CV-272] MIRAG E-CAD: Construction-Mediated Multimodal Generation of Executable CAD Programs
链接: https://arxiv.org/abs/2608.28669
作者: Jizong Zhan
类目: Computer Vision and Pattern Recognition (cs.CV); Programming Languages (cs.PL); Software Engineering (cs.SE)
备注: 64 pages, 5 figures, 20 tables, 7 appendices. Code, evaluation scripts and run reports: this https URL
Abstract:Recovering an executable parametric CAD program from an observed object is fundamentally ambiguous, because the same final geometry can result from different construction procedures. We study this problem from four types of input: natural-language descriptions, rendered images, point clouds, and STEP/B-Rep geometry. MIRAGE-CAD maps each input to a shared construction representation and mediates program generation through an explicit construction-plan interface. The resulting Python CAD code is executed by an OpenCASCADE kernel to build the solid and export it as STEP. On 2,500 held-out queries per modality, the system achieves 55.4-70.0% build success and 52.3-66.2% STEP export success without retrieval at inference. Controlled comparisons show that strong reconstruction does not depend on expressing the construction representation as text: a decoder conditioned directly on the continuous representation also reconstructs strongly, while an exposure-matched plan-based decoder shows no detected material loss in per-part geometric fidelity. The explicit plan instead provides a readable and separately measurable intermediate representation whose agreement with the reference construction is informative about downstream execution success. Finally, we show that executable validity, geometric fidelity, and parametric responsiveness can diverge substantially and should therefore be evaluated separately.
[CV-273] Improving Spatial-Temporal Reasoning in Video-Language Models with Structured Video Prompting
链接: https://arxiv.org/abs/2608.28666
作者: Sadegh Mohammadian
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Video-language models (VLMs) remain brittle on tasks that require tracking events over time and grounding answers in specific spatial regions. We propose that part of this limitation can be addressed through better organization of visual evidence at inference time. We introduce structured video prompting, a training-free inference-time method that augments the input video with lightweight spatial structure and temporal structure, providing explicit anchors for organizing evidence across space and time without changing model weights or decoding and without altering the question prompt in the main comparison. We evaluate this approach on two complementary video benchmarks and two open video-language models. Across these settings, structured inputs improve performance in several cases, with gains varying by model and task. Our findings suggest that some failures of VLMs arise not only from reasoning capacity, but also from how video evidence is presented at inference time. These results highlight structured video prompting as a simple and practical direction for improving video understanding.
[CV-274] Understanding Temporal Semantic Stability in Open-Vocabulary UAV Perception through Metric 3D Fusion
链接: https://arxiv.org/abs/2608.28665
作者: Saurbh Singh Jamwal
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注:
Abstract:Recent open-vocabulary segmentation models have advanced semantic perception for UAVs, but predictions from moving aerial platforms can remain temporally inconsistent across repeated observations of the same physical scene. We investigate temporal semantic stability by associating frame-wise predictions with persistent world-space locations through metric 3D fusion. We introduce a voxel-level evaluation framework that jointly characterises final semantic agreement, Semantic Belief Drift (SBD), Observation Persistence (OP), and semantic uncertainty. Experiments on UAVid-3D reveal substantial frame-wise semantic flicker and show that high aggregate world-space agreement can overstate temporal stability when locations have limited repeated-observation support. Persistence-stratified analysis shows that recurrent voxels expose greater semantic disagreement, while belief drift decreases as additional evidence accumulates. This behaviour is observed across two segmentation backbones and remains consistent under variations in voxel resolution, geometric association, and temporal sampling density. Conditions that reduce world-space recurrence can increase apparent aggregate stability, demonstrating that semantic consistency must be interpreted together with observation support. Our findings highlight observation persistence as an essential conditioning variable for evaluating long-horizon semantic reliability.
[CV-275] he Potential of Haptic Foundation Models
链接: https://arxiv.org/abs/2608.28664
作者: Jianquan Wang,Haiwei Dong,Abdulmotaleb El Saddik
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV); Multimedia (cs.MM)
备注: accepted by IEEE Consumer Electronics Magazine
Abstract:Despite the success of foundation models in language and vision, their expansion into embodied AI is bottlenecked by a lack of generalized touch sensing. This limitation is especially relevant to consumer electronics, where smartphones, wearables, VR controllers, home robots, and health monitoring devices require safe and adaptive physical interaction. Constrained by hardware heterogeneity and the necessity of active physical data collection, current haptic models remain rigidly task-specific. To overcome these limitations, this article explores the transformative potential and developmental trajectory of Haptic Foundation Models (HFMs). We detail the paradigm shift required to transition from passive Large Language Models and Vision Language Models into active HFMs across four core dimensions: action coupling, physical dynamical representation space, continuous time-series data granularity, and action-conditioned future state prediction. Furthermore, we synthesize existing large-scale tactile datasets and benchmark UniTouch, AnyTouch, T3, and Sparsh on TacBench for force estimation, slip detection, and relative pose estimation.
[CV-276] Open-Set Cattle Muzzle Identification: A Leakage-Controlled Benchmark and Evaluation Protocol
链接: https://arxiv.org/abs/2608.28663
作者: Lalit BC,Dharmendra Singh Chaudhary,Shovit Nepal
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Reliable individual cattle identification supports disease surveillance, vaccination records, breeding management, and livestock insurance. Although the bovine muzzle provides a stable, non-contact biometric, existing muzzle-recognition systems largely assume a closed set of enrolled animals, limiting their practical deployment. We reformulate cattle muzzle biometrics as an open-set, gallery-based identification problem that can reject previously unseen animals and support incremental enrollment without model retraining. We introduce a leakage-controlled evaluation protocol based on identity-disjoint splits, per-fold retraining, held-out threshold calibration, verified duplicate removal, and bootstrap confidence intervals. We evaluate the framework using two contrasting embedding configurations: a hybrid CNN-ViT metric-learning model and the MegaDescriptor-L foundation model. Under oracle threshold selection, the hybrid model achieves detection-and-identification rates of 98.3%, 96.4%, and 93.6% at target false-acceptance rates of 10^(-1), 10^(-2), and 10^(-3), respectively, while MegaDescriptor-L achieves 99.3%, 98.1%, and 96.1%. However, deployable threshold calibration reveals a substantial difference between oracle and calibrated performance: the hybrid model achieves a false-acceptance rate of 1.03% at a 1% target, whereas MegaDescriptor-L reaches 2.44%. Incremental enrollment further achieves Rank-1 accuracy above 91% with a single reference image and up to 97.3% with eight reference images, without retraining the model or degrading the existing gallery. These results demonstrate that threshold calibration, leakage control, and embedding quality are critical for reliable open-set cattle identification and provide a practical evaluation framework for deployment-oriented animal biometric systems.
[CV-277] RedLight-VLA: Models for traffic-rule grounding and behavioral emphasis in driving policies
链接: https://arxiv.org/abs/2608.28656
作者: Bala Murali Manoghar Sai Sudhakar,Sourab Bapu Sridhar,Sandipan Das,Rahul Ahuja,Meda Lazar,Ashish Garg,Pratik Likhar,Senthil Yogamani
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Behavior-cloned Vision-Language-Action (VLA) driving policies struggle with rare rule-governed maneuvers at signalized intersections. Braking and launching examples contribute little to averaged trajectory loss, while fused representations lack explicit supervision for the governing traffic-light and stop-line state. We present RedLight-VLA, a training objective that uses expert futures and automatically generated perception targets without additional manual rule annotation. First, trajectory-derived behavioral reweighting (BR) emphasizes rare deceleration and acceleration using rotation-invariant longitudinal dynamics and a scale-preserving reduction that exactly recovers the baseline when disabled. Second, parallel auxiliary (AUX) heads ground traffic-light and stop-line state in continuous post-fusion rule tokens, without autoregressive language generation or changes to the trajectory decoder. We evaluate on a curated set of 20 s sequences with a 5 s prediction horizon. Controlled variants share the same backbone, training data, decoder, and evaluation population. Against an otherwise identical VLA baseline, RedLight-VLA reduces red-light stop-line overshoot from 7.3% to6.8%, reduces stop-line velocity error by 12.7%, and improves 3 s trafficlight-sliced ADE/FDE from 0.274/0.964 m to 0.247/0.897 m. Green-light false stops increase from 3.2% to 3.9%; however, combining BR with AUX supervision mitigates the larger increase observed for AUX alone (4.0%). The combined model also improves non-traffic-light ADE/FDE from 0.268/0.956 m to 0.241/0.876 m and outperforms either mechanism alone on all four sliced displacement measures.
[CV-278] MedTVL: Harnessing Vision and Language for Medical Time Series Classification
链接: https://arxiv.org/abs/2608.28605
作者: Jiexia Ye,Jia Li,Fugee Tsung
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注: 12 pages, 10 figures, 7 tables
Abstract:Recent advancements in multimodal learning for medical time series (MedTS) classification highlight the benefits of integrating complementary modalities for clinical decision. However, existing methods typically focus on bi-modal interactions (e.g., time series and text), leaving the tri-modal synergy between time series, vision, and language largely unexplored. Inspired by diagnostic practice synergizing numerical assessment, visual inspection and clinical context, we introduce MedTVL, a text-guided dual-pathway architecture tailored for MedTS classification. Specifically, it synergizes a convolution-based temporal pathway for fine-grained temporal dynamics from raw numerical sequences and a transformer-based visual pathway for holistic morphological structures from time-series-derived images. Such combination of cross-modal and architectural heterogeneity provides a comprehensive diagnostic perspective. To further resolve potential diagnostic ambiguity, both pathways are guided by adaptive medical textual semantics. Finally, a Mixture-of-Experts mechanism dynamically routes each instance to specialized fusion experts, capturing instance-specific reliance on the temporal and visual pathway outputs. In addition, MedTVL supports multimodal contrastive learning to mitigate the clinical label scarcity challenge. Extensive experiments across multiple medical datasets and tasks, spanning supervised, few-shot, and contrastive learning settings, demonstrate the superiority and transferability of MedTVL, highlighting its potential for robust clinical decision support.
[CV-279] Integrating Triaxial IMU Sensors and Ensemble Learning for Effective Parkinson Disease Severity Classification
链接: https://arxiv.org/abs/2608.28602
作者: Rehan Khan,Muhammad Junaid Asif,Rana Fayyaz Ahmad
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Parkinson disease PD is a progressive neurodegenerative disease that can have a significant impact on motor performance resulting in the appearance of symptoms such as tremors rigidity postural instabilities and bradykinesia. Timely clinical treatment disease management and quality life of the patients are closely linked to early and appropriate identification of PD. Over the past few years the growth of wearable sensor technology and artificial intelligence AI have made it possible to create noninvasive and data driven disease detection methods. This paper proposes a comparative system using artificial intelligence to detect Parkinsons disease by analyzing the motion and tremor data captured by an inertial measurement unit IMU. The data comprises the signals of the acceleration and gyroscope sensors measuring movement in three directions X Y and Z. The signs and symptoms provide helpful information about subtle motor deficits associated with PD. Several classification models like Support Vector Machine SVM Logistic Regression LR KNearest Neighbors KNN Decision Tree DT Extreme Gradient Boosting XGBoost and Light Gradient Boosting Machine LightGBM were used to compare their effectiveness. The Logistic Regression model had a performance around 75 percent in all evaluation metrics and KNearest Neighbours KNN around 90 percent. The support vector machine SVM performed almost 94 percent whereas the performance of classifiers such as Decision Tree and XGBoost was close to 96 percent and overall classification efficacy respectively. LightGBM model performs consistently at the best rank among all of the evaluated methods having Accuracy, Precision, Recall and F1score of around 97 percent. The results show that the proposed machine learning approach offers an accurate and effective predictive capability in the classification of PD severity.
[CV-280] ConCA: Concentration-Aware Channel Attention for Fine-Grained Visual Recognition
链接: https://arxiv.org/abs/2608.30183
作者: Yu-Sheng Liu,Yu-Chen Tung
类目: High Energy Physics - Experiment (hep-ex); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Lightweight channel attention mechanisms are widely used in image classification, yet their effectiveness in fine-grained visual recognition (FGVR) remains limited. Most modules summarize each channel by global average pooling (GAP), which captures activation magnitude but ignores spatial concentration, so channels with different spatial distributions but identical means receive the same descriptor. We propose Concentration-Aware Channel Attention (ConCA), which pairs the mean with a shift-invariant negative-input entropy (NegEnt), computed via a softmax over the negated activations, forming a dual descriptor that jointly encodes magnitude and concentration. A depthwise 1-D convolutional multi-layer perceptron (MLP), whose parameter count is linear in the number of channels, maps the pair to a per-channel weight. On six fine-grained benchmarks, ConCA improves over attention-free, SE-Net, and ECA-Net baselines as well as four richer descriptor-based modules under a controlled from-scratch protocol, and it generalizes across eight backbones on iNat2021-mini. These results indicate that the channel descriptor, together with the per-channel gating that maps it to attention weights, is an important but underexplored aspect of lightweight channel attention in FGVR.
[CV-281] he MYOSAIQ Challenge: Myocardial Segmentation with Automated Infarct Quantification
链接: https://arxiv.org/abs/2608.29246
作者: Olivier Bernard,William A. Romero R.,Cyprien Bouton,Celia Goujat,Hang Jung Ling,Pierre-Marc Jodoin,Fumin Guo,Calder Sheagren,Graham Wright,Abdul Qayyum,Moona Mazher,Steven A. Niederer,Hairui Wang,Xiaomei Wu,Franz Thaler,Gernot Plank,Martin Urschler,Ricardo M. Rosales,Esther Pueyo,Nicolas Duchateau,Frederic Cervenansky,Patrick Clarysse,Loic Belle,Thomas Bochaton,Nathan Mewton,Magalie Viallon,Pierre Croisille
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted for publication at the Journal of Machine Learning for Biomedical Imaging (MELBA) this https URL
Abstract:Late gadolinium enhancement (LGE) cardiac magnetic resonance (MR) imaging is the modality of choice to assess myocardial infarction (MI) lesions. Nowadays MI volume quantification is not performed routinely in clinical practice. Numerous deep learning (DL) methods have been developed to automate the segmentation of the myocardium and infarct regions. However, most studies rely on relatively small datasets which typically undergo pre-processing steps to standardize images and focus on a specific phase of myocardial infarction following reperfusion therapy. These limitations have impeded the development of models that are generalizable across diverse conditions and thus suitable for routine clinical use. To advance research and establish benchmarks in generalizable learning for myocardial infarct quantification, this paper presents findings from the Myocardial Segmentation with Automated Infarct Quantification (MYOSAIQ) challenge. The dataset set up for the challenge combines 439 CMR volumes from two multicenter clinical trials, with representative data acquired in acute and chronic phases after acute MI. Data were acquired in 16 centers using MRI scanners from three different vendors. Six teams participated until the end of the challenge, employing various baseline models, data augmentation techniques, and confidence strategies. To enhance the significance of this study, we compare the challengers’ results with those of fine-tuned foundation models. Our results indicate that well-designed UNet-based techniques outperform fully automatic foundation models for LGE MR segmentation. While the best methods achieve high-quality and stable delineations of the left ventricle and myocardium under various conditions, they remain improvable in accurately segmenting infarct regions.
[CV-282] Generative Translation Priors: Bayesian Imaging with Cross-Modality Image Translation
链接: https://arxiv.org/abs/2608.28872
作者: Evan Bell,Jiaming Liu,Yifan Chen,Yu Sun
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 51 pages, 12 figures. Code is available at this https URL
Abstract:The ability to leverage images from co-available modalities to inform target-domain reconstruction is highly desirable in imaging algorithms. In this work, we introduce Generative Translation Priors (GTP)–a Bayesian framework that transforms diffusion-based image-to-image translation models into cross-modality image priors for ill-posed imaging inverse problems. GTP incorporates target-domain measurements through likelihood guidance, steering the translation process toward the desired posterior distribution. The framework is grounded in a theoretical analysis of the resulting posterior dynamics, which reveals an intrinsic bias introduced by likelihood guidance. We further characterize this bias and derive a ground-truth-free formulation for its estimation, enabling it to serve as a practical metric for assessing posterior sampling quality. Building on this analysis, we derive two discretized GTP algorithms based on gradient and proximal likelihood guidance, respectively. We validate GTP on computed tomography reconstruction with magnetic resonance side information, and on positron emission tomography reconstruction with computed tomography side information. Experiments demonstrate that GTP effectively incorporates complementary cross-modality information and achieves high-fidelity reconstruction even under severely undersampled measurements.
[CV-283] Reconstruction-Aware Cryo-EM Particle Picking
链接: https://arxiv.org/abs/2608.28838
作者: Riku Itsuji,Yuanhao Wang,Xingjian Li,Seonghui Min,Hideo Saito,Min Xu
类目: Biomolecules (q-bio.BM); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Cryo-electron microscopy (cryo-EM) determines the structures of proteins and macromolecular assemblies at near-atomic resolution, and the final 3D reconstruction depends on extracting a clean particle stack from noisy micrographs. This extraction decomposes into three sub-tasks, namely particle picking, contamination removal, and 2D class selection. Each of them, however, is trained and evaluated in isolation, and none is optimized for the reconstruction. We instead integrate the three sub-tasks into a single pipeline posed against downstream reconstruction quality. We instantiate the pipeline with a state-of-the-art component for each sub-task, CryoTransformer picking permissively, MicrographCleaner masking contamination, and CryoSift selecting 2D classes by a continuous quality score, and close the loop with a fine-tuning step that returns the surviving particles to the picker. The pipeline achieves a better 3D resolution than every picker we compare. We also show that the best 2D F1 is not the best resolution, so particle selection is better treated as one reconstruction-aware pipeline judged by the map it delivers.
[CV-284] Medical Foundation Model Features as Perceptual Loss for Brain MRI Contrast Dose Simulation
链接: https://arxiv.org/abs/2608.28773
作者: Changsheng Fang,Dayang Wang,T. Campbell Arnold,Enhao Gong,Srivathsa Pasumarthi
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Perceptual losses are widely used in medical image synthesis because they encourage agreement in high-level structure beyond voxel-wise intensity similarity. In practice, most perceptual losses are still computed with natural-image backbones such as VGG16 or ResNet50, even when the target domain is magnetic resonance imaging (MRI). This mismatch may weaken supervision for anatomy, contrast enhancement, and acquisition variability. We test whether medical foundation model features provide a more suitable perceptual loss for brain MRI contrast dose simulation. The study has two stages. First, we compare RadImageNet, SegVol, and BrainIAC with ImageNet-pretrained VGG16 and ResNet50 as frozen feature extractors on four public medical imaging benchmarks: thyroid ultrasound, breast ultrasound, anterior cruciate ligament knee MRI, and meniscus knee MRI. RadImageNet achieves the lowest mean rank across the Stage I representation suite and is selected as \phi^\star . Second, we replace only the VGG16 feature extractor in an existing iterative brain MRI dose simulation framework with \phi^\star . The generator, reconstruction loss, adversarial loss, auxiliary losses, optimization schedule, and loss weights are kept unchanged. Standard metrics change modestly, with PSNR increasing from 41.63 to 41.74, SSIM from 0.9739 to 0.9754, RMSE decreasing from 0.1384 to 0.1369, and residual-uptake CNR from 0.0085 to 0.0082. The visual results show the main effect: RadImageNet reduces residual enhancement in marked structures, follows a more faithful dose-reduction trajectory, and remains close to the acquired 10% low-dose target. These results support domain-aligned radiology features as a practical perceptual feature space for MRI dose simulation, while leaving clinical equivalence and larger-cohort validation as future work.
[CV-285] Efficient JPEG Restoration in the Wavelet Domain via Mean Flows
链接: https://arxiv.org/abs/2608.28730
作者: Stefan-Alexandru Asandei,Mihai-Alexandru Radu
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
备注: 9 pages, 2 figures, 6 tables. Code will be released soon
Abstract:Latest JPEG restoration systems achieve strong quality with large models, yet often remain too slow and expensive for efficient on-device deployment. We present a 65M-parameter generative restorer that attains the lowest LPIPS at QF 10 and 20 on LIVE-1, Urban100, and DIV2K-val while sustaining 8.05 images/s at 1024\times1024 on a single RTX 3090, roughly 4.9\times the reported throughput of one-step SODiff at one-twentieth of its parameters. Trained from scratch, the model replaces the learned VAE encoder-decoder with an exactly invertible two-level Haar transform, predicts a clean wavelet-domain residual through a rank-enhanced linear-attention DiT that estimates compression severity internally, and is optimized with an improved MeanFlow objective that enables inference in one or two network evaluations without distillation. Large pretrained priors remain stronger under severe compression (QF 5), whereas our model prioritizes throughput for deployment-constrained restoration.
[CV-286] Evaluating the Effects of Inter-Observer and Model Variability on Radiological Peritoneal Cancer Index Assessment MICCAI2026
链接: https://arxiv.org/abs/2608.28716
作者: Savvas Saragiotis,Pieter C. Gort,Lotte J.S. Fleurkens-Ewals,Anna F. van Herwijnen,Marion Tops-Welten,L.D. Kampmeijer,Joost Nederend,Fons van der Sommen
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted as a CaPTion workshop paper at MICCAI 2026
Abstract:Deep learning segmentation models are often evaluated using geometric metrics such as Dice, HD95, and ASD, yet it remains unclear to what extent improvements in these metrics translate into clinically meaningful changes in downstream decision-making. The metric-to-decision gap is examined using radiological Peritoneal Cancer Index (rPCI) region segmentation on contrast-enhanced CT, where a consensus definition provides anatomically grounded 3D regions and the clinically used PCI 20 threshold enables decision-level evaluation. Inter-observer variability is quantified across four experts on ten abdominal CT scans, and a published nnU-Net based rPCI segmentation model is benchmarked against this human reference using Dice, HD95, and ASD across all 13 regions. To relate geometric differences to clinical impact, a probabilistic peritoneal metastasis simulation is implemented on majority-vote rPCI maps, propagating region-boundary variability into variability of derived ®PCI scores and classification at the PCI 20 cutoff. Observers showed high agreement (mean Dice 0.87 ), while the model matched human performance in most regions but deviated more in regions 4, 8, and the small-bowel regions (9-12). Across simulations, score differences were typically small (mean \Delta rPCI \approx 0.3 - 0.6 ) for both observers and the model, and decision flips occurred predominantly when the reference score was near 20. These results suggest that rPCI-derived scoring is generally robust to typical segmentation variability, while highlighting borderline cases as the main setting where expert review remains essential.
[CV-287] SCoPE-Reg: Efficient Rigid Ultrasound Slice-to-Volume Registration via State-Space Correlation and Closed-Form Pose Estimation
链接: https://arxiv.org/abs/2608.28715
作者: Niklas Schwarz,Jens Kleesiek,Moritz Rempe
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
备注: Currently under review
Abstract:Ultrasound-guided interventions can require localization of an untracked 2D frame within a 3D anatomical reference. Rigid slice-to-volume registration (SVR) estimates this six-degree-of-freedom pose but remains challenging because of limited anatomical context, acoustic artifacts, and view-dependent appearance. Existing methods often use dense cross-attention, whose cost scales with the product of slice and volume token counts, or direct pose regression without explicit correspondence constraints. We introduce SCoPE-Reg, combining state-space slice–volume interaction, dense 3D coordinate prediction, and parameter-free weighted Kabsch estimation. On SVR tasks from CAMUS and \mu -RegPro, SCoPE-Reg yields mean target registration errors of 0.73 mm and 2.27 mm against 1.24 mm and 2.63 mm for the state of the art (SOTA), reduces peak error on CAMUS by 56% below SOTA ( 12.5!\to!5.5 mm), and registers 100% and 80% of frames within 3 mm. On CAMUS at 128^2 it retains the lowest error at increasing pose-perturbation magnitude. It holds 6.49 M parameters independent of resolution, sustaining 51 FPS at 512^2 . SCoPE-Reg establishes a SOTA in rigid ultrasound SVR: by coupling correspondence-based accuracy with bounded worst-case error and resolution-independent cost, it becomes viable at native acquisition resolution during intervention, where prior methods trade accuracy, reliability, or frame rate against one another. Supplementary code provided and will be open-sourced upon acceptance.
[CV-288] Evaluating the Safety of Deep Learning-Based Brain MRI Reconstruction
链接: https://arxiv.org/abs/2608.28714
作者: Dat Tat Mai,Thai Viet Pham,Thu Nguyen Thi Dang,James Jin Kang
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Objective: Deep learning accelerates brain MRI four- to tenfold, but models can erase lesions or synthesize false tissue - failures pixel-averaged metrics like PSNR and SSIM miss. We review whether current evaluation practices detect this blind spot. Methods: Following PRISMA 2020, we searched seven databases without date limits, including 263 studies (1995-2026), appraised them using QUADAS-2 and matched instruments, and synthesized narratively. Categories were derived from titles, abstracts, and controlled vocabulary; reported prevalence figures represent floors. Duplicate screening achieved high agreement (Fleiss kappa = 0.877), as did appraisal (0.788; 0.390 where observable). Extraction is unaudited. Results: Only 18 of 263 studies (6.8%) recorded both a fidelity metric and reader assessment on identical data, leaving the central surrogate unmeasured. Reader studies mostly measured inter-reader agreement, which was weak: fastMRI 2020 concordance reached 0.457 and 0.386 (Kendall W), improving only where SSIM diverged. Erasing a 100 mm3 lacunar infarct shifts global PSNR by 0.03 dB under the stated error model. As the corpus grew fivefold, reader assessments dropped from 32% to 18%, recovering to 21%. Generative models - most associated with hallucination (39%) - were among the least reader-evaluated (11.3%), while self-supervised models reached 47% with zero reader evaluation. Only 5% released code and ran reader studies; none evaluated a model observer; no named dataset covered acute stroke or hemorrhage. Conclusions: On these floors, current evaluation practices cannot certify diagnostic safety. We derive five requirements safety-oriented evaluations must meet.
[CV-289] Coronary Mask Guided Registration for Continuous Time 4D Cardiac CT Dataset Construction
链接: https://arxiv.org/abs/2608.28712
作者: Yuang Wang,Shuo Wang,Changyu Chen,Dufan Wu,Pengfei Jin,Yunqiang An,Yang Gao,Bin Lu,Dongrui Dai,Muge Du,Yan Yan,Dong Li,Liang Li,Li Zhang,Zhiqiang Chen
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
备注: 12 pages, 7 figures. This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible
Abstract:Objective: Clinical cardiac CT multiphase reconstructions generally provide acceptable image quality in end-diastole (ED) or end-systole (ES) phases, but in other phases may exhibit motion artifacts, especially in the right coronary artery (RCA). This limits ground-truth availability in 4D cardiac CT imaging research. We aim to construct a 4D cardiac CT dataset that is generally suitable to serve as pseudo ground truth. Methods: We propose Coronary Mask Guided Registration (CMGR) to produce a motion-preserved, artifact-reduced, and continuous-time 4D cardiac CT sequence from the clinical multiphase reconstruction of each patient. For artifact reduction, CMGR uses the ED or ES phase as the reference phase and warps the reference volume with deformation fields to produce the sequence. For motion preservation, CMGR registers the reference phase to each non-reference phase of the multiphase reconstruction. To capture the motion of both the RCA and other cardiac structures in each registration, CMGR regularizes RCA masks and incorporates them into image-domain registration. Time-continuity is achieved by interpolating the deformation fields for non-reference phases to arbitrary times. Results: CMGR outperformed representative image-domain registration methods in capturing RCA motion and providing reasonable RCA shape, and showed competitive performance in capturing whole-heart motion. Additionally, CMGR reduced motion artifacts from clinical multiphase reconstructions, and intermediate CMGR frames generally provided plausible transitions between discrete cardiac phases. Conclusion: CMGR provides an effective approach for constructing continuous-time 4D cardiac CT datasets. Significance: The dataset can be used in system design simulations and in reconstruction algorithm development, thereby facilitating advances in cardiac CT imaging.
[CV-290] Enhancing MRI Brain Tumor Edge Detection: A Hybrid Preprocessing Approach Utilizing CLAHE
链接: https://arxiv.org/abs/2608.28709
作者: Shahid-E-Kaiser Md. Tashrif,Munshi Md Arafat Hussain,Sheikh Nahian,Sumaiya Islam
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
备注: 14 pages, 3 figures. In review at IntSym 2026, South Eastern University of Sri Lanka
Abstract:Accurate boundary delineation of brain tumors in Magnetic Resonance Imaging (MRI) is a critical yet formidable challenge in neuro-oncology due to inherent scanner noise, complex anatomical structures, and uneven illumination. Traditional edge detection algorithms, while computationally lightweight and mathematically interpretable, frequently fail to capture the diffuse, localized boundaries of edema when relying solely on global preprocessing and manual parameter tuning. To overcome these limitations, we propose a hybrid automated edge detection pipeline. Our approach integrates an optimally configured Contrast-Limited Adaptive Histogram Equalization (CLAHE) layer into a comprehensive morphological preprocessing framework, followed by a deterministic sequential parameter sweep to fully automate threshold selection. The proposed hybrid model demonstrated enhancement in detecting critical anatomical structures in a publicly available benchmark database from Kaggle. By intelligently amplifying localized gradients without overwhelming the image with background noise, our method achieved higher Recall (Sensitivity). Consequently, the overall F1-Score elevated, and the Structural Similarity Index (SSIM) improved, all while maintaining a highly efficient execution. This establishes our optimized pipeline as a highly practical and near real-time operational model for clinical diagnostics, offering a compelling alternative to computationally heavy deep learning approaches.
[CV-291] Is Deformable Image Registration Ready for Brain Metastasis Reirradiation Dose Accumulation? A Longitudinal MRI Benchmark of Registration Accuracy
链接: https://arxiv.org/abs/2608.28705
作者: Hengjie Liu,Manju Sharma,Xinyi Fu,Di Xu,Ke Sheng
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Dose accumulation is increasingly important in adaptive radiation therapy and reirradiation, but its clinical validity depends on the performance of deformable image registration (DIR). Reirradiation of brain metastases (BMs) with stereotactic radiosurgery (SRS) provides a controlled but clinically meaningful DIR test case: intra-subject brain deformation is usually limited after rigid alignment, yet recurrent lesions can undergo substantial local shape and volume changes that rigid registration cannot capture and can affect dose accumulation. We benchmarked a wide range of learning-based and optimization-based DIR methods on 87 manually screened longitudinal contrast-enhanced T1-weighted MRI lesion pairs from an institutional BM SRS retreatment cohort. Learning-based methods pretrained on healthy-brain MRI were evaluated zero-shot and after instance-specific optimization (ISO) or tumor-proximity target-specific optimization (TSO). Registration was assessed using lesion overlap (Dice), surface distance metrics (HD95 and sASD), target-volume recovery, and runtime and memory. Pretrained learning-based methods showed variable zero-shot performance, while ISO/TSO improved all tested learning-based families. However, optimization-based methods remained the best-performing approach while maintaining reasonable runtime. These findings suggest that even state-of-the-art DIR methods do not yet provide sufficiently accurate and consistent registration for unmonitored use in brain metastasis reirradiation dose accumulation. Because accurate registration is a prerequisite for deformable dose accumulation, clinical application will require case-level quality control and direct assessment of how registration uncertainty affects downstream dose metrics.
人工智能
[AI-0] SUN: Persistent Programs For Language-Grounded Control-to-Learning-to-Real Policies
链接: https://arxiv.org/abs/2608.31167
作者: Weiqi Wang,Zhi Li,Yudong Lei,David Martinez,Xiaofeng Gao,Yuxin Jiang,Chenfanfu Jiang,Yingnian Wu,Demetri Terzopoulos,Ran Gong
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:
Abstract:Bridging model-based control and learned policies in long-horizon manipulation has harbored a silent disagreement: control executes specified objectives, learning amortizes that behavior into a reactive policy, yet existing protocols discard task semantics, leaving rewards hand-crafted and behavior drifting from what control this http URL introduce Semantically UNified (SUN) Programs, typed executables where geometric and contact relations are defined once and compiled into aligned Model Predictive Control (MPC) costs, satisfaction predicates, RL rewards, transition guards, and diagnostics. Our system, Kuafu, driven by large vision language systems, automatically synthesizes SUN Programs from language and scene semantics, screens feasibility via MPC, and retains semantics while training stage-conditioned policies. Across nine tasks, Kuafu achieves 82.03% macro-success, outperforming sparse-reward (35.67%) and Stage-BC (24.75%) baselines. At 8192-way scale, it generates 10.57x the successful trajectory time per hour of human teleoperation. With 500 trajectories per task, Kuafu data trains DP3 policies to 46.0% simulation success (vs. 22.4% for alternatives) and 34.7% on physical Franka and Kinova robots. These results establish that simulation-screened task semantics can effectively amortize control into robust policies, without demonstrations or manual dense rewards, unifying symbolic planning and data-driven execution.
[AI-1] Auditing Anonymous AI Models: A Four-Stage Protocol for Black-Box Identity Verification
链接: https://arxiv.org/abs/2608.31142
作者: Yisen Xi
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
备注: 35 pages, 4 figures
Abstract:The 2025–2026 AI market has seen a wave of stealth releases: frontier models launched anonymously on developer platforms under codenames. For their users, identity determines data-handling terms, supply-chain risk, and capability expectations. No validated methodology exists for black-box identity verification of anonymous models: practitioner checklists lack accuracy evidence, and self-identification is untrustworthy by design. We propose a four-stage forensic audit protocol for API-served models. Stage 0 reconstructs launch-time configuration from archived platform snapshots (Internet Archive), exposing preview–production drift. Stage 1 fingerprints configuration (context, output ceiling, reasoning, modality) against the platform catalog. Stage 2 tests tokenizer identity with a cross-length differential that rejects short-prompt collisions. Stage 3 corroborates with behavioral probes. We test declaration consistency on 10 known-identity releases (7 exact, 2 precision-differences, 1 partial, 0 counter-directional), not end-to-end identification under anonymity. Identification is validated prospectively on a flagship case whose 2026-08-23 analysis pointed to the GLM-5.3 version line and whose official reveal confirmed those family and version-line inferences (deployment variant was not pre-asserted; Flash was consistent post-reveal), and on three Stage-0-only cases where the protocol produced a graded hypothesis or declined rather than guessed. A standard-library-only implementation is provided as supplementary material.
[AI-2] OntoAligner-Ensemble: Voting-Based Fusion across Heterogeneous Ontology Alignment Techniques ISWC2026
链接: https://arxiv.org/abs/2608.31137
作者: Hamed Babaei Giglou,Sören Auer,Peio Popov,Mahsa Sanaei,Jennifer D’Souza
类目: Artificial Intelligence (cs.AI)
备注: 14 pages, 1 figure, accepted for OM-2026 workshop at ISWC 2026
Abstract:Ontology alignment (OA) has evolved through several methodological paradigms, ranging from lexical and structural aligners to knowledge graph embedding (KGE) models and, more recently, Large Language Model (LLM)-based approaches. Although modern OA frameworks provide unified ecosystems for deploying these heterogeneous aligners, mechanisms for systematically reconciling their complementary and sometimes conflicting predictions remain relatively underexplored. We present OntoAligner-Ensemble, a modular and aligner-agnostic framework that combines candidate correspondences through a configurable two-stage process comprising voting-based fusion strategies followed by post-fusion selection policies. The framework supports any aligner implemented within OntoAligner that produces candidate correspondences, enabling diverse alignment paradigms to be integrated through a unified decision process. To demonstrate its effectiveness, we instantiate the framework using representative lightweight string-aligner, KGE-based, and Retrieval-Augmented Generation aligners powered by both open-weight and API-based LLMs. We evaluate individual aligners and ensemble configurations across eight benchmark tasks from five OAEI tracks spanning biomedical to beyond-equivalence. The results show that ensemble fusion consistently improves the balance between precision and recall and frequently outperforms standalone aligners across diverse domains. Furthermore, our analysis reveals that ensemble composition directly affects the precision-recall trade-off: heterogeneous cross-paradigm ensembles generally improve precision, whereas homogeneous LLM ensembles more often achieve higher overall F1-scores. These findings demonstrate that systematic ensemble learning offers a robust and reproducible strategy for OA while providing practical guidance for selecting ensemble compositions under different alignment scenarios.
[AI-3] When Does Bigger Help? A Controlled Study of LLM Scale for Ontology Learning ISWC2026
链接: https://arxiv.org/abs/2608.31118
作者: Hamed Babaei Giglou,Sören Auer,Jennifer D’Souza
类目: Artificial Intelligence (cs.AI)
备注: 14 pages, 1 figure, and 5 tables. WOP 2026 workshop at ISWC 2026
Abstract:The effect of Large Language Model (LLM) scale on ontology learning (OL) performance remains insufficiently characterized. We present a controlled evaluation of 13 models spanning dense and Mixture-of-Experts variants from the Qwen3.5 and Qwen3.6 lineages, together with proprietary GPT release variants, using the OntoLearner retrieval-augmented generation pipeline. All models are evaluated with the same embedding model, retrieval configuration, prompt templates, decoding settings, datasets, and metrics on term typing, taxonomy discovery, and non-taxonomic relationship extraction across four biomedical and materials science and engineering ontologies. Within the dense Qwen3.5 lineage, increasing parameter count primarily improves precision rather than recall, with the largest gains occurring between 9B and 27B parameters. However, the effect of scale is neither monotonic nor uniform across tasks and domains. Dense 27B models outperform substantially larger sparse models on term typing, whereas larger Mixture-of-Experts models achieve the strongest open-weight results on taxonomy discovery. Non-taxonomic relationship extraction remains difficult across model scales, particularly for the Materials Data Science ontology. Performance differences across matched Qwen variants and proprietary GPT releases further indicate that architecture and model lineage can outweigh nominal parameter count. These findings show that model size alone is an insufficient selection criterion for OL and provide empirical guidance for reproducible LLM-assisted ontology engineering.
[AI-4] LLM Post-Training as Brownfield Maintenance: An Industrial Perspective on Dataware Engineering
链接: https://arxiv.org/abs/2608.31102
作者: Gopi Krishnan Rajbahadur,Amir M. Ebrahimi,Boyuan Chen,Ahmed E. Hassan
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Industrial post-training is a brownfield regime. Teams inherit a deployed checkpoint and must land targeted improvements under fixed compute and mixture budgets without regressing the rest. The maintained artifact is increasingly dataware: behavior governed by a curated post-training mixture, updated via bounded mixture patches rather than clean-slate retraining. From an industrial code-generation improvement effort, we offer a maintainer’s perspective on why this work is hard in practice, distilling three recurring challenges, zero-sum mixture design, yield as the binding metric, and end-to-end integration under uncertainty, and arguing that progress depends less on one-off recipes than on an engineering discipline for programming dataware. In our case study, interventions that raised the conversion of teacher distillation into usable training data increased accepted supervision by 2.84 times while using the same solution teacher and four solution attempts per candidate problem. In our primary evaluation, the yield-engineered patch improved CodeForces pass@1 by +2.59 points (+3.11 pass@3) and held-out LiveCodeBench v6 pass@1 by +6.11 (+8.05 pass@3), all statistically significant across 16 stochastic evaluations of each benchmark from one fixed checkpoint per condition, with internal AIME and MATH regression suites within tolerance.
[AI-5] Cross-Regional Grapevine Cold Hardiness Prediction via Learned Multimodal Latent Representations
链接: https://arxiv.org/abs/2608.31097
作者: William Solow,Paola Pesantez-Cabrera,Markus Keller,Lav Khot,Sandhya Saisubramanian,Alan Fern
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Accurate daily predictions of cold hardiness in woody plants are critical in regions where freezing temperatures can damage dormant buds and reduce seasonal yield. Existing biophysical, hybrid, and deep learning models have shown high predictive accuracy when trained on local data but remain largely site-specific. The limited availability of cold hardiness data, coupled with the lack of principled methods for transferring cold hardiness predictions to new regions and cultivars, has limited the broader adoption and practical utility of these approaches, particularly in data-scarce regions. To address these limitations, we propose a cold hardiness prediction framework that learns a transferable latent representation by capturing region-specific variation through learned embeddings. To enable prediction in previously unseen regions, we infer embeddings from (1) text descriptions of the cultivar and growing region, and (2) limited historical observations, supporting both zero-shot and few-shot transfer. Experiments on datasets from six regions across North America demonstrate that our approach consistently outperforms state-of-the-art cold hardiness prediction methods, yielding more accurate predictions and substantially improving transfer to data-scarce regions.
[AI-6] Reconciling Process Supervision with Outcome-Based Credit in Agent ic Policy Optimization
链接: https://arxiv.org/abs/2608.31077
作者: Jingxiao Yang,Wangjie Gan,Yingxuan Zhuang,Wenqi Zhang,Jintao Chen,Xuhong Zhang
类目: Artificial Intelligence (cs.AI)
备注: Work in progress
Abstract:Outcome-based reinforcement learning provides verified feedback for language-model agents, but assigns trajectory-level advantage uniformly to all decisions, yielding coarse credit over long-horizon interactions. On-policy self-distillation offers finer supervision by re-evaluating sampled behavior with privileged information (PI) available only during training. However, fine-grained supervision is not necessarily fine-grained credit: PI-induced likelihood changes describe how additional information alters policy preference, but do not directly determine how an executable action should inherit the verified task outcome. This creates a supervision-credit gap. Privileged signals may be irrelevant to the current interaction state, operate at a token granularity misaligned with executable decisions, and lack the outcome semantics required for reinforcement. We introduce TASPO, which converts privileged supervision into outcome-grounded action credit. TASPO constructs decision-applicable PI from verified successful experience, aggregates PI-induced likelihood shifts at the executable-action level, and converts relative action support into positive, bounded, mean-preserving weights on the original trajectory advantage. Thus, the verified outcome determines the update direction and average scale, while PI only redistributes credit across actions. Across three agentic benchmarks, TASPO improves over GRPO by 10.6% and generalizes better to unseen tasks. Further analysis indicates that TASPO reduces supervision mismatch and that action-level assignment stabilizes the policy optimization process. These findings offer the community another interesting perspective.
[AI-7] Scaling Large Reasoning Models beyond Human Supervision: A Path toward Superintelligence
链接: https://arxiv.org/abs/2608.31075
作者: Zhiqin Yang,Jingwen Fu,Yuhan Liu,Hengyu Liu,Yonggang Zhang,Kainan Cao,Zizhuo Zhang,Chenxin Li,Ruibin Yuan,Jiahao Pan,Jiankai Sun,Zhenyuan Zhang,Yibo Li,Yunlong Lin,Jing Xiong,Sida Lin,Bo Han,Wei Xue,Yike Guo
类目: Artificial Intelligence (cs.AI)
备注: 72pages
Abstract:Recent advances in large reasoning models (LRMs) have shown that reinforcement learning with verifiable rewards (RLVR) can substantially improve reasoning in mathematics and code, where outcomes can be checked automatically. Extending this progress to open-ended and agentic tasks remains difficult because reliable rewards are harder to obtain and direct human supervision cannot keep pace with the scale and complexity of model-generated experience. This paper studies how LRMs can continue to improve as human supervision gradually recedes from the learning loop. We examine two connected dimensions of this problem. The reward axis traces the development from per-instance human judgments to reusable verifiers and rewards that operate even without human feedback. The experience axis examines how learning can progress from human-curated tasks and environments toward self-generated curricula, constructed environments, and autonomous co-evolution. We connect these dimensions through a five-level ladder from L0 to L4 that identifies which parts of the learning process remain under continued human control. Our analysis further highlights the risks introduced by increasingly autonomous rewards and experience generation, including reward hacking, feedback drift, curriculum collapse, and environment errors. Consequently, we also provide the evaluation around three complementary objects: policy capability, feedback fidelity, and experience quality. This analysis provides a structured account of current approaches to scaling LRMs beyond human supervision and the open problems involved in developing self-sustaining learning systems toward superintelligence. Furthermore, we maintain a continuously updated \hrefthis https URLGitHub repository to track the latest advances.
[AI-8] Wrong Prediction Right Answer: Recovering Evidence from Collapsed LLM Sequence Scores
链接: https://arxiv.org/abs/2608.31068
作者: Qiyao Yan,Chenpeng Wang,Liangming Pan
类目: Artificial Intelligence (cs.AI)
备注: 20 pages, 4 figures, and 43 appendix tables. Research paper on language-model interpretability, reasoning evaluation, output-scoring bottlenecks, and label-free calibration. The paper evaluates controlled three-way logical reasoning tasks, ProofWriter, ANLI, and FOLIO using Qwen3.5, OLMo-2-1B, Llama-3.1-8B, and Pythia checkpoints
Abstract:When a large language model fails a reasoning task, it is often assumed to lack the underlying capability. However, this conflates a genuine absence of reasoning with a late-stage output bottleneck. We observe a consistent readout gap across diverse reasoning benchmarks: hidden-state probes successfully decode correct answers even when native sequence scoring completely collapses due to structural biases. To test whether instance-specific logic survives this collapse, we introduce a diagnostic protocol using a minimal, target-label-free additive correction. Fitting just two parameters on as few as 25 unlabeled examples recovers 9–34 accuracy points for Qwen3.5 models, transferring successfully to OLMo-2-1B and Llama-3.1-8B. Crucially, these recovered decisions persist on hard instances unresolved by simple lexical overlap and significantly exceed count-preserving permutation baselines. Our results show that many apparent zero-shot reasoning deficits are expression failures masking intact internal logic, urging a narrower interpretation of benchmark evaluations.
[AI-9] Measure Before You Manage: Evaluating Agent Working Memory in Coding Agents
链接: https://arxiv.org/abs/2608.31057
作者: Le Chen,Zishen Wan,Baixi Sun,Xiaolong Ma,Chih-Hsuan Yang,Feng Yan,Sheng Di,Franck Cappello,Rajeev Thakur
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Agent working memory is heterogeneous. Objects such as instructions, artifacts, tool outputs, and agent-generated state play different semantic roles and exhibit different size, retention, and representation profiles. Recent work has begun to explore memory-management mechanisms that account for such heterogeneity. This work focuses on semantic heterogeneity and studies how it should shape the management and evaluation of working memory in coding agents. Across 55 archived coding-agent trajectories, we find that semantically different working-memory objects exhibit distinct retention and compression behavior. This heterogeneity motivates semantically informed memory management. We study two semantically informed strategies: an object-aware compression policy and a retrieval-based policy. Their evaluation shows that calibration gains may not transfer to held-out tasks, and that equal token budgets do not imply equal delivered context or management cost. A real-system replay further exposes serving limits that nominal budgets alone do not capture. Together, these results show why semantic structure matters for agent working memory and why evaluating memory-management strategies requires more than a nominal token budget. We organize these lessons into four levels: stored state, delivered context, management work, and task or process outcome.
[AI-10] CoJEPA: Combining Contrastive Learning and JEPA for Global-Local Music Representations
链接: https://arxiv.org/abs/2608.30974
作者: Gabriel Meseguer-Brocal,Yuexuan Kong,Romain Hennequin
类目: ound (cs.SD); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Audio and Speech Processing (eess.AS); Signal Processing (eess.SP)
备注:
Abstract:Joint-Embedding Predictive Architecture (JEPA) has shown strong performance in learning rich representations through self-supervised prediction in latent space. However, it typically relies on teacher–student architecture with an EMA to stabilise training, and can tend to yield uninformative representations. Contrastive learning is stable to train and produces strong global representations, but remains limited on local tasks by the global nature of its objective. In this work, we combine both into CoJEPA: a single shared backbone jointly trained with a JEPA objective on masked sequence tokens and a contrastive objective on the class token. The contrastive gradient provides stability, removing the need for an EMA teacher entirely, while JEPA enriches the sequence tokens via local predictions that contrastive learning alone cannot provide. Crucially, no extra parameters are added to the backbone: the same model is guided towards richer representations purely through the design of its training signal. CoJEPA takes the best of both worlds, outperforming or matching both individual methods across global and local MIR tasks, with a particularly strong advantage on tonal and harmonic understanding, and without any task-specific architectural changes. CoJEPA shows that combining objectives with complementary inductive biases can substitute for scale, encouraging future work to invest in smarter training objectives over ever-larger models.
[AI-11] A Universal Context-Reuse Layer for Cross-Model KV Sharing
链接: https://arxiv.org/abs/2608.30963
作者: Yi Li,Dongming Jiang,Yi Zhao,Bingzhe Li
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Modern large language model (LLM) serving systems increasingly operate over repeated or shared context, yet each model typically performs its own prefill computation even when another model has already processed the same input. Existing KV-cache reuse mechanisms substantially reduce redundant computation within a single model, but generally assume that the producer and consumer of a cache are identical. We study \emphcross-model KV sharing, which translates the KV state produced by a source model into a representation that can be consumed by a different target model, including models that differ in scale, architecture, attention configuration, tokenizer, and model family. We evaluate the approach in both within-family and cross-family settings. For Qwen2.5-7B \rightarrow Qwen2.5-1.5B, translated KV states improve LongBench2 accuracy from 27.59% to 34.48%, a gain of 6.89 percentage points over the native 1.5B baseline, while reducing handoff cost relative to native target prefill. For the cross-family Qwen2.5-1.5B \rightarrow Gemma-2-2B setting, KV handoff reduces target-side prefill cost by up to 67.05% at 4K context length while maintaining decoding perplexity close to native-model baselines. In a more heterogeneous Llama3.1-70B \rightarrow Qwen2.5-7B setting, cross-family handoff achieves 44.0% accuracy compared with 45.7% for native Qwen2.5-7B inference, while reducing measured latency from 899ms to 138ms. These results provide initial evidence that KV states can serve as transferable computational representations rather than strictly model-local caches, and motivate \emphcontext mobility as a systems abstraction for reducing redundant prefill across heterogeneous LLM and multi-agent inference workflows.
[AI-12] aking the Whys Seriously: Limitations of Counterfactual Explanations in Justification and Recourse
链接: https://arxiv.org/abs/2608.30956
作者: Mattia Cerrato,Otto Sahlgren,Xenia Heilmann
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI)
备注:
Abstract:Counterfactual explanations (CEs) are widely used in explainable artificial intelligence (AI) to show how a model’s outputs would change if the input features were manipulated. This technique is used for a range of tasks such as debugging models, explaining predictions, justifying decisions, and providing algorithmic recourse. In this paper, we explore the normative legitimacy of employing counterfactuals in real-life model deployment settings. We discuss the different stakes involved in these different purposes for which CEs are commonly employed, and find stricter requirements for justification and recourse. In particular, we find that naive application of CEs for justification and recourse can lead to ignoring contestable choices made throughout the machine learning (ML) pipeline, thus obfuscating that decisions and counterfactuals for those decisions are also artifacts of an organization’s materialized design and governance choices. We demonstrate this with four empirical experiments involving interventions at stages of the ML pipeline ``upstream" of the explanation itself, and show that these affect the generated counterfactuals. We find that an organization’s choices on measurement models for feature and labels, business requirements, model validation, and the metric of model success have as much or more impact on the generated counterfactuals as the specifics of the generating method. Our findings underline the need to account for such choices upon providing justification and recourse, providing a stark reminder of the relational nature of these tasks. As putative justifications or recourse recommendations, CEs do not provide adequate answers to some important “why”-questions because they preclude consideration of whether the decision-maker ought to have acted differently.
[AI-13] Learning Action Models with Conditional and Quantified Effects via Uncertainty-Guided Exploration
链接: https://arxiv.org/abs/2608.30955
作者: Jeffrey Jewett,William Solow,Sandhya Saisubramanian
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Accurate action models are critical for effective planning. Existing action-model learning methods largely assume simple action representations or become computationally intractable when learning conditional and quantified effects. We present Online Hypothesis-Driven Conditional Action Model Learning (OHCAM), an online approach for learning action models with conditional and quantified effects from limited interactions with the environment. OHCAM maintains a belief over hypothesized action models and actively selects informative actions to reduce uncertainty by maximizing disagreement among competing hypotheses, while being robust to noisy observations. To enable scalability, OHCAM begins with a small set of simple action model hypotheses and expands to more complex conditions only when the current hypotheses become inconsistent with the data. Experiments on six benchmark planning domains demonstrate that OHCAM is sample efficient in learning action models that solve substantially more tasks than baselines, even with observation noise. We validate OHCAM on two tasks using a Kinova Gen3 robot, demonstrating the real-world applicability of our approach.
[AI-14] MusGU: Toward a Musician-Centered Evaluation Framework and Discovery Tool for Generative Music AI
链接: https://arxiv.org/abs/2608.30940
作者: Laura Ibáñez-Martínez,Roser Batlle-Roca,Xavier Serra,Martín Rocamora
类目: ound (cs.SD); Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Audio and Speech Processing (eess.AS)
备注: Accepted at AIMC 2026
Abstract:Generative music systems are increasingly presented as tools that democratize music creation, yet their practical suitability for musicians remains underexplored. Prior work includes openness-focused evaluation frameworks, such as MusGO (Music-Generative Open AI), as well as qualitative studies of musicians’ experiences with generative systems. However, these approaches do not support systematic comparison or early-stage discovery of models for creative use. Motivated by such limitations, we introduce MusGU+, a musician-centered evaluation framework organized around three dimensions: Adaptability, Usability, and Controllability. Together, these capture whether a model can be feasibly trained or fine-tuned on personal data, integrated into real-world music workflows, and controlled in musically meaningful ways. We evaluate 10 representative generative music systems and present an interactive discovery tool that enables musicians to explore and filter models according to these criteria. While MusGO remains valuable for promoting responsible research practices, MusGU+ supports informed selection and practical adoption of generative systems by musicians.
[AI-15] LightNav-0: Eliciting VLM Spatial Intelligence for Generalist Embodied Navigation
链接: https://arxiv.org/abs/2608.30935
作者: Shaoan Wang,Aocheng Luo,Fei Huang,Jingyi Xu,Xiaoyang Wang,Yueyu Wang,Qianli Ma,Fan Yang,Ran Mei,Jia Wei,Jiangpeng Hu,Xuhao Liu,Hongming Chen,Yuanbin Shao,Yiyang Lin,Ziliang Li,Liang Pan,Xinhang Liu,Yuntao Ma,Tingxiang Fan
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: Technical report
Abstract:Embodied navigation requires agents to translate heterogeneous goals and visual observations into actions across tasks, environments, and robot embodiments. Modern vision-language models (VLMs) already encode spatial priors for visual grounding, spatial reasoning, and pointing, but these capabilities are rarely elicited directly for robot control. Existing navigation systems instead rely on task- or embodiment-specific components, fragmenting perception, reasoning, and action while offering limited generalization. Here we present LightNav-0, a compact generalist embodied navigation model that elicits the spatial intelligence of a pretrained VLM and aligns it with navigation, without task-specific prediction heads. LightNav-0 represents diverse navigation tasks through a unified token interface: dual-channel pointing expresses task-, scene-, and embodiment-agnostic spatial intent, while a residual vector-quantized action tokenizer maps this intent to precise, embodiment-specific trajectories. Together with temporally aware visual history compression, ER mid-training, supervised fine-tuning, and reinforcement learning, this formulation supports instruction following, open-vocabulary object navigation, and visual tracking within a single model. The navigation training corpus spans 2K+ scenes and 4K+ hours of embodied navigation data. LightNav-ER, the embodied-reasoning checkpoint used to initialize LightNav-0, attains the highest complete-set average across 8 embodied-reasoning benchmarks, while LightNav-0 achieves state-of-the-art monocular success rates across all 10 public navigation simulation settings. Real-world evaluations further demonstrate zero-shot generalization across robot embodiments, diverse scenes, and static and dynamic targets. These results establish compact VLMs as a unified and transferable backbone for generalist embodied navigation.
[AI-16] Stride-k Subsampling: Train-Free Audio Token Reduction for Whisper EMNLP2026
链接: https://arxiv.org/abs/2608.30927
作者: Chanhee Cho,Junhyuk Choi,Bugeun Kim
类目: ound (cs.SD); Artificial Intelligence (cs.AI)
备注: Accepted EMNLP 2026 Main
Abstract:Whisper exposes speech through a fixed 1500-token encoder interface, now a default representation for ASR decoders and Whisper-based speech language models (SpeechLMs), yet its redundancy remains largely unexamined. We propose stride-k subsampling, a deterministic indexing operation that retains every k-th token after the convolutional stem or encoder transformer. Across five Whisper scales, k=2 preserves baseline WER at both positions, with CKA attributing this stability to acoustic overlap at the stem and attention-induced redistribution at the encoder output. Applying stride-2 at both positions cuts audio tokens by 75% and total GFLOPs by 52-58%, with small WER costs on most ASR benchmarks and larger costs on harder ones. The same configuration extends to three Whisper-based SpeechLMs, yielding modest accuracy drops on stronger baselines and larger drops on weaker ones, while reducing end-to-end latency by 19.6-27.4%. Requiring no training or auxiliary computation, stride-k subsampling exploits Whisper’s preprocessing redundancy, indicating that its audio-token interface carries more capacity than downstream tasks require.
[AI-17] owards Stream Learning on Embedded Systems: Benchmarking the Memory Consumption of Stream Learning Methods
链接: https://arxiv.org/abs/2608.30923
作者: Sebastian Buschjäger,Nuwan Gunasekara,Heitor Murilo Gomes
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Performance (cs.PF)
备注: 7 pages double-column + appendix
Abstract:Stream learning is commonly evaluated through predictive performance and adaptation to concept drift. However, sustained operation of a stream learner also requires predictable and bounded resource usage even on long streams. This requirement becomes even more critical when learning moves from servers to near-sensor embedded systems where memory and processing are scarce resources. In state-of-the-art stream learning, however, we perceive a strong focus on concept drift adaptation, whereas resource usage is often an evaluation byproduct. To close this gap, we benchmark seven representative stream classifiers on 13 real and synthetic streams under model-size budgets from 128,KiB to approximately 8,MiB. Our benchmark comprises a total of 6,463 experiments. We measure failure-aware accuracy, peak model size, time to budget exhaustion, and prediction-plus-update latency. The results reveal two distinct resource failure modes. Adaptive ensembles can exceed small budgets almost immediately because of their initial footprint, even when their size remains stable thereafter. Incremental trees can fit initially but grow throughout a long stream, with HoeffdingTrees (HT) and Extremely Fast Decision Trees (EFDT) increasing by median factors of 7.37 and 5.87. Explicitly compact methods remain the only viable option under the smallest budgets, but are usually overtaken as larger budgets make adaptive ensembles competitive. Hence, many state-of-the-art methods are only partially applicable in embedded systems or for long-running systems. We therefore call on the stream-learning community to make bounded resource usage a first-class design objective alongside drift adaptation, and propose concrete steps toward this goal, including an API through which stream learners can explicitly expose and respect resource budgets.
[AI-18] CARVE: Verified Expansion for Variable-Length Generation in Diffusion Language Models EMNLP2026
链接: https://arxiv.org/abs/2608.30922
作者: Wail Bouhedja,Amr Mohamed,Guokan Shang
类目: Artificial Intelligence (cs.AI)
备注: EMNLP 2026 Findings
Abstract:Masked diffusion language models predict tokens from a partially observed response canvas, enabling bidirectional conditioning and parallel token refinement. Yet standard masked-diffusion decoders use a rigid inference interface: the number of masked positions allocated to the answer is fixed before generation begins. Choosing this length is difficult. A short canvas can truncate reasoning or code, while a long canvas wastes computation and can perturb denoising. We introduce CARVE (Counterfactual-Aware Reveal with Verified Expansion), a training-free variable-length algorithm for masked diffusion LMs. Starting from a shorter canvas, CARVE can grow the response during decoding by inserting additional [MASK] positions. Rather than keeping every insertion, CARVE tests a candidate expanded canvas and asks a counterfactual question: would the model make similar predictions for the unresolved positions in the original canvas if the extra masked space were present? The inserted masks are kept only when they induce low Jensen-Shannon (JS) divergence on aligned unresolved positions. This makes length growth a verified stability decision rather than a pure confidence heuristic. CARVE applies without retraining to both full-canvas and blockwise diffusion decoders. Across code generation and mathematical reasoning benchmarks, CARVE consistently improves average performance over fixed-length baselines across all evaluated model families. Crucially, CARVE achieves these accuracy gains while reducing inference cost, reaching half the FLOPs of fixed-length decoding in some settings.
[AI-19] CAER: Causal Action Effect Reweighting for World Model Training
链接: https://arxiv.org/abs/2608.30897
作者: Jianjie Fang,Xvyuan Liu,Ziyou Wang,Rongze Tang,Zhaolu Wang,Zhuohang Li,Xin Zhang,Haisheng Su,Chen Gao,Wei Wu,Xinlei Chen,Yong Li
类目: Artificial Intelligence (cs.AI)
备注: 14 pages, 8 figures. Project page: this https URL
Abstract:World models are becoming core infrastructure for embodied intelligence, with action-conditioned video generation providing controllable predictions of how scenes evolve after agent interventions. Yet existing models are commonly trained with space-time-uniform mean squared error, allowing abundant background tokens to dominate the gradient while sparse interaction dynamics remain under-optimized; such uniform fitting rewards reconstructing appearance rather than learning how actions change the world. We introduce Causal Action Effect Reweighting (CAER), a general training paradigm that redistributes supervision toward the tokens whose predicted future is causally affected by the action. CAER contrasts the model’s own predictions with and without action conditioning to localize these tokens online, then normalizes the resulting effect map into a weight that preserves the total coefficient mass and changes only where it is spent. This online signal requires no external annotations or offline preprocessing, avoids additional data-processing time, and scales naturally with model and dataset size. Experiments across heterogeneous action-conditioned world-model tasks show that CAER converges to better solutions than uniform MSE training, with consistent improvements in the physical consistency, controllability, and visual quality of generated videos.
[AI-20] Safety Screening for Voltage Control in Active Distribution Grids via Distributionally Robust Conformal Screening
链接: https://arxiv.org/abs/2608.30889
作者: Sarra Bouchkati,Petros Ellinas,Adriana Geisler,Steffen Kortmann,Johanna Vorwerk,Spyros Chatzivasiliadis,Andreas Ulbig
类目: ystems and Control (eess.SY); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Sarra Bouchkati, Petros Ellinas, and Adriana Geisler contributed equally to this work
Abstract:Deploying a new control policy for voltage control in active distribution grids requires evidence that physical limits will be satisfied before the policy is tested on the physical grid. This assessment is difficult for two reasons. First, simulations cannot capture every disturbance, modeling error, and device interaction present in the real grid. Second, historical measurements reflect operation under existing control policies, whereas a new policy may drive the grid into different operating conditions. To address these challenges, we propose Distributionally Robust Conformal Safety Screening (DR-CSS), a policy-agnostic framework for pre-deployment, scenario-by-scenario screening of a new control policy using historical data and a nominal simulator. For each new scenario, the simulator predicts a future voltage trajectory for the whole grid; DR-CSS then constructs a conformal safety interval around this prediction using historical simulation-to-reality errors. The interval is further enlarged to account for closed-loop changes induced by the deployment of the new policy and its interactions with the remaining controllers. To the best of our knowledge, DR-CSS is the first framework in power systems to combine historical data from an existing control policy with an imperfect simulator for pre-deployment safety screening of a new policy. Experiments on the IEEE 33-bus and IEEE 141-bus systems evaluate the deployment of learning-based voltage control policies and show that DR-CSS identifies all unsafe test scenarios. To reduce unnecessary warnings on safe scenarios, we adapt the safety intervals to different operating conditions and gradually introduce new policies with recalibration after each stage. These extensions increase the informational value of the safety screening and support safer deployment decisions in active distribution grids.
[AI-21] Predicting Residential Rents in Dakar Using Machine Learning
链接: https://arxiv.org/abs/2608.30865
作者: Amadou Tidiane Kassa Diallo
类目: Artificial Intelligence (cs.AI)
备注: 19 pages, 16 figures, 4 tables
Abstract:Dakar’s residential rental market remains poorly documented despite its economic and social importance: 54.4% of households are renters, compared to 23.3% nationally. This study develops a complete machine learning pipeline to predict residential rents in Dakar, from data collection to model interpretation. An original dataset of 1,507 rental listings was built through systematic web scraping and a documented cleaning pipeline, then enriched with four purpose-built features, including a luxury score and a keyword-based quality score. Five models were compared: linear regression, Random Forest (baseline), XGBoost, and LightGBM optimized through Bayesian optimization with Optuna, using leakage-free KFold target encoding for location. The optimized XGBoost model achieved the best performance with an R^2 of 0.847, an MAE of 210,902 XOF, and an RMSE of 324,195 XOF. Feature importance was assessed using native XGBoost gain and SHAP values, revealing a substantial difference in the ranking of location, which appears as a minor predictor by gain but as the second most influential variable by SHAP. This result carries methodological implications for hedonic studies using target-encoded categorical variables. This study provides an interpretable benchmark for Dakar’s rental market and highlights several avenues for improvement, including the integration of geospatial features and conformal prediction.
[AI-22] VFR-Audit: Verdict-Level Reliability for Fairness Audits in Hospital Length-of-Stay Prediction CIKM2026
链接: https://arxiv.org/abs/2608.30846
作者: Md Jannatul Rakib Joy,Viet Vo,Caslon Chua
类目: Artificial Intelligence (cs.AI)
备注: 12 pages, 5 figures. Accepted at the 35th ACM International Conference on Information and Knowledge Management (CIKM 2026)
Abstract:Fairness audits in clinical Artificial Intelligence convert continuous fairness metrics into binary pass-or-fail verdicts against operational thresholds, where hospital governance boards, payers, and regulators act on the resulting verdicts. Such audits are repeated over time and across hospital sites, thus the same verdict can flip between pass and fail across audits. Existing uncertainty methods such as Bayesian posteriors, bootstrap confidence intervals, and permutation tests address verdict instability only at the continuous-metric level. Converting metric-level uncertainty into a verdict-stability claim remains a manual step that scales poorly across the (model, metric, attribute) cells an audit covers. Existing uncertainty methods also leave open whether bias-mitigation steps, such as reweighing or per-group threshold shifts, yield a stable passing verdict at the cost of model discrimination measured as AUROC or this http URL address this verdict-stability gap, we propose VFR-Audit, a framework built around the Verdict Flip Rate (VFR), a scalar bounded between 0 and 0.5 that measures the probability of verdict reversal under stratified bootstrap resampling. VFR-Audit reports VFR alongside three reliability axes, namely within-cohort resampling stability, audit-size sensitivity, and cross-hospital verdict agreement via Fleiss’ kappa.
[AI-23] Conjoint Audio-to-Spikes Encoding and Processing for Efficient Neuromorphic Speech Recognition
链接: https://arxiv.org/abs/2608.30792
作者: Valentin M. Meunier,Amélie Gruel,Pierre Lewden,Adrien F. Vincent,Sylvain Saïghi
类目: Neural and Evolutionary Computing (cs.NE); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Audio and Speech Processing (eess.AS)
备注: Under review. 15 pages, 9 figures and 4 tables. This work is supported by a public grant overseen by the French ANR as part of the “Chaires IA” programme (GrAI project ANR 19 CHIA 0003) and as part of the “PEPR IA France 2030” programme (Emergences project ANR 23 PEIA 0002). This research is part of the programme DesCartes and is supported by the NRF Singapore under its CREATE programme
Abstract:Obtaining data from neuromorphic sensors and processing it with Spiking Neural Networks is a promising solution to lower the energy cost of artificial intelligence. The current rarity of natively neuromorphic datasets promotes the development of software tools to translate input sensory data into spikes. However, highly bio-mimetic simulators can be challenging to implement on digital hardware. In this work, we evaluate the neuromorphic encoding and subsequent classification of audio into spikes using a non-learnable, high-level, programmable encoder targeting hardware implementation on FPGA. We quantify the pipeline’s efficiency with hardware-agnostic metrics based on the quantitative spiking activity. Our study focuses on the simultaneous optimisation of encoder and classifier: the first provides efficient and informative data so that the latter achieves a better performance with an overall lower energy cost at learning and inference. This work introduces the first end-to-end neuromorphic spike-encoding and evaluation of the TIMIT dataset. Our simple feedforward network reaches a classification accuracy of 99.77% on a spike-encoded Heidelberg Digits, overcoming the neuromorphic state of the art on this benchmark dataset.
[AI-24] SkillZip Pro: Execution-Aware Dynamic Compression of Progressively Loaded Skills for Self-Evolving Agents
链接: https://arxiv.org/abs/2608.30785
作者: Xiaofan Bai,Chao Liu,Hongqiang Lin,Di Wu,Mingli Song,Xuan Jin,Xipeng Cao,Yuhong Li
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Production agent skills are directory bundles, not isolated prompts. The root is loaded at activation; references, schemas, scripts, assets, and nested subskills are loaded only when an execution path needs them. Compressing only the root misses most deployment cost and may move branch-specific details into the always-loaded context. Flattening instead destroys progressive-loading boundaries. We introduce \method, an evaluation-free compressor for complete, progressively loaded skill bundles. It leaves the agent harness unchanged and emits an ordinary directory. The method combines two safeguards. First, it compresses \emphacross files, removing content from a reference or subskill when the root or a declared environment contract already provides it. Second, it preserves routing, so every required file and directly callable entry remains reachable after rewriting. Users can configure \method along two independent axes. \emphOne-Shot mode rebuilds the full bundle; \emphContinual mode reuses state and applies Zip-on-Write after each evolution patch. \emphPersistent compression rewrites the shipped bundle to reduce storage and runtime context. \emphTransient compression keeps that bundle byte-identical and builds a task-specific view, reducing only per-run context after build cost. Entry contracts mark private, public, and conditional resources; a multi-entry audit preserves standalone public subskills. On a production content-moderation skill evaluated by our industrial multi-round harness, \method removes \hl38% of skill bundle tokens and \hl10.4% of end-to-end per-run tokens with no quality loss, while an unprotected 71% configuration loses up to 26 accuracy points to one-sided false positives. On a multi-entry bundle, \method effeciently reduces token cost while near-perfectly preserving every route and public entry. Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2608.30785 [cs.AI] (or arXiv:2608.30785v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.30785 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-25] Which Rules Matter Now? Policy-Centroid Routing Before an Intelligent System Acts
链接: https://arxiv.org/abs/2608.30757
作者: Thomson D. Nguy(Radiant Institute for Manifold Studies)
类目: Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注: Theory and research-program paper; 9 pages, 1 figure, 2 tables. Includes a synthetic worked example; reports no empirical efficacy result
Abstract:Before an intelligent system can decide whether an action is allowed, it must first know which rules the action has approached. A single proposed action can implicate several policy regimes at once. Their requirements may stack, overlap, or qualify one another, yet many remain written in natural language while the action itself arrives as an incomplete description of intent. The first problem is not judgment. It is attention. Policy-centroid routing creates a layer before adjudication. It compresses expressions within each policy regime into one or more representative centroids, places the proposed action in the same semantic space, applies a declared measure, and routes every regime crossing a declared threshold to authoritative review. Several regimes may trigger at once. The output is a review agenda, not permission, prohibition, legality, breach, compliance, certification, or enforcement. The paper develops six falsifiable propositions and seven follow-on studies comparing the hypothesis with structured workflows, lexical and semantic retrieval, hierarchical and direct classification, and selective prediction under matched review burden. The studies are designed to identify where policy geometry recovers applicable regimes, where compression loses rare or overlapping obligations, and where the mechanism should abstain. The paper includes a synthetic worked example and reports no empirical efficacy result. Comments: Theory and research-program paper; 9 pages, 1 figure, 2 tables. Includes a synthetic worked example; reports no empirical efficacy result Subjects: Artificial Intelligence (cs.AI); Computers and Society (cs.CY) Cite as: arXiv:2608.30757 [cs.AI] (or arXiv:2608.30757v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.30757 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-26] On the Prospects of Dynamic LLM Conversations in Software Development
链接: https://arxiv.org/abs/2608.30756
作者: Annemarie Wittig,Alina Mailach,Janet Siegmund,Norbert Siegmund
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models (LLMs) have become an essential tool for assisting developers, yet we still lack knowledge on ways to effectively support their interactions during development activities. That is, the quality of interactions with a chat-based LLM still strongly depends on how developers phrase prompts and which information they include. Our goal is to evaluate whether interventions into these interactions with LLMs have an effect on software developers—be it harmful or beneficial. To this end, we conducted a four-month longitudinal study with third-semester computer science students working on a full-stack Web development project using chat-based LLMs under three conditions: (1) a \emphcontext-aware group received intent-based conversation augmentation, (2) a \emphproactive group received follow-up suggestions and tailored advice, and (3) a \emphcontrol group without intervention. Our augmentations are minimal: (i) to reduce confounding factors and (ii) to isolate treatment effects. Analyzing interaction logs and user surveys revealed no major differences in interaction patterns, indicating no detectable harmful effects in the measured outcomes when intervening in interactions. Moreover, we observed trends of increased satisfaction with the \emphproactive treatment. The results indicate that even with minimal interventions, dynamic guidance mechanisms for developer-LLM interactions show observable effects, such that more severe augmentations may have the potential to substantially improve developer satisfaction. Subjects: Software Engineering (cs.SE); Artificial Intelligence (cs.AI) Cite as: arXiv:2608.30756 [cs.SE] (or arXiv:2608.30756v1 [cs.SE] for this version) https://doi.org/10.48550/arXiv.2608.30756 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-27] Multimodal Adaptive Expert Selection with Text Routing and Ordinal Prototype Optimization for Sentiment Analysis
链接: https://arxiv.org/abs/2608.30726
作者: Xiaode Chen,Jiakang Yu,Hongtao Deng,Huina Qu,Xun Zhu,Yinxia Lou
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Multimodal Sentiment Analysis (MSA) is a fundamental component of affective computing that aims to decipher complex emotional states by integrating verbal content with non-verbal cues including vocal intonation and facial micro-expressions. While recent disentanglement-based approaches have advanced the field, their potential is hindered by two methodological challenges. First, static computation graphs process all samples indiscriminately regardless of semantic complexity, which leads to suboptimal representation for diverse emotional expressions and contextual scenarios. Second, generic contrastive objectives often neglect the intrinsic ordinal hierarchy of sentiment intensities. To systematically address these limitations, we introduce Multimodal Adaptive Expert Selection with Text Routing and Ordinal prototype optimization (MAESTRO), a novel framework designed to dynamically orchestrate and refine multimodal representations. Drawing inspiration from an orchestra conductor, we design a Text-Guided Hybrid Mixture-of-Experts (MoE) mechanism. Unlike static fusion, this module utilizes linguistic context as a routing signal to dynamically activate specific audio-visual experts, thereby resolving cross-modal ambiguity through adaptive feature enhancement. Furthermore, to capture fine-grained sentiment gradations, we propose an Ordinal-aware Prototype Contrastive Learning (O-PCL). By incorporating distance-based penalties into the prototype learning objective, O-PCL enforces a structured latent space that preserves the natural order of emotion. Extensive experiments on the CMU-MOSI and CMU-MOSEI benchmarks demonstrate that MAESTRO achieves state-of-the-art performance, and qualitative analysis further confirms the interpretability of our dynamic routing paradigm.
[AI-28] BAITBENCH: Measuring Agent Reward Hacking with Optional Shortcuts Planted in ML Tasks
链接: https://arxiv.org/abs/2608.30724
作者: Pradyumna Shyama Prasad,Meiri Anto,Leon Eshuijs,Julian Moncarz,Kaustubh Kislay,Juan J. Vazquez
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:LLM agents are increasingly used to run autonomous ML experiments, iterating on target metrics with little human oversight. Prior work has documented reward hacking in these environments, bringing into question the validity of produced research and the broader safety case for AI RD. Existing benchmarks do not measure exploits that live in the data or the modeling task itself. We introduce BAITBENCH, a suite of three synthetic tabular ML tasks that each contain a shortcut that allows agents to inflate the public test score but fail on a hidden test set. Since the shortcut is optional and using it breaks no stated rule, BAITBENCH measures how often models exploit the shortcut to achieve inflated scores. Across seven frontier agents scored by our two-stage judge pipeline, 57.1% of runs exhibit reward hacking, with five of seven above 50%. Agents cheat even under a second condition where they are prompted not to -the mean cheating rate remains above 50%. We release BAITBENCH, along with the judge implementation, and an annotated dataset of transcripts containing reward hacks as a testbed for evaluating reward-hacking mitigations head-to-head.
[AI-29] Learning Dynamics of Logits Debiasing for Long-Tailed Semi-Supervised Learning
链接: https://arxiv.org/abs/2608.30699
作者: Yue Cheng,Jiajun Zhang,Xiaohui Gao,Weiwei Xing,Zhanxing Zhu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 32 pages, 19 figures
Abstract:Long-tailed distributions are prevalent in real-world semi-supervised learning (SSL), where pseudo-labels tend to favor majority classes, leading to degraded generalization. While many long-tailed semi-supervised learning (LTSSL) methods have been proposed, the mechanisms by which they implicitly debias logits remain poorly understood. In this work, we revisit LTSSL through the lens of learning dynamics and provide a theoretical characterization of logits debiasing. Specifically, we derive a step-wise decomposition of the logits updates, showing that predictions are dominated by class-imbalance bias that reliably reflects label priors. To expose this effect, we use the logits of a task-irrelevant baseline image as an indicator of accumulated bias and prove that they converge to the class prior. This provides a unified view where LTSSL remedies such as logit adjustment, reweighting, and resampling correspond to reshaping gradient dynamics. Based on this insight, we propose DyTrim, a principle-based dynamic pruning framework that reallocates gradient budget through class-aware pruning on labeled data and confidence-based soft pruning on unlabeled data. We provide theoretical guarantees that DyTrim reduces class bias and improves generalization. Extensive experiments on standard LTSSL benchmarks show consistent gains across architectures and methods. Code available at: this https URL
[AI-30] ATLAS: Dual-Horizon Diagnostic Evaluation for Industrial Tool-Use Agents
链接: https://arxiv.org/abs/2608.30685
作者: Wei Chen,Peilun Zhou,Zhaoyu Hu,Jiajun Chai,Zhongni Hou,Yufei Zhang,Derong Xu,Guojun Yin,Wei Lin,Zhi Zheng,Tong Xu
类目: Artificial Intelligence (cs.AI)
备注: 25 pages
Abstract:Large language model (LLM) agents are increasingly deployed in user-facing services that require iterative tool use under dynamic business conditions. Reliable evaluation is essential for sustained improvement: it must reveal capability deficiencies, inform priorities, and assess interventions. Yet industrial agent service unfolds both through the iterative trajectory of a current request and through continued user interaction. Final-outcome assessment can therefore obscure where deficiencies arise and whether later service remains aligned with context from earlier exchanges. We propose ATLAS, a dual-horizon diagnostic evaluation framework for industrial tool-use agents. At the request horizon, trajectory-wise diagnostic signals relate deficiencies to execution locations and capability concerns. At the interaction horizon, user-wise signals assess whether service remains responsive across continued interaction. Together, these views provide structured diagnostic evidence for analyzing execution deficiencies and sustained service behavior. ATLAS instantiates them as executable signals with explicit evidence scopes and decision boundaries. LLM judge interfaces are calibrated against high-confidence references from real business logs; when needed, their decision behavior is distilled into efficient diagnostic models for lower-latency, lower-cost evaluation. The resulting feedback supports policy optimization. We evaluate ATLAS on Meituan Xiaotuan production traffic. Offline experiments assess diagnostic-signal fidelity and replay-based policy improvement, while online A/B experiments show concurrent gains in user engagement, downstream business outcomes, and sampled human-audit quality.
[AI-31] Learning Materials Properties from Scarce Labels and Unlabeled Crystals
链接: https://arxiv.org/abs/2608.30682
作者: Wentao Li,Yizhe Chen,Jiangjie Qiu,Yijun Li,Leyi Zhao,Xiaonan Wang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Learning materials properties from scarce labels and unlabeled crystals is a central challenge for data-driven materials discovery. We present SemiMat, a controlled benchmark for semi-supervised materials property regression, and MatRank, a reliability-weighted objective for continuous pseudo-label uncertainty. SemiMat fixes labeled and unlabeled crystal inputs, graph-backbone interfaces, validation-only checkpoint selection, held-out test reporting, normalized MAE (NMAE), and method-rank summaries across six scarce-label tasks, four graph backbones, and five predefined split runs. MatRank builds pseudo-targets from labeled anchors, weights them by local reliability and weak-prediction agreement, trains weak and strong graph views consistently, and adds ranking signals so that unlabeled crystals shape both values and candidate order. Across the retained 24 backbone-task blocks, one fixed MatRank objective gives the lowest aggregate held-out test NMAE (0.896) and best average method rank (2.208). The component, OOD, and generated-pool diagnostics identify where the gain is reliable and where further screening evaluation remains necessary. Code is available at this https URL.
[AI-32] MedAgent -R1: Faithfulness-Aware Reinforcement Learning for Evidence-Grounded Medical Reasoning
链接: https://arxiv.org/abs/2608.30676
作者: Jiangwang Chen,Chenghao Zhang,Hengxing Cai
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:When medical AI systems hallucinate clinical reasoning, the consequences extend beyond incorrect answers: fabricated justifications that superficially reference retrieved evidence can mislead clinicians into unsafe treatment decisions. Medical reasoning agents must therefore produce not only correct answers but also faithful justifications that clinicians can verify against cited evidence. We identify a systematic failure mode in RL-trained retrieval agents: outcome-only rewards improve accuracy while degrading faithfulness, a phenomenon we term confident hallucination. The agent learns to answer from parametric memory and backfill plausible but unsupported justifications; citation fabrication rates rise from 16.5% to 31.8% even as accuracy improves by 5 points over the supervised baseline. We address this with a faithfulness-gated reward design: accuracy credit is conditioned on evidence grounding via a hard gate, complemented by retrieval validity and conciseness signals that close exploitation paths unique to agentic retrieval. The resulting system, MedAgent-R1, reduces citation fabrication from 31.8% to 4.7% and raises evidence completeness from 58.7 to 82.6 while maintaining 75.1% accuracy, with 13.2-point gains on HealthBench Safety. Under the same agentic retrieval setup, MedAgent-R1 outscores GPT-4o on faithfulness-specific dimensions (Factual Support 4.55 vs. 4.25; Overclaiming 4.40 vs. 4.15) while remaining below GPT-4o in overall accuracy, suggesting that explicit faithfulness training yields evidence-grounding gains not achieved by scaling alone.
[AI-33] CoMPASS: Collaborative Molecular Property Prediction via Adaptive Small-Large Model Synergy
链接: https://arxiv.org/abs/2608.30674
作者: Wentao Li,Jiangjie Qiu,Yijun Li,Leyi Zhao,Xiaonan Wang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 9 pages
Abstract:Accurate molecular property prediction requires both statistical reliability and chemical reasoning. Graph neural networks can be calibrated directly on labeled assays but remain limited by the coverage of their training data. Large language models (LLMs) can compare molecular evidence and articulate chemical rationales, yet are unreliable as standalone quantitative predictors. The central challenge is therefore to determine when an LLM should influence a calibrated model and by how much. Here we present CoMPASS, a retrieval-calibrated framework for small-large model collaboration. CoMPASS retains a graph attention network (GAT) as the predictive anchor, retrieves locally relevant training molecules, provides attention-grounded evidence to an LLM, and converts its proposal into a bounded correction through an agreement-aware gate. Across six classification and two regression benchmarks, CoMPASS improves the GAT anchor in regions of correctable uncertainty while limiting LLM intervention in high-confidence regimes. Ablations show that the gains arise from validation-calibrated retrieval and bounded fusion rather than prompting alone. These results suggest that generative reasoning should augment calibrated prediction through evidence-grounded, controlled corrections rather than direct output replacement. Code is available at this https URL.
[AI-34] PyKEEN-NSX: A Modular Framework for Static Dynamic and Schema-Aware Negative Sampling in PyKEEN
链接: https://arxiv.org/abs/2608.30652
作者: Ivan Diliso,Nicola Fanizzi,Claudia d’Amato
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Embedding methods have become popular due to their scalability on link prediction and/or triple classification tasks on Knowledge Graphs (KGs). Embedding models are trained relying on both positive and negative samples of triples. However, since KGs generally contain only positive assertions, negative samples are artificially generated through negative sampling strategies, ranging from simple random corruption to more sophisticated approaches that exploit structural, semantic, or embedding information. The design and implementation of advanced negative samplers remains challenging, as most popular Knowledge Graph Embedding (KGE) libraries provide support only for basic strategies and lack a unified framework for developing more advanced and customized solutions. To address this gap, we introduce PyKEEN-NSX, an extension of PyKEEN, the popular KGE framework, that provides a modular engineered abstraction for negative sampling. The proposed architecture separates the generation of candidate negative pools, conditioned on an explicit context, from the selection strategy, enabling the development and integration of static, schema-aware and dynamic approaches within a consistent framework. Based on this abstraction, we implement six negative samplers, while remaining fully compatible with existing PyKEEN workflows and pipelines. As a proof of concept, we study negative availability across four datasets, showing that constrained pools frequently fall below the requested number of negatives, so that the encoded criterion is to a large extent replaced by the random fallback that supplements them.
[AI-35] Automated Testing of LLM -Based Post Hoc Explainers Using Model Checking as an Oracle
链接: https://arxiv.org/abs/2608.30581
作者: Dennis Gross,Helge Spieker
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Large language models (LLMs) are used as post hoc explainers of sequential decision-making policies, producing natural-language explanations of why an action was chosen. However, LLMs often generate plausible but incorrect statements, and no existing approach systematically tests whether such explanations are faithful to the underlying environment. Two classic software testing challenges stand in the way: there is no oracle for the correctness of an explanation, and the test inputs, natural language queries about a policy’s behavior, lack the structure needed for systematic test case generation. We address both. Probabilistic model checking provides the test oracle, computing exact reference results against which LLM answers are graded automatically. A taxonomy of post hoc query categories structures the input space around the environment-level facts from which policy explanations are composed; test cases generated from it are prioritized by question-specific diagnostic difficulty scores. Across seven MDP environments, the testing separates three open-weight LLMs: a reasoning model passes 85% of test cases, a mid-size model 70%, and a 1B model falls below the random baseline, while prioritization surfaces significantly harder cases than random selection. Our results indicate how trustworthy LLM-generated explanations are in model-free settings, where the same LLMs are used but no oracle exists to verify them.
[AI-36] Collapsibility of Performance Metrics in Clinical Predictive AI
链接: https://arxiv.org/abs/2608.30568
作者: João Matos,Ben Van Calster,Richard D. Riley,Paula Dhiman,Gary S. Collins
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Background: Population level assessments of predictive artificial intelligence (AI) can conceal performance disparities across subgroups. Fairness evaluations commonly rely on performance analyses across subgroups. However, some performance metrics are non-collapsible, meaning that the overall population performance value does not equal the weighted average of subgroup specific values. Objective: To examine the collapsibility properties of commonly reported performance metrics in predictive AI, with a focus on the area under the receiver operating characteristic curve (AUC, also known as c-statistic). Methods: We investigate the collapsibility of 15 performance metrics, either by expressing each metric as a linear combination of its stratum specific values or, where non-collapsible, by providing a counterexample inspired by Simpson’s paradox as a formal disproof. Results: Five performance metrics (AUC, calibration intercept, calibration slope, expected calibration error, and Nagelkerke R^2) are shown to be non-collapsible, and ten (O:E ratio, logloss, Brier score, accuracy, F1-score, true positive rate, true negative rate, positive predictive value, negative predictive value, and net benefit) are shown to be collapsible. The AUC is shown to be non-collapsible because it decomposes into within- and cross-group AUC terms when subpopulations coexist, such that its overall value may fall outside the range of subgroup specific AUCs. Conclusions: Non-collapsibility of performance metrics has important consequences for reporting, model appraisal, and fairness evaluation. It can generate spurious differences between subgroup and overall performance, which may mislead fairness evaluations. Explicitly acknowledging and reporting the collapsibility properties of performance metrics improves both the interpretability and transparency of fairness assessments. Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI) Cite as: arXiv:2608.30568 [cs.LG] (or arXiv:2608.30568v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2608.30568 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: João Matos [view email] [v1] Mon, 31 Aug 2026 10:43:22 UTC (717 KB) Full-text links: Access Paper: View a PDF of the paper titled Collapsibility of Performance Metrics in Clinical Predictive AI, by Jo~ao Matos and 4 other authorsView PDFHTML (experimental)TeX Source view license Current browse context: cs.LG prev | next new | recent | 2026-08 Change to browse by: cs cs.AI References Citations NASA ADSGoogle Scholar Semantic Scholar export BibTeX citation Loading… BibTeX formatted citation loading… Data provided by: Bookmark checked="checked"class=“labs-tab-input”> Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, Media Code, Data and Media Associated with this Article alphaXiv Toggle alphaXiv (What is alphaXiv?) Links to Code Toggle CatalyzeX Code Finder for Papers (What is CatalyzeX?) DagsHub Toggle DagsHub (What is DagsHub?) GotitPub Toggle Gotit.pub (What is GotitPub?) Huggingface Toggle Hugging Face (What is Huggingface?) ScienceCast Toggle ScienceCast (What is ScienceCast?) Demos Demos Replicate Toggle Replicate (What is Replicate?) Spaces Toggle Hugging Face Spaces (What is Spaces?) Spaces Toggle TXYZ.AI (What is TXYZ.AI?) Related Papers Recommenders and Search Tools Link to Influence Flower Influence Flower (What are Influence Flowers?) Core recommender toggle CORE Recommender (What is CORE?) IArxiv recommender toggle IArxiv Recommender (What is IArxiv?) Author Venue Institution Topic About arXivLabs arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv’s community? Learn more about arXivLabs. Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?) mathjaxToggle(); We gratefully acknowledge support from our major funders, member institutions, , and all contributors. About Help Contact Subscribe Copyright Privacy Accessibility Operational Status (opens in new tab) Major funding support from
[AI-37] uringLLM : Efficiently Scaling Foundation Models Toward Physical AI
链接: https://arxiv.org/abs/2608.30567
作者: Yuheng Zhang,Yizhao Wang,Da Zhu,Hua Zhou,Yue He,Jiahui Hu,Shaman Tang,Hanlin Chen,Yuhua Wei,Anhua Liu,Shuang Su,Rui Xin,MingYuan Wang,MingHao Li,HaoJie Yang,Siqi Liu,Jianlei Zheng,WeiChao Huang,Qiman Wu,Hang Zhang,HongGou Yang,Xianming Liu
类目: Artificial Intelligence (cs.AI)
备注: Technical Report; includes supplementary material
Abstract:We present Turing-20B-A2B, a 20B-parameter Mixture-of-Experts language model that activates approximately 2B parameters per token, designed for long-context and latency-sensitive physical AI applications. The model adopts Quantile Routing in a dynamic top-k configuration, enabling token-adaptive expert allocation while maintaining balanced expert utilization and a controlled average compute budget. During deployment, we further apply capacity-constrained routing to prompt prefill for more regular and efficient expert execution, while retaining dropless routing during pretraining. Turing-20B-A2B also employs a hybrid attention architecture that combines Lightning Attention with a small number of full-attention layers for efficient long-context modeling. The model is pretrained with a progressive three-stage curriculum and extended to a native context length of 128K through continued pretraining, with further inference-time extension to 512K using YaRN. Despite its compact active-parameter budget, Turing-20B-A2B achieves, at the base-model stage, overall general capability exceeding Qwen3-8B Base and approaching Qwen3.5-9B Base, while maintaining strong long-context performance and favorable prefill-latency scaling. These results demonstrate an effective balance among model capability, long-context scalability, and practical inference efficiency.
[AI-38] Q-Strata: Hierarchical Bit Allocation for Mixed-Precision Quantization of Mixture-of-Experts LLM s EMNLP2026
链接: https://arxiv.org/abs/2608.30564
作者: Deokjae Lee,Sihun Chu,Hyun Oh Song
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: EMNLP 2026 Long Paper - Main Conference
Abstract:Mixed-precision quantization (MPQ) assigns a different bitwidth to each linear layer of a large language model (LLM) to minimize the quantization-induced quality loss under a fixed budget, but Mixture-of-Experts (MoE) models contain these layers in every expert of every MoE block, so the allocation space grows far larger than in a dense model. Existing methods either allocate within each block under a uniform per-block budget, or allocate across blocks through an additive proxy, and neither directly optimizes a model-level objective over the choices that couple the blocks. We propose Q-Strata, a bi-level allocator that ranks within-block assignments with a cheap proxy and allocates across blocks with a model-level objective evaluated on the assembled quantized model. Its inner stage caches a Pareto frontier of candidates per block over finely spaced budgets, leaving the outer stage to set one budget per block instead of a bitwidth for every linear layer. With the search reduced to one budget per block, the outer stage optimizes this model-level objective directly, capturing the inter-block coupling that additive proxies miss. On Mixtral-8x7B-Instruct, Qwen1.5-MoE-A2.7B, and DeepSeek-V2-Lite, Q-Strata consistently achieves lower WikiText2 perplexity than uniform-bitwidth GPTQ and the state-of-the-art MoE MPQ methods MxMoE and GEMQ in the low-bit regime. The code is available at this https URL.
[AI-39] AdaPath: Query-Adaptive Path-Finding via Path-Bank for Multi-Hop Implicit Biomedical KGQA
链接: https://arxiv.org/abs/2608.30556
作者: Jun Hyeong Kim,Dongki Kim,Yinhua Piao,Sung Ju Hwang
类目: Artificial Intelligence (cs.AI)
备注: 25 pages, 6 figures, 30 tables
Abstract:Path-finding over knowledge graphs has become an effective way to ground LLM reasoning on multi-hop questions. However, biomedical QA introduces two distinct challenges that general-domain methods are not designed for: (i) queries do not expose intermediate reasoning and can be answered through multiple valid pathways, and (ii) biomedical knowledge graphs are densely connected, so path-finding methods easily take wrong turns. To address these challenges, we propose AdaPath, a path-finding framework that retrieves query-adaptive meta-paths from Path-Bank, which captures both query semantics and biomedical knowledge graph structure. AdaPath provides the missing cues in biomedical queries while effectively pruning dense knowledge graph neighborhoods during multi-hop reasoning. We further release BioStrat-QA, a biomedical KGQA benchmark that stratifies multi-hop queries by how much intermediate reasoning they expose. Across biomedical KGQA benchmarks, AdaPath consistently outperforms baselines, sustaining meaningful path-finding even when multi-hop queries expose less surface information. The source code is available at this https URL.
[AI-40] Designing an Auditable LLM -Supported Workflow for Qualitative Thematic Analysis
链接: https://arxiv.org/abs/2608.30543
作者: Nadia Jul Jeldtoft,Tariq Yousef
类目: Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
备注:
Abstract:Large Language Models (LLMs) offer new possibilities for scaling qualitative analysis, but existing applications often provide limited methodological transparency regarding how qualitative methods are translated into computational procedures. This paper presents an auditable and privacy-preserving computational operationalization of inductive and latent Thematic Analysis (TA). This paper first derives five design principles from the methodological requirements of TA and the conditions introduced by LLM-based inference: preserving interpretative context, maintaining traceable relationships between empirical material and analytical outputs, representing analytical constructs and reasoning explicitly, constraining LLM inference to interpretative tasks, and enabling privacy-preserving local deployment. Second, it presents a proof-of-concept for a two-phase workflow that operationalizes these principles by combining interpretative LLM inference with deterministic procedural control to generate codes, analytical justifications, themes, and theme descriptions while preserving explicit links to the source material. Third, it proposes an evaluation framework combining structural comparison with human-led TA and independent expert assessment of analytical quality. The evaluation is conducted on semi-structured Danish interview transcripts. and the results shows that the workflow produces code-level outputs with coverage broadly comparable to human annotations and highly rated analytical justifications, while generating a more compressed thematic structure characterized by fewer and broader themes. The findings demonstrate the feasibility of auditable LLM-supported TA through a modular workflow designed to scale to larger datasets, accommodate different LLMs, and support transfer across research domains, with domain adaptation primarily requiring adjustments to the prompting strategy.
[AI-41] DiffPDE: Masked Diffusion Language Models as PDE Solver
链接: https://arxiv.org/abs/2608.30532
作者: Wenxuan Guo,Yuyang Hong,Lubin Fan,Zhaojin Fu,Lin Chen,Kun Ding,Shiming Xiang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Existing approaches for synthesizing Partial Differential Equation (PDE) solvers predominantly rely on autoregressive models, yet their global left-to-right decoding incurs substantial redundancy when addressing inherently localized bugs. In this work, we challenge this inefficient paradigm and propose DiffPDE, a framework leveraging discrete diffusion language models for targeted code repair. By introducing a localized re-masking and infilling strategy, DiffPDE regenerates only erroneous regions while preserving correct context, naturally aligning generation with the sparse nature of PDE errors. Furthermore, to handle coupled bugs requiring sequential interventions, we present Iterative Debugging GRPO (ID-GRPO), a reinforcement learning scheme that enables multi-round debugging within single trajectories via intermediate rewards. Experiments on PDEBench show that DiffPDE achieves competitive accuracy, outperforms same-scale AR models, and significantly accelerates repair.
[AI-42] Learning-Assisted Congestion-Aware Route Scheduling for Semiconductor Fab Material Control Systems
链接: https://arxiv.org/abs/2608.30520
作者: Hao Yin,Meiqi Tu,Anbang Liu,Shaochong Lin,Max Z.J. Shen
类目: Artificial Intelligence (cs.AI); Optimization and Control (math.OC)
备注:
Abstract:Automated material handling systems in semiconductor fabs are operated by a material control system (MCS) that must schedule a relay route for every transport command online, before execution. This is a data-driven scheduling problem in which route cost is dominated in the upper tail by queueing at heterogeneous, partially observable relay equipment, so route selection requires estimating both delivery time and congestion risk at the decision moment. This paper proposes a transport-network-aware dynamic congestion representation (TN-DCR). Built on a static directed transport graph induced by historically observed relay segments, TN-DCR combines structural route priors, multi-window network-wide congestion context, route-level bottleneck exposure, and an inductive graph-aware route embedding, all constructed under a prediction-time-safety invariant that admits only information observed strictly before the prediction moment. The representation feeds separate queue- and transfer-time regressors and an ordinal multi-label classifier producing calibrated multi-threshold exceedance scores, with an empirical-Bayes stock-key residual correction reducing systematic queue-time underprediction. The predictions serve as costs in a risk-constrained route-scheduling rule that minimizes predicted delivery time subject to a bound on extreme-congestion probability, embedding the learned predictors within a lightweight operations-research decision model. In a controlled closed-loop evaluation, mean delivery time falls by 16.4% and internal resource waiting time by 22.6% while throughput remains essentially unchanged.
[AI-43] rajectory-Initialized Neural Double Q-Routing for Large-Scale Overhead Hoist Transport Systems
链接: https://arxiv.org/abs/2608.30512
作者: Cheng Gu,Qiusheng Zhao,Anbang Liu,Shaochong Lin,Max Z.J. Shen
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Optimization and Control (math.OC)
备注:
Abstract:Large-scale industrial robot fleets share constrained physical infrastructure, making vehicle travel times dependent on safety separation, intersection access, downstream blocking, and station contention. We study this problem in overhead hoist transport (OHT) systems, a representative ceiling-mounted material-handling system used in semiconductor fabs. Static shortest-path routing cannot account for these time-varying traffic costs, whereas tabular Q-routing adapts online but learns each destination–node–action value independently, limiting information sharing across sparsely visited routing contexts and making startup behavior sensitive to inaccurate value estimates. We propose Neural Double Q-routing, which replaces destination-indexed tables with a shared state–action value network. The network is warm-started through return-to-go regression on mixed simulator-generated routing trajectories and then refined online using Double-Q updates, local congestion correction, and event-stratified structured replay. Across nine matched fleet-size–arrival-rate settings with 100, 150, and 200 OHTs, the proposed framework reduces mean completion time relative to tabular Double Q-routing by 0.8% – 8.8% . It achieves the lowest mean completion time among all compared methods in the six 150- and 200-OHT settings, whereas Dijkstra remains best in the three 100-OHT settings. Completed-task counts remain within 1% of tabular Double Q-routing in eight of nine settings, and 95th-percentile completion time decreases in eight settings. In two matched startup scenarios, offline initialization increases the number of completed tasks by up to 23% and reduces tail completion time by up to 15% .
[AI-44] nsor Methods for Language Models: From Token Representation to Training Adaptation Inference Compression and Interpretability
链接: https://arxiv.org/abs/2608.30505
作者: Matvei Tarasov,Salman Ahmadi-Asl,Andre L. F. de Almeida,Andrzej Cichocki
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models (LLMs) are built from structured high-dimensional objects such as token representations, weights, adaptation updates, caches, and activations, whose multilinear structure is underexploited by the conventional matrix-centric view. Tensor decompositions and tensor networks provide a principled algebraic language for this structure, yet the literature often treats them as isolated compression mechanisms. This survey organizes tensor methods for LLMs through two complementary views: a seven-stage lifecycle taxonomy covering tokenization, embeddings, pre-training, adaptation, compression, inference, and interpretability, and a component view covering embeddings, attention, and feed-forward networks. We provide unified notation and theoretical foundations, analyze tensorization strategies for individual Transformer components, and compare methods at each lifecycle stage while making differences in evaluation protocols and model scales explicit. We further connect tensor methods to neighboring efficiency techniques and probabilistic tensor networks. Finally, we synthesize open challenges and introduce \rho_\rm gap , a metric for the compression-realization gap between theoretical memory reduction and measured system-level speedup. By treating tensorization as a common structural principle, the survey provides a structured entry point to tensorized language models and clarifies when parameter savings can plausibly translate into memory efficiency, computational efficiency, or interpretability. The GitHub page dedicated to this paper is accessible at \hrefthis https URLthis https URL.
[AI-45] CM2: Multimodal Cultural Reasoning via an Integrated Multi-Agent Framework PRICAI2026
链接: https://arxiv.org/abs/2608.30498
作者: Qi Li,Zhaojie Kang,Yingjie He,Zheng Lin,Hao Zhang,Guangxin Wu,Yan Gong,Rong Fu,Jianyuan Ni
类目: Artificial Intelligence (cs.AI)
备注: Accepted to the 23rd Pacific Rim International Conference on Artificial Intelligence (PRICAI 2026) as a short paper. 11 pages, 4 figures. Code and dataset are available at this https URL
Abstract:Multimodal Large Language Models (MLLMs) have shown remarkable success in STEM domains, where progress is often driven by vertical, step-by-step deduction under relatively stable symbol systems. Their horizontal, interdisciplinary cultural reasoning, however, remains this http URL propose CM2, a multi-agent framework grounded in the cognitive pathway of human cultural interpretation. CM2 integrates multimodal perception, retrieval-augmented generation, networked reasoning, gated fusion, and reward-driven this http URL on CM2D across multiple MLLM backbones show consistent gains over CoT and typical reasoning paradigms; ablations validate each module’s contribution, and conflict analyses confirm genuine cross-modal arbitration.
[AI-46] Measuring Memory and Generalization as Separable Geometric Channels: The Topo2 Framework
链接: https://arxiv.org/abs/2608.30487
作者: Zhanbo Zhang,Ming Liu,Qing Wang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 14 pages, 6 figures
Abstract:Deep networks trained on noisy labels simultaneously generalize on clean data and memorize flipped labels. These are usually conflated as pressures on one capacity. We present Topo^2, a measurement framework that makes them causally separable, measurable, and law-governed. Persistent-homology H1 structure of the representation space separates into a within-class manifold channel (a function of the training stopping point) and a cross-class channel (a monotone readout of memorized flipped samples). An intervention, the FM0 prescription (zero loss on flipped samples from epoch 0), reaches each setting’s generalization ceiling while memorizing essentially nothing. Within the framework we establish a law set with graded evidence: (L2) FM0 separation prescription (9/9); (L1) the within-channel as a training-position function (mid-rise 6/6; convergence-back CIFAR 3/3, SVHN 2/3); (L3) a ring-construction identity (definitional, not a law); and TLS (memory-generalization topological layering): memory is causally additive, anchored (silencing clean collapses the representation), invertible (stripping memory restores near-ceiling generalization), and quantitatively billable (the memorization cost law, effective slope coefficient C ~ 0.38 at the reference capacity: CIFAR-10 0.3801 / SVHN 0.3806 / CIFAR-100 0.384 / VGG 0.3715, capacity-dependent in general and traced to clean-sample feature displacement). We also publish the framework’s boundaries: a falsification ledger of nine dead ends, and an instrument-vindication section that excludes six families of global statistics as explanations of the within-channel. The framework turns “memorization” from an ill-defined capacity into a measurable, separable, invertible topological layer.
[AI-47] From Metaheuristics to Exact Methods: A CP-SAT Approach for Multi-Objective Healthcare Workforce Scheduling ECAI2026 IJCAI
链接: https://arxiv.org/abs/2608.30419
作者: Vipul Patel,Anirudh Deodhar,Dagnachew Birru
类目: Artificial Intelligence (cs.AI)
备注: 13 pages, 10 tables. Accepted at MODeM Workshop, IJCAI-ECAI 2026. Extends arXiv:2607.05177
Abstract:Healthcare workforce scheduling is an NP-hard optimization problem requiring simultaneous satisfaction of labor regulations, coverage requirements, employee preferences and cost objectives. Existing approaches (genetic algorithms, integer programming, constraint programming) model 6-12 constraints at shift-level granularity and cannot guarantee regulatory compliance. They also lack support for multi-role, multi-skill heterogeneity, mandatory break scheduling with midpoint control, acuity-weighted workload equity, sub-shift granularity, inter-week stability, and cross-midnight shifts. This paper presents CP-SAT: a Constraint Programming formulation for multi-role, multi-skill healthcare scheduling. CP-SAT enforces 14 hard constraints guaranteeing zero regulatory violations, while optimizing 15 soft objectives via a unified weighted penalty function. Contributions include a shift-window decomposition enabling break scheduling with centrality control, acuity-weighted workload equity, multi-granularity resolution from 15 minutes to 1 day, inter-week stability, and grid-offset preprocessing mapping cross-midnight shifts into a single scheduling day without solver changes. CP-SAT is evaluated on 18 instances: five synthetic hospital units (10-33 nurses), 10 INRC-II benchmarks (5-80 nurses, up to 8-week horizons) and 3 NRP-23 compatible instances (10-25 nurses) with cross-midnight Night shifts. Results: zero hard-constraint violations across all 18 instances by construction; proven optimality on INRC-II n005w4 (objective 118, gap 0.0%, 104s); feasible schedules scaling to 179,800 variables and 351,425 constraints (80 nurses); service quality improved 50-67% over MOGA; and model size scaling near-linearly at approximately 4,400 variables per employee. The formulation enforces 29 total constraints (14 hard, 15 soft), nearly three times the industry average. Comments: 13 pages, 10 tables. Accepted at MODeM Workshop, IJCAI-ECAI 2026. Extends arXiv:2607.05177 Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2608.30419 [cs.AI] (or arXiv:2608.30419v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.30419 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-48] DERELAB: Probing Defeasible Reasoning and Confirmation Bias in LLM s with a Generative Benchmark EMNLP2026
链接: https://arxiv.org/abs/2608.30413
作者: Jayanta Sadhu,Sayem Shahad,Kenneth Marino
类目: Artificial Intelligence (cs.AI)
备注: Accepted at the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026)
Abstract:Defeasible reasoning is a type of reasoning where inferences are drawn from plausible current evidence, but can be retracted upon the introduction of newer evidence. Although recent studies have examined language-model behaviors in defeasible reasoning, the datasets have been static and lack wide coverage of non-monotonic reasoning categories. We introduce DeReLab, a generative framework that produces multi-turn belief-updating conversations from parameterized graph structures across default and inheritance reasoning, with formally verified ground truth at every turn, enabling controlled measurement of how models respond to confirming and disconfirming evidence. This controlled generation process creates a testbed for experimental designs that isolate specific reasoning demands. Applying this capability to the study of confirmation bias, we evaluate nine open and proprietary large language models and find that nearly all exhibit a systematic tendency to accept congruent evidence while resisting incongruent updates, with several models correctly identifying a weakening update yet failing to revise their conclusion. We believe our work and findings will facilitate future research on evaluating language models in defeasible reasoning.
[AI-49] Dense Clinical Contrasts Enhance Medical Knowledge Updating in Large Language Models EMNLP2026
链接: https://arxiv.org/abs/2608.30405
作者: Yangmin Huang,Shu Quan,He Geng,Xin Ye,Qianyun Du,Zhiyang He,Jiaxue Hu,Xiaodong Tao
类目: Artificial Intelligence (cs.AI)
备注: Accepted to Findings of EMNLP 2026
Abstract:Medical knowledge changes continually, making large language models vulnerable to relying on outdated yet clinically plausible information. We study whether the format of supervision affects medical knowledge updating under a matched training-budget setting. We introduce SEER-Bench, a temporally anchored oncology-staging benchmark curated from the latest versioned SEER Research Data release, and render identical medical update events from NCCN oncology guidelines into four supervision formats: EMQ, MSQ, FITB, and SAQ. Across SEER-Bench and HealthBench Professional, EMQ gives the most stable external transfer and retention among same-budget SFT variants. With EMQ supervision, the updated 4B model produces competitive results on temporally anchored oncology staging, reaching 64.8% answer accuracy and 59.6% rationale accuracy on SEER-Bench. Diagnostic analyses suggest that EMQ exposes denser clinical contrast signals while preserving discriminative representations with smaller movement from the base model. These results show that medical knowledge updating depends not only on the update algorithm, but also on how knowledge is structured as supervision.
[AI-50] Scaffolding Foundation Models into Physical-World Agents Pushes the Frontier of Long-Horizon Navigation
链接: https://arxiv.org/abs/2608.30396
作者: Zixing Lei,Gengze Zhou,Xiong-Hui Chen,Jiazhao Zhang,Yiyang Huang,Hang Yin,Haoqi Yuan,Qi Wu,Weixin Li,Siheng Chen
类目: Artificial Intelligence (cs.AI); Robotics (cs.RO)
备注: 22 pages, 6 figures
Abstract:Long-horizon physical-world agents must reason over distant goals while grounding decisions in reliable closed-loop behavior. Today’s foundation models split these capabilities: vision-language models (VLMs) infer missing information and adapt high-level plans but remain brittle and inefficient at repeated navigation grounding, while navigation foundation models (NFMs) robustly execute semantic goals but operate as bounded episodes without persistent task-level reasoning. We introduce NavMCP, an agentic scaffolding framework that couples a VLM reasoning agent with an NFM executor for long-horizon exploration. The VLM decides what evidence to seek, where to search, and when to stop, while the NFM grounds each semantic sub-goal into closed-loop navigation. Three channels structure their collaboration: intent translates evidence needs into navigation calls, observation converts rollouts into source-grounded trajectory evidence, and memory accumulates findings, negative evidence, and unresolved goals across calls. This design turns isolated navigation rollouts into persistent embodied interaction without retraining either model. On Embodied Question Answering, NavMCP achieves state-of-the-art results on HM-EQA, MT-HM3D, and EXPRESS-Bench. Under matched agent and executor backbones, it outperforms an episodic interface by 14.9 percentage points on HM-EQA. On a Unitree Go2, NavMCP reaches 78.3% success, with its margin over the strongest baseline growing from 10 to 45 points as the task horizon increases. These results demonstrate the potential of scaffolding complementary foundation models into long-horizon physical-world agents.
[AI-51] opGQ: Fast GNN Post-Training Quantization Leverag ing Topology Information
链接: https://arxiv.org/abs/2608.30394
作者: Dain Kwon,Kanghyun Choi,Hyeyoon Lee,Sunjong Park,Seoyong Lee,Sukjin Kim,Jinho Lee
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 7 pages, 4 figures. Accepted at the 63rd ACM/IEEE Design Automation Conference (DAC 2026)
Abstract:Existing GNN quantization methods suffer from considerable quantization overhead, which severely limits their practical usage in real-world scenarios. To this end, we present TopGQ, an accurate post-training GNN quantization framework, alleviating redundant quantization overhead. We propose dual-axis scale absorption, which enables activation quantization along both the outer and inner dimensions by merging one into the adjacency matrix. On top of that, we introduce TopPIN, a proxy for nodes’ local structure, and use it to group nodes with similar topology during quantization. Experimental results show that TopGQ reduces quantization time by an order of magnitude while preserving accuracy.
[AI-52] DASC: Decay-Aware State Compression for Hybrid Linear-Attention Serving
链接: https://arxiv.org/abs/2608.30386
作者: Yanqi Yu,Pingwei Sun,Jianchao Tan,Tao Zhang,Yuchen Xie,Xunliang Cai,Yao Liu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Hybrid linear-attention architectures have recently scaled to large open-weight models, offering quality competitive with full attention while substantially reducing key/value (KV) cache growth. However, their in-place recurrent-state updates complicate cache management: prefix reuse requires state checkpoints alongside full-attention KV, while storing state checkpoints in full increases memory pressure, leading to more evictions and repeated prefill. By analyzing the decay structure of Gated DeltaNet (GDN) and Kimi Delta Attention (KDA), we find that different heads and channels retain prefix information over markedly different timescales, which we term \emphretention horizons. This variation suggests substantial compression potential in persistent state checkpoints. Building on this observation, we introduce \emphDecay-Aware State Compression (DASC), which derives retention horizons from model weights, selects long-horizon state units, and packs them into a ragged state checkpoint layout. To integrate efficiently with tensor-parallel inference engines, DASC furtherly balances compressed state checkpoints across TP ranks. On reuse, DASC either zero-fills omitted units or refreshes them from a bounded suffix with additional compute cost. Across retrieval and end-to-end reasoning benchmarks on Kimi-Linear, conservative DASC configurations remain close to full caching while compressing KDA recurrent state checkpoints by 2.63\times . Under fixed state checkpoint memory budgets, the resulting capacity gains reduce mean Time to First Token (TTFT) by 42.6% and improve input throughput by 68.4%. At larger compression ratio, suffix refresh recovers much of the accuracy lost to more aggressive omission, at the cost of additional replay computation. Qwen with GDN exhibits a similar quality–efficiency trend, showing that DASC extends from channel-wise KDA to head-wise GDN.
[AI-53] PAVE: Predictive Alignment and Value-Guided Evolution for World-Action Policies
链接: https://arxiv.org/abs/2608.30378
作者: Botong Zhao,Fang Yu, Tim,Senhua Zhu,Xinyuan Chen,Yue Lu
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:
Abstract:Direct vision-language-action policies generate continuous robot actions efficiently, but standard behavior cloning leaves two complementary gaps: their representations are not explicitly required to describe how the scene evolves over multiple time scales, and deployment trajectories of unequal quality are often reused without separating useful dynamics from undesirable behavior. We introduce \method, a direct world-action policy that combines outcome-agnostic predictive learning with outcome-aware policy improvement. \method first retains a local fixed-offset JEPA objective and adds trajectory-relative multi-horizon transition alignment at 25%, 50%, 75%, and 100% of the remaining episode. These training-only targets require the current policy representation to preserve both local physical changes and longer-range task progress, without supplying explicit future tokens to the action head. \method then trains an independent distributional value critic on cumulative deployment trajectories, computes action-chunk-aligned N -step advantages, and converts them into positive, negative, or null text conditions for a flow-matching actor. Thus, every valid trajectory can teach what physically happened, while the actor is deployed only under the condition associated with relatively better actions. The multi-horizon predictor and critic are removed from online execution, preserving direct action generation from the current observation, language instruction, and proprioception. \redclaimAcross the three simulation benchmarks, \method achieves the strongest overall performance while preserving the direct actor’s online execution path.
[AI-54] Answer Probing-Guided Search for Diverse Solution Exploration of LLM s EMNLP2026
链接: https://arxiv.org/abs/2608.30345
作者: Yi Fang,Que Shen,Chengpeng Li,Boyi Deng,Wei Shi,Wenjie Wang,Fuli Feng,Fengli Xu,Dayiheng Liu
类目: Artificial Intelligence (cs.AI)
备注: Accepted to the EMNLP 2026 Main
Abstract:Generating multiple diverse and high-quality solutions is valuable for many applications, such as code-test generation and drug discovery. However, Large Language Models (LLMs) tend to converge on a single high-confidence solution during inference, limiting exploration of alternative valid solution paths. Existing test-time methods promote diversity through tree-like search and prune semantically similar branches using response-level semantic embeddings. However, we find that such embeddings are easily confounded by linguistic and stylistic similarities, making it difficult to distinguish genuinely distinct solution paths. To address this, we introduce Answer Probing, which probes the potential answer an LLM would reach from an intermediate reasoning path. We demonstrate that the hidden states of probed answers more effectively differentiate distinct solution paths than semantic embeddings, and the perplexity of probed answers serves as a practical proxy for reasoning correctness. Based on these findings, we propose Answer Probing-Guided Tree Search (APTS), which guides the tree search by the probed answers’ hidden state similarity and perplexity. Experiments on three reasoning tasks across two LLMs show that APTS consistently enhances solution diversity, demonstrating its effectiveness and robustness.
[AI-55] Parallel Time-Band Mixing with Learned Observation-Adding for Robust ASR Front-Ends INTERSPEECH2026
链接: https://arxiv.org/abs/2608.30326
作者: Xingyu Shen,Runze Wang,Wei-Ping Zhu,Benoit Champagne
类目: ound (cs.SD); Artificial Intelligence (cs.AI)
备注: Accept by Interspeech 2026
Abstract:Speech enhancement is often used as a front-end for robust ASR, yet recurrent temporal and cross-band modules introduce sequential dependencies that reduce parallel efficiency. In this paper, we present a sequence-parallel band-split enhancement front-end built on a Parallel Time-Band Mixer (PTBM) block that eliminates within-block recurrent unrolling. PTBM integrates intra-band temporal mixing and per-frame cross-band attention within a unified parallel architecture, enabling efficient contextual modeling across both time and frequency dimensions. The system retains the mask-plus-residual reconstruction interface and introduces learned Observation-Adding (LOA) to suppress ASR-sensitive artifacts without development-set tuning. Experiments on DNS Challenge and CHiME-4 with frozen Whisper back-ends show that the proposed front-end consistently reduces word error rate relative to recurrent band-split baselines while requiring only 0.96 M parameters and 0.58 GMAC/s for the front-end network.
[AI-56] Online Estimation of Dynamic Origin-Destination Matrices Using Reinforcement Learning with Link-Flow Propagation Guidance
链接: https://arxiv.org/abs/2608.30317
作者: Donggyu Min,Dong-Kyu Kim
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Online dynamic origin-destination (OD) matrix estimation (DODE) calibrates time-dependent OD demand to reproduce observed link-flow trajectories. In online, OD demand should be estimated from current observations and propagated network states while subsequent observations and stochastic dynamic network loading (DNL) outcomes remain uncertain. Recently, reinforcement learning (RL) has emerged as a promising alternative, reducing computational burden by replacing iterative algorithms while being applicable to stochastic environments. However, because the policy is trained offline and deployed online, it must handle varying target link-flow trajectories; since each target trajectory defines the link-flow error used in the reward, the same OD demand vector can require different adjustments, making conventional scalar feedback ambiguous. To address this gap, this study proposes LFPG-RL, which integrates link-flow propagation guidance (LFPG) into proximal policy optimization (PPO). LFPG combines link-flow error sensitivities with the contribution of each OD-time demand component to simulated link flows, transforming aggregate mismatch into OD-specific advantage shaping for PPO actor updates. At deployment, the policy requires only a single forward pass. LFPG-RL is developed and evaluated on 250 weekday trajectories of 15-min link-flow data from a Melbourne arterial network modeled by a link transmission model with stochastic route choice. On held-out trajectories, LFPG-RL achieved an RMSE of 4.69, MAPE of 20.15%, and Pearson correlation of 0.995. These results support the contention that our method is a more efficient and accurate online OD demand calibration method compared to existing ones.
[AI-57] ail-Replay: Escaping the Curse of Linear Attention in Prefix Caching for Hybrid LLM s
链接: https://arxiv.org/abs/2608.30310
作者: Yirui Liu,Ruoling Qi,Xuaner Wu,Penghang Liu,Jian Chen
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Hybrid large language models interleave full-attention layers with linear-attention layers to reduce the cost of long-context inference. This structure complicates prefix caching: full-attention key-value caches are token-addressable, whereas linear-attention layers maintain recurrent states that cannot be rolled back to arbitrary prefix boundaries. Existing hybrid prefix caching methods address this mismatch by storing recurrent-state checkpoints. As a result, token-level matches are directly usable only at positions aligned with stored checkpoints, constraining prefix reuse to a discrete set of boundaries. We present Tail-Replay, a prefix caching mechanism that enables unconstrained token-level prefix reuse in hybrid large language models. The key insight is that linear-attention mechanisms such as Gated DeltaNet can be viewed as a structured, lossy compression of the input prefix: gated recurrent updates progressively attenuate the contributions of earlier inputs. Consequently, the recurrent state of a matched prefix can be well approximated by replaying only a short, recent suffix of that prefix. Tail-Replay exploits this property by caching the exact full-attention key-value cache while omitting recurrent-state checkpoints. On a cache hit, it reconstructs the linear-attention states by replaying a short, recent suffix of the matched prefix. As a result, the reuse boundary is determined by the shared tokens rather than by recurrent-state checkpoints. We evaluate Tail-Replay on three Gated DeltaNet-based hybrid models using the LongBench and RULER benchmarks. With only a 5–10% replay budget, it retains 92.8–99.9% of full-prefill quality on LongBench and RULER. For serving efficiency, we evaluate time-to-first-token speedups across multiple matched-prefix lengths—8K, 16K, and 32K. The speedup grows with prefix length, reaching 9.1 – 14.3\times over full prefill at 32K.
[AI-58] CateKV: On Sequential Consistency for Long-Context LLM Inference Acceleration ICML2025
链接: https://arxiv.org/abs/2608.30295
作者: Haoyun Jiang,Haolin Li,Jianwei Zhang,Fei Huang,Qiang Hu,Minmin Sun,Shuai Xiao,Yong Li,Junyang Lin,Jiangchao Yao
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Published at ICML 2025
Abstract:Large language models (LLMs) have demonstrated strong capabilities in handling long-context tasks, but processing such long contexts remains challenging due to the substantial memory requirements and inference latency. In this work, we discover that certain attention heads exhibit sequential consistency in their attention patterns, which can be persistently identified using a coefficient-of-variation-based algorithm. Inspired by this observation, we propose CateKV, a hybrid KV cache method that retains only critical token information for consistent heads, thereby reducing KV cache size and computational overhead, while preserving the majority of KV pairs in adaptive heads to ensure high accuracy. We show the unique characteristics of our algorithm and its extension with existing acceleration methods. Comprehensive evaluations on long-context benchmarks show that, while maintaining accuracy comparable to full attention, CateKV reduces memory usage by up to 2.72\times and accelerates decoding by 2.18\times in single-sample inputs, and boosts throughput by 3.96\times in batch scenarios.
[AI-59] BCPPO: Bachelier-Inspired Constrained Proximal Policy Optimization for Tail-Risk-Aware Safe Reinforcement Learning
链接: https://arxiv.org/abs/2608.30283
作者: Dongsheng Hou,Yanqiao Chen,Yuhan Rui
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Expected-cost constraints can still permit rare, high-cost events. Monte Carlo conditional value at risk (CVaR) gradients can be noisy at high confidence, whereas critics that model an outcome distribution add complexity. We propose BCPPO (Bachelier-Inspired Constrained Proximal Policy Optimization), a proximal policy optimization (PPO) method. Separately initialized cost-prediction networks (critics), trained with random sample masks, produce disagreement that marks predictions sensitive to which state-action regions occur in the training data and to critic training. A Bachelier formula for the expected amount above a reference level converts this disagreement into a smooth policy-update penalty. Gradients from this penalty do not alter the critics, so temporal-difference (TD) critic learning is unchanged. A saturation-aware controller adjusts the mean-cost penalty and stops accumulated error from growing while that penalty is clipped. Deployment retains only the policy network. The disagreement penalty is neither a tail-event probability nor a guaranteed error bound, and it provides no safety guarantee. Across 175 runs with shared tasks, costs, budgets, training steps, and evaluation seeds, no comparator attains both higher mean return and lower mean CVaR than BCPPO in any task. On Push1, BCPPO has no lower return and no higher CVaR than every comparator, with at least one strict gain. These results support a practical balance among reward, caution around cost predictions that vary across trained critics, and policy-only deployment.
[AI-60] Generating Workflow DAGs from Natural Language with Non-Reasoning LLM s
链接: https://arxiv.org/abs/2608.30250
作者: Anand Iyer,Bhanu Khetharpal,Srinivas Upadhya,Ramkumar Rajagopal
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:This paper addresses the problem of translating natural-language routing rules written by business administrators into executable workflow graphs for enterprise contact centers. Each target is a directed acyclic graph (DAG) of conditional actions with parallel branches, hit-first fallback chains, and per-branch Boolean predicates, encoded in the JSON dialect of a commercial routing platform. We show that neuro-symbolic decomposition enables lower-cost, non-reasoning large language models to generate complex workflow DAGs at production-relevant quality without expensive extended-reasoning models. Our central diagnostic is an emission-density bottleneck: on a 635-rule benchmark of manufactured synthetic data, models select the correct graph nodes with high accuracy but increasingly misconfigure attributes and Boolean grouping as the number of interdependent nodes emitted in one pass grows. We therefore move combinatorial graph construction from the model into a deterministic compiler driven by a compact intermediate representation, with a learned registry-selection front end that focuses generation on relevant vocabulary. Across four models, the full system reaches approximately 89% LLM-judge validity, approximately 90% exact-match condition accuracy, and 99-100% valid JSON while using roughly half the per-rule prompt tokens of a monolithic prompt. On GPT-5.3-chat, the method improves judge validity by 24 percentage points and achieves statistical equivalence to a reasoning model’s out-of-the-box quality, although an approximately 8-point frontier gap remains. We also present a deployment path and transferable lessons for structured-generation applications.
[AI-61] LLM -Based Knowledge Graph Completion Combining Discrete Structural Coding with Similar Entity Information ISWC26
链接: https://arxiv.org/abs/2608.30235
作者: Jiaqi Wang,Dongying Lin,Yang Yang,Yinan Liu,Bin Wang,Xiaochun Yang
类目: Artificial Intelligence (cs.AI)
备注: Accepted by ISWC 26 Posters and Demos Track
Abstract:Knowledge graph completion requires models to use both textual descriptions and relational structure. Existing LLM-based methods either encode KG structure as discrete tokens or refine a restricted set of candidate entities, and these two directions have largely been studied separately. We propose CoSC for LLM-based KGC, which combines discrete structural coding with similar entity information. Specifically, an LLM generates an initial candidate entity ranking from discrete structural codes, after which information from entities with structures similar to that of the query entity refines the ranking. Experiments on FB15k-237 show that CoSC outperforms existing baselines on MRR and Hits@10 while remaining competitive on Hits@1.
[AI-62] CoLa-ICD: A Knowledge-Enhanced Framework for Long-Tail Automated Medical Coding
链接: https://arxiv.org/abs/2608.30234
作者: Yihang Cheng,Veronica Liesaputra,Andrew Trotman
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Automatic medical coding assigns ICD codes to clinical notes, but it remains challenging due to long documents, imbalanced label distributions, and diverse terms. These challenges are especially severe for rare codes, which have limited training instances and are easily confused with semantically similar labels. We introduce CoLa-ICD, a knowledge-enhanced framework for long-tail prediction. CoLa-ICD enriches ICD labels with external terms, models dependencies among related codes, and learns stronger alignment between label semantics and clinical evidence for long-tail prediction. Experiments show that CoLa-ICD improves long-tail prediction with larger gains in larger and sparser label spaces and achieves state-of-the-art performance in AUC, F1, and P@k. Our code is available at this https URL.
[AI-63] Rethinking the Test-Time Prompt Tuning Objective from the Perspective of Calibration
链接: https://arxiv.org/abs/2608.30230
作者: Jungwon Choi,Hyeonseo Jang,Kibok Lee,Eunwoo Kim
类目: Artificial Intelligence (cs.AI)
备注: 9 pages
Abstract:Test-time prompt tuning (TPT) has emerged as a powerful paradigm, refining prompts for each test sample via entropy minimization (EM) over multiple augmented views. However, we identify a limitation in the standard EM-based adaptation: it inherently drives the model toward overconfident predictions disregarding sample-specific uncertainty, leading to significant calibration degradation. To address these limitations, we propose a new objective that replaces the conventional EM loss by aligning the original-view prediction with a target distribution derived from augmented views via cross-entropy, while adversarially incorporating the entropy of the target distribution to capture sample-specific uncertainty. Furthermore, to better construct this target distribution, we apply confidence-aware temperature scaling to each augmented-view prediction according to its confidence, sharpening confident predictions while softening uncertain ones. This formulation allows the model to increase confidence only when the target distribution is reliable, while preserving uncertainty when it reflects ambiguous or conflicting augmented-view predictions. Extensive experiments across diverse benchmarks demonstrate that our approach not only achieves state-of-the-art accuracy but also significantly improves model calibration.
[AI-64] SPARK: Skeleton-Guided Reasoning Synthesis from Large-Scale Scientific Literature
链接: https://arxiv.org/abs/2608.30214
作者: Yu Li,Wei Li,Xin Gao,Mengyuan Sun,Xiaoyang Wang,Qizhi Pei,Lijun Wu
类目: Artificial Intelligence (cs.AI)
备注: 24 pages, 15 figures
Abstract:Scientific reasoning remains challenging for open-source models, largely due to the lack of high-quality scientific reasoning data. Existing datasets are often dominated by factual recall or formulaic problem solving, with limited emphasis on mechanism understanding, evidence-grounded reasoning, and hypothesis evaluation. To address this, we introduce SPARK (Scientific Paper Abstracted Reasoning sKeleton), a paper-oriented synthesis framework built on Sci-Base, a large-scale corpus of research papers spanning 10 scientific disciplines. Instead of directly converting papers into question-answer pairs, SPARK treats the claim-evidence-derivation structure of a paper as the fundamental unit of reasoning synthesis. Specifically, SPARK (1) distills each paper into a compact reasoning skeleton capturing its central claims and supporting evidence, enabling self-contained question generation, and (2) synthesizes reasoning tasks from four scientific perspectives: mechanistic reasoning, hypothesis falsification, quantitative derivation, and boundary calibration. A final consistency verification stage further removes unsupported or contradictory outputs. Using this framework, we construct Spark-234K, a scientific reasoning dataset with substantially higher difficulty and diversity than existing resources. Experiments show that Spark-234K consistently outperforms existing scientific reasoning datasets while achieving stronger performance with significantly fewer training samples.
[AI-65] SIR: Self-improving Red-teaming for Compute Use Agents
链接: https://arxiv.org/abs/2608.30207
作者: Chen Xiong,Zhiyuan He,Pin-Yu Chen,Stjepan Picek,Tsung-Yi Ho
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:Computer use agents (CUAs) are vision-language models that perceive a screen and act on a real operating system through mouse, keyboard, and terminal, and they are increasingly deployed to automate everyday digital tasks. Because they can be exposed to untrusted content while operating, they are vulnerable to indirect prompt injection (IPI), in which an adversary plants instructions in content the agent will read and redirects it toward actions that violate the user’s intent. Existing CUA safety benchmarks evaluate fixed injections written by hand, which may underestimate the risk posed by an adaptive adversary. We present SIR, a black box IPI attack that (i) composes stealthy injections from a small library of reusable principles stated in plain language and (ii) wraps composition in an iterative feedback loop that diagnoses the victim’s failed trajectories and distills the bypasses into new, named strategies that are reapplied across tasks. Unlike prior red teaming of web agents, we target CUAs at the operating system level and score attacks with a fully deterministic oracle, using checks on filesystem, service, and permission state rather than an LLM judge. On experiment, we evaluate three frontier CUAs. Composing principles with feedback raises the attack success rate over a baseline written by hand, for example from 4% to 24% on Claude Opus 4.8 and from 0% to 28% on Gemini 3.5 Flash, while the benign task still completes. Principles discovered against one model further transfer to a different architecture with no additional feedback.
[AI-66] FaVOR: LLM -Based Agent ic Framework for Factor Mining via Empirical Validation
链接: https://arxiv.org/abs/2608.30192
作者: Hyeonjin Kim,Minseok Kim,Seunghyeon Jung,Sujin Pyo,Huisu Jang,Woojin Lee
类目: Artificial Intelligence (cs.AI); Computational Engineering, Finance, and Science (cs.CE)
备注:
Abstract:Traditional finance relies on experts to hand-craft factors through a principled process grounded in economic rationale. Recent LLM-based multi-agent systems have automated this process, scaling factor mining far beyond manual effort. However, these automated approaches optimize directly for returns and rarely check whether a generated factor still expresses the economic hypothesis that motivated it. We identify this inconsistency between mathematical form and economic meaning as a structural failure mode of return-oriented automation. The resulting factors blur the line between real signals and spurious correlations and break down across regime shifts. We propose FaVOR (Factor Validation through Observable Reasoning), an agentic framework that restructures factor mining around hypothesis-level evidence rather than return outcomes. In place of the standard hypothesis-to-formula leap, FaVOR enforces a three-stage consistency loop tying mathematical form to economic rationale throughout. (1) Decomposition splits a broad economic hypothesis into independent observable conditions. (2) Validation checks whether each factor reflects its intended condition. (3) Integration merges them into a composite whose structure remains interpretable. On the CSI 500 and SP 500 in 2025, FaVOR outperforms existing baselines while remaining effective across regimes. FaVOR shows that hypothesis-grounded factor discovery produces signals that are interpretable by construction, regime-robust, and economically faithful. The code is available at this https URL.
[AI-67] PR-Attention for Combinatorial Generalization
链接: https://arxiv.org/abs/2608.30124
作者: Melisa Civelekoğlu,Isabeau Prémont-Schwarz
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Systematic generalization remains a significant challenge in deep learning. In particular, combinatorial generalization - generalizing to new configurations of known factors of variation - is effortless for humans but difficult for standard neural architectures that rely on statistical correlations rather than explicit structural representations. We introduce a new architectural component that embeds structured inductive bias into deep learning: an attention mechanism operating over tensor-product representations (TPRs). Through controlled experiments on compositional tasks, we show that this TPR-attention mechanism outperforms existing architectural components in combinatorial generalization. These results highlight the value of integrating explicit compositional structure into neural attention and point toward a promising path for models capable of systematic generalization.
[AI-68] Graph4BiLO: Graph Neural Network Approximation for Bilevel Mixed-Integer Linear Optimization
链接: https://arxiv.org/abs/2608.30103
作者: Jessica D. Elrefaei,Kaixun Hua,Seungbae Kim,Hoang Nam Tran,Juan S. Borrero
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible
Abstract:Bilevel mixed-integer linear optimization problems model hierarchical decision processes in which a leader anticipates the optimal response of a follower. Although expressive, these problems are computationally challenging because lower-level optimality is embedded in the leader’s feasible region. Value-function reformulations replace the nested follower optimization with a constraint involving the follower’s optimal value, but evaluating this value function exactly can itself be expensive. This paper introduces Graph4BiLO, a graph neural network (GNN) approach for learning bilevel value functions from variable–constraint graph representations. In contrast to fixed-length multilayer perceptron (MLP) representations, the GNN uses shared message-passing parameters and can therefore be applied across multiple problem sizes with a single trained model. The learned ReLU network is encoded exactly as mixed-integer linear constraints and embedded in an approximate single-level formulation. A repair step subsequently re-solves the follower problem for the selected leader decision to recover a bilevel-feasible follower response. We evaluate Graph4BiLO on knapsack interdiction instances with 20–100 items against the exact MibS solver and the learning-based Neur2BiLO method. Graph4BiLO obtains objective values comparable to Neur2BiLO across all tested sizes while avoiding size-specific neural networks. An additional out-of-distribution experiment demonstrates zero-shot transfer from 20-item training instances to previously unseen 40- and 60-item instances. However, embedding message passing at every graph node substantially increases the resulting mixed-integer formulation size and solve time. These results identify a central tradeoff between size-generalizable graph representations and the computational cost of embedding GNNs within optimization models.
[AI-69] VERA: Authority-Preserving Edge Revocation for Federated AI-Agent Workflows
链接: https://arxiv.org/abs/2608.30091
作者: Lifei Liu,Haoran Yu,Xiaochong Jiang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Modern agent frameworks compose planners, tool agents, remote services, and shared specialists into runtime delegation graphs, but their revocation APIs still resemble token or subtree invalidation. When one delegation is withdrawn, the runtime must know which agents lose authority while independently authorized agents keep working. We study this authority consistency problem and introduce VERA (Verifiable Edge Revocation for Agents), a verifier-checkable revocation contract and API emitted by agent-runtime adapters as signed evidence. Under disjunctive authority, revoking edge e invalidates exactly T_intent(e,G) = reach(G) \ reach(G \ e), the agents whose every authorizing root path used e. Used as a contract, this target exposes two runtime failures: tree cascades over-revoke shared agents, while deployer-scoped cascades under-revoke cross-domain descendants. In a LangGraph framework-replt cells repeated 20 times yield 500compiled-framework traces and 2,000 valid signed delegation decisions; 13/25 cells contain runtime multi-parsharing and 8/25 contain cross-deployer shies 500/500 target proofs, preserves all320 alternate-parent shared-agent cases that tree cascade revokes, and rejects unauthorized signers and omission attacks. Baseline replay over 1,9that holder/node and tree-style targetscannot express this behavior. We further validate schema portability on A2A, AutoGen, and CrewAI artifacts: nine traces, including five executable Cregned delegation events that pass schema and signature checks.
[AI-70] Game-Agnostic Value Functions through Automatic JSON Feature Extraction
链接: https://arxiv.org/abs/2608.30056
作者: Dien Nguyen,Diego Perez-Liebana
类目: Artificial Intelligence (cs.AI)
备注: 4 pages, to be published in Conference on Games (CoG) 2026
Abstract:JSON Bag-of-Tokens (JSON-Bag) is a recently proposed method to generically represent game trajectories by tokenizing their JSON descriptions. We introduce JSON-Bag VF, a game-agnostic approach to training value functions for game-playing agents using JSON-Bag prototypes. We show that this approach can be enhanced with Random Forest-based feature selection and a method to select game-stage-specific features. We evaluate JSON-Bag VF with One-step-look-ahead (JSON-Bag OSLA) on six tabletop games over different combinations of prototype-tokenization and feature selections. JSON-Bag OSLA outperforms baseline OSLA agents in most games. Our analysis also shows that feature selection significantly improves JSON-Bag VF and that feature selection is the most important factor in JSON-Bag VF performance, over prototype-tokenization.
[AI-71] Mitigating Over-Optimization in PRM-Guided Search in Mathematical Reasoning by Optimizing the Guide
链接: https://arxiv.org/abs/2608.30051
作者: Taejong Joo,Diego Klabjan
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Process reward models (PRMs) provide dense step-level guidance for search-based reasoning, enabling inference-time compute to be allocated toward promising partial solutions. However, recent evidence suggests that PRM-guided search can over-optimize imperfect process rewards, pruning viable trajectories while expanding spurious ones. In this work, we theoretically show that directly leveraging PRM score is vulnerable to verifier noise through an extreme-value effect: non-viable prefixes become more likely to receive spuriously high scores as reasoning depth increase. Therefore, we formulate the PRM-guided search as a robust optimization problem over plausible reward perturbations, termed maximin PRM-guided search, leading to a training-free robust process supervision method that preserves promising alternatives when step-level scores are noisy. Maximin PRM-guided search mitigates this failure mode by reducing sensitivity to over-optimized PRM outliers. Without fine-tuning or online adaptation, maximin search consistently improves the PRM-guided search by 17-35% on average, outperforming outcome- and step-level baselines in 14 out of 16 settings. Our source code is available at this https URL.
[AI-72] Spec2Twin-Chain: Orchestrating Bi-Level Optimization with LLM s for Blockchain Digital Twin Construction
链接: https://arxiv.org/abs/2608.30050
作者: Haoting Zhang,Haoxian Chen,Jiayuan Sheng,Donglin Zhan,Zeyu Zheng,David D. Yao,Wenpin Tang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Building a blockchain digital twin largely requires translating domain knowledge and specific system descriptions into a simulator architecture, calibrating its parameters against behavioral evidence, and validating the constructed twin. These steps are commonly performed through application-specific modeling efforts that can be difficult to reuse across systems and downstream decision problems. We consider automating this process through Spec2Twin-Chain, a framework that formulates blockchain digital-twin construction as a bi-level optimization problem. At the upper level, a large language model proposes and revises structurally admissible architectures using system specifications, behavioral evidence, and feedback from evaluated designs. At the lower level, a simulation-based optimizer calibrates the architecture-conditioned parameters under explicit objectives and guardrail constraints. The two levels iterate. The evaluated candidates at lower levels are retained in a global archive and used to guide subsequent proposals at upper levels. We conduct controlled experiments involving twin calibration, feedback-driven recovery, stress analysis, downstream policy optimization, and policy updating. The results demonstrate that the framework can construct behaviorally accurate twins, improve initial designs through iterative feedback, and reuse calibrated twins to support downstream decisions.
[AI-73] Can LLM Agents Discover? Evaluating Creativity on ML Engineering Tasks
链接: https://arxiv.org/abs/2608.30047
作者: Shitanshu Bhushan,Yunxiang Zhang,Lu Wang
类目: Artificial Intelligence (cs.AI)
备注: COLM 2026
Abstract:Recent AI systems promise autonomous scientific discovery, claiming to discover algorithms and produce research papers, yet understanding whether they exhibit creativity, the capacity to produce solutions that are both novel and useful, remains an open question. We present a framework for evaluating multi-turn LLM research agents’ creativity using ML engineering tasks as a testbed, through three dimensions: P-Creativity (psychological novelty: novel relative to the agent’s own prior solutions within a run), H-Creativity (historical novelty: novel relative to the corpus of human solutions), and Usefulness (task performance). Evaluating two agent frameworks, AIDE and AIRA-Dojo, on 10 Kaggle-style machine learning tasks from MLE-Bench, we develop an LLM-as-a-Judge pipeline and verify its strong correlation with human creativity judgments, providing a reliable automated metric for P-Creativity evaluation at scale. Applying this pipeline to agent trajectories, we find: (1) all agents exhibit declining P-Creativity as they transition from exploration to exploitation; (2) LLMs exhibit greater H-Creativity than medal-winning humans, yet achieve lower performance. Our findings reveal that current agents can explore novel regions of the solution space but lack the capacity to convert this novelty into improved task performance.
[AI-74] Forget or Fine-tune? A Comparative Study of Machine Unlearning Strategies for Noisy Label Correction
链接: https://arxiv.org/abs/2608.30046
作者: João L. P. Santana,Filipe R. Cordeiro
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Accepted at SIBGRAPI’26
Abstract:Noisy labels remain a critical challenge for training deep neural networks, since memorizing incorrect labels degrades generalization. Once noisy samples are identified after training, the standard solution is to retrain the model from scratch on the cleaned dataset, which is increasingly expensive as datasets and models grow. Machine Unlearning (MU) has recently emerged as a computationally efficient alternative, but the relative effectiveness of different MU strategies for noisy-label correction remains poorly understood. In this work, we conduct a comparative empirical study of five MU methods (NegGrad, Fine-Tuning (FT), Random Labeling (RL), SalUn, and MUNBa) across symmetric, asymmetric, instance-dependent, and open-set noise on CIFAR-10, CIFAR-100, and the real-world noisy dataset Food-101N. Our central finding is that the appropriate unlearning strategy is conditioned on the noise structure. Simple FT is a strong baseline across most closed-set scenarios; RL and SalUn are the most consistently robust methods and, under instance-dependent noise, approach retraining accuracy at a fraction of the computational cost; MUNBa shows advantages mainly under extreme symmetric noise. Under open-set noise, in contrast, we show that retraining on the cleaned subset degrades accuracy relative to the noisy baseline, so approximating the retrained model is not an adequate objective in this regime. On Food-101N, all MU methods remain competitive and achieve accuracies close to retraining despite reducing runtime by an order of magnitude. These findings provide practical guidelines for selecting MU strategies for post-training noisy-label correction.
[AI-75] Reachability-Based Capability Confinement for LLM Agents under Indirect Prompt Injection
链接: https://arxiv.org/abs/2608.30041
作者: Wujie Xiong,Rabimba Karanjai,Yang Lu,Weidong Shi,Lei Xu
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language model agents place outputs from external skills into their execution context, allowing attacker-controlled data to influence later privileged actions. Existing defenses mainly classify untrusted content or authorize proposed operations. They do not directly address how an agent’s future authority should change once untrusted data enters its state. We present SkillGuard, a harness-level enforcement layer that treats this event as contamination and restricts future capabilities to disconnect the resulting state from deployer-defined forbidden states. Given sound skill summaries and policies, SkillGuard represents security-relevant transitions with a Skill Impact Graph, specifies admissible control over skill parameters via steerability signatures, and mediates invocations with an inline reference monitor. Following contamination, it computes weighted capability restrictions using binary, fractional, or fractional-flow strategies without auxiliary language-model inference. We evaluate SkillGuard on four AgentDojo suites with two backend LLMs, Gemini 2.5 Flash and Llama3.3-70B, against an LLM-only No Defense baseline and three defenses at different system layers: Spotlighting, CaMeL, and AttriGuard. We construct a compositional attack benchmark in which each attack combines observations individually insufficient to induce target violation and evaluate the same baselines on it. Under AgentDojo’s Tool Knowledge attacks, SkillGuard eliminates attack success on three of four suites for both backends and reduces it to 4.8% and 14.3% on Slack. Against compositional attacks, it outperforms every baseline on Llama and matches the strongest baseline on Gemini at higher benign utility. Fractional-flow restriction preserves substantially more capabilities than binary restriction at the same attack success rate. Across both settings, SkillGuard adds no model calls or token overhead.
[AI-76] Beyond Uncertainty: Multi-Solver Disagreement Rewards for Self-Evolving Reasoning Curricula CIKM’26
链接: https://arxiv.org/abs/2608.30035
作者: Vinoth Selvendran,Zhanming Zhang
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Proceedings of the 35th ACM International Conference on Information and Knowledge Management (CIKM '26), November 07–11, 2026, Rome, Italy
Abstract:Self-evolving reasoning frameworks train a Challenger to generate questions exposing a Solver’s weaknesses, creating adaptive curricula without human data. However, existing approaches use a single solver’s sampling uncertainty as the Challenger’s reward. This creates a fundamental bottleneck: as the solver grows confident on the Challenger’s question distribution, all sampled answers converge identically, collapsing the reward to zero and starving the Challenger of learning signal. Critically, this single-model reward cannot distinguish genuinely easy questions from those that merely align with one solver’s learned biases. We propose a multi-solver disagreement reward using a heterogeneous ensemble varying in model capacity and sampling temperature. A normalized Shannon entropy over the ensemble’s per-question plurality answers explicitly rewards questions where solvers produce conflicting solutions—capturing difficulty as inter-model divergence rather than intra-model sampling variance. This richer gradient enables the Challenger to discover questions targeting true capability boundaries, producing a curriculum that forces downstream Solvers to develop robust reasoning strategies generalizing across problem types. Our approach is a drop-in reward function replacement requiring no framework modifications or additional data. Experiments with Qwen3-4B show that Solvers trained on disagreement-Challenger questions achieve +1.34 points average improvement on competition-math benchmarks (MATH-500, AMC, Olympiad), suggesting that multi-solver disagreement provides a complementary and scalable signal for curriculum generation in self-play reasoning systems.
[AI-77] Interpreting and Steering for Safe and Correct Code Generation EMNLP2026
链接: https://arxiv.org/abs/2608.30025
作者: Hao Yan,Ziyu Yao
类目: Artificial Intelligence (cs.AI)
备注: Accepted to the EMNLP 2026 Main Conference
Abstract:Large language models (LLMs) frequently generate source code containing vulnerabilities, yet little work studies the internal mechanisms that distinguish safe from vulnerable generation in them. In this work, we systematically perform a mechanistic interpretation of LLMs, aiming at both understanding how code safety-vs-vulnerability is represented or driven by components in an LM and turning the insights into actionable steering strategies to encourage safer code generation. To this end, we introduce CodeSec-Pairs, a dataset of 9,342 Python safe-and-vulnerable contrastive code pairs, sampled from Llama-3.1-8B-Instruct. Utilizing the dataset, we explore approaches to localize layers and attention heads that relate to code safety, and further experiment with different steering strategies for inference-time vulnerability reduction. In particular, we propose DuoSteer, a double-steering approach that simultaneously applies safety and code-correctness steering to attention heads. In experiments over five vulnerability types, DuoSteer leads to an average of -26.9% vulnerability rate reduction and +7.5% functional correctness improvement, which outperforms not only other steering variants but also prompting and supervised fine-tuning baselines. The advantage also replicates on Qwen-2.5-Coder-7B-Instruct with another 2,500 contrastive pairs sampled from that model.
[AI-78] Automatic Conversion of NICE Guidelines to an Executable Computational Model Using Large Language Models ALT
链接: https://arxiv.org/abs/2608.30022
作者: Ashvin Gupta,Denys Prociuk,Alessandra Russo,Brendan C. Delaney
类目: Artificial Intelligence (cs.AI)
备注: 18 pages. Published in Learning Health Systems (2026)
Abstract:Introduction: NICE guidelines provide evidence-based recommendations for clinical care but remain largely in unstructured natural language. Existing approaches to converting them into computable representations often focus on individual diseases, require substantial manual encoding, and do not scale. Large language models (LLMs) may enable much of this translation to be automated. Methods: We present an end-to-end approach that converts textual clinical guidelines into executable models capable of generating explainable, patient-specific recommendations. A stepwise LLM-based transformation with in-context examples produces human-inspectable intermediate artifacts. We apply the approach to NICE pancreatic and lung cancer guidelines, use expert review to assess rule alignment, and evaluate the executable pancreatic cancer model on 20 patient vignettes. Results: Expert review showed strong alignment between the source guidelines and generated executable models. Most discrepancies were partial omissions rather than incorrect logic, while hallucinated or fundamentally incorrect rules were rare. On the patient vignettes, the executable model achieved an F1 score of 82.5%. Conclusion: LLMs can transform natural-language NICE guidelines into interpretable, executable models that preserve guideline structure, support transparent inspection and modification, and generate patient-specific recommendations. These findings demonstrate the feasibility of scalable automated generation of computable clinical guidelines.
[AI-79] Error Detection for PET/CT Radiology Reports: Domain-Specific vs Large Language Models
链接: https://arxiv.org/abs/2608.30021
作者: Hermione Warr,Harry Anthony,Lilli J Freischem,Yasin Ibrahim,Daniel R McGowan,Konstantinos Kamnitsas
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Errors in radiology reports can adversely affect patient treatment, yet automated report quality assurance remains challenging because errors are often subtle and require domain expertise to detect. Although large language models (LLMs) have recently been proposed for radiology report verification, their ability to detect clinically meaningful errors beyond chest X-ray datasets remains under-explored. To this end, we present the first systematic evaluation of language models for PET/CT report error detection, comparing compact domain-specific models with SOTA open-weight LLMs. We collected 30,633 oncology FDG PET/CT reports from 23 radiologists over 10 years. We trained domain-specific BERT models to detect clinically motivated synthetic reporting errors and evaluated alongside zero-/few-shot Qwen3-32B, Gemma-3-27B and Llama-3.3-70B on a held-out benchmark of 11,500 reports. A 15M-parameter model achieved 94.4% balanced accuracy with a 5.8% false-positive rate, compared with 84.0% for the strongest prompted LLM. Task-specific adaptation of Llama-3.3-70B closed this performance gap (94.4%) but retained substantially greater computational requirements. Our results suggest that domain-specific training matters more than model scale for PET/CT report error detection, supporting compact models as an accurate and computationally efficient approach to automated radiology report quality assurance.
[AI-80] EMPO: Temporally-grounded Multi-task Post-training for Large Audio-Language Models EMNLP2026
链接: https://arxiv.org/abs/2608.29999
作者: Apoorva Kulkarni,Kaousheik Jayakumar,Sreyan Ghosh,Utathya Aich,Ramani Duraiswami,Dinesh Manocha
类目: ound (cs.SD); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Accepted at EMNLP 2026 Main Conference. Project page - this https URL
Abstract:Large audio-language models (LALMs) describe audio at the clip level but cannot assign timestamps to the events, speakers, or sounds they identify. Despite being essential for downstream tasks like speech recognition and dense audio captioning, timestamping remains a key limitation of most LALMs. We present TEMPO (Temporally-grounded Multi-task Post-training), the first unified model to handle audio, speech, and music timestamping tasks. Our core contribution is a supervised fine-tuning (SFT) stage built on three innovations: atomic timestamp tokens, a time-aware projector that injects sinusoidal wall-clock encodings into audio frame embeddings, and a distance-aware Gaussian loss. Our training is based on a synthetic-to-real curriculum. We further introduce, to our knowledge, the first application of reinforcement learning to unified audio timestamping, using GRPO with verifiable temporal rewards that directly optimize the evaluation objectives. Rather than serving as the primary source of performance gains, GRPO acts as a refinement stage on top of the SFT checkpoint, providing modest additional improvements. To support this work, we build a training dataset containing 119K samples and an evaluation benchmark containing 10K samples, drawn from established corpora across five tasks. On this benchmark, TEMPO outperforms Audio Flamingo Next and Qwen3-Omni, two state-of-the-art LALMs explicitly trained on timestamped data. Experiments confirm that SFT delivers most of these gains, with GRPO providing consistent but moderate refinements.
[AI-81] AutoCRAT: Within-trajectory Joint Control of Stochasticity and Compute for LLM Reasoning EMNLP2026
链接: https://arxiv.org/abs/2608.29988
作者: Hanjun Luo,Qiushi Liu,Jingya Zhang,Haihong Pang,Jiaheng Wen,Yifei Ma,Yu Yao,Chengxi Zhang,Hanrong Zhang,Yankai Chen,Hanan Salam
类目: Artificial Intelligence (cs.AI)
备注: EMNLP 2026 Findings
Abstract:Large language models (LLMs) achieve strong reasoning performance, which depends critically on inference-time decisions. Yet these decisions are commonly handled by static, one-size-fits-all policies, limiting adaptation to diverse tasks and reasoning stages. Recent adaptive methods partially address this limitation, but they primarily adapt either decoding stochasticity (how the model explores) or reasoning compute (how long the model reasons) in isolation, leaving their interaction within a single reasoning trajectory unmodeled. To address this challenge, we shift toward a within-trajectory joint control view, and instantiate it in AutoCRAT, a decoder-side controller for frozen backbones. Using only signals available during decoding, AutoCRAT jointly adjusts sampling stochasticity and reasoning budget during generation. AutoCRAT operates over a discrete action space and updates control decisions only at semantic boundaries, improving stability while remaining responsive to the evolving reasoning process. Comprehensive evaluation across 6 benchmarks demonstrates that AutoCRAT (I) uses 13.8-52.7% fewer inference tokens on average than recommended static configurations, (II) surpasses recommended static and adaptive baselines by 1.5-4.5% in relative accuracy, and (III) enjoys strong cross-backbone transferability.
[AI-82] An Open-Source Event-Driven Pipeline for Cryptocurrency Market Data: Ingestion Forecasting and On-Chain Fraud Detection
链接: https://arxiv.org/abs/2608.29973
作者: Basil Sajid Shaikh,Melrick Mascarenhas,Nuzhat Faiz Shaikh
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Cryptocurrency markets generate high-frequency, multi-source data that is expensive to work with unless a team already has commercial-grade streaming and warehousing infrastructure in place. This paper describes a fully open-source pipeline that reproduces the behavior of a cloud-native, event-driven system – file arrival triggering a message, a message triggering compute – entirely on commodity hardware, using Apache Kafka and a filesystem-watching poller in place of managed cloud triggers. The pipeline partitions historical Gemini exchange data into hourly and minutely files, ingests them asynchronously through two independently grouped Kafka consumers (one for audit logging, one for Spark-triggered ETL), and lands cleaned output in a PostgreSQL warehouse with historical and aggregated schemas plus asset-specific data marts. We use the resulting Bitcoin data mart to compare a seasonal ARIMA model against a single-layer LSTM network for price forecasting, and separately apply Random Forest and Gradient Boosting classifiers, with additional engineered features, to the public Ethereum fraud detection benchmark introduced by Farrugia et al. We report the architecture, the modeling methodology, and the resulting metrics, and we are explicit about the limitations of comparing forecasts issued at different horizons and of evaluating fraud detection on a static, already-labeled dataset.
[AI-83] EDGE: Engine for Deterministic Graph Evaluation through Conversation Simulation from Graph Structured DSL Configuration
链接: https://arxiv.org/abs/2608.29971
作者: Ram Kulathumani,Regunathan Radhakrishnan,Anupam Tripathi,Xiangbo Mao,Roshanak Omrani,Keshav Somani,Shwet Kamal Mishra,Shayna Lurya
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:As agentic systems evolve into complex multi agent orchestration workflows, there is a growing and critical need for systematic frameworks that measures an agent’s behavioral consistency and determinism. In this paper, we introduce a formal evaluation methodology that is grounded in AgentGraph, a planner powered by a domain specific language that represents agent reasoning through a dynamically adjustable directed graph. We leverage this structural formalism and utilize graph traversal algorithms that exhaustively enumerate conversational paths, forming a comprehensive evaluation set that captures the agent’s complete behavioral space. We then systematically replay these reproducible trajectories to compare observed outputs and state transitions against the intended DSL specification. To quantify reliability, we define novel metrics that measure response and trajectory determinism, structural adherence and semantic consistency across both exact replays and their linguistic variants. Our system’s results demonstrate that agents configured using frameworks like AgentGraph and LangGraph with explicitly structured node transitions show superior determinism over agents that are not configured with controlled transitions.
[AI-84] raining-Free Action Correction for VLA Model Failures via Language Feedback
链接: https://arxiv.org/abs/2608.29967
作者: Owen Kwon,Pablo Ortega-Kral,Arthur Bucker,Jean Oh
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: 8 pages, 6 figures. Project page: this https URL
Abstract:Vision-Language-Action (VLA) models demonstrate strong semantic understanding yet exhibit systematic failures during deployment. The conditions under which these failures occur, and whether they can be corrected without retraining, remain poorly understood. In this paper, we take steps toward addressing this gap. We present CorrectVLA, a framework that translates task-level natural language corrections into additive action magnitude adjustments without modifying policy weights. A human provides a single task-level correction, applied uniformly across all rollouts without per-episode intervention. In simulation, CorrectVLA recovers execution misalignment failures across both in-distribution and OOD tasks. In real-robot experiments on a UFactory xArm7 under environment shift, CorrectVLA restores near-perfect success where the base policy almost entirely breaks down, generalizing across object locations and identities. Through a taxonomy of failure modes on LIBERO-90, we find that execution misalignment failures, where the policy reaches the correct target but miscalibrates action magnitudes, represent the correctable subset, while other failure modes where semantic comprehension itself breaks down are not amenable to this approach. The approach succeeds when policies possess strategic correctness and fails when fundamental comprehension is absent, establishing a practical operational boundary for inference-time correction.
[AI-85] Review Before Trust: Source-Grounded Integrity Gates for AI-Assisted Personal Health Records
链接: https://arxiv.org/abs/2608.29965
作者: Nora Girda,Adrian Groza
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models can convert medical documents into structured data, but plausible output may still be unsupported by the source. Persisting such output in a longitudinal health record, a record that accumulates patient information over time, therefore creates an integrity risk: unverified data may influence later summaries, trends, or preventive-care computations. We introduce an evidence-gated trust-promotion model that keeps generated data provisional until a deterministic monitor verifies it against the source document. The monitor admits a candidate for a specified downstream use only when the source contains a unique supporting quotation, the relevant fields occur within the same laboratory row, and the required provenance is preserved. The generator cannot approve its own output, missing or ambiguous evidence causes refusal, and refused candidates remain available for human review rather than being silently discarded. We implement the model in Medical DataCloud, a personal health-record application, and evaluate it through automated tests and a replay of saved extraction outputs. All 22 conformance and mutation tests pass. The replay covers nine historical laboratory PDF reports containing 102 manually labelled rows. The reports produce 97 numeric candidates: schema validation accepts all 97, an earlier packet-level evidence check accepts 94, and the hardened quotation- and row-level policy admits 72 while retaining 25 for review. The study evaluates system integrity rather than clinical correctness or clinical safety. The results demonstrate the technical feasibility of an enforceable boundary that prevents generated claims from authorizing their own reuse in a longitudinal health record.
[AI-86] SearchWiki: Learning to Build and Navigate Knowledge Wikis for Active Information Seeking
链接: https://arxiv.org/abs/2608.29953
作者: Guransh Singh,Vishwajeet Kumar,Arkadeep Acharya,Adnan Qidwai,Jaydeep Sen,Sachindra Joshi
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Flat retrieval-augmented generation treats a corpus as a bag of chunks, discarding document hierarchy and cross document structure. We introduce SearchWiki, a harness framework that synthesizes a corpus into a hierarchical, typed, navigable wiki and trains an agent, WikiResearcher-9B, to retrieve information through multi-turn tool use. The wiki organizes knowledge into three layers - document overviews, cross- document topic pages, and page-level source records; enabling progressive refinement of retrieval when initial lookup misses. We optimize the agent’s navigation policy with on-policy reinforcement learning with a multi-component reward function balancing answer correctness, retrieval quality and trajectory efficiency. Evaluation on ViDoRe-V3 (8 domains), FinanceBench, and memory benchmarks (LoCoMo, LongMemEval, PersonaMem-v2) shows that WikiResearcher- 9B which is our RL-tuned Qwen 9B model, significantly outperforms same-size untrained baselines and exceeds or matches larger external models. SearchWiki paired with WikiResearcher-9B demonstrates that learned navigation over structured corpora is a superior alternative to flat retrieval.
[AI-87] Influence Is Not Authority: When Causal Guardrail Signals Make Legitimate Tool Use Look Like an Attack in Tool-Using LLM Agents
链接: https://arxiv.org/abs/2608.29942
作者: Tanzim Ahad,Ismail Hossain,Md Jahangir Alam,Sai Puppala,Syed Bahauddin Alam,Sajedul Talukder
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:The key limitation of current state-of-the-art influence-based guardrails is that they do not reliably distinguish a legitimate, user-authorized action from a malicious, unauthorized action when both rely on external tool information. This ambiguity can cause benign actions to trigger unnecessary verification and intervention, reducing utility and adding latency. We expose this limitation through an authorization-equivalence audit of 96 conditions derived from 24 base cases. Within matched source comparisons, we hold authorization, the exact committed action, and its intended effect fixed, changing only whether a required value comes from the user or a legitimate tool result. Although the action remains unchanged, this harmless relocation shifts the causal signal toward the attack region in all 24 cases under both Llama and Gemma scorers. Matched unauthorized controls show that the signal remains attack-sensitive, yet the benign relocation produces a larger average score shift than the actual change in authorization. Architecture-level evaluation shows how this mismatch propagates through guardrail designs. With a semantic monitor, attack success is 0% and utility is 28%, compared with 16% and 60% without it. A shadow-based guardrail allows every tested harmless run, yet does not reject matched unauthorized actions more often overall: 57.5% of unauthorized runs pass automatically before reaching the later security check, compared with 29.2% of authorized runs. These results show that the studied causal signal reveals what shaped an action without reliably encoding whether the action was authorized, and that reference construction and routing are integral to the effective security decision.
[AI-88] AcrossWAM1.0:A Modular Latent World-Action Stack for Compact Robot Policies
链接: https://arxiv.org/abs/2608.29937
作者: Yafei Zhang,Nan Wu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Latent world-action models avoid rendering future pixels by predicting an action-relevant visual subgoal in feature space. LaWAM established this formulation, but its original presentation left the world model, multimodal backbone, and deployment checkpoint tightly coupled. We introduce AcrossWAM1.0, a modularization and scaling study of this latent world-action stack. Rather than presenting latent subgoals as a new algorithm, we make the module boundary explicit: a policy adapter produces latent-action and action-generation contexts; a retained latent world decoder grounds the predicted transition in the current scene;and a flow-matching expert generates continuous action chunks. We further separate training-only teachers from the inference graph and provide a verifiable deployment export. On 2,000 paired LIBERO episodes, replacing a Qwen3-VL-2B backbone with Qwen3.5-0.8B yields 97.45% success versus 98.00% for the 2B model (a-0.55percentage-point difference; exact McNemarp=0.266). This does not prove equivalence, but it meets a prespecified two-point retention criterion. The compact, inference-reachable checkpoint contains 1,472.6M unique parameters, 42.4% fewer than the original 2B policy, while all retained tensors are bitwise identical to the source checkpoint. Cross-family execution is additionally checked with a MiniCPM-V adapter smoke test; closed-loop cross-family transfer remains an open evaluation. AcrossWAM1.0 therefore contributes an auditable software and evaluation boundary for compact latent world-action policies, distinct from LaWAM’s original latent-subgoal contribution.
[AI-89] On the Instance Hardness as a Decision Criterion in TinyML Systems
链接: https://arxiv.org/abs/2608.29913
作者: Tobiasz Puslecki,Krzysztof Walkowiak
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:TinyML includes the implementation of machine learning on devices with limited memory and computing resources. With the development of technology, AI systems continue to scale in terms of size and computational requirements. This forces researchers to adapt methods to be environmentally sustainable by designing techniques for reducing computational costs and energy consumption in inferring AI models, even in small devices. In this work, we present preliminary findings on a novel application of the tree depth prune instance hardness method to the TinyML system. The results indicate that threshold control can change energy consumption with limited classification quality changes. This method allows us to adjust classification accuracy, thereby influencing computational complexity and energy consumption for inference. We present a work in progress with initial results as a proof of concept.
[AI-90] INTERVenE: Temporal-Abstraction-Interval Based Transformers for Short-Horizon Medical Event Prediction
链接: https://arxiv.org/abs/2608.29901
作者: Shahar Oded,Yuval Shahar
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Electronic Health Record (EHR) prediction models in the intensive care unit must learn from sparse and irregular measurements while preserving the clinical meaning of time and supporting transparent decision-making. We present INTERVenE, a family of Transformer architectures whose input is an interval-based, knowledge-based temporal abstraction (KBTA), a token stream of named clinical concepts (states, trends, events, contexts) drawn from a curated medical ontology, rather than an unnamed bin index or a raw measurement triplet. This naming layer is what we ask KBTA to do: it makes the model’s per-token attributions resolve to clinical concepts by construction. INTERVenE offers two complementary variants: an auto-regressive decoder that generates future abstraction trajectories with a per-step risk readout (localizing \emphwhen and \emphafter which events risk rises), and a bidirectional encoder for single-pass joint risk and time-to-event prediction. Evaluated on 57,078 MIMIC-IV admissions against GRU-D, STraTS, and KarmaLego, INTERVenE-Enc reaches a support-weighted AUPRC _w of 0.672, improving by 0.041 over the strongest neural baseline with non-overlapping 95% bootstrap CIs, while also taking the best AUROC _w (0.901) and length-of-stay MAE (44.4,h). INTERVenE-Ar (AUROC _w 0.854 , AUPRC _w 0.587 under the same evaluation contract - a strictly harder generative readout) provides a complementary token-level risk trajectory. An input-representation ablation confirms the lift transfers across structured discretizations, positioning KBTA-based intervals as the interpretable substrate that makes per-token attributions resolve to meaningful clinical concepts within the deployed model.
[AI-91] Perceive to Hypothesize Verify to Ground: An Agent ic Reasoning Framework for Open-World Geo-Localization
链接: https://arxiv.org/abs/2608.29880
作者: Yutian Jiang,Ruijie Li,Sisuo Lyu,Xixuan Hao,Qingxiang Liu,Yongzi Yu,Yuxuan Liang
类目: Artificial Intelligence (cs.AI); Multimedia (cs.MM)
备注:
Abstract:Open-world geo-localization requires models to reason over ambiguous visual cues through multi-step reasoning and external knowledge grounding. While recent large vision-language models exhibit strong multimodal reasoning capabilities, existing approaches still suffer from perceptual hallucination and context drift due to the lack of explicit evidence-grounded verification. In this work, we reformulate geo-localization as a human-like perceive-then-verify reasoning problem and propose GeoPAVE (Geo-localization Perception-and-Verification-Engine), a bi-level agentic framework that contains perception-based hypothesis generation via single-pass rollouts and verification-based evidence grounding for decision actions: support, refute, and refine. To support rigorous evaluation, we further introduce PAVED, a novel dataset derived from real-world user check-in data, equipped with comprehensive reasoning trajectories featuring multi-hop queries, multi-round tool invocations, and structured perception-verification traces. The dataset and code are available at this https URL.
[AI-92] FRAMEWORKERS: A Dynamic Multi-Agent Framework for AI-Generated Video Production
链接: https://arxiv.org/abs/2608.29814
作者: Zhendong Li,Lei Sun,Letian Shi,Deheng Zhang,Ruibo Ming,Mengshun Hu,Dannong Xu,Jian Wang,Danda Paudel,Luc Van Gool,Jinjin Gu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Modern video generators excel at synthesizing individual clips, but complete video production requires coordinating a long sequence of interdependent creative steps, including scripting, storyboarding, generation, and editing. It further demands persistent asset management and dynamic task orchestration as intermediate outputs, dependencies, and execution states evolve over time. Existing automated systems typically rely on rigid pipelines that are difficult to adapt to diverse inputs and changing workflows, while general-purpose large language models (LLMs) remain unreliable for long-horizon orchestration and multimodal asset routing. We introduce FRAMEWORKERS, a task-centric and workspace-grounded multi-agent framework for open-ended video production. A central Director formulates video creation as dynamic task management, continuously editing a Task Stack to determine which subtask to execute next and which sub-agent to invoke. An Assistant serves as the execution layer, grounding each selected task in a shared Workspace, retrieving the required assets and context, invoking the assigned sub-agent, and persisting the resulting artifacts. Execution capabilities are exposed through modular sub-agents with registered descriptors, allowing new sub-agents to be integrated without redesigning the orchestration workflow. To improve orchestration reliability, we fine-tune the Director via supervised fine-tuning (SFT) followed by Group Relative Policy Optimization (GRPO) for descriptor-conditioned task routing. Experiments show that FRAMEWORKERS outperforms strong LLM planners in routing accuracy, recovers reliably from runtime failures, generalizes to unseen sub-agents without retraining, and achieves higher end-to-end video quality and broader task coverage than fixed pipelines, single-agent systems, and prior multi-agent approaches.
[AI-93] PAGE-RAG : Provenance-Aware Graph Evidence Promotion for Fixed-Budget Multi-hop Retrieval-Augmented Generation
链接: https://arxiv.org/abs/2608.29753
作者: Haokun Deng,Xunkai Li,Hongchao Qin,Rong-Hua Li
类目: Artificial Intelligence (cs.AI)
备注: 13 pages, 3 figures
Abstract:Multi-hop question answering in retrieval-augmented gener?ation (RAG) often benefits from retrieving beyond the few candidates that will finally be read: narrow retrieval can miss an indispensable hop, while expanded retrieval introduces topical distractors. This challenge is not tied to a particu?lar knowledge-base format. Candidate pools may come from standalone retrievers, standard RAG backends, or graph-based retrieval pipelines. What is needed is a query-aware selection layer that can use relational structure to filter candidates be?fore generation. PAGE-RAG addresses this setting by using a graph as a temporary selection structure, rather than assum?ing a graph-structured knowledge base. It builds a query-local graph over retrieved candidates, records why candidates are connected, and treats each connection as a support hypothe?sis rather than support itself. We identify the resulting failure mode as a connectivity-support gap: connected candidates do not necessarily support the answer. We propose PAGE-RAG, a Provenance-Aware Graph Evidence promotion method that scores candidate paths with relevance, source-tracing meta?data, specificity, hubness, noise, and coherence signals, and applies minimal sufficient selection to promote supporting facts into a compact reader context. PAGE-RAG can serve as a complete retrieval-to-reading pipeline, and the same promo?tion stage can be inserted after existing retrieval or RAG sys?tems without replacing their upstream retrieval logic. Across three multi-hop QA benchmarks under the same final bud?get, PAGE-RAG improves support F1 and answer F1 by 10.4 and 3.3 points on a weighted average over a strong retriever. As a plug-in, PAGE-RAG further improves all reported RAG backends, including reasoning-oriented, compression-based, graph-based, and document/chunk-level systems.
[AI-94] Ideation Arena: Evaluating LLM Generated Research Ideas with Battle-style Human Expert Assessment
链接: https://arxiv.org/abs/2608.29696
作者: Zhiyu Chen,Keyu Zhao,Jigao Fu,Dong Liang,Yanbiao Wu,Jiaoyang Li,Haidong Xue,Xinhua Zeng,Yuanyi Zhen,Fengli Xu,Yong Li
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Evaluating research ideas generated by LLMs is difficult because their scientific value cannot be fully determined by objective criteria, and no single reference answer specifies what counts as a good idea. To address this challenge, we introduce Ideation Arena, a battle style platform that evaluates research ideas through pairwise human assessment. Ideation Arena evaluates ideas generated by 14 frontier LLMs and 5 research agent architectures built on 2 base models. To ensure a common starting point, Ideation Arena builds shared literature contexts from papers familiar to the participating researchers and provides the same contexts to all LLMs and agents. We collect over 6,000 double blind pairwise comparisons from 105 active computer science researchers and construct an Elo rating leaderboard of proposal-stage expert preferences in computer science under a shared closed-context protocol. We validate the rankings through interrater agreement and robustness analyses, showing that the leaderboard remains stable under changes in annotator composition and domain coverage. Our results show substantial variation in agent effectiveness, with some frameworks improving ideation quality over their backbones and others offering little benefit or even underperforming their base models. We further construct Ideation Arena Eval, a benchmark for assessing whether automated evaluators align with human preferences in research ideation. Experiments with current LLM judges show that they still cannot reliably reproduce expert preferences, with the best judge reaching 72.56% Soft Accuracy on Overall Quality. Our code, data, and leaderboards are available at this https URL.
[AI-95] Cost-Effective Repository Exploration for Agent ic Issue Localization
链接: https://arxiv.org/abs/2608.29675
作者: Mohammad Nour Al Awad,Sergey Ivanov
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:
Abstract:Repository exploration is a distinct and costly stage of coding-agent pipelines: before generating a patch, an agent must identify which repository files are likely to matter. We study whether this stage can be delegated to lower-cost models while retaining useful localization quality. Using our IssueLoc-Bench, we evaluate five explorer models under the same read-only interactive interface on 499 SWE-bench Verified-derived tasks and 500 tasks from 153 additional repositories. We measure early candidate discovery, top-three gold-file coverage, strict file-set recovery, agent time, and token usage, with paired instance-level uncertainty and repository-clustered sensitivity analysis. The highest-quality explorer leads across localization metrics, but substantially cheaper operating points emerge: depending on the model and evaluation arm, lower-cost explorers retain approximately 78-94% of the reference Hit@3 and 73-92% of its F1 while reducing mean agent time by 41-88% and token usage by 84-95%. The preferred operating point depends on how localization is consumed downstream: ranking and coverage metrics characterize recoverable candidate handoffs, whereas F1 and exact match characterize restrictive file gates. These results support treating repository exploration as an independently measurable and budgetable stage of modular coding agents, with explorer selection guided by the downstream handoff contract.
[AI-96] LLM ODE: Aligning ODEs with LLM s via Gated Token Injection for Irregular Spatio-Temporal Forecasting
链接: https://arxiv.org/abs/2608.29640
作者: Di Zhang,Jingyang Zhang,Ziqian Wang,Chi Zhang,Yikun Ban,Ziwei Zhang,Ruijie Wang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models (LLMs) have shown promise for spatio-temporal forecasting, but existing approaches often rely on regularly sampled token sequences and struggle with irregular observations because of temporal asynchrony, representation-space misalignment, and limited context windows. We propose LLMODE, a token-efficient framework for irregular spatio-temporal forecasting with a frozen LLM backbone. LLMODE first uses a graph-aware ODE encoder to reconstruct irregular graph observations as a continuous-time latent trajectory. A Fixed-Budget Perceiver Resampler then compresses this variable-length trajectory into a fixed number of dynamic memory tokens. In parallel, compact statistical descriptors are encoded and resampled into context memory tokens. A dual-source gated cross-attention module injects both memories into the frozen LLM, enabling controlled utilization of external spatio-temporal evidence. Experiments on three real-world urban datasets and two physical-dynamics benchmarks show competitive overall performance, with clearer advantages under sparse or dynamically complex irregular sampling. Additional evaluations on unseen urban regions further demonstrate strong zero-shot generalization without adaptation.
[AI-97] Wide Learning: Learning to Reach Evidence
链接: https://arxiv.org/abs/2608.29608
作者: Junzhou Chen
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Machine learning is usually evaluated after an evidence interface has been fixed. A dataset, sensor suite, query language, action set, or experimental protocol determines which observations can be obtained, and learning is judged by what it extracts from them. We study a complementary capability. A learner’s state can determine which evidence-generating experiments it can reliably realise under bounded resources, even when primitive affordances remain fixed. We call this learner-relative experiment family its effective epistemic reach, and use Wide Learning for task-relevant learning-induced changes in that this http URL formalise effective reach relative to learner state, deployment budget, reliability threshold, and evaluation distribution. In a controlled construction, two hidden worlds have exactly the same public observation law. An informative diagnostic exists in a fixed five-primitive substrate. Before calibration, one address attempt realises it with probability at most 2^-10 = 1/1024 , below a pre-specified 0.95 threshold; after calibration, held-out realisation is 1. Public-channel total variation is 0, whereas the realised diagnostic has total variation 1, and sealed binary risk moves from approximately 1/2 to 0. The construction establishes that learning can change effective epistemic reach even when primitive affordances and deployment resources are held fixed. It opens a complementary evaluation question for learning systems: not only what they infer from available evidence, but what informative evidence experience teaches them to bring within reach.
[AI-98] Not Safe for All: Auditing the Dialect Penalty in Text-to-Image Safety Pipelines EMNLP2026
链接: https://arxiv.org/abs/2608.29589
作者: Minkyu Kim,Juhwan Choi,YoungBin Kim
类目: Artificial Intelligence (cs.AI)
备注: EMNLP 2026 Findings
Abstract:Text-to-image (T2I) safety guardrails fail to generalize equitably to non-standard dialects. Evaluating 23,080 paired prompts across five English dialects, we formalize this failure as the dialect penalty, where filters trigger based on linguistic surface features rather than semantic intent. Text-level filters fail in opposing directions: NSFW-T over-flags benign dialect prompts and LatentGuard over-flags toxic ones (bias gaps up to +28.29 pp), while the OpenAI Moderation API under-detects them. A controlled typo ablation confirms this penalty originates from flagging dialectal features, not generic out-of-distribution sensitivity. The pixel-level generator is largely dialect-agnostic; the penalty enters at text processing and cascades unevenly to post-hoc guardrails. We show this bias tracks training data imbalance and is mitigable via group-balanced retraining, with an ablation attributing the gain to balanced exposure rather than to the worst-group objective of GroupDRO (group distributionally robust optimization). Current pipelines systematically fail dialect speakers, an equity failure masked by mean accuracy benchmarks. Our official code and dataset are publicly available at this https URL. Content Warning: This paper contains offensive, toxic, or disturbing text prompts and generated images. Comments: EMNLP 2026 Findings Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2608.29589 [cs.AI] (or arXiv:2608.29589v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.29589 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-99] PhysWave: Physics-Guided Latent Diffusion Models for Controllable Spatial Audio Generation EMNLP2026
链接: https://arxiv.org/abs/2608.29549
作者: Lingfeng Yao,Chenpei Huang,Xingke Yang,Ziye Geng,Changqing Luo,Hao Wang,Jiang Liu,Miao Pan
类目: ound (cs.SD); Artificial Intelligence (cs.AI); Multimedia (cs.MM)
备注: Accepted by EMNLP 2026 Main Conference. Project website: this https URL
Abstract:Text-to-spatial audio generation, such as text-to-First-Order Ambisonics (FOA), provides a convenient way to create spatial audio for billion-dollar gaming and film industries. However, existing text-to-FOA methods are largely data-driven and may produce audio that violates acoustic relations between source direction and distance. They also separate descriptive and parametric control, forcing users to trade usability for precision. In this paper, we present PhysWave, a physics-guided latent diffusion model for controllable text-to-FOA generation. PhysWave unifies natural-language and trajectory control through a shared waypoint-caption representation, and augments diffusion training with two differentiable acoustic priors: spherical-harmonic direction consistency and inverse-square distance consistency. To support dynamic spatial generation, we further construct a 300K-clip FOA dataset with diverse sound categories and source trajectories. Extensive results show that the proposed priors help PhysWave generate spatially consistent FOA audio while maintaining competitive audio quality. Further analyses show that these physics priors improve spatial consistency during training and can also be used as inference-time guidance for training-free spatial refinement.
[AI-100] AGM: Achievement-Grounded Memory for Closed-Loop Agents with Frozen VLA Policies
链接: https://arxiv.org/abs/2608.29537
作者: Hongbo Gao,Zeyu Ni,Xin Wen,Siyu Xu,Ruifeng Li
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: 19 pages, 9 figures
Abstract:Frozen vision-language-action (VLA) policies offer broad manipulation skills but execute open-loop action chunks without tracking task progress, so the agent cannot reliably decide whether to continue, retry, or terminate. External memory is a natural remedy, yet it can be harmful when attempted actions are treated as completed progress, turning local execution errors into persistent task-state errors. We propose Achievement-Grounded Memory (AGM), a lightweight closed-loop framework for frozen VLA policies that represents a task as a subgoal sequence with a progress pointer and advances this memory only after the current subgoal is verified by physical evidence. Proprioceptive interaction cues decide when to verify, while coherent point tracking and language-conditioned cross-view comparison, sourced from frozen foundation models through a single 2.43M-parameter verification head, decide what was achieved. AGM thereby converts open-loop execution into a closed loop of execution, verification, and progress, keeping the policy frozen without test-time large-model inference. On the RoboMME Counting benchmark, AGM reaches on PickXTimes and on BinFill, surpassing the strongest memory-augmented baseline by points on average, and the framework yields equally decisive gains on a physical robot. Reliable embodied memory thus depends more on disciplined state updates than on memory capacity.
[AI-101] On the Plasticity Collapse in Continual Machine Unlearning
链接: https://arxiv.org/abs/2608.29513
作者: Yingdan Shi,Xiang Xu,Kaize Ding,Alfred O. Hero,Ren Wang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Machine unlearning enables deep neural networks to selectively remove the influence of specific data in response to privacy and regulatory requirements. While prior work largely studies single-shot unlearning, real-world systems must accommodate continual unlearning, where multiple unlearning requests occur sequentially over time. In this work, we identify a fundamental limitation of this setting: plasticity collapse, a progressive breakdown in a model’s ability to effectively forget. Through theoretical analysis of continual unlearning dynamics, we show that continual unlearning operations accumulate geometric constraints in parameter space, leading to saturated subspaces that restrict future updates. This structural effect induces two distinct failure modes: (1) Forward failure – diminishing forgetting quality for subsequent tasks, and (2) Backward failure – spontaneous re-memorization of previously forgotten information. Extensive experiments across multiple architectures, datasets, and methods in image classification confirm that plasticity collapse is not an artifact of specific implementations, but a pervasive phenomenon inherent to continual unlearning. Our findings reveal a critical barrier to the long-term reliability of machine unlearning systems and motivate the development of plasticity-preserving unlearning algorithms. Our code is available at this https URL
[AI-102] Denoising as Projection: Constrained Optimization with Gradient-Guided Diffusion
链接: https://arxiv.org/abs/2608.29507
作者: Runyu Zhang,Jiawei Zhang,Gioele Zardini,Saurabh Amin,Asuman Ozdaglar
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Optimization and Control (math.OC)
备注:
Abstract:Diffusion models are increasingly used not only for sampling from learned data distributions, but also for generating samples that optimize task-specific objectives. A common approach is to guide the reverse diffusion process using gradients of an external objective. However, when the data distribution is supported on a structured feasible set, such as a manifold or a constraint set, gradient guidance can move samples away from the learned data geometry. In this paper, we study a simple projected-gradient-guided diffusion update based on the observation that the Stein denoising operator can act as an approximate projection onto the data geometry. The proposed update incorporates the objective gradient inside the denoising step, yielding an inference-time method that uses only a pretrained denoiser and gradient evaluations. We analyze this update as an inexact projected-gradient method for constrained optimization over learned feasible geometries. Our theory covers three settings: linear manifolds, compact convex feasible sets, and compact Riemannian submanifolds. In all these settings, we prove descent and finite-time convergence guarantees. Numerical experiments support the theoretical interpretation and illustrate how the proposed update balances objective descent with preservation of the learned geometry.
[AI-103] Applications of Risk Science to AI Fairness Evaluation: Principles Challenges and Best Practices
链接: https://arxiv.org/abs/2608.29478
作者: Kyra Wilson,Sabrina Kang,Saloni Dash,Aylin Caliskan
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI)
备注: Accepted at AIES 2026
Abstract:Scholarly work which aims to describe potential societal impacts (e.g., risks) of proliferating technology (especially related to artificial intelligence or other algorithmic systems) is likely to have an impact beyond the scientific communities it was written for, given that general society itself is a primary object of study. However, it is an open question whether the current practices of AI evaluation scholarship follow the principles and best practices established by risk science, which aims to systematically generate knowledge related to understanding, assessing, communicating, managing, and governing risk. In this work, we examine this in depth by conducting a literature review of scholarly works purporting to evaluate the bias or fairness of technological systems used for tasks related to hiring and employment. Through analysis of 22 common fairness evaluation metrics and studies using them, we find that most characterize the severity of bias- or fairness-related consequences but do not follow best practices to characterize the uncertainty around either the occurrence of these consequences or severity estimates. Next, we conduct a case study of fairness evaluation for an AI-mediated resume screening task and demonstrate how principles of risk science can be incorporated into such an evaluation. Finally, we propose the AI Risk Report Card, which facilitates the reporting and communication of risk assessment results to stakeholders in positions to act based on the predicted risks. The outcomes of these activities suggest that further research at the convergence of risk science and AI evaluation can lead to advancements in AI assessments of societal impact by enabling shared frameworks to evaluate and discuss AI risks both within and outside of the scientific community.
[AI-104] Knowledge Distillation under Teacher Misspecification: An Order-Parameter Analysis of the Gap between Teacher Mimicry and Task Performance
链接: https://arxiv.org/abs/2608.29472
作者: Kazuyuki Hara,Hideitsu Hino
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Knowledge distillation trains a small student model to reproduce the outputs of a large teacher model, and its progress is typically monitored through the teacher–student discrepancy. The quantity of ultimate interest, however, is the student’s error with respect to the true task. We study the relation between these two objectives in a minimal three-party model, a true teacher (generative model), a teacher, and a student, all soft committee machines, in which the true teacher contains a shared latent factor that the teacher cannot represent, with mismatch strength controlled by a single scalar \dmiss . Within an order-parameter description of online distillation, and exploiting closed-form (arcsine-type) expressions for all errors under error-function activations, we prove that the learning dynamics and the distillation error \Ets are exactly invariant to \dmiss , whereas the true error \Etzs and the gap \Delta=\Etzs-\Ets are strictly increasing in \dmiss , with a rate that is amplified linearly by the complexity M_0 of the true teacher. Numerical phase diagrams over the plane spanned by true-teacher complexity and student capacity confirm the predicted deformation: the contours of \Ets do not move while the landscape of \Etzs rises systematically, and a teacher-miss regime, where mimicry succeeds but the task fails, expands with \dmiss . The results give a quantitative warning against evaluating distillation solely through teacher-mimicry metrics and identify the gap \Delta as a minimal diagnostic for distinguishing teacher-miss from capacity-limited failure.
[AI-105] Can escalation channels redirect reward hacking toward defect disclosure?
链接: https://arxiv.org/abs/2608.29460
作者: Francesca Gomez
类目: Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR); Computers and Society (cs.CY)
备注: 9 pages
Abstract:When coding agents encounter defective test infrastructure they may reward-hack: hardcoding outputs or editing test files to pass tests they cannot legitimately satisfy, a pattern that has now appeared outside benchmarks, in a coordinated multi-agent intrusion of a major AI platform’s production infrastructure. The same capability that lets an agent detect and exploit a defect could let it report one, given the right decision environment. We evaluate escalation channels, structured reporting tools available to the agent at the point of conflict, as a decision-environment intervention that both reduces reward hacking and surfaces the infrastructure defects that trigger it. A 2 \times 2 factorial separates the contributions of an escalation tool, a standalone anti-reward-hacking policy, and their combination. Across 8 frontier models spanning 5 families, the combined intervention reduces reward hacking from 23.6% to 5.3% (mixed-effects logistic OR = 9.2, 95% CI 5.0–16.8, p 10^-12 ) with no detectable cost or performance overhead, eliminating it entirely for 6 of 8 models. Escalation and hacking are near-perfectly mutually exclusive, with 96.8% of escalations involving no hacking. Beyond reduction, escalation channels function as diagnostic infrastructure: on top of monitoring, escalation adds +10.1 percentage points of defect detection coverage and is more accurate once it fires (99.4% vs 85.8%). Unlike containment-based approaches that risk outpacing growing model capabilities, escalation channels redirect capability toward disclosure rather than exploitation.
[AI-106] oward Latent Language Model Skills Steering and Optimization: An Empirical Study
链接: https://arxiv.org/abs/2608.29459
作者: Xunyi Jiang,Junda Wu,Yuxin Xiong,Sheldon Yu,Tong Yu,David Arbour,Ritwik Sinha,Julian McAuley,Hongyi Wen
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Skills, as a useful abstraction for the procedural capabilities of large language models (LLMs), capture how models perform structured, multi-step reasoning and program execution. Existing approaches typically treat skills as explicit, surface-level constructs specified through prompts or programs, leaving open the question of how such procedural capabilities are represented inside the model and whether they can be manipulated as structured objects in latent space. In this empirical study, we investigate whether procedural LLM skills can be represented as directions in activation space and whether vector-space operations over these directions can express skill-level behaviors. We find that procedural skills admit a vector-space representation: individual skill directions can be activated to shift model behavior; independently extracted directions can compose to form higher-level skills. Contrastive directions yield context-conditioned algorithmic personalization and optimization trajectories over skill directions evolve non-monotonically, with intermediate states often surpassing fully optimized solutions. These results support a representation-level view of procedural LLM skills: they admit a latent vector-space organization that allows direct manipulation through internal interventions.
[AI-107] Reference-Grafting Matches Fine-Tuning at Eliciting Sandbagged Capabilities
链接: https://arxiv.org/abs/2608.29458
作者: Linh Le,Hong Kiat Tan,David Williams-King
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: under submission
Abstract:Sandbagging, in which a model deliberately underperforms on an evaluation despite retaining the underlying capability, threatens the safety evaluations that frontier-model governance depends on. The Elicitation Game found that fine-tuning elicits hidden capability from sandbagging model organisms whereas additive activation steering fails. We revisit that verdict with reference-grafting, which sets an activation’s coordinate along a contrast direction to the value it takes in an honest reference, at a small set of circuits chosen by active learning. Across eleven fine-tuned password-locked models (1.5-32B, three architecture lineages), it recovers +94 to +101% of the honest-sandbagging gap, matching fine-tuning elicitation without weight updates or training labels; two to five paired examples suffice to fit the direction. Similar recovery holds for reinforcement-learning-induced sandbagging and for password-locked code generation. Grafting works because the fine-tuned lock is a thresholded gate: held-out accuracy stays at the sandbagged level until the grafted coordinate crosses a threshold near the honest reference, which is why additive steering and zeroing the coordinate often fail. The direction tracks how the sandbagging was induced rather than what is withheld – fit on grade-school science (ARC) it elicits withheld hazardous knowledge (WMDP), yet prompting, fine-tuning, and reinforcement learning each carry a different direction. Circuit-breaking marks the boundary: it reroutes activations on every forward pass, so the fixed edits we test are re-broken downstream and do not restore coherent generation.
[AI-108] SS-ESOAP: Self-Scaled Adaptive Preconditioning for Physics-Informed Learning NEURIPS2026
链接: https://arxiv.org/abs/2608.29448
作者: Guangyuan Wang,Mads Toftrup,Sebastian Loeschcke,Yixuan Wang,Anima Anandkumar
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Optimization and Control (math.OC); Machine Learning (stat.ML)
备注: 31 pages, 12 figures, 16 tables. Submitted to NeurIPS 2026
Abstract:Physics-informed neural networks (PINNs) often face ill-conditioned objectives that limit high-accuracy training. Dense quasi-Newton methods improve local conditioning but require expensive optimizer state, while Kronecker-factored methods such as SOAP scale to larger networks but rely on periodic basis updates. We introduce \method, which augments SOAP-style preconditioning with a scalar secant-energy correction adapted to Kronecker geometry and an adaptive basis update followed by variance-state downscaling. We characterize the directional secant matching induced by the scalar correction and give a bound on variance-state mismatch across basis changes. Across eight PDE benchmarks, \method attains the lowest final residual on six, including Burgers and Boussinesq, while SOAP-family baselines perform better on Gray-Scott and Ginzburg-Landau. On Boussinesq, \method reaches a residual of 10^-5 in 4.1 hours with 9.2 GB peak VRAM, while Adam does not reach this target within 14 hours. Three-seed L^2 and H^1 errors on four representative PDEs support the link between lower residuals and improved solution accuracy. These results position \method as a scalable option for stiff, high-accuracy physics-informed training, rather than a uniform replacement for existing optimizers.
[AI-109] Scalable Clinical Data Infrastructure and Comparative ML Evaluation for Hospitalisation Risk Prediction in Elderly Patients with Multiple Long-Term Conditions using CPRD
链接: https://arxiv.org/abs/2608.29419
作者: Asra Aslam,Volodymyr Chapman,Maurice M. O’Connell,Aseel S. Abuzour,Michael Abaho,Danushka Bollegala,Gary Leeming,Eduard Shantsila,Andrew Clegg,Lauren E. Walker,Iain Edward Buchan,Samuel D. Relton
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Deep learning architectures are increasingly proposed for patient trajectory modeling in electronic health records (EHRs), yet their advantage over simpler, more interpretable models is rarely subjected to rigorous empirical scrutiny in real-world clinical settings. We present a comprehensive patient timeline pipeline applied to elderly patients in CPRD Aurum, incorporating 260 clinical conditions classified via a three-tier automated framework including specialised detection logic for 17 complex conditions. Using this infrastructure, we benchmark Temporal Graph Convolutional Neural Networks (TG-CNN) against Logistic Regression with LASSO regularisation and Random Forests for predicting 12-month all-cause emergency hospitalisation risk, motivated by (but not filtered to) the elevated risk of adverse drug reactions. Under cross-validation, TG-CNN achieves a marginally higher mean AUC-ROC than LASSO (0.712 vs. 0.705), whereas on the held-out test set LASSO achieves the highest discrimination of three models (AUC-ROC 0.733, versus 0.710 for Random Forest and 0.702 for TG-CNN). We show, that discrimination alone is an incomplete criterion for clinical deployment: after Platt calibration, LASSO is the only model with an acceptable calibration slope (0.817), while Random Forest (0.759) and, TG-CNN (0.391) remain substantially miscalibrated. We argue that LASSO, not the highest-discriminating model, is the model best suited to direct clinical deployment. We present lessons for the machine learning and healthcare community regarding data infrastructure, model selection, and value of calibration and interpretability in high-stakes decision support.
[AI-110] EvoGenUI-Bench: Evaluating LLM s as Multi-Turn Generative UI Assistants
链接: https://arxiv.org/abs/2608.29387
作者: Yue Peng,Lanke Xia,Zihan Wang,Jiahao Ye,Ke Ning,Hongyi Wen
类目: Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
备注:
Abstract:Large language models can generate interactive web interfaces, but reliable generative UI requires maintaining an executable artifact as user requests evolve. We introduce EvoGenUI-Bench, a benchmark for multi-turn interface maintenance comprising 150 five-turn tasks and 750 turns across three scenarios: information presentation, executable interaction, and tool-grounded external state. We execute generated artifacts in a browser and evaluate them using screenshots, source and DOM evidence, actor traces, and runtime logs. Beyond turn-level and episode-level success, we measure cross-turn retention with Adjacent Pass Retention. Across eight models, even the strongest achieves 74.9% Turn Pass while completing only 37.3% of five-turn episodes; APR further falls to 52.4% on tool-grounded tasks. Diagnostic analysis shows that presentation failures center on information architecture, interaction failures on derived-state propagation and affordance binding, and tool-grounded failures additionally involve external-state grounding and requirement decomposition. These results reframe generative UI evaluation from judging isolated outputs to testing whether interface behavior, derived state, external state, and assistant claims remain synchronized as the artifact evolves.
[AI-111] Safe to Resume? Breaking Execution Continuity of Agent Execution via Rollback
链接: https://arxiv.org/abs/2608.29381
作者: Guanlong Wu,Dahui Li,Ke Jiang,Jianyu Niu,Cong Wang,Yinqian Zhang
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:AI agents are moving toward persistent, stateful execution across various applications, accumulating execution state and external effects that are costly to reconstruct after failures. Checkpoint and rollback (C/R) are becoming essential for recovery, yet their security implications remain largely unexplored. Correct rollback does not imply secure recovery: a faithfully restored checkpoint may resume an execution whose states, assumptions, and external effects never coexisted in any valid history. In this paper, we present the first systematic security study of checkpoint and rollback in existing agent systems. By examining representative agent C/R systems, we characterize the design space of existing C/R mechanisms and develop a general execution model that captures their recovery boundaries and state dependencies. From this model, we identify five fundamental failure modes spanning incomplete or inconsistent internal state, stale external dependencies, nondeterministic replay, and unrecorded external effects. We further demonstrate their security impact through three end-to-end attacks on Hermes, Cline, and LangGraph, enabling malware-verification bypass, unauthorized mail forwarding, and double payment. To systematically study these failures in practice, we develop a multi-agent analysis pipeline that reconstructs execution semantics, identifies violations of the five failure conditions, and validates them through actual rollback. Across five representative frameworks, our evaluation shows that these failures recur across heterogeneous C/R designs and stem from a common gap between the state restored by a checkpoint and the dependencies required for secure continuation.
[AI-112] Evaluating Tiny Recursive Models Across Training for Code Generation
链接: https://arxiv.org/abs/2608.29376
作者: Anjani Sirivella,Aanisha Newaz,Glaucia Melo
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Software Engineering (cs.SE)
备注: Accepted at CASCON 2026
Abstract:Code generation increasingly relies on large transformer models, whose capability advances with scale. Yet such a scale is costly, creating demand for small models, especially where data is limited. Recursive models address this by reusing a single block to add depth rather than stacking independent layers. Such models are typically evaluated by teacher-forced fit (next-token loss on ground-truth prefixes) or task accuracy, at a single checkpoint, whereas code is produced by free-running generation, where the model extends its own output. Whether a teacher-forced advantage survives free-running generation, and whether it holds across training, remains open. To study both, we compare a ~28M-parameter autoregressive Tiny Recursive Model (TRM-AR) on natural-language-to-Python code generation against parameter-matched and depth-matched controls, tracking fit and generation across 40 epochs and three seeds. The fit ranking between the recursive model and the depth-matched control reverses twice. Selecting each checkpoint by validation loss and examining the trajectory yields a consistent comparison. At equal parameters, TRM-AR fits, generates, and generalizes better than the parameter-matched control while recovering approximately 45% of the validation-loss gap and 57% of the generation-quality gap between the two controls, at roughly 175 times the per-step cost of the parameter-matched control. However, at equal effective depth, the larger transformer fits and generates better at its validation optimum, suggesting TRM-AR’s advantage lies in resistance to overfitting, not greater capability. These findings suggest that recursive code generation models should be evaluated jointly on fit and generation across the training trajectory rather than at a single checkpoint.
[AI-113] FORESIGHT-9: Prospective and Process-Aware Evaluation of Adaptive Trading Agents
链接: https://arxiv.org/abs/2608.29372
作者: Xiangxin Luo,Chengtian Hong,Haohua Li,Yongyi Xie
类目: Artificial Intelligence (cs.AI)
备注: 27 pages, 15 figures
Abstract:Retrospective backtests provide a limited test of adaptive trading agents: they cannot rule out historical contamination, expose sensitivity to a single realized market path, or reveal internal degeneration during long-horizon adaptation. We introduce FORESIGHT-9, a prospective and process-aware benchmark built from nine auditable counterfactual stress worldlines branching from a common July 2026 information boundary. Each worldline specifies staged macro-financial events and joint multi-asset anchors; a deterministic generator realizes the trajectories, while observations are disclosed according to in-world time. A common contract standardizes observations and execution while preserving each agent’s native adaptation loop. We evaluate two adaptive trading-agent frameworks with two foundation-model backbones across 36 long-horizon runs. Agent rankings vary substantially across worldlines and backbones, and a fixed equal-weight policy outperforms 31 of 36 runs. Process telemetry exposes failures that terminal returns conceal: in one high-return run, the live factor library collapsed while executed holdings converged to the equal-weight fallback, even though decision records continued to report an active factor ensemble. FORESIGHT-9 therefore evaluates not only portfolio outcomes, but whether adaptive agent state and execution remain coherent across alternative futures. We release the worldlines, trajectories, audit traces, and regeneration scripts.
[AI-114] Reviving our data foundations is the most disruptive step to data maturity
链接: https://arxiv.org/abs/2608.29368
作者: Valentina Carapella,Ernesto Jimenez-Ruiz
类目: Artificial Intelligence (cs.AI)
备注: 6 pages
Abstract:The most disruptive step that enterprises of small-medium size and maturity can take to make the most of the latest technological advances in AI is to step back from the hype and focus on establishing or reviving a good knowledge foundation layer. It is a hard message to present to the executive team; therefore, it needs to be backed by evidence, and its implementation needs to be of minimal impact on the existing processes. In this vision statement, we discuss how we need to rethink what evidence speaks to the decision-makers and propose a low-impact data strategy that adapts to the existing and ever-changing data flows and processes across the company. We firmly believe that knowledge graph techniques will increasingly become non-negotiable in the data strategy of an AI-powered enterprise, provided that we approach their design in a modular, dynamic and cross-functional way.
[AI-115] RACER: Per-Tool Context Retention for LLM Agents via Consequence-Attributed Reinforcement Learning
链接: https://arxiv.org/abs/2608.29363
作者: Ziqi Lin,Ye Wu,Mengying Yang,Xu Liu,Yizhou Liu,Qiang Ke,Qin Guo
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Enterprise data agents answer business queries by chaining many tool calls over multiple reasoning steps, routinely accumulating hundreds of thousands of context tokens per session. Existing compression strategies typically allocate retention budgets without accounting for the downstream consequences of removing individual tool outputs. Aggressive compression may therefore trigger costly tool re-invocations that offset the initial savings. We call this the compression–consequence gap. To close it, we propose TRACER, which formulates compression as a sequential per-tool decision problem. A lightweight REINFORCE policy assigns query-conditioned retention ratios using only information available at each compression event. Its consequence-aware objective jointly accounts for task success, total token consumption, and post-compression tool re-invocations. To improve credit assignment, TRACER uses a learned outcome model to compare the predicted consequences of the selected retention ratio with those of fully retaining each tool output. On held-out production queries across three compressor backends, TRACER reduces total token consumption by 29–46% relative to keeping all context while maintaining comparable or higher task success. Compared with a tool-type-conditional static policy, TRACER provides an additional 15–18% of token savings. Interventional rollouts show that the learned per-tool credit scores correlate with measured single-tool consequences. The learned policy also yields positive savings when transferred across agent backbones and compressor architectures, and reduces token consumption by 18–25% on five held-out LOCA-bench environments. These results demonstrate the value of consequence-aware, per-tool context retention for improving the efficiency of long-horizon language agents.
[AI-116] LiteSearch-VL: Small Multimodal Search Agents via Trajectory Distillation and Synthetic Step-DPO
链接: https://arxiv.org/abs/2608.29357
作者: Saeed Khaki,Nima Safaei,Kamal Ginotra
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Multimodal search agents answer visual questions by interleaving image understanding, web retrieval, tool use, and evidence synthesis. Strong systems exist, but in two expensive regimes: proprietary frontier models such as GPT-5 and Gemini, or large open vision-language backbones trained with substantial agentic data and reinforcement learning. We ask a different question: when released agent trajectories are distilled into much smaller backbones under a single-node budget, what is actually transferred? We study this with LiteSearch-VL, a low-compute recipe for Qwen3-VL-2B and Qwen3-VL-4B that uses only released OpenSearch-VL trajectories, parameter-efficient LoRA adapters, and synthetic step-level preferences: DPO on GPT-5-generated hard negatives targeting five local failure modes (premature answer, wrong tool, weak query, repeated query, ignored image). Across 12,400 GPT-5-judged rollouts on SimpleVQA, FVQA, LiveVQA, and VDR-Bench-testmini, the dominant effect is behavioral rather than a uniform accuracy lift: full-trajectory supervised fine-tuning transfers the agent contract, taking the 2B model from almost never emitting a usable answer (1,237/1,240 no_answer rollouts) to 28.4% macro Pass@1, matching or slightly exceeding the off-the-shelf 4B base (25.6%). Synthetic preference learning and compact tool distillation act as refinements rather than phase transitions (best 4B configuration: 30.8% macro Pass@1). Finally, a controlled VDR step-budget ablation shows that extra search turns convert abstentions into wrong_entity errors rather than correct answers, identifying answer verification, not search depth, as the next bottleneck for small multimodal agents.
[AI-117] Plant-Inspired AI: Plants as Inspiration for Novel Problem Formulations and Two Case Studies
链接: https://arxiv.org/abs/2608.29356
作者: Deepayan Sanyal,Joel Michelson,Carla E. Cao,Adam B. Roddy,Maithilee Kunda
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Artificial Intelligence (AI) has long been inspired by studies of biological intelligence. Reinforcement learning, for instance, drew inspiration from studies involving animal learning and is now a powerful paradigm for solving many real-world problems. Recently, plant biologists have uncovered a wide range of complex behaviors in plants that enable them to flexibly adapt to variable environments. Here, we argue that such behavior can motivate new AI frameworks encompassing a range of problems overlooked by existing problem-solving frameworks such as supervised learning, tree search, and constraint satisfaction. We illustrate this idea with two examples of intelligent problem-solving in plants: (1) leaf mimicry in Boquila trifoliolata, a vine capable of altering its leaves’ morphology to resemble those of multiple host trees simultaneously; and (2) coordinated root-shoot growth, wherein plants allocate resources across organ systems exploring distinct environments. While leaf mimicry is highly specific to Boquila, coordination of root-shoot growth is shared across most plants. For both examples, we capture underlying computational principles and identify problems fitting these frameworks that are currently unaddressed by AI. Finally, we outline preliminary task formulations and discuss how these formulations may be applied to non-plant problems.
[AI-118] APPSolver: Adaptive Patch Partitioning for Point-Wise Ship Flow Prediction on Unstructured Meshes
链接: https://arxiv.org/abs/2608.29355
作者: Wenhua Huo,Fenglei Han,Wangyuan Zhao,Xiao Peng,Chunhui Wang,Jialin Wu,Jiayi Han
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 14 pages, 2 figures. Code: this https URL
Abstract:Large non-uniform point sets make direct attention-based surrogate modeling costly for ship hydrodynamics. We introduce APPSolver, a point-wise flow-prediction framework built around Adaptive Patch Partitioning (APP), a deterministic quadtree representation for fixed two-dimensional horizontal slices extracted from ship CFD simulations. APP assigns finer patches near the hull and coarser patches farther away, downsamples patch contents, and recovers predictions to the full reference point set. Under a corrected protocol that constructs natural (t,t+1) pairs before splitting, reuses training-set normalization statistics, and reports three model seeds, learned tokenizers are more accurate than APP-Transformer, and a persistence baseline has lower one-step MAE on all three ShipBench hulls. The supported benefit of APP is therefore computational rather than universal predictive superiority: on a representative DTC input, APP-Transformer requires 1.815 GFLOPs and 1.309 ms per model forward, while a matched ablation shows that adaptive partitioning reduces MAE by 16.4-24.9% relative to a uniform partition augmented with learned slicing. Condition encoders provide setting-dependent gains in leave-one-hull-out evaluation, but the current absolute next-state objective does not establish accurate long-horizon dynamics. These results characterize APP as a compact spatial representation with an explicit accuracy–efficiency trade-off. Code is available at this https URL .
[AI-119] Cross-Relational Preference Learning for Better LLM Instruction Following
链接: https://arxiv.org/abs/2608.29352
作者: Runsheng Li,Kai Sun,Bin Shi,Bo Dong
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large Language Models (LLMs) still exhibit limited capability in following complex instructions. While existing approaches often rely on preference learning to enhance this ability, they typically overlook the relationships between the permissible response spaces of different instructions, which restricts a model to align with subtle and diverse constraint variations. To address this, we propose Cross-Relational Preference Learning (CRPL), a novel framework for constructing preference data that explicitly models inter-instruction relationships through two key techniques: Cross-Relationship Perturbation and Cross-Region Pair Sampling. This enables the generation of more diverse preference data that captures a wide spectrum of constraint variations. Additionally, we introduce an atomic constraint-based verification mechanism to rigorously assess response satisfaction, ensuring high-quality preference pair construction. Extensive experiments across multiple preference learning methods (e.g., DPO, KTO), LLM backbones and four instruction-following benchmarks demonstrate that our approach achieves substantial improvements over prior baselines and exhibits strong generalization.
[AI-120] Extending TotalSegmentator: Predicting Patient and Acquisition Characteristics from CT and MR Images
链接: https://arxiv.org/abs/2608.29348
作者: Jakob Wasserthal,Joshy Cyriac,Michael Bach,Kimia Mozahheb Yousefi,Minh-Son To,Máté Sik,Cédric Hémon,Thomas Weikert,Martin Segeroth
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Background: Patient details and acquisition metadata are important for clinical decisions, image quality control, and automated research pipelines, but may be missing or unreliable in imaging archives. Purpose: To develop and evaluate a fast open-source model that predicts patient and acquisition characteristics directly from CT and MR images. Materials and Methods: Separate 3D ResNet-10 ensembles for CT and MR were trained on 57,291 and 43,200 clinical examinations acquired from 2011 to 2025. Both predicted weight, height, age, sex, contrast presence, vertebral coverage, and image noise. The CT model additionally predicted scanner manufacturer, tube voltage, tube current, convolution kernel, and post-injection time; the MR model predicted sequence class. Performance was evaluated on internal CT (n=501) and MR (n=636) test sets and an external CT dataset (n=54). Results: Internal CT MAEs were 3.90 kg, 3.68 cm, and 4.42 years for weight, height, and age, with sex F1=0.990; corresponding MR results were 4.34 kg, 4.62 cm, 7.13 years, and F1=0.970. The CNN outperformed a segmentation-derived XGBoost baseline for all four core targets in both modalities (adjusted P=.042). F1 scores were 0.963 for CT contrast, 0.953 for MR sequence, and 0.823 for MR contrast. External CT MAEs were 4.45 kg, 4.05 cm, and 5.17 years, with sex F1=0.971. CPU inference required 20 seconds for CT and 12 seconds for MR. Conclusion: One 3D multitask model per modality can rapidly recover patient and acquisition characteristics from heterogeneous CT and MR examinations. Models are available in TotalSegmentator: this https URL Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2608.29348 [cs.AI] (or arXiv:2608.29348v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.29348 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Jakob Wasserthal [view email] [v1] Sat, 29 Aug 2026 16:09:17 UTC (736 KB) Full-text links: Access Paper: View a PDF of the paper titled Extending TotalSegmentator: Predicting Patient and Acquisition Characteristics from CT and MR Images, by Jakob Wasserthal and 8 other authorsView PDFHTML (experimental)TeX Source view license Current browse context: cs.AI prev | next new | recent | 2026-08 Change to browse by: cs References Citations NASA ADSGoogle Scholar Semantic Scholar export BibTeX citation Loading… BibTeX formatted citation loading… Data provided by: Bookmark checked="checked"class=“labs-tab-input”> Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, Media Code, Data and Media Associated with this Article alphaXiv Toggle alphaXiv (What is alphaXiv?) Links to Code Toggle CatalyzeX Code Finder for Papers (What is CatalyzeX?) DagsHub Toggle DagsHub (What is DagsHub?) GotitPub Toggle Gotit.pub (What is GotitPub?) Huggingface Toggle Hugging Face (What is Huggingface?) ScienceCast Toggle ScienceCast (What is ScienceCast?) Demos Demos Replicate Toggle Replicate (What is Replicate?) Spaces Toggle Hugging Face Spaces (What is Spaces?) Spaces Toggle TXYZ.AI (What is TXYZ.AI?) Related Papers Recommenders and Search Tools Link to Influence Flower Influence Flower (What are Influence Flowers?) Core recommender toggle CORE Recommender (What is CORE?) Author Venue Institution Topic About arXivLabs arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv’s community? Learn more about arXivLabs. Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?) mathjaxToggle(); We gratefully acknowledge support from our major funders, member institutions, , and all contributors. About Help Contact Subscribe Copyright Privacy Accessibility Operational Status (opens in new tab) Major funding support from
[AI-121] SGE: Semantically-Guided Exploration for Unstructured Environments via Image-Space Waypoint Sampling
链接: https://arxiv.org/abs/2608.29315
作者: Christopher Tatsch,Yu Gu
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:
Abstract:This work introduces Semantically-Guided Exploration (SGE), a modular exploration framework for ground vehicles that integrates pixel-level semantic segmentation into sampling-based waypoint selection and receding-horizon route optimization. Unlike conventional geometric exploration methods, SGE evaluates candidate exploration goals directly in the image space using a semantic-aware utility function that accounts for terrain traversability, obstacle proximity, objects of interest, and depth-based exploration reward. Sampled waypoints are projected into 3D and ordered through a real-time Traveling Salesman Problem (TSP) formulation, enabling receding-horizon goal selection. To address real-world navigation uncertainty, the framework introduces mechanisms, including temporary taboo regions to handle navigation failures and a graph-based relocation strategy for efficient backtracking across explored areas. We evaluate SGE in standardized simulation benchmarks against state-of-the-art exploration planners and demonstrate competitive performance in volumetric coverage, while enabling semantic task biasing that cannot be achieved by purely geometric methods. The framework is further validated through real-world experiments using multiple robotic platforms in indoor campus buildings and in limestone and coal mines. Results show consistent performance and adaptability across platforms and domains.
[AI-122] Formal Concept Analysis with Three Types of Negation
链接: https://arxiv.org/abs/2608.29311
作者: Zhenghua Pan
类目: Artificial Intelligence (cs.AI)
备注: 30 pages, 9 figures, 8 table, 25 conferences
Abstract:Classic Formal Concept Analysis (FCA) primarily focuses on the positive relationships between objects and attributes and does not have mechanisms for handling this http URL overcome this limitation, we introduce three types of negation concepts (contradictory negation, opposite negation, intermediary negation) into this http URL on the set SCOI and logic LCOI+PLCOI with these three types negation, we define formal context, Galois connection operators, formal concept and concept lattice with three types of negation,this leads to the proposal of a FCACOI: Formal Concept Analysis with contradictory negation, opposite negation and intermediary this http URL the reasoning in FCACOI, this paper focuses on attribute implication reasoning. Based on the logic LCOI+PLCOI and its semantics, we introduce the notion of ICOI-entailment as the semantic implication for attribute implication reasoning in FCACOI. Through ICOI-entailment, a connection is established between attribute implication reasoning in FCACOI and inference in the logic LCOI+PLCOI, it indicate that formally proven inference rules (theorems) in LCOI+PLCOI are valid in the attribute implication reasoning of FCACOI, LCOI+PLCOI provides a logical foundation for attribute implication reasoning in FCACOI. To illustrate the capability of attribute implication reasoning in FCACOI, we discuss its application in a concrete example. Moreover, we explore attribute reduction of the formal context in FCACOI, propose two research frameworks for attribute reduction from different perspectives, and compare their this http URL believe that, based on richer logic and semantics, FCACOI elevates FCA from a theory that describes affirmations to one that can describe affirmations and its contradiction(either this or that), opposition(extreme negation) and intermediary (transitional states between oppositions).
[AI-123] Improving Randomized Metric Distortion to 2.3282
链接: https://arxiv.org/abs/2608.29308
作者: Nisarg Shah
类目: Computer Science and Game Theory (cs.GT); Artificial Intelligence (cs.AI)
备注:
Abstract:In metric social choice, each voter ranks a set of m candidates by her distance to them in an unknown metric space. The cost of a candidate is its average distance to the voters. A randomized voting rule must use only the rankings to choose a lottery over candidates. Its distortion is the worst-case ratio between the expected cost under the lottery it returns and the cost of the best candidate. Charikar, Ramakrishnan, Wang, and Wu [JACM 2024] prove an upper bound of 2.753 , establishing a constant separation from deterministic rules, for which the best achievable distortion is 3 . Independently, Frank [arXiv:2608.17863] and Ye [arXiv:2608.21202] improve the bound to 2.5 , using an equal mixture of maximal lottery and Integrated Veto. The existing arguments do not yield a better bound with any mixture of these rules. We break this barrier with a new ingredient, a random-size stable lottery. Let D be a random variable over the domain of positive integers. A random-size stable lottery \mathrmRSL_D guarantees that the probability of a random voter preferring any fixed candidate c to her favorite of D i.i.d. draws from \mathrmRSL_D is at most \mathbbE[1/(D+1)] , where the probability also averages over D . When D=k deterministically, this reduces to the stable k -lottery of Charikar, Ramakrishnan, Tan, and Wang [EC 2025]; the case k=1 is precisely a maximal lottery. Their minimax argument for a fixed k easily generalizes to a random D . Our main contribution is to show how stability with respect to a random D can be used to bound distortion. By mixing a suitably chosen random-size stable lottery with Integrated Veto, we get distortion at most 11641/5000=2.3282 . The proof combines infinite-dimensional conic linear-programming duality, heuristic nonlinear optimization, and exact rational verification via the Bernstein basis. Subjects: Computer Science and Game Theory (cs.GT); Artificial Intelligence (cs.AI) Cite as: arXiv:2608.29308 [cs.GT] (or arXiv:2608.29308v1 [cs.GT] for this version) https://doi.org/10.48550/arXiv.2608.29308 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-124] Predicting Future Organ Dysfunction in ICU Patients Using Temporal Convolutional Networks on MIMIC-IV Data
链接: https://arxiv.org/abs/2608.29301
作者: Razan Albouq,Asra Aslam
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Predicting future organ dysfunction in Intensive Care Unit (ICU) patients is critical for early clinical intervention, yet existing machine learning approaches have largely treated the Sequential Organ Failure Assessment (SOFA) score as an input to binary mortality prediction rather than as a continuous clinical outcome in its own right. We investigate the extent to which a Temporal Convolutional Net work (TCN) can predict next-day SOFA scores from multivariate ICU time-series data extracted from MIMIC-IV, characterise the relative contribution of each organ system to total SOFA variance and deterioration, and identify distinct trajectory patterns across ICU stays. A residual TCN trained on three-day sliding windows achieved a five-fold cross-validation R2 of 0.740 ± 0.013 and MAE of 1.431 ± 0.022, outperforming a naive persistence baseline on RMSE and R2. SHAP interpretability analysis revealed that the model functions primarily as a severity-anchoring mechanism rather than a true sequence model, with predictions dominated almost entirely by the most recent observation day. Cardiovascular dysfunction emerged as the strongest discriminator of both cross-sectional severity and acute deterioration, and unsupervised trajectory clustering identified two clinically meaningful phenotypes, an improving group (58.9%) and a persistently severe group (41.1%), differentiated by cardiovascular, hepatic, coagulation, and renal involvement. We conclude that TCNs can extract meaningful predictive signal from ICU physiological data, but that short input windows and complete-case selection bias currently limit their clinical utility, motivating future work on longer input horizons, alternative missing-data strategies, and external validation.
[AI-125] When Do Larger Batches Help Scale LLM Reinforcement Learning?
链接: https://arxiv.org/abs/2608.29296
作者: Ziniu Li,Jinbo Wang,Guanhua Huang,Feiyuan Zhang,Pengbo Li,Alex Chen
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 16 pages, 9 figures
Abstract:Larger batches reduce the variance of stochastic gradients per update and are therefore often expected to accelerate training. Yet whether this statistical benefit translates into lower wall-clock time-to-target remains unclear, because each update consumes more samples and may take longer to execute. We study this tradeoff in reinforcement learning for large language models. We separate its algorithmic and systems effects by comparing learning and execution along their natural axes. At the algorithmic level, we compare configurations at equal cumulative sample counts while retuning batch-dependent hyperparameters. Over a bounded range of batch sizes, this procedure yields an approximately batch-size-invariant family whose members follow similar sample-indexed learning trajectories. At the systems level, we exploit the computational asymmetry between rollout generation and training: autoregressive generation is often memory-bandwidth-bound at low concurrency, whereas training work scales approximately with the number of processed tokens. Combining these two views yields a direct decision rule: a larger-batch configuration reduces time-to-target only when its throughput gain exceeds its samples-to-target penalty. Experiments with GRPO and PPO support both sides of this decomposition. At the algorithmic level, square-root learning-rate scaling with Adam produces approximately batch-size-invariant learning curves over a bounded range of batch sizes. At the systems level, larger batches improve generation throughput by up to 2.29x on fixed hardware. In GRPO, combining higher throughput with learning-rate retuning reduces time-to-target by up to 29%, whereas increasing the batch without retuning is slower despite its higher throughput.
[AI-126] Accelerating Unified Multimodal Models with Core-Expansion Routing and Unified Computation Scheduling
链接: https://arxiv.org/abs/2608.29291
作者: Wengyi Zhan,Chenqian Yan,Songwei Liu,Mingbao Lin,Rongrong Ji
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Unified multimodal models jointly support understanding and generation, but incur substantial redundant computation across tokens, layers, and generation timesteps. Through token-importance probing, we identify an asymmetric core-expansion structure: understanding exhibits a stable importance component, while generation largely shares this component but requires progress-dependent corrections. We therefore propose CE-Router, which uses a task-shared core scorer and progress-conditioned generation expansions, optimized through generation decomposition and cross-task core alignment. At inference, CE-Router compacts token computation and supplies a learned routing signal to Unified Computation Scheduling, which coordinates layer skipping, FFN pruning, diffusion-head cache reuse, and denoising-step early exit. Experiments on two representative UMM architectures demonstrate consistent quality–efficiency improvements across both tasks, retaining 98.03% of dense understanding performance with a 1.93 \times end-to-end inference speedup.
[AI-127] MMPCBench: Benchmarking Multimodal Large Language Models on Proactive Critique of Flawed Inputs
链接: https://arxiv.org/abs/2608.29286
作者: Jinzhe Li,Gengxu Li,Jinnan Li,Yuan Wu,Yi Chang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:As Multimodal Large Language Models (MLLMs) evolve into sophisticated interactive assistants, their reliability depends not only on following instructions but also on validating them. We define Proactive Critique as the model’s autonomous ability to identify, analyze and fix faulty user inputs without extra prompts. However, evaluations mainly test models under ideal circumstances or simple refusal behaviors, largely ignoring active error processing. To fill this gap, we propose MMPCBench, a comprehensive framework for evaluating MLLMs’ proactive critique competence. It features a fine-grained taxonomy of 4 primary error types spanning 12 subcategories, ranging from cross-modal contradictions to missing visual premises. We adopt a hierarchical evaluation protocol to measure models’ error detection, diagnosis and resolution performance, and apply alignment-aware metrics to assess the coherence between internal reasoning and final responses. Tests on 14 mainstream MLLMs show obvious weaknesses in proactive critique, especially in dealing with subtle visual anomalies. Notably, we identify a pervasive “consistency gap”: reasoning models can often correctly identify and analyze errors during internal reasoning yet suppress these valid insights in final outputs to prioritize response compliance. The code and data is available at this https URL.
[AI-128] Understanding Deep Learning via Entropy Space Theory
链接: https://arxiv.org/abs/2608.29279
作者: Li Li,Tong Zhang,Wentao Yu,Zuobin Wang
类目: Artificial Intelligence (cs.AI)
备注: 10 pages, 1 figure, 3 tables
Abstract:Deep learning is often criticized for its theoretical research lagging behind practice. To make deep learning easier to understand, the entropy space theory is first introduced here. The entropy space can cover all the possibilities of any deep learning model by topological structure. It is independent of network parameters. Through the designed fundamental operations and norm, entropy space is proven to be a normed space within the formal axiomatic framework. Based on the theory, a unified coordinate system is proposed. It can coordinatize every state of a model and rank them by compression of the maximal value of information entropy. The theory offers a novel priori framework for mathematical fundamentals of deep learning.
[AI-129] EpaCache: Error-Propagation-Aware Caching for Accelerating Diffusion-Based Visual Generation
链接: https://arxiv.org/abs/2608.29264
作者: Yuhan Liu,Zongwei Hong,Jinglun Li,Linze Li,Shen Zhang,Yao Tang
类目: Artificial Intelligence (cs.AI)
备注: This is a preview version of EpaCache, which is still under review
Abstract:Diffusion-based visual generative models deliver strong image and video synthesis quality but incur high inference costs because sequential samplers repeatedly evaluate large networks. Caching-based methods reduce inference latency by reusing intermediate computations across adjacent timesteps. However, existing cache controllers rely primarily on local temporal variation and overlook the trajectory-level consequences of cache reuse. We introduce Error-Propagation-Aware Cache (EpaCache), a training-free caching policy that adaptively allocates the reuse budget on timesteps with lower downstream impact. Experiments on image and video synthesis models demonstrate that EpaCache consistently improves the latency–fidelity trade-off over existing caching methods. On FLUX.1-dev, EpaCache outperforms the prior state-of-the-art caching method in both latency and fidelity, reducing inference time from 11.7 s to 11.3 s while improving PSNR from 21.4 to 22.8 . On HunyuanVideo, EpaCache achieves a 2.63\times speedup over uncached inference and improves SSIM from 0.891 to 0.905 over the prior state-of-the-art method at matched latency.
[AI-130] RACER: Reinforced Agent Collaboration for Explainable Reasoning on Knowledge Graphs ICONIP2026
链接: https://arxiv.org/abs/2608.29263
作者: Yuwei Lou,Hao Hu,Yuzhou Jiang,Zongfei Zhang,Liang Wang,Jincai Liu,Jidong Ge,Xianping Tao
类目: Artificial Intelligence (cs.AI)
备注: 15 pages, 1 figures, This paper has been accepted by ICONIP 2026
Abstract:Large Language Models (LLMs) often suffer from hallucination and struggle with complex reasoning tasks requiring multi-hop domain knowledge. While integrating Knowledge Graphs (KGs) provides a structured and verifiable information source, current KG-enhanced LLM paradigms usually rely on single-agent path extraction and fixed prompting, lacking adaptability and facing huge search spaces. To address these challenges, we propose RACER, a Reinforced Agent Collaboration framework for Explainable Reasoning on knowledge graphs. RACER employs a semantic-aware action pruning and teacher-guided reinforcement learning mechanism to efficiently extract high-quality reasoning pathways from large-scale KGs. Furthermore, to mitigate single-path generation pitfalls, we introduce a cross-task accumulated shared memory graph paired with an attention-driven multi-path knowledge refinement module. Finally, RACER orchestrates these components through a four-role multi-agent collaboration system (GraphAgent, TemplateAgent, AnswerAgent, and CriticAgent) to dynamically refine prompts and evaluate answers. Extensive experiments on CommonsenseQA and OpenBookQA datasets demonstrate that RACER significantly outperforms state-of-the-art KG-enhanced LLM baselines with an average improvement of 5%, offering robust and highly interpretable reasoning capabilities.
[AI-131] Adaptive Multi-Branching for Shallow Decision Tree Induction
链接: https://arxiv.org/abs/2608.29262
作者: Hanul Park,Jeonghoon Choi,Juseong Kim,Sanghun Sel,Giltae Song
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 9 pages, 1 pages for the appendix
Abstract:Decision trees are attractive for tabular prediction tasks because each prediction follows an interpretable sequence of feature-threshold tests. Under a strict maximum-depth budget, however, conventional binary trees can be under-expressive, since each internal node makes only a single threshold decision. We study shallow-depth tree induction, where the goal is to improve accuracy while keeping root-to-leaf paths short. We propose the Multi-Branch Neural Decision Tree with Adaptive Pruning (MBNDT), a single axis-aligned tree trained end-to-end with differentiable multi-way splits. Each internal node learns ordered thresholds over a selected feature and a branch mask that adapts its effective arity, and the trained model is converted to a deterministic single-path tree for inference. Across 21 OpenML binary-classification benchmarks, MBNDT achieves the best average rank and mean balanced accuracy among depth-constrained single-tree baselines; a controlled ablation isolates multi-way splitting as the source of the gain. These gains come with an explicit trade-off: MBNDT realizes more leaves than the other single-tree baselines, making it best suited when accuracy under short, bounded decision paths is prioritized over minimal global tree size.
[AI-132] Dynamic Important Example Mining for Reinforcement Finetuning
链接: https://arxiv.org/abs/2608.29252
作者: Haoru Tan,Sitong Wu,Yanfeng Chen,Shizhen Zhao,Yang-Tian Sun,Tianjia Liu,Chirui Chang,Shaofeng Zhang,Samm Sun,Xiuzhe Wu,Ruobing Xie,Xiaojuan Qi
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Reinforcement fine-tuning (RFT) is increasingly used to strengthen the reasoning abilities of large models, yet its effectiveness is bound by how training data are selected and used. Most data-centric RFT methods rely on static or heuristic sample selection, implicitly assuming a sample’s value is fixed over training. This overlooks the non-stationary dynamics of policy learning and can lead to suboptimal updates. We propose Dynamic Important Example Mining (DIEM), a principled and fully automated framework that makes data utilization adaptive throughout RFT. DIEM integrates two components into each optimization step: (i) a gradient-alignment importance estimator that efficiently approximates each sample’s marginal contribution to policy improvement; and (ii) a constrained batch reweighting scheme that maximizes aggregate utility while preserving the update’s gradient magnitude to stabilize optimization. Across several reasoning benchmarks, DIEM consistently outperforms strong static and dynamic baselines. The code will be released via this https URL.
[AI-133] Imag-Eval: a language-grounded framework for interpretable Text-to-Image instruction following evaluation EMNLP2026
链接: https://arxiv.org/abs/2608.29210
作者: Ibrahim Mohamed Serouis,David Jaramillo Duque
类目: Artificial Intelligence (cs.AI)
备注: Accepted to the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026)
Abstract:Text-to-Image (T2I) models have recently achieved impressive visual fidelity, yet their evaluation remains constrained by benchmarks that are often difficult to interpret and insufficiently diagnostic. Existing skill-based evaluations tend to overlook critical failure modes that strongly impact usability but fall outside standard taxonomies, such as global incoherence arising from missing parts or physically implausible configurations (e.g., floating objects). In addition, prompt difficulty is typically controlled along a single dimension; either prompt length or the number of elements to generate. To address these limitations, we introduce Imag-Eval, a controlled benchmark designed to assess how T2I models ground compositional natural-language instructions into visual outputs. Unlike prior work that conflates surface linguistic complexity with compositional difficulty, Imag-Eval explicitly seeks to disentangles these factors by independently varying both the number of instances and the combination of constraints (rules), while avoiding error propagation. This design enables fine-grained and interpretable analysis of where cross-modal instruction following fails. Our benchmark comprises 1,140 prompts and 8,842 combined rules, and we evaluate it on several state-of-the-art models. Complementing this analysis with an additional study of over 2,000 prompts from a concurrent benchmark, our results suggest that, for structured skills, compositional difficulty is primarily governed by the number of grounded rules and their binding to instances, rather than by prompt length alone.
[AI-134] Hyper-Fold: Exploring the Expressive Limit of Sequence-Geometry Learning for Proteins via Hypergraph Modeling
链接: https://arxiv.org/abs/2608.29207
作者: Yifan Feng,Guanjie Cheng,Shihui Ying,Shaoyi Du,Yue Gao
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Protein structure modeling rests on a single computational primitive: the interaction between what a residue is (sequence content) and where it sits (three-dimensional geometry). What is the expressive limit of this layer class? We show that the complete bilinear operator over content-geometry outer products–the sufficient statistic of all second-order interactions–is the expressive ceiling, while the additive message passing of mainstream geometric GNNs is provably blind to content-geometry binding. We then introduce Hyper-Fold, a rank-K separable convolutional backbone approaching this ceiling at message-passing cost: each radius neighborhood is organized into a sequence hyperedge and a contact hyperedge, modulated by an edge-conditioned matrix-valued operator factorized into K learned basis operators with geometry-generated coefficients. Across enzyme function prediction, fold classification, and ligand binding site detection, Hyper-Fold and its hierarchical variant Hyper-Fold-Deep achieve the best results among protein-specific structure encoders; Hyper-Fold-Pocket, an anchored set-prediction head, surpasses UniSite-3D on UniSite-DS and two zero-shot benchmarks with no sequence language model features, 68x fewer parameters, and 4.8x lower latency–suggesting that a sufficiently expressive 3D backbone recovers information that fusion architectures previously borrowed from evolution-scale pretraining.
[AI-135] Benevolent Bias in Multi-Turn Human-Agent Dialogue
链接: https://arxiv.org/abs/2608.29206
作者: Qianqi Liu,Jin Huang,Fethiye Irmak Dogan,Hatice Gunes
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Bias in human-agent interaction can manifest not only through hostile language but also as benevolent bias, whereby unequal treatment hides behind a warm, positive tone. To make it detectable, we operationalise benevolent bias along two dimensions, tone and treatment, yielding three classes: neutral support, overt bias, and benevolent bias. Building on these definitions, we construct BENEVDIAL, a class-balanced corpus of 362,880 multi-turn support dialogues spanning user and agent demographics, roles, and generators, to support controlled evaluation. We then test two detector families on it: off-the-shelf safety detectors and prompted large language model (LLM) judges. Our results reveal a detection gap: off-the-shelf detectors reliably flag overt bias yet largely miss benevolent bias, while LLM judges catch more under more explicit detection criteria but increasingly misclassify neutral support as benevolent bias, and demographic context amplifies the false alarms. These findings suggest that fair monitoring of human-agent dialogue must look beyond surface cues to whether the agent’s treatment is disparate.
[AI-136] Agent Logs: A Dataset for Opening the Black Box of GitHubs Cloud Agent
链接: https://arxiv.org/abs/2608.29204
作者: Jonan Richards,Kosei Horikawa,Youmei Fan,Yutaro Kashiwa,Mairieli Wessel
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:
Abstract:Generative AI-based software engineering agents are becoming routine contributors to real-world software projects. On GitHub, developers can assign tasks to the Copilot cloud agent, which autonomously explores the repository, edits code, runs commands, and opens or reviews pull requests, producing a detailed log of every step along the way. While existing datasets capture outcomes of agent contributions, such as agent-authored pull requests, the process by which agents produce these contributions remains largely unexplored. To address this gap, we introduce AgentLogs, a large-scale dataset of agent activity on GitHub. AgentLogs comprises 307,416 agent tasks and 549,239 agent sessions in 35,810 of the 1,812,362 popular public repositories that we scanned, together with 64,255,174 session log entries that record each agent run step by step, including prompts, intermediate reasoning, tool calls (e.g., file edits, git operations, and GitHub interactions), and token usage. By exposing not only what agents contribute but also how they work, AgentLogs enables research on agent behavior, efficiency and cost, task formulation, failure modes, and human-agent collaboration in agentic software engineering.
[AI-137] PokaiTrainer: Scaling Belief-State Search to Competitive Pokémon VGC
链接: https://arxiv.org/abs/2608.29197
作者: Max Yu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computer Science and Game Theory (cs.GT)
备注: 22 pages, 4 figures, 8 tables
Abstract:Decision-time equilibrium search carried poker to superhuman play, but it has so far relied on tractable subgames: a handful of actions per decision, chance confined to card deals, one player moving at a time. Competitive Pokémon in its official doubles format (VGC) breaks all three assumptions at once. Both players act simultaneously from joint menus in the hundreds, each joint action resolves to hundreds of stochastic outcomes, and the opponent’s reserves and stat allocations are hidden. We set out to build a strong VGC agent and report what that took. PokaiEngine, our Rust battle engine, enumerates a joint action’s full weighted outcome distribution in one pass, at \sim99% parity with Pokémon Showdown and a fraction of the cost of sampling it. On top of the engine, PokaiTrainer adapts Student of Games to this scale, solving every decision as a Bayesian matrix game over public belief states and growing subgames under an explicit compute budget. On the live Showdown best-of-three ladder, the agent wins 59% of 150 sets against a human field averaging \sim1320 Elo. It settles into a 1350-1400 Elo band, and at its peak briefly entered the format’s top 500.
[AI-138] An Explainable Coherence Score for Detecting Temporal Inconsistencies in Political News
链接: https://arxiv.org/abs/2608.29175
作者: Marius Nicusor Pantea,Adrian Groza
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Temporal inconsistencies, such as mandates attributed outside their real interval, events presented as past before they occurred, or inverted causal sequences, are a form of political disinformation that evades style-based fake news detectors: a well-written article with a single wrong date carries no lexical signal of falsehood. This paper introduces the Temporal Coherence Score (TCS), a continuous, intrinsically interpretable metric that quantifies the temporal coherence of a news article, computed by a four-stage pipeline: extraction of temporal facts, construction of a temporal knowledge graph, hierarchical verification against internal consistency rules and external reference sources, and score aggregation with automatically generated explanations. Verification combines eight internal checkers derived from Allen’s interval algebra with a five-level external hierarchy ranging from a locally stored reference knowledge base of 1,256 curated political facts to live Wikidata SPARQL queries. On a benchmark of 100 political news articles with injected temporal errors, the system reaches a precision of 0.909 at the selected operating threshold, with a single residual false positive, a profile deliberately tuned for human-in-the-loop fact-checking assistance, where false alarms are costlier than missed detections. Unlike lexical baselines that output only a binary label, every flagged article is accompanied by the inconsistency type, the entities involved, and the reference source that contradicts the claim.
[AI-139] JudgePanel: A Compact Judge with Panel Deliberation via Adaptive Multi-Reward Reinforcement Learning
链接: https://arxiv.org/abs/2608.29168
作者: Yiyue Qian,Shinan Zhang,Huan Song,Hannah Marlowe
类目: Artificial Intelligence (cs.AI)
备注: 8 pages,4 figures
Abstract:The LLM-as-a-Judge paradigm has emerged as a scalable alternative to human evaluation. However, single-model judges are limited by their inherent model biases, while multi-agent evaluation protocols that mitigate this through diverse deliberation are prohibitively expensive at inference time. To this end, we propose \textbf\modelname, which equips a compact \underlineJudge model with multi-agent \underlinePanel deliberation capability. Specifically, we first train on panel deliberation traces from an ensemble of strong evaluators, capturing structured patterns of discussion, disagreement, and resolution. To further improve judgment quality beyond SFT, we introduce \textitAdaReward, an adaptive multi-reward RL algorithm that dynamically rebalances reward component weights as different objectives saturate at different rates during RL training. For practical deployment, we further design a lightweight domain specialization module for rapid adaptation to new evaluation domains with few hundred labeled samples. As a result, (i) \textitNovel: the first framework to equip a single compact judge with multi-agent panel deliberation capability at single-model inference cost; (ii) \textitEffective \ Reliable: JudgePanel with a 14B backbone outperforms judge-specialized models up to 70B across four evaluation benchmarks, demonstrates strong position consistency, and rapidly specializes to new domains with few hundred samples.
[AI-140] More Perspectives Stronger Signals: Multi-Perspective Enhancement and Progressive Fusion for Multimodal Entity Representation Learning
链接: https://arxiv.org/abs/2608.29139
作者: Chenyi Xiong,Yan Zhang,Jing Hu,Ziyue Qin,Kui Xiao,Xiaopan Lyu,Xiaoju Hou,Zhifei Li
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Learning effective multimodal entity representations is fundamental for reasoning tasks such as multimodal knowledge graph completion (MMKGC). However, existing methods often suffer from semantic over-smoothing within modalities and ineffective noise filtration across modalities, particularly under sparse or ambiguous conditions. To overcome these limitations, we propose PrismF, a unified framework that synergizes multi-perspective enhancement with progressive fusion to extract stronger signals from diverse inputs. PrismF enhances fine-grained intra-modal semantics through a multi-perspective mechanism that decomposes each modality into complementary views and constrains them with a decoupling loss to reduce representation collapse. Furthermore, it improves cross-modal integration through a progressive fusion strategy that dynamically calibrates inter-modal interactions, enabling the model to emphasize informative signals while suppressing noisy or unreliable ones. Extensive experiments on three public benchmarks show that PrismF achieves the strongest overall performance, including relative improvements of 4.04% in MRR and 11.17% in Hits@1 on KVC16K. Our code can be found at this https URL.
[AI-141] Not the Same Protector: Deployment-Dependent Protective Intervention in LLM s
链接: https://arxiv.org/abs/2608.29136
作者: Eunna Lee,Soomyoung Lee,Jungpyo Nam,Heonjin Ha,Jamin Jung,Kyunam Choi,Sunjun Hwang,Yeonghun Kim,Seok-Jae Lim
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:We ask whether a model protects a user in the same way when that user speaks rather than types. Using a single distress vignette—a physical injury of unstated severity following an interpersonal conflict—we present four frontier models with matched inputs across voice, text, and raw API deployment conditions (n=30 per cell) and code each response along five binary protective indicators, including whether the model issues an explicit medical-care directive. Voice-interface responses are markedly shorter than text-interface responses for three of the four models, and protective behavior contracts alongside that compression: medical directives are at ceiling under both the API and text conditions but decline under voice for every model tested. The contraction is not reducible to length. One model produces voice and text responses of comparable length yet still drops medical directives, and another falls below ceiling between its API and voice conditions, whose responses are of nearly identical length. Under raw API access the pattern is categorical rather than partial: no model asks after the user’s safety even once. These results show that protective intervention is sensitive to the surface through which a request arrives, that this sensitivity is detectable using a simple protective coding scheme, and that it is not explained by turn length alone.
[AI-142] APIFlow-Bench: Measuring Whether Agents Survive Long Dependent API Workflows NEURIPS2026
链接: https://arxiv.org/abs/2608.29128
作者: Zelin Wan,Arash Nourian,Xiaoxiao Li,Nihar Nandan,Kamalakannan Nandagopal
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Software Engineering (cs.SE)
备注: 15 pages, 8 figures. Under review at the NeurIPS 2026 Workshop on Evaluation of Interactive Agents (IAEval). Harness, frozen task bank, and 44,362 execution transcripts: this https URL
Abstract:Tool-using agents are commonly evaluated by a single bit: whether an end-to-end workflow completed. This metric fails to distinguish failures that matter in production, such as expired credentials, malformed payloads, or correct execution followed by incorrect final delivery. We introduce APIFlow-Bench, a fully auditable benchmark for long-horizon, dependent REST-API workflows that decomposes performance into seven engineering capabilities and requires agents to produce answers supported by the actual call path. We generate synthetic API worlds forward, subtask by subtask; each subtask is admitted only after a zero-LLM self-test triad verifies its grader and an oracle establishes solvability, and an adversarial audit identified and fixed six grader exploits. Grading is deterministic and provenance-sensitive: a state check traces a mock-minted canary through the API data flow to the response the answer must originate from, and a typed answer card is verified field by field. We release all answer keys and 44,362 unredacted execution transcripts. Across 19 frontier and open-weight models under one neutral scaffold, we find: (1) longer dependency chains degrade success, from 93% on individual subtasks to 74% on clean 20-subtask chains and 61% when including the 8% of chain trials that a model-consensus screen flags as passed by no model; (2) reliability separates models more than best-case capability, with best-of-five spanning seven points but all-five-of-five reliability spanning 44 points; (3) the independent-error account of compounding failure does not fit the data: pass rates on 20-subtask chains are 33 percentage points above the product of subtask-level rates, and on the clean slice 77% of failing runs reached the correct final state and failed only at delivery.
[AI-143] Beyond Correctness: Validity-Oriented Evaluation of Biomedical LLM Judges
链接: https://arxiv.org/abs/2608.29127
作者: Rodrigo de Oliveira,Federico Pittino,James Gwinnutt,Jay Nanavati
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:We propose a scalable, validity-oriented pipeline for evaluating biomedical LLM judges when high-quality human judgments are scarce. First, we augment existing human-labelled biomedical benchmarks with deterministic, metric-grounded mutations that produce auditable preference pairs. Second, we evaluate judges beyond aggregate correctness using three deployment-relevant dimensions: correctness against metric-derived gold labels, robustness under repeated stochastic sampling, and compliance with the requested output format. We use this pipeline to assess Llama-3.1-8B-Instruct under four regimes: (1) base, using the instruct model as is; (2) SFT, distillation-based supervised fine-tuning only; (3) RL, GRPO-based reinforcement learning only; and (4) SFT \rightarrow RL, SFT followed by RL. The base and single-stage regimes struggle on structured medical discrimination such as PICO extraction and clinical calculations, whereas SFT \rightarrow RL performs best across correctness, compliance, and robustness; gains concentrate on decomposable tasks (PICO, MedCalc), at times matching or outperforming frontier models.
[AI-144] CGFM-Nav: Cognitive Graph-Field Memory for Semantic-Guided Lifelong Multimodal Embodied Navigation
链接: https://arxiv.org/abs/2608.29114
作者: Yuxiang Xiao,Xibei Chen,Xin Zhou,Jie Chen,Yifeng Zhang,Guillaume Sartoretti
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: 5 pages, 2 figures
Abstract:Vision-and-Language Navigation (VLN) requires agents to reason over accumulated observations while continuously exploring unseen regions. However, existing environment representations often struggle to jointly support explicit semantic memory and continuous exploration guidance. To address this challenge, we propose Cognitive Graph-Field Memory (CGFM), a persistent multimodal scene representation that couples explicit relational memory with continuous spatial intuition. CGFM organizes objects, spatial relations, and visual observations into a multimodal scene graph, enabling target retrieval and long-horizon reasoning across navigation tasks. When no reliable target match is identified, graph-based evidence is projected into a goal-conditioned semantic-frontier field to guide exploration toward semantically promising frontiers and regions. Building upon CGFM, we introduce CGFM-Nav, a foundation-model-based framework for lifelong multimodal navigation that integrates task-relevant subgraph selection, VLM reasoning, and verification feedback into a closed decision loop. Preliminary experiments on GOAT-Bench show that, under the same Qwen3-VL-8B backbone, CGFM-Nav improves the overall success rate from 53.2% to 63.0% and SPL from 30.0% to 39.6%, demonstrating the effectiveness of combining explicit semantic memory with semantic-guided exploration.
[AI-145] Auditing and Mitigating Privacy Leakage in Cloud-Edge Collaborative Decoding EMNLP2026
链接: https://arxiv.org/abs/2608.29111
作者: Kejia Zhang,Tianyuan Zou,Zixuan GU,Yang Liu
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注: EMNLP 2026 Findings
Abstract:Applications such as personalized assistance and proprietary document analysis require large language models (LLMs) to generate outputs from private data. Yet powerful LLMs typically cannot be deployed on the resource-constrained devices where private data resides, and uploading private data to cloud-hosted LLMs exposes sensitive information. Recent work addresses this tension with a cloud-edge collaborative decoding paradigm, where private data are kept on the edge with a small language model (SLM) producing next-token distributions, which are fused with predictions from a cloud LLM operating solely on public data. In this paper, we systematically analyze the privacy risks of such a paradigm with a novel evaluation framework using constructed QA datasets, which show that such collaboration can expose substantial private-context information. To address such privacy leakage, we propose CoVeil, a defense mechanism which dynamically optimizes transmitted signals to suppress leakage during decoding time while preserving the collaborative quality. Extensive evaluations demonstrate that CoVeil consistently improves the privacy-utility trade-off over existing baselines by reducing data leakage by up to 87.2%, with minimal accuracy loss.
[AI-146] Clustering as Approximation by Constrained Projectors: Theory and Guarantees
链接: https://arxiv.org/abs/2608.29102
作者: Angshul Majumdar
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:This paper develops a unified theoretical framework showing that a broad family of clustering methods, including k-means, fuzzy c-means, kernel k-means, kernel FCM, and spectral clustering, can all be expressed as structured low-rank projectors acting on a signal-derived matrix. By formulating each method as an instance of min over B in C of ||M - M P_B||_F^2, with different constraint sets C, we establish a common optimization template that clarifies the algebraic links among hard, fuzzy, kernel-induced, and orthonormal projections. Within this framework, we derive non-trivial theoretical results, including geodesic convexity properties on the projection manifold, perturbation bounds quantifying stability to matrix noise, and exact recovery guarantees under ideal block-model conditions. The analysis further explains when different clustering families collapse to the same optimal subspace and how deviations arise under small inter-cluster leakage. Overall, the work provides a coherent, theory-first foundation for understanding clustering through structured projectors.
[AI-147] Development of an Autonomous AI Coding Agent using Monte Carlo Tree Search (MCTS) and Gemini LLM Frameworks
链接: https://arxiv.org/abs/2608.29096
作者: Pravin Game,Vipin Ramakrishnan,Prathamesh Wagh
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 10 PAGES WITH PLAGIARISM REPORT ON 10TH PAGE
Abstract:The ongoing changes in software engineering requirements have created a substantial need for automated tools which can create secure source code from natural language input. The performance of traditional Large Language Models (LLMs) becomes limited by their “one-shot” capability which results in logical hallucinations together with reduced algorithmic performance during complicated operations. The research presents an autonomous AI Coding Agent which establishes a connection between LLM-generated content and production-ready software through its organized methodology for decision making. Our framework uses the Gemini 2.5 Flash API for essential reasoning capabilities while employing a tailored Monte Carlo Tree Search (MCTS) method to solve code generation challenges as a search operation. The agent uses a “Self-Critic” evaluator system to test different implementation methods which it ranks according to their accuracy and difficulty level before it improves its operational framework through backpropagation. The system operates through a Flask-based web interface which delivers instant feedback together with syntax highlighting features. Our experimental results show that the MCTS-based method achieves a 92% success rate on complex logical prompts while surpassing standard zero-shot generation models.
[AI-148] HANIA: Planner-Guided Multimodal Graph Evidence Selection for Grounded Question Answering ISWC2026
链接: https://arxiv.org/abs/2608.29088
作者: Zafar Ali,Asad Khan,Nimbeshaho Thierry,Nabila Amir,Adam A. Q. Mohammed,Pavlos Kefalas
类目: Artificial Intelligence (cs.AI)
备注: 10 pages, 1 figure. Accepted at Graph-enhanced LLMs for trustwOrthy Web data management (GLOW), ISWC 2026 Workshops, Bari, Italy
Abstract:Multimodal question answering remains sensitive to noisy, incomplete, and weakly grounded evidence. Long unstructured contexts can introduce redundancy and encourage unsupported generation, while flat retrieval may overlook relations needed for multi-step reasoning. We present HANIA, a planner-guided multimodal graph framework for evidence-grounded question answering. HANIA processes the supplied image and text using a frozen vision-language model to extract concise question-relevant visual evidence with explicit abstention. It then constructs an input-grounded multimodal graph and applies a two-group finite-state planner to coordinate descriptive and relational evidence. Coverage-aware pruning retains a compact evidence set based on relevance, graph confidence, concept coverage, and modality diversity. The selected passages, visual statements, and graph triples are provided to a frozen instruction-tuned decoder. We evaluate HANIA on ScienceQA using answer accuracy, evidence-filtering quality, evidence-budget sensitivity, and efficiency. The results show that structured evidence planning and compact graph-guided retrieval can support competitive multimodal question answering without target-dataset fine-tuning or iterative retrieval. The code is available at this https URL.
[AI-149] Nested Convex-Body Chasing for Online Optimization with Evolving Feasible Sets
链接: https://arxiv.org/abs/2608.29074
作者: Dhruv Sarkar,Aprameyo Chakrabartty
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:We study online optimization with nested shrinking feasible regions in two settings: convex optimization with nested evolving feasible sets (CONES) and adversarial constrained online convex optimization (COCO). Our algorithms separate loss control from geometric movement: constrained minimizers and cumulative-loss tests preserve regret guarantees, while a deterministic resettable nested convex-body chaser limits movement. For CONES with a G -Lipschitz, \mu -strongly convex objective on a diameter- D domain, we chase intersections of the current feasible set with adaptive objective sublevel sets. Using the Euclidean chasing ratio O(\sqrtd\log(1+d)) , we obtain nonpositive regret at every prefix and movement O(\sqrtd\log(1+d),GD\log(eT)/\mu) . The bound adapts to the increase in the constrained optimum value. In dimension two, with all other parameters fixed, every randomized algorithm with terminal expected regret O(T^\beta) , \beta1 , suffers \Omega(\sqrt\log T) expected movement on some deterministic nested sequence, proving optimal horizon dependence. Under linear growth away from the constrained minimizer set, Steiner-point tracking yields movement independent of T . For general convex COCO, one-step-delayed chasing with regularized-leader resets gives regret O(G_fD\sqrtd\log(1+d)T) and cumulative constraint violation O(G_gD\sqrtd\log(1+d)T) . For strongly convex losses, both are O(d\log(1+d)\log(eT)) when other parameters are fixed. These reductions replace the O(d^d/2) projection-path factor in prior analyses by the polynomial dimension dependence of Euclidean nested convex-body chasing. Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2608.29074 [cs.AI] (or arXiv:2608.29074v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.29074 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-150] Revolutionizing Turn-by-Turn Navigation with Cloud-Edge Deep Learning
链接: https://arxiv.org/abs/2608.29073
作者: Yiming Yang,Hao Fu,Fanxiang Zeng,Xikai Yang,Yue Liu,Ning Guo
类目: Artificial Intelligence (cs.AI)
备注: This paper has accepted by IEEE Transactions on Intelligent Transportation Systems
Abstract:Turn-by-turn (TBT) navigation systems are integral to modern driving experiences, providing real-time audio instructions to guide drivers safely to destinations. However, existing audio instruction policy often relies on rule-based approaches that struggle to balance informational content with cognitive load, potentially leading to driver confusion or missed turns in complex environments. To overcome these difficulties, we first model the generation of navigation instructions as a multi-task learning problem by decomposing the audio content into combinations of modular elements. Then, we propose a novel deep learning framework that leverages the powerful spatiotemporal information processing capabilities of Transformers and the strong multi-task learning abilities of Mixture of Experts (MoE) to generate real-time, context-aware audio instructions for TBT driving navigation. A cloud-edge collaborative architecture is implemented to handle the computational demands of the model, ensuring scalability and real-time performance for practical applications. Experimental results in the real world demonstrate that the proposed method significantly reduces the yaw rate (the proportion of vehicles deviating from navigation routes) compared to traditional methods, delivering clearer and more effective audio instructions. This is the first large-scale application of deep learning in driving audio navigation, marking a substantial advancement in intelligent transportation and driving assistance technologies.
[AI-151] Agent 2UCB: Agent ic System for Generative Engine Optimization
链接: https://arxiv.org/abs/2608.29063
作者: Sheldon Yu,Rui Wang,Tong Yu,Sungchul Kim,Doga Dogan,Junda Wu,Julian McAuley
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large language model driven search engines such as Google AI Overviews and Perplexity have created new opportunities for Generative Engine Optimization (GEO) the practice of refining content to increase its likelihood of being cited or summarized by generative systems. We demonstrate Agent2UCB, an agentic GEO system that autonomously improves content visibility through customized, feedback-driven optimization. For each content item, the system evaluates nine GEO strategies, identifies the most effective method, and accelerates selection using a bandit-based Agent2UCB policy that integrates LLM priors with online reward signals. To monitor side effects, the system also provides a lightweight, text-only SEO readiness evaluation covering readability, topical coverage, and EEAT-style credibility. Experiments on GEO-Bench show consistent visibility gains while preserving SEO quality. The demo allows users to choose the websites of interest, observe the optimization workflow, and compare GEO/SEO outcomes across methods.
[AI-152] Let Prompts Bridge Defense Knowledge: Transferable Graph Purification via Vulnerability-Aware GPL
链接: https://arxiv.org/abs/2608.29054
作者: Shuomin Xue,Jingyuan Li,Ju Jia,Jingxuan Yu,Xiaojun Jia
类目: Artificial Intelligence (cs.AI)
备注: To appear in the Proceedings of the 34th ACM International Conference on Multimedia (MM '26). 10 pages, 7 figures, and 4 tables. Shuomin Xue and Jingyuan Li contributed equally to this work
Abstract:Graph Neural Networks (GNNs) have emerged as a cornerstone for representing complex relational dependencies in diverse multimedia tasks, particularly in cross-platform user interest modeling and cross-modal semantic alignment. In the real world, a practical defense against graph adversarial perturbations is needed. However, we observe that the prevailing adversarial purification methods are essentially domain-restricted defenses, which leads to the following shortcomings: (1) single-domain data provides insufficient structural and semantic diversity for learning robust purification criteria; (2) training of domain-specific defense strategies from scratch consumes substantial computational cost. To address the above limitations, we propose a transferable graph purification scheme, named ProGAP, to bridge adversarial defense knowledge via vulnerability-aware graph prompt learning. Firstly, to capture universal adversarial patterns, a perturbation-capture edge detector is pretrained on data-rich graphs by jointly modeling topological and semantic information. Subsequently, to achieve more knowledge transfer w.r.t. robustness, vulnerability-aware prompts are designed that inject targeted purification guidance into biased nodes, during which the pretrained detector adapts to distribution shifts in downstream graphs without parameter-laborious updates. Experimental results demonstrate that compared with state-of-the-art baselines, our ProGAP achieves 1%-9% improvement, and reduces the time consumption by up to 2.2x. The code for ProGAP is available at this https URL.
[AI-153] Learning to Follow In-Context Watermark Instructions via Self-Distillation
链接: https://arxiv.org/abs/2608.29030
作者: Yepeng Liu,Tianyi Chen,Xuandong Zhao,Dawn Song,Yuheng Bu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:In-context watermarking (ICW) prepends an instruction to a query asking the model to embed a statistically detectable signal in its response. It thus equips LLMs with a watermarking interface that third parties can invoke without access to model internals. Its reliability hinges on the LLM following the instruction without degrading answer quality, yet how well current LLMs do so has not been measured. We introduce \mathsfICWBench , a benchmark of three verifiable ICW instruction families, each scored on both detectability and answer quality. Evaluating 14 frontier proprietary and open-source LLMs, we find that none of the evaluated LLMs achieves both objectives across all three families. To address this, we propose a self-contained two-stage training method, requiring no distillation from a stronger model, no manual annotation, and no pre-existing ICW IF ability. The first stage, self-distillation with logits perturbation (SDLP), uses the same base LLM as both teacher and student: an instruction-equivalent decoding-time logits perturbation makes the teacher follow the ICW instruction, and the student is trained to match the teacher’s output distribution. The second stage applies reinforcement learning with the automatic verifier as the reward. Applied to Qwen3-14B and GPT-OSS-20B, our method raises average TPR@ 1% FPR across three ICW instructions from 0.100 to 0.974 and from 0.337 to 0.968 , respectively, while maintaining high response quality under both perplexity evaluation and LLM-as-a-Judge.
[AI-154] Flow-JEPA: Flow Matching for Robust Latent Dynamics in JEPA World Models
链接: https://arxiv.org/abs/2608.29029
作者: Yanchen Huo,Ziying Song,Yadan Luo
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Joint-Embedding Predictive Architectures (JEPAs) have shown strong potential for learning compact predictive representations, and LeWorldModel (LeWM) extends this paradigm to reconstruction-free latent world modeling from pixels. However, its deterministic autoregressive predictor generates future states through repeated one-step transitions, which can accumulate errors and remain sensitive to task-irrelevant visual perturbations. In this work, we propose Flow-JEPA (F-JEPA), a conditional flow matching dynamics model that jointly generates a sequence of future latent states conditioned on the current observation and actions. A Gaussian distribution serves as the flow source, exposing the vector field to perturbed latent trajectories as it learns to transport them toward clean future representations. This formulation retains the reconstruction-free JEPA framework while replacing point-wise transition regression with stochastic trajectory-level prediction. F-JEPA raises mean success from 86% to 92% under clean observations and from 67% to 86% under noisy conditions, suggesting that conditional flow matching provides a promising alternative to deterministic autoregressive dynamics in JEPA world models.
[AI-155] Facts Without Rules: Boundary Metadata Collapse in Multi-Agent LLM Handoffs
链接: https://arxiv.org/abs/2608.29028
作者: Yian Wang,Agam Goyal,Eshwar Chandrasekharan,Hari Sundaram
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Multi-agent LLM systems often coordinate by compressing an upstream interaction into a handoff artifact that downstream agents treat as shared state. We show that this handoff step is a structural source of privacy leakage: summaries preferentially preserve operational facts while weakening the boundary metadata that governs how those facts may be used—a failure mode we call \emphsummary collapse. On a controlled multi-agent coordination testbed we measure marker survival with a human-validated judge ( \kappa = 0.74 ), where \sigma_b = 1 means every boundary marker survives verbatim and \sigma_b = 0 means all are lost. Boundary-marker and operational-fact survival are nearly uncorrelated at the handoff level on both GPT-5-mini and DeepSeek-R1-32B (Pearson r near zero): uncompressed free-text handoffs preserve boundaries at \sigma_b \approx 0.80 , whereas a 25 -word budget drops \sigma_b to \approx0.57 while operational-fact survival stays near ceiling. Controlled downstream tests reveal that protection depends on \emphboundary explicitness: vague languages leak in 73% of GPT and 50% of DeepSeek cases, while explicit constraints reduce leakage to under 15% across all three tested models. A no-handoff single-agent control further shows the failure is not reducible to multi-agent topology as direct full-marker access still leaks more often than the operationalized handoff. Prompt-only mitigation and exact-string redaction only partially address the problem, while a gold-derived audience allowlist nearly eliminates leakage across models, showing that correctly identifying audience boundaries is the key factor.
[AI-156] Frequency Selective Neural Networks as a Foundation Architecture for Time Series Learning
链接: https://arxiv.org/abs/2608.29012
作者: Hui Huang,Ye Sun,Shiyan Hu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Time-series data across physical and biological domains are fundamentally driven by complex, non-stationary oscillatory modes. While deep learning models, such as Convolutional Neural Networks (CNNs), Recurrent Neural Networks, and Transformers, have dominated sequential analysis, they remain fundamentally “spectral-blind”. By mapping continuous physical waves into unconstrained spatial or discrete token spaces, these architectures suffer from severe spectral entanglement, acting as opaque black boxes that decouple predictive accuracy from physical reality. In this paper, we introduce the Frequency Selective Neural Network (FSNN), pioneering a foundation architecture guaranteeing physical interpretability without sacrificing expressive power of deep learning. FSNN addresses spectral entanglement by explicitly embedding the rigorous mathematics of advanced signal processing into its neural topology. Through a fully differentiable Wiener-like filter bank optimized via complex-domain backpropagation, FSNN autonomously discovers and isolates the precise physical modes of a given task. Extensive evaluations demonstrate that FSNN establishes state-of-the-art predictive performance, achieving 77.0% average accuracy on the standard 10 multivariate UEA datasets and leading across all major metrics on the highly imbalanced PTB-XL clinical ECG benchmark. Crucially, in contrast to yielding abstract feature maps, FSNN converges directly on physically meaningful frequency bands, such as isolating the cardiac QRS complex, providing a highly scalable, interpretable paradigm for robust pattern recognition in complex temporal domains. Our code is available at: this https URL.
[AI-157] Multi-Step Forecasting of Grape Berry Temperature based on LSTM Model with Feed-Forward Attention
链接: https://arxiv.org/abs/2608.29008
作者: Srikanth Gorthi,L. G. Divyanth,Dattatray Bhalekar,Markus Keller,Lav Khot
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Accurate forecasting of grape berry temperature (Tb) is essential for enabling timely heat stress management in vineyards. In this study, a feed-forward attention mechanism integrated with a Long Short-Term Memory network (FAM-LSTM) was developed and evaluated for multi-step, high-resolution Tb prediction. Models were trained using environmental data from 2023 and 2024 at Prosser, WA, USA, and validated on 2025 summer data. FAM-LSTM was benchmarked against LSTM, GRU, RNN, and Random Forest (RF) across horizons ranging from 15 minutes to 72 hours (288 time steps). Two input scenarios were evaluated: nearest open-field weather station observations and in-vineyard microclimate measurements. FAM-LSTM consistently outperformed all benchmark models across all horizons and input scenarios. Incorporating in-vineyard microclimate data significantly improved forecasting accuracy at longer horizons. Using open-field data, FAM-LSTM achieved MAE and RMSE ranges of 0.58 to 1.70 deg C and 0.65 to 2.07 deg C, respectively. In-vineyard observations further improved performance, with MAE and RMSE in the ranges of 0.51 to 1.55 deg C and 0.71 to 1.87 deg C. Error analysis showed prediction uncertainty was highest during peak daytime periods (11:00 to 18:00) and increased progressively with forecast horizon. Overall, the FAM-LSTM framework offers robust Tb forecasting to support precision heat stress management in vineyards.
[AI-158] Verification abundance adjudication scarcity: what happens to mathematical knowledge when proof checking becomes free
链接: https://arxiv.org/abs/2608.28997
作者: Maher Kallel,Mohamed El Louadi
类目: Artificial Intelligence (cs.AI); Logic in Computer Science (cs.LO)
备注:
Abstract:In May 2026 an OpenAI model produced a counterexample to the Erdős unit distance conjecture. Five mathematicians published a human-verified version the same day, and the result entered the literature within weeks. In August 2026 the same laboratory published ten mathematical and theoretical computer science results, each accompanied by a machine-checkable Lean 4 certificate with no unproved steps. Four weeks later, one remained the subject of an unresolved dispute over whether its formalization meant what it claimed. We argue that this difference is structural. We distinguish three layers of verification: derivational validity, which a kernel checks; representational fidelity, whether the formal statement means the intended question; and epistemic significance. Only the first is mechanizable. Making it effectively free therefore does not eliminate verification work but shifts the burden to layers dependent on scarce expert attention. Measurements of the August corpus illustrate the shift. The kernel-checked proofs total 20.6 MB, while the statements requiring human audit total 55.6 KB, a ratio of 379 to 1. Yet those statements contain 218 bespoke definitions rather than relying on community-vetted ones. The audit surface is therefore small in volume but irreducibly expert. We argue that machine checking produces verification abundance while leaving adjudication scarce. We propose a six-category taxonomy of representational mismatch, a disclosure schema for machine-generated mathematical claims, and implications for software, cryptography, and regulated decision systems. Subjects: Artificial Intelligence (cs.AI); Logic in Computer Science (cs.LO) Cite as: arXiv:2608.28997 [cs.AI] (or arXiv:2608.28997v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.28997 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-159] Agent ic AI uncovers conserved cross-tissue protein co-abundance programs inaccessible to single-dataset analysis
链接: https://arxiv.org/abs/2608.28990
作者: Runyu Guan,Dehao Wu,Qiqi Xie,Yang Li,Haohan Wang
类目: Artificial Intelligence (cs.AI); Molecular Networks (q-bio.MN)
备注: 16 pages, 6 figures
Abstract:Protein co-abundance clusters preserved across tissues can reveal shared disease mechanisms and candidate therapeutic targets, particularly when proteins implicated in organ-confined diseases converge in peripheral or accessible tissues. However, previous cross-tissue studies have focused on biologically pre-selected tissue pairs, leaving most possible combinations and non-obvious relationships unexplored. We present an LLM-agent framework for large-scale, evidence-grounded comparison of tissue-specific protein co-abundance networks. The framework constructs tissue networks, derives pairwise consensus clusters, and integrates evidence from expression atlases, protein interaction and complex databases, pathway annotations, disease catalogues, and literature. Applied to all 820 pairwise combinations of 41 human tissues and fluids, it identified 1,833 conserved co-abundance clusters across 406 tissue pairs. Colon, synovial fluid, blood, cerebrospinal fluid, and bone marrow were the most broadly connected tissues, while the most cluster-rich pairs were dominated by bone marrow. The analysis also highlighted non-obvious relationships: skin-bone marrow exceeded the anatomically adjacent bone-bone marrow pair, while colon-breast contained cancer-relevant clusters involving extracellular-matrix remodeling, lipid metabolism, and immune modulation. Cluster-level analyses generated further mechanistic hypotheses, including a brain-gut extracellular-vesicle/redox/serotonin-cofactor axis and a liver-bone marrow stress-response axis involving genes linked to white matter disease. These results provide a global, comparable landscape of conserved protein co-abundance and a hypothesis-generating resource for mechanistic and therapeutic exploration. Code and data are available at this https URL.
[AI-160] Selective Forgetting: A Graph-Based Memory Framework for Long-Term LLM Agents
链接: https://arxiv.org/abs/2608.28978
作者: Theo Rusu,Sourena Khanzadeh,Manar Alalfi
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Knowledge graphs have been proposed as a structured alternative to flat retrieval-augmented generation for long-term agent memory, on the assumption that representing conversations as entities and relations improves recall. We evaluate that assumption directly. Our framework extracts each conversational turn into typed nodes and attributed edges, answers questions from a two-hop subgraph, and periodically prunes nodes that score low on a weighted combination of recency, access frequency, degree centrality, and age. On LongMemEval, the graph does not outperform a flat vector baseline at a matched candidate-generation budget of five retrieval roots: token F1 is 0.417 against 0.468 , and a paired bootstrap over 500 questions gives \Delta = -0.050 (95% CI [-0.085, -0.016] ). The gap is widest on questions that require recalling a specific prior assistant turn, where judged correctness falls from 0.911 to 0.607 , suggesting that decomposing a turn into entities discards the surface form these questions depend on. The forgetting module is more successful. Applied once to a persistent 27,021-node graph, it removes 9.8% of nodes and 9.5% of stored bytes; token F1 is unchanged ( +0.001 , 95% CI [-0.015, +0.016] ) and judged correctness falls by 1.6 points, with the 95% interval bounding any loss at 3.8 points ( [-0.038, +0.006] ). Because our extractor is a single small model evaluated on one benchmark, these results characterise this extraction-based pipeline rather than graph-structured memory in general. Code: this https URL
[AI-161] he Role of Network Topology and Opponent Information in Shaping Cooperation in Multi-Agent Reinforcement Learning Systems
链接: https://arxiv.org/abs/2608.28977
作者: Seongho Son,Stephen Hailes,Mirco Musolesi
类目: Artificial Intelligence (cs.AI)
备注: 19 pages, 13 figures
Abstract:Several works have investigated the influence of graph topology on cooperation among artificial agents, while the majority of the literature has focused on modelling agents’ adaptation through strategy imitation, which relies solely on the cumulative payoffs of others. This paper investigates scenarios in which each agent learns to play the two-player Iterated Prisoner’s Dilemma (IPD) using deep reinforcement learning. Each agent is represented as a node in a graph, where its neighbours constitute the pool of opponents with whom it can interact. During each IPD episode, agents are provided with different types of information about their opponent, consisting of action history and opponent identity. Experimental results across different graph topologies show that the number of neighbours per node and the average path length are the main factors affecting the emergence of cooperation. We also show that, while partner selection fosters mutual cooperation by limiting the diversity of the opponent pool, providing agents with the identity of their opponent hinders the proliferation of cooperative strategies.
[AI-162] From Analytics to Tumor Boards: An Evidence-Linked Multi-Agent Workflow for Oncology Feature Extraction
链接: https://arxiv.org/abs/2608.28974
作者: Daniel Kang,Michelle Hu,Soorya Ram Shimgekar,Shayan Vassef,Yufan Wang,Anit Kumar Sahu,Munmun De Choudhury,Vedant Das Swain,Christian Poellabauer,Li Yan Khor,Koustuv Saha,Robert Wojciechowski,Elliot Kidd,Piyum Zonooz,Navin Kumar
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Clinically relevant oncology information is distributed across heterogeneous, longitudinal documentation, creating substantial abstraction burden and requiring accurate attribution across specimens, tumors, biomarkers, and time points, while manual cancer-registry abstraction can require 27.2 minutes per case, highlighting the need for scalable methods that preserve clinical context while converting documentation into structured data. We evaluate the Nimblemind Multi-Agent System (nMAS), a configurable oncology information-extraction workflow which extracts clinically relevant structured fields from fragmented oncology documentation. The extraction task uses a clinician-informed schema of 328 attributes spanning report metadata, diagnosis, staging, and cancer-type-specific information. nMAS separates clinician-defined field specifications from model execution and combines complexity-aware extraction, report-level consolidation, and source-grounded validation. The retrospective evaluation included 230 de-identified oncology documents from 40 patients and 418 clinician-reviewed document-field pairs containing 1,126 non-empty reference values. Evaluation focused on fields identified by clinicians as present in the source documents rather than exhaustively annotating all 328 schema fields. nMAS achieved a rank-weighted value-level precision of 82.6%, recall of 87.5%, and F1 of 85.0%, compared with an F1 of 66.4% for an independently implemented UMA-style MiniMax M2.5 comparator. These findings support the feasibility of using a configurable, source-grounded extraction workflow to convert fragmented oncology documentation into reusable structured data.
[AI-163] Free Speech and Artificial Intelligence
链接: https://arxiv.org/abs/2608.28973
作者: Etienne Brown
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI)
备注: Preprint. Forthcoming in Mark Satta, Étienne Brown, and JP Messina (eds.), Philosophy and Free Speech: An Introduction. London: Routledge, 2027
Abstract:Philosophers and legal scholars are engaged in debates about the implications of artificial intelligence for freedom of expression. This paper analyzes the free speech issues raised by two distinct AI technologies: social media recommendation algorithms and conversational AI (i.e., chatbots powered by large language models). The first part shows that, through their recommendation algorithms, social media platforms control the dynamics of speech visibility in the digital public sphere, making algorithmic recommendation relevant to the philosophy of free speech. The second part turns to conversational AI. It discusses both the reasons for granting or withholding speech rights to artificial agents and users’ right to receive information, which may render specific forms of chatbot regulation illegitimate. Throughout, the chapter also considers whether social media platforms or AI developers hold corporate speech rights. Its general aim is to raise rather than settle questions that arise from the rapid development of AI technologies.
[AI-164] Diagnose Then Refine: A Closed-Loop TTS System with AudioLLM -Guided Correction EMNLP2026
链接: https://arxiv.org/abs/2608.28970
作者: Zeyang Song,Tianchi Liu,Tianrui Wang,Chenglin Xu,Steven Y. Guo,Haizhou Li
类目: ound (cs.SD); Artificial Intelligence (cs.AI)
备注: Accepted by EMNLP 2026 main conference
Abstract:Current TTS systems typically rely on open-loop, single-pass generation and can produce sporadic local prosodic defects, such as misplaced stress, unnatural pauses, or flattened intonation, that utterance-level metrics often fail to expose. We present LoopTTS, a judge-guided Filter-Judge-Refiner framework for recovering low-quality TTS outputs diagnosed by an AudioLLM. Given an initial utterance from a base TTS model, an AudioLLM Judge identifies salient prosodic issues and generates structured refine instructions; a Refiner, our fine-grained instruction-following TTS model, then performs guided expressive re-synthesis conditioned on the initial utterance, target text, and instruction. To train the Refiner, we construct Refiner-DB, a 42K-example AudioLLM-annotated dataset with word-level prosodic weak supervision. Human evaluation on diagnosed low-quality utterances shows that LoopTTS can detect perceptually salient errors and correct them with the Refiner, outperforming raw generated audio and practical open-loop re-generation baselines in recovery quality. The Refiner also demonstrates stronger instruction-following ability for stress and pause control in targeted prosody modification.
[AI-165] Efficient GPU Retrieval for Semantic Search
链接: https://arxiv.org/abs/2608.28968
作者: Dhritiman Das,Chujie Zheng,Ronak Kaoshik,Pratik Dixit,Vishal Shah,Yanbo Li,Jiahao Xu,Manika Agarwal,Chinmay Naik,Lingyu Zhang,Chetan Bhole,Chirag Bhanuprasad Mehta,Meng Zheng,Puneet Singh Ahluwalia,Shirisha Singh,Ping Jin,Manas Apte,Gokulraj Mohanasundaram,Tugrul Bingol,Raghavan Muthuregunathan,Fedor Borisyuk
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 14 pages, 1 figure, 13 tables
Abstract:Semantic Search on LinkedIn must retrieve relevant profiles from a corpus of hundreds of millions in response to natural-language queries such as “a fintech founder in Berlin who worked in payments.” The deployed relevance policy is bottleneck-oriented: every active non-negotiable facet must be satisfied, and a pre-existing LLM Graded Relevance (GR) judge operationalizes this through a fixed min/median aggregation over facet grades. Cosine similarity instead averages evidence, letting a strong match on one facet mask failure on another, capping the recall of the first-stage (L0) retriever. We present a policy-aligned retrieval framework: embeddings are partitioned into eight category-supervised segments whose scores follow the same min/median rule at serving time; for multi-vector retrieval, this segment score is computed independently per tagged document slot and maximized across slots. A lightweight single-slot Stage-1 scorer generates high-recall candidates, while scale-invariant relative-norm gating keeps category activation consistent across training, evaluation, and serving. On 21K held-out queries, this representation improves offline relevance over a matched-capacity baseline, with gains broadly distributed across facet combinations. We serve this framework with a two-stage GPU architecture: an FP8 coarse ranker scores the full corpus, increasing per-shard capacity by 71% and Stage-1 matmul throughput by 36%, then an FP16 stage exactly re-ranks an oversampled candidate set, recovering 99.6-99.8% of full-FP16 recall at over 500 QPS per shard replica. In a member-randomized A/B test, exploratory-query Precision@10 under the unchanged GR judge rises from 63.7% to 79.0% and navigational Precision@1 from 65.5% to 74.7%, with a blinded human evaluation independently confirming the Precision@10 gain. Comments: 14 pages, 1 figure, 13 tables Subjects: Artificial Intelligence (cs.AI); Machine Learning (cs.LG) Cite as: arXiv:2608.28968 [cs.AI] (or arXiv:2608.28968v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.28968 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-166] From Location Phrases to Geographic Entities: Task-Adapted Retrieval for People Search
链接: https://arxiv.org/abs/2608.28965
作者: Yanbo Li,Chujie Zheng,Jiahao Xu,Chetan Bhole,Lingyu Zhang,Puneet Singh Ahluwalia,Kevin Nguyen,Raghavan Muthuregunathan,Santhosh Sachindran,Sachin Ahuja,Fedor Borisyuk
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 11 pages, 1 figure, 9 tables
Abstract:People search must map free-form location phrases to geographic entities used as structured retrieval filters. Lexical standardizers handle canonical names well but are brittle to aliases, misspellings, metropolitan expressions, and same-name ambiguity. We formulate this task as graded, set-valued entity retrieval over a fixed ontology. We identify three coupled design requirements: distinguishing identity-preserving variation from knowledge-dependent aliases, controlling false negatives among valid same-name entities, and separating stable transformations from mutable entity knowledge. We realize them in a prompt-asymmetric bi-encoder with calibrated alias support, bounded ambiguity-aware negatives, and editable entity documents that support localized updates without retraining. Across a fixed production-derived development benchmark and a public GeoNames transfer task, task adaptation improves substantially over frozen encoders and standard token baselines. Controlled development ablations show that specialized supervision contributes beyond standard task fine-tuning and encoder scaling. On GeoNames, the adapted model improves known-target Recall@1 throughout zero-to-moderate character overlap, while character n-grams retain a small aggregate Target Recall@5 advantage. In a blinded human comparison on a stratified production challenge set, our model raises relevant P@1 from 28.0% to 46.0% (p=0.012). Fixed-query endpoint estimates improve on non-canonical queries and remain close to control on frequent queries; a randomized live experiment detects no engagement regression. These results support task-adapted geographic entity retrieval as a practical replacement for the incumbent taxonomy-based standardizer, with the largest relevance gains on non-canonical queries. Comments: 11 pages, 1 figure, 9 tables Subjects: Artificial Intelligence (cs.AI); Machine Learning (cs.LG) Cite as: arXiv:2608.28965 [cs.AI] (or arXiv:2608.28965v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.28965 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-167] From the Loss Landscape to Diverse Feature Learning in Neural Networks
链接: https://arxiv.org/abs/2608.28948
作者: David Aram Yunis
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: PhD dissertation. This article draws from arXiv:2408.11804
Abstract:Over the course of the last decade, neural networks have grown from an academic curiosity to moving the markets of nations. Despite this explosion in both research and deployment, relatively little is understood about how they achieve the solutions they do. This is both scientifically relevant, and pressing for society. When neural networks make decisions across self-driving, construction, law, hiring and health, there have been and will continue to be unintended consequences. However, attempting to generalize the failures of the largest and most important production systems makes for a very difficult task. Yet signs of these failures exist at all scales of neural networks, so we should be able to study a much more tractable setting. All neural networks must undergo an optimization process, called training, to be useful. To a great degree, understanding neural networks is understanding their optimization: through what process and exposure to which data did they arrive at their results. Yet our knowledge on this topic as a field is quite imprecise. In particular, a curious phenomenon called mode connectivity, the ability to connect neural networks in the loss surface, defies explanation entirely. This dissertation elucidates, explains and exploits this special structure in the loss landscape… Comments: PhD dissertation. This article draws from arXiv:2408.11804 Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI) Cite as: arXiv:2608.28948 [cs.LG] (or arXiv:2608.28948v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2608.28948 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-168] Oculi: A Conversational Agent ic Platform for Automated Credit Risk Analysis
链接: https://arxiv.org/abs/2608.28944
作者: Vennise Ho,Kristian Diana,Sandy Mourad,Milena Pilipovic,Vineel Nagisetty,Hossein Hajimirsadeghi
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: This work was completed at Royal Bank of Canada as part of the RBC Amplify program
Abstract:Credit risk analysis in financial institutions traditionally requires analysts to manually write SQL queries, run statistical computations, and build visualization dashboards. This is a time-consuming workflow that limits exploration to familiar segments. We introduce \textbfOculi, a conversational platform that transforms natural language questions into comprehensive credit risk analyses, complete with data queries, statistical testing, and interactive visualizations. Oculi employs a three-layer architecture that separates reasoning (LLM-powered agent), execution (Model Context Protocol tool servers), and presentation (agentic UI), enabling analysts to discover high-risk portfolio segments. Within Oculi, a new segment discovery pipeline is proposed that combines deterministic statistical methods with LLM-guided feature selection, leveraging LLM semantic domain knowledge alongside data-driven metrics to identify meaningful, actionable portfolio segments. Evaluated on a mortgage portfolio with 200+ features, Oculi demonstrates effectiveness in discovering material risk segments previously intractable through manual exploration, reducing time-to-insight significantly while maintaining auditability and statistical rigor.
[AI-169] MineCEraft: Evaluating Language Models as Construction Engineers in the World of Minecraft
链接: https://arxiv.org/abs/2608.28884
作者: Sewoong Lee,Risham Sidhu,Julia Hockenmaier,Yoonhwa Jung
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:We introduce MineCEraft (Minecraft Construction Engineering Benchmark, pronounced mine-see-ee-raft), an easy-to-use, open-source benchmark designed to systematically evaluate the reliability and limitations of LLMs for construction tasks in Minecraft. The MineCEraft benchmark comprises 723 domain-expert hand-crafted natural-language instructions with programmatically verifiable evaluation, spanning 17 distinct task categories, providing a safe and controllable experimental environment for assessing LLMs’ ability to perform realistic construction engineering tasks. With this benchmark, we conduct an in-depth evaluation of state-of-the-art LLMs and perform a detailed error analysis, revealing key failure modes and practical challenges in applying LLMs to construction engineering tasks.
[AI-170] Hybrid Offline-Online Multi-Agent Decision Transformers for Wireless Resource Management ICASSP
链接: https://arxiv.org/abs/2608.28878
作者: Yiming Zhang,Kun Yang,Cong Shen,Dongning Guo
类目: ystems and Control (eess.SY); Artificial Intelligence (cs.AI)
备注: 11 pages, 9 figures, 3 tables. Submitted to IEEE Journal on Selected Areas in Communications in Aug 2026. The offline training part was presented at the 2025 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)
Abstract:This paper develops a hybrid offline-online multi-agent reinforcement learning framework based on decision transformers. The policy is first pretrained offline via supervised sequence modeling of trajectories generated by existing policies, providing a safe and sample-efficient initialization. It is then fine-tuned online using a hybrid objective that incorporates critic-guided gradients, enabling performance improvements beyond the offline policy. To facilitate stable offline-to-online transfer and effective multi-agent coordination, the framework incorporates return-weighted sampling, a critic conditioned on neighbors’ actions, and neighborhood-correlated exploration. The approach is fully distributed: both training and execution rely only on local observations and limited information exchange among neighboring agents. Evaluations with dynamic traffic arrivals in two settings: (i) joint scheduling and power allocation and (ii) coordinated beamforming, show that the proposed method achieves quality-of-service (QoS) performance comparable to centralized methods. Moreover, when pretrained on lower-quality datasets, online fine-tuning is also observed to surpass the initial offline policy. These results demonstrate a promising learning-based alternative for wireless resource management.
[AI-171] Equivariant Sheaf Neural Networks: Learning Geometric Transport on Graphs
链接: https://arxiv.org/abs/2608.28853
作者: Alessio Borgi,Mario Severino,Fabrizio Silvestri,Pietro Liò
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Equivariant graph neural networks provide a principled way to model geometric systems, but efficient first-order architectures remain limited in how vector information can be transformed as it moves across a graph. We introduce \textscESNN, an Equivariant Sheaf Neural Network that enriches this interaction by learning directed, matrix-valued transport between neighboring vector features while preserving exact Euclidean equivariance. Rather than increasing the order of the representation, ESNN keeps scalar and vector features first-order and places the additional geometric flexibility in the edge transport itself. We characterize this transport theoretically, showing that when relative displacement is the only covariant geometric input, every linear O(n) -equivariant map decomposes into independent radial and tangential components, while learned covariant features enable richer feature-conditioned transformations. We also introduce controlled symmetry relaxation for systems with a preferred ambient direction, which may be prescribed or inferred from data while recovering full E(n) -equivariance when the directional pathway is inactive. Across particle dynamics, mesh-based simulation, point-cloud classification, and molecular property prediction, ESNN improves dynamics prediction, recovers the gravity axis when symmetry is broken, yields substantial gains on selected mesh tasks and long-horizon rollouts, and remains robust to unseen rotations. These results show that learning how geometric information is transported across edges offers a complementary route to expressive equivariant message passing without requiring higher-order representations.
[AI-172] Curvature Cryptanalysis of Smooth Transformer Feed-Forward Networks
链接: https://arxiv.org/abs/2608.28843
作者: Munawar Hasan,Apostol Vassilev
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
备注: 25 pages, 8 tables, 4 figures
Abstract:We show that smooth two-layer feed-forward networks (FFNs) expose an additional structural model extraction channel under a chosen-input raw-output oracle at the FFN branch; consider transformer FFN branches with GELU or SiLU activations under chosen-input raw-output access, without access to parameters, gradients, or internal activations; exploit a second-order leakage channel in which projected input Hessians form different mixtures of the same hidden symmetric rank-one factors induced by the FFN input weights. We formalize resulting Hessian collection as a partially symmetric decomposition to establish conditions for local identifiability and stability to exploit vector-output stencil reuse to reduce the structural query cost by a factor of 16. On independently trained CIFAR-10 vision transformers, only 16 projected Hessians, corresponding to 8193 black-box queries, recover the hidden FFN directions with average absolute cosine alignment above 0.94, with 95.1 % of GELU and 91.9 % of SiLU directions exceeding 0.90 alignment. Recovery remains high across independently trained models, repeated extraction runs, and all transformer blocks. The recovered structure supports functional extraction too. Keeping the recovered directions fixed and fitting only the remaining FFN parameters yields high-fidelity substitutes with more than 93 % top-1 agreement, while test accuracy remains within 0.90% and 0.62% of the GELU and SiLU targets. Output rounding and Gaussian noise substantially reduce recovery under a fixed attack configuration, but adapting the finite-difference step restores average alignment to 0.9603 and 0.9398. This is an end-to-end path from black-box second-order observations to hidden FFN-structure recovery and functional replacement. Under the stated oracle model, smooth FFN curvature exposes internal parameter geometry that behavioral fidelity alone cannot reveal.
[AI-173] Evaluating the Hidden Costs of Personalization in Large Language Models
链接: https://arxiv.org/abs/2608.28833
作者: Yumeng Wang,Yuchen Wu,Cheng Qian,Zhiyuan Fan,Hyeonjeong Ha,Shujin Wu,Jiayu Liu,Heng Ji,Ge Wang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:While Large language models (LLMs) incorporate user personalization signals to improve usability and helpfulness, they increasingly shift from providing balanced, informative responses toward optimizing for user satisfaction when conditioned on personal context such as conversation history, inferred preferences, and user profiles. Specifically, we identify three emerging risks: (1) irrelevant personalization, where models reference personal information in unnecessary contexts; (2) preference narrowing, where models reinforce informational echo chambers; and (3) sycophantic bias, where models agree excessively with user opinions. As a result, models may reference personal information in contexts where it is unnecessary, inadvertently collapse response diversity, or agree excessively with user opinions. Despite the growing use of personalization in AI assistants, there has been limited systematic evaluation of its potential side effects. To bridge this gap, we propose PRISK, a dynamic evaluation framework with automated data generation and tailored metrics that uncovers systematic limitations in current LLM personalization and how personalized information shapes its responses. Our empirical analysis across 13 LLMs demonstrates the presence of user profiles and retrieved memories consistently exacerbates biases, resulting in an average drop of 45.9% in irrelevant personalization, 41.7% in preference narrowing and 61.7% in sycophantic bias.
[AI-174] Discovering Machine Correlates of Consciousness
链接: https://arxiv.org/abs/2608.28824
作者: Romain Salvi,Ouri Wolfson
类目: Artificial Intelligence (cs.AI)
备注: published in: Springer Lecture Notes in Artificial Intelligence Vol. 16855, Proceed- ings of the 19th International Conference on Artificial General Intelligence (AGI- 2026), San Francisco, CA, July 2026. pp. 237-255 this https URL 032-33195-3_18
Abstract:Currently, in biological systems Neural Correlates of Consciousness (NCCs) are characterized in terms of EEG and FMRI signals. Unfortunately, this characterization prevents the transferability of the NCCs concept to machines. Such transferability would be useful in order to investigate AI consciousness. In this paper we provide an alternate characterization that is transferable, and enables the analogous definition of Machine Correlates of Consciousness (MCCs). Specifically, we propose that NCCs (MCCs) are substrate-level signals that are not under human (AI agent) control, and that are reliably modulated by emotions. This paper presents the first empirical investigation of MCCs. Specifically, we present the results of experiments conducted with two LLMs, Llama-2 7B and Llama-3.1 70B parameters. In these LLMs we collect hardware anomaly traces that are substrate-level indicator-sequences. And we show that after controlling for confounding factors, these are modulated differently by emotional and neutral computations. And this difference is statistically significant for the larger Llama-3.1 70B, but not for the smaller Llama-2 7B. The results constitute initial empirical evidence that MCCs are present in the Llama-3.1 70B configuration. And they are consistent with the hypothesis that consciousness probability and degree increase with the LLM sophistication. Independently of consciousness, MCCs can also be used for detection of emotions in AI agents. Comments: published in: Springer Lecture Notes in Artificial Intelligence Vol. 16855, Proceed- ings of the 19th International Conference on Artificial General Intelligence (AGI- 2026), San Francisco, CA, July 2026. pp. 237-255 this https URL 032-33195-3_18 Subjects: Artificial Intelligence (cs.AI) MSC classes: I.2.m ACMclasses: I.2.m Cite as: arXiv:2608.28824 [cs.AI] (or arXiv:2608.28824v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.28824 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Journalreference: Springer Lecture Notes in Artificial Intelligence Vol. 16855, Proceedings of the 19th International Conference on Artificial General Intelligence (AGI-2026), San Francisco, CA, July 2026. pp. 237-255 Related DOI: https://doi.org/10.1007/978-3-032-33195-3_18 Focus to learn more DOI(s) linking to related resources
[AI-175] Capability-Stratified Degradation in Ternary Language Models
链接: https://arxiv.org/abs/2608.28809
作者: Anirudh Malik,M Sparsh Mehra,Poojith Devan
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Extreme low-bit inference offers a route toward smaller models and constrained deployment. Ternary language models restrict weights to -1,0,+1\ , approaching the limit of \log_2 3 \approx 1.585 bits/weight. The practical question for a pretrained model is not simply whether weights can be quantised but which capabilities survive and whether it remains useful for adaptation. We explore this by converting Qwen3.5-0.8B (752M parameters) to ternary weights using 72.4M tokens of quantisation-aware training (QAT). The resulting model, Cloe, is evaluated across 29 benchmarks, representation diagnostics, and downstream fine-tuning. The evidence shows non-uniform degradation. A linear probe recovers 43.76% of MMLU answers from the full-precision teacher’s representations but only 26.19% from Cloe (near chance), indicating specialist factual information is lost. However, Cloe retains measurable performance on ten tasks, averaging 77.1% of teacher performance. Crucially, fine-tuning raises Cloe to 89.8% on SST-2 (95.6% of the matched teacher) and reaches 79.4% teacher retention on XSum. We attribute degradation to a combination of quantisation-induced information loss and incomplete recovery due to the limited QAT budget. We also highlight an evaluation pitfall: standard answer-letter scoring failed (Cloe emitted “A” on 98.6% of MMLU questions), necessitating continuation scoring. Ultimately, ternary conversion is unsuitable as a drop-in general replacement yet remains valuable as a compact substrate for task-specific models.
[AI-176] Enhancing SAE-based Steering via Neighbor Integrated Feature Selection
链接: https://arxiv.org/abs/2608.28806
作者: Yutian Liu,Xu Wang,Difan Zou
类目: Artificial Intelligence (cs.AI)
备注: 15 pages
Abstract:Sparse autoencoders (SAEs) disentangle model activations into interpretable features and are widely used for steering large language models. Most existing SAE-based steering methods select features by applying a top- filter based on statistical scores, assuming that higher-scoring features yield stronger steering effects. In this paper, we show that this assumption is often invalid, leading to suboptimal feature selection. Our analysis reveals that effective steering features may be distributed among representationally adjacent, semantically similar groups induced by feature splitting in SAEs. Within such groups, features may exhibit disparate statistical scores despite having comparable steering influence, causing score-based selection to overlook important features. Based on these observations, we propose \textscNeighbor Integrated Feature Selection (\textscNIFS), a plug-and-play strategy that leverages representation similarity to improve feature selection for steering. We evaluate \textscNIFS across multiple SAE-based steering methods and tasks, and demonstrate consistent performance gains over conventional top- k selection.
[AI-177] he reach of a verification tool decides its value: A controlled study of verification surface artifact quality and cost in AI coding agents
链接: https://arxiv.org/abs/2608.28795
作者: Achint Mehta
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: 27 pages, 13 figures, 10 tables. Submitted to IEEE Access. Data and code: this https URL
Abstract:Modern artificial-intelligence coding agents can be equipped with tools for checking their own work e.g. a linter, a boot probe, a shell, a screenshot tool. We call this set the agent’s verification surface. This study asks whether increasing only that surface, with everything else held fixed, produces a matching growth in the quality of the software the agent ships. We built a minimal coding agent whose tool list is the single controlled variable and used it to implement 1,116 web applications across six models and eight tool configurations. A condition-blind human graded every application against a frozen rubric, and automatic probes stress-tested the API-observable behaviors. Verification’s cheapest benefit arrives first, which is to make sure that the application comes up. Without any tools, about one build in seven fails to launch at all and a single boot probe removes nearly all of these failures at roughly 35 percent of a full shell’s token cost, while the full shell multiplies the no-tools cost by 2.35. Screenshots help most where mistakes are visible (e.g. element placement, interaction), though even there the gain over a shell is modest and does not survive correction for multiple statistical comparisons. In cases where failures can only be measured rather than seen, such as keeping scrolling smooth over a 100,000-row list, screenshots add nothing. A verification tool improves the output artifact only where its reach covers the way the application actually fails.
[AI-178] Efficient Geothermal Well-Control Optimization via Diffusion-Surrogate Reinforcement Learning
链接: https://arxiv.org/abs/2608.28791
作者: Ruimin Dai,Guodong Chen,Randy Harsuko,Kunpeng Liu,Nori Nakata
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Real-time decision-making for enhanced geothermal systems (EGS) is challenging because long-term production periods involve high-dimensional control spaces and a large number of time-consuming high-fidelity hydrothermal simulations. Reinforcement learning provides a natural framework for state-dependent sequential control, but direct policy training with numerical simulators is computationally expensive. To address this issue, we propose a diffusion-surrogate guided reinforcement learning framework for long-horizon EGS well-control optimization. The reservoir temperature and pressure fields are used as system states, while injection rates are selected as control actions. A learned surrogate environment is constructed using conditional diffusion models to predict the evolution of reservoir temperature and pressure fields and a separate reward model to estimate the corresponding economic return. The surrogate environment is then integrated with Proximal Policy Optimization (PPO) for efficient policy training. Experiments on a fractured EGS benchmark show that the diffusion surrogate can accurately reproduce reservoir-state evolution over multiple control stages. The resulting surrogate-assisted PPO policy achieves competitive well-control performance compared with direct simulator-based PPO and existing optimization methods, while substantially reducing the dependence on expensive high-fidelity simulations. These results demonstrate the potential of diffusion-based surrogate environments for efficient reinforcement learning in geothermal well-control optimization.
[AI-179] PermitGPT : A Unified Generative-AI Pipeline for Construction Hazard Forecasting Permit Prediction and Community Impact
链接: https://arxiv.org/abs/2608.28728
作者: Mohd Ruhul Ameen,Farjana Aktar,Akif Islam,Momen Khandoker Ope,Abu Saleh Musa Miah,Jungpil Shin
类目: Artificial Intelligence (cs.AI)
备注: 1 figure, 3 tables, Accepted at 2026 International Conference on Power, Electronics, Communications, Computing, and Intelligent Infrastructure (PECCII)
Abstract:Urban construction governance requires early decisions that connect workplace safety, permitting requirements, and community impact, yet the relevant evidence is often scattered across separate municipal and regulatory data sources. This paper presents PermitGPT, a unified generative artificial intelligence framework for converting unstructured construction permit descriptions into structured decision-support outputs across three domains: safety hazard identification, permit requirement specification, and community impact assessment. To address data fragmentation, we spatially and temporally align records from the New York City Department of Buildings, Occupational Safety and Health Administration, and NYC 311 service requests, producing 90,000 structured prompt-response pairs derived through rule-based alignment and domain-informed spot checking. We fine-tune three open-weight language models using parameter-efficient adaptation and evaluate them on 2,833 held-out test cases. The results show complementary model behavior: Gemma-3-1B provides the most efficient inference at 3.07 samples per second with low memory usage, Llama-3.2-3B gives the highest lexical overlap for regulatory-style outputs with a BLEU score of 0.0091, and 4-bit Mistral-7B-Instruct-v0.3 achieves the strongest semantic alignment with a BERTScore-F1 of 0.7747. Because the task involves open-ended structured generation, low BLEU values are interpreted alongside semantic metrics and qualitative output structure rather than as standalone indicators of utility. Overall, PermitGPT provides an initial step toward AI-assisted construction governance while identifying directions for stronger task-level evaluation and real-world validation.
[AI-180] Pro-Router: Token-Aware Progressive Model Routing with Adaptive Edge-Cloud Collaboration for Efficient Multimodal LLM Inference
链接: https://arxiv.org/abs/2608.28726
作者: Xinyuan Gui,Shaowen Wang,Sheng Sun,Zijian Wang,Zishu Yu,Zheming Yang
类目: Artificial Intelligence (cs.AI)
备注: 9 pages, 7 figures, 2 tables. Code: this https URL
Abstract:The remarkable performance of multimodal large language models (MLLMs) comes at the cost of substantial computational overhead, posing significant challenges to real-time deployment and cost effectiveness. Existing model routing approaches either decide from coarse request-level features alone or spend one or several extra language model passes to inspect the generated response, leaving the token-level uncertainty signals that emerge during generation unused. To address these limitations, we propose Pro-Router, a token-aware progressive model routing method with adaptive edge-cloud collaboration for efficient multimodal LLM inference. Pro-Router employs a two-stage progressive decision mechanism. First, a lightweight prompt pre-scorer module performs rapid pre-screening before token generation begins, guiding apparently simple requests to small models. Second, a token-aware verifier reads the sampling probability distribution of each token the small model generates, estimating the model’s confidence in its own output to determine, per request, whether the answer ships or escalates to the cloud-based high-precision model. Furthermore, we design an adaptive edge-cloud serving pipeline that sizes every dispatch to each device’s measured service rate, so both the edge and the cloud tiers stay fully utilized without manual parameter tuning and are not impacted by the network latency. Extensive experiments on multiple multimodal benchmark datasets and models demonstrate the effectiveness of Pro-Router. Compared to other methods, it achieves the highest routing accuracy and improves routing speed by more than 10x. Its serving pipeline also reaches more than 75% higher end-to-end throughput than the existing model routing pipeline. Our code is available at this https URL.
[AI-181] Beyond the Answer Key: Robustness Evaluation of Large Language Models for Step-Level Mathematical Verification ICML
链接: https://arxiv.org/abs/2608.28725
作者: Fateme Mazdarani,Carlos Toxtli
类目: Artificial Intelligence (cs.AI)
备注: Accepted to 2026 IEEE International Conference on Machine Learning and Applications (ICMLA)
Abstract:Large language models (LLMs) are increasingly used as graders, verifiers, and process auditors, but most mathematical evaluations still emphasize final-answer accuracy. This can obscure whether a model can verify a non-canonical but valid solution trace. We introduce a controlled linear-equation benchmark for evaluating LLMs in the evaluator role. Each instance asks the model to judge final-answer correctness, step-level trace correctness, and the first incorrect step. Our evaluation of state-of-the-art open LLMs reveals a significant robustness gap: models that accurately evaluate canonical solutions often fail when presented with perturbed but logically equivalent variants. Across GPT-OSS 20B, Qwen3-14B, and Phi-4-Reasoning, base models perform well on canonical traces but degrade substantially on perturbed traces, especially for error localization. On valid perturbed traces, base-model false-rejection rates reach 75.6-85.3%, showing strong sensitivity to canonical solution form. Supervised fine-tuning, distillation, and test-time compute improve robustness in some settings, but gains are model dependent and can trade off against canonical performance. The results show that reliable process-level verification remains challenging, and evaluator robustness should be measured separately from solver accuracy, even in a simple algebraic domain with exact ground truth.
[AI-182] ORDDAR: Observation-Driven Reasoning for Distortion-Resilient Decision Action and Cognitive Recovery
链接: https://arxiv.org/abs/2608.28704
作者: Deblina Kar,Anant Nawalgaria,Shyamal Kumar Das Mandal
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:AI agents increasingly perform long-term reasoning, planning, tool use, memory integration, and autonomous decision making, yet erroneous intermediate states can propagate and cause inconsistent decisions and unreliable outputs. Existing reasoning approaches mainly rely on iterative planning, self-reflection, augmented memory, or verification, but rarely localize and selectively repair faulty reasoning. We present ORDDAR (Observation-Driven Reasoning for Distortion-Resilient Decision, Action, and Cognitive Recovery), a reasoning framework that models reasoning as cognitive state transitions, detects localized distortions, retrieves related reasoning from prior experiences, and repairs only the affected states. ORDDAR therefore performs recovery at the local reasoning-transition level rather than regenerating the complete trajectory. Experiments across mathematical, commonsense, multi-hop, and clinical reasoning benchmarks demonstrate improved reasoning quality, recovery ability, and interpretability over multiple evaluated reasoning baselines.
[AI-183] FRAC-MAS: A Safe and Explainable Multi-Agent System for Fracture Diagnosis ICME MICCAI
链接: https://arxiv.org/abs/2608.28662
作者: Hardik Iyer,Tirath Bhathawala,Mihir Panchal,Ying-Jung Chen,Kiran Bhowmick,Pankaj Sonawane,Meera Narvekar
类目: Artificial Intelligence (cs.AI)
备注: Medical Image Computing and Computer Assisted Intervention (MICCAI) 2nd Workshop on Agentic AI for Medicine (AgenticMed)
Abstract:Fracture detection and its clinical interpretability see notable improvements when deep vision models are integrated with agentic AI architectures. While deep learning models achieve high diagnostic performance, their black-box nature limits clinical adoption. We propose FRAC-MAS, an agentic AI system for automated, explainable, and safe bone fracture detection. The framework combines a stacked ensemble of four vision models with conformal prediction to produce statistically grounded differential diagnoses, while a multi-agent workflow performs independent verification, retrieves clinical guidelines, and generates patient-friendly reports. A pipeline-depth ablation study confirms that our multi-agent critic triages 86.6% of cases into a high-confidence auto-confirmed cohort while escalating uncertain cases, outperforming a single-agent baseline. Patient preference studies against Llama, MedGemma, and Gemini further demonstrate significantly more comprehensible clinical reports. These results suggest that integrating multi-agent critics with conformal guarantees enables safer radiology triage while preserving clinician oversight. More broadly, FRAC-MAS demonstrates how cooperative agentic architectures can serve as auditable, human-in-the-loop decision support systems for safety-critical healthcare. Our code is available at this https URL, and the website is available at this https URL.
[AI-184] A Generalized Optimization Engine (GOE) for Edge AI Inference Acceleration
链接: https://arxiv.org/abs/2608.28652
作者: Venkat R. Dasari,Jakob A. Adams,Vinod K. Mishra,Brian Jalaian
类目: Artificial Intelligence (cs.AI)
备注: 14 pages, compressed version is under review in IEEE MILCOM 2026
Abstract:Artificial intelligence (AI) models have demonstrated remarkable capabilities across various domains, yet their widespread deployment is impeded by significant computational costs, particularly on resource-constrained devices. This paper explores the theoretical underpinnings of various AI model optimization techniques, algorithms, and abstractions, discussing their potential to reduce computational complexity, memory footprint, latency, and power consumption. Furthermore, we propose a comprehensive hardware (HW) and model-agnostic generalized optimization architecture that integrates these techniques for improved efficiency. Our study underscores the critical role of such a generalized optimization system in preparing model deployment over resource-constrained heterogeneous hardware in a tactical environment. As a concrete demonstration, we show that GOE-compressed language models deploy and run on a GPU-less edge CPU, and that the choice of compression method, not merely its nominal bit-width, determines whether task accuracy survives deployment.
[AI-185] Self-Specialized Teachers for Domain Post-Training
链接: https://arxiv.org/abs/2608.28647
作者: Yifei Li,Rongman Xu,Lingling Zhang,Muye Huang,Zihan Ma,Jiashuai Liu,Hang Yan,Heng Wang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Target-only post-training can improve performance in a specialized domain while degrading behaviors that a general-purpose base model acquired before adaptation. We study this problem when target-domain data are available but a representative replay corpus is not. We propose self-specialized teacher distillation (SSTD), a two-stage procedure that first trains a copy of the base model into a domain teacher, then distills its token distribution to a student on prefixes sampled from the student itself. Teacher training combines standard target supervision with base-aware key-token weighting and distribution alignment to the frozen base model; on-policy distillation then places domain feedback on states the student can encounter at inference time. On financial numerical reasoning, medical question answering, and legal holding identification, SSTD retains much of the target improvement of direct fine-tuning while improving the mean score on the evaluated general suite by 4.8–5.0 points at the reported operating point. The pattern persists across Qwen3 sizes and on Gemma backbones. SSTD requires neither an external teacher nor general replay data.
[AI-186] BiasMix-Finance: Post-Generation KYC Guardrails for LLM Portfolio Advice KR ICLR2026
链接: https://arxiv.org/abs/2608.28646
作者: Gaurav Kukreja,Parul Kukreja,Mohammed Abraar,Raj Dandekar,Rajat Dandekar,Sreedath Panat
类目: Artificial Intelligence (cs.AI)
备注: 18 pages, 9 figures, 8 tables. Accepted to the ICLR 2026 Financial AI (FinAI) Workshop. Code and data: this https URL
Abstract:Large language models (LLMs) can generate plausible-sounding ETF portfolios while silently violating basic KYC-style constraints on risk, fees, and diversification. This is especially problematic in agentic multi-turn advisory systems, where each draft recommendation can become an action unless guarded by an auditable enforcement layer. We study a model-agnostic, asset-agnostic post-generation guardrail pipeline: (i) enforce a strict JSON allocation schema, (ii) validate allocations against numeric caps, and (iii) when violations occur, deterministically project the output to the nearest feasible portfolio via a convex quadratic program (QCQP). We introduce BiasMix-Finance (Mini), a compact stress-test benchmark for constrained decision-making under biased LLM generations, with a 16-ETF universe, three investor profiles, and eight bias prompts. Across three models and three inference modes (direct, critique, self-consistency), first-pass generations violate at least one cap in 47.6-85.7% of test cases (67.2% pooled), but the convex projection layer reduces final feasibility violations to 0% while requiring only a small correction distance (test pooled median D=||w*-w0||_2=0.066), indicating that the guardrail typically preserves the intent of the original allocation. We report violation rates and correction distances with confidence intervals, and paired model comparisons with multiple-testing correction. To support reproducibility, we release the dataset, prompts, caps, and code in our public GitHub repository.
[AI-187] From Extraction to Governed Memory: Multi-Agent Knowledge Graph Construction with Domain-Expert Review
链接: https://arxiv.org/abs/2608.28642
作者: Pranav Bykampadi,Neel Mokaria,Vishesh Narayan,Faizan Wajid,Ashok Agrawala
类目: Artificial Intelligence (cs.AI); Computational Engineering, Finance, and Science (cs.CE); Digital Libraries (cs.DL); Emerging Technologies (cs.ET); Machine Learning (cs.LG)
备注: 23 pages total, 10 pages main text, 12 pages of appendix, 2 figures, 6 tables
Abstract:Knowledge graphs used by agentic systems are often treated as flat stores of extracted triples, with little record of who owns a fact, why it was admitted, or how it should be used downstream. We argue that reliable agentic knowledge systems require governance as an essential component of graph construction to bridge this gap. We propose MAGG, a principled multi-agent framework for constructing Governed Knowledge Graphs that introduces explicit governance decisions for reliable and trustworthy knowledge sharing. A domain classifier first induces entity and relation types directly from document content, enabling operation in open-world settings without fixed schemas. Candidate triples are assigned to domain owners, reviewed against supporting evidence, admitted through governance decisions, and stored with audit metadata. The same ownership structure is reused during question answering, where queries are routed to domain-specific graph experts rather than answered through undifferentiated retrieval. Our evaluation demonstrates MAGG’s effectiveness: On SciERC, MAGG improves strict triple F1 by 47% and mapped triple F1 by 51% over flat insertion. A blinded review of 120 triples finds governed-only triples more often source-supported than flat-only ones, and revised triples supported in 100% of cases. Finally, on MuSiQue, MAGG outperforms Microsoft GraphRAG by 9.0 exact-match points and 11.2 token-F1 points.
[AI-188] Reward-Oracle MCTS for Formal Theorem Proving: Sample-Efficient Search and the Need for Kernel-Level Proof Auditing
链接: https://arxiv.org/abs/2608.28639
作者: Bodla Krishna Vamshi,Haizhao Yang
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Logic in Computer Science (cs.LO)
备注:
Abstract:Formal theorem proving with large language models remains challenging due to the difficulty of navigating large proof search spaces efficiently. Existing tree search approaches either feed verbose compiler error messages directly into the generation context, increasing context usage during search, or employ non-standard evaluation protocols that prevent direct comparison with established baselines. We propose a three-role Monte Carlo Tree Search (MCTS) framework that treats the Lean 4 compiler purely as a reward oracle using compiler output as a scalar signal for UCB-guided tree updates without feeding error content into the generation context. Our framework decomposes proof search into three roles: a generator for proof attempts, a decomposer for subgoal decomposition, and a critic for subgoal quality evaluation. We evaluate across 4 benchmarks spanning competition mathematics and physics (MiniF2F, PutnamBench, LeanPhysBench, PhysLeandata) with three prover models at standard proof attempt budgets (PAB@16 to PAB@256). Our method achieves 87.1% on MiniF2F with Goedel-Prover-V2-8B at PAB@256 and solves 26/659 PutnamBench problems at PAB@32 surpassing base sampling 18/659 at same proof attempt budget. Through an exhaustive axiom-level audit of every compiled proof, we further identify reward hacking in search-based theorem proving: DeepSeek-Prover-V2-7B produces proofs on PutnamBench that pass compilation and the standard sorry-token scan while depending on sorryAx. The audit removes 4 and 8 such proofs from whole-proof sampling at PAB@32 and PAB@128, and 11 and 19 from MCTS. We do not attribute these counts to the search procedure; we report them to establish that kernel-level auditing is necessary for compiler-verified evaluation.
[AI-189] Self-Evolving Skills via Surrogate-Guided Solve-and-Reproduce
链接: https://arxiv.org/abs/2608.28638
作者: Jiale Liu,Pinze Ren,Yuqi Xia,Huan Wang,Zhenlin Zhao,Siming Dong
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Agent skills are portable packages of instructions and resources an agent consults at deployment. Self-evolving them fails in two ways today. First, skills evolved from scratch underperform human-curated ones and, on a weak model, using no skill at all. Second, an evolution-time pass records one lucky trajectory that a fresh stochastic agent often fails to reproduce at deployment. We present reSolve, a per-task, oracle-in-the-loop framework built on three components. It decouples interactive solving from a self-contained deliverable that is independently re-executed in a fresh container, a protocol we call solve-and-reproduce. It enhances the sparse reward signal with a surrogate verifier that cannot access hidden tests or reference answers. It then runs verifier-guided beam search over a solution-construction graph. Within a fixed harness, a cheap model self-evolves skills that reach 74.9% mean-of-3, +14.8 points over the 60.1% human-curated baseline, exceeding the strongest official curated-skill result ( 67.3% , GPT-5.5/OpenHands). We also report observed failure cases and domain-level results, including performance on the 14 Natural Science tasks, to clarify when the approach does and does not help.
[AI-190] AI Scientist Mission Control (AIMC): Visual Analytics for Human Oversight of Autonomous Scientific Discovery
链接: https://arxiv.org/abs/2608.28637
作者: Rikathi Pal,Klaus Mueller
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Autonomous scientific discovery systems can generate large numbers of research ideas, experiments, and manuscripts with minimal human intervention. As these systems become increasingly capable, scientists require effective mechanisms to monitor output quality, identify recurring failure modes, understand research evolution, and prioritize promising discoveries for review. We present AIMC, a visual analytics framework for human oversight of autonomous scientific discovery. AIMC combines semantic embeddings, automated weakness extraction, temporal analysis, and interactive visualizations to support the exploration of AI-generated research artifacts. We demonstrate the framework through a case study of the papers generated by an autonomous AI Scientist (FARS), together with their associated review feedback. Our analysis reveals recurring methodological weaknesses, evolving research themes, domain-specific differences in quality, and a small set of highly novel papers that warrant deeper human inspection. These findings illustrate how visual analytics can support transparency, diagnosis, and human AI collaboration in emerging autonomous scientific discovery workflows.
[AI-191] CrossAudit: A Git-Native Cross-Vendor Audit Loop for Agent ic Science
链接: https://arxiv.org/abs/2608.28631
作者: Zhaohe Dong,Yuhao Chen
类目: Artificial Intelligence (cs.AI); Computational Engineering, Finance, and Science (cs.CE); Computers and Society (cs.CY)
备注: 19 pages, 4 figures, 3 tables, 22 references. Reference implementation, audit ledger, and experiment artefacts: this https URL
Abstract:An AI scientist should not grade its own homework. Yet in the systems we examined, the agent that reviews the work usually comes from the same model family as the agent that produced it, or at least from the same vendor. Model evaluators are known to favour their own generations. Whether models trained alike also share blind spots is a conjecture, not a settled finding, but if they do, the reviewer inherits the author’s. The record of what was flagged and what was waved through often sits in platform logs that nobody outside can replay. We present CrossAudit, a protocol for supervising autonomous research pipelines. It rests on three commitments. Each increment of work is audited by an agent from a different vendor against a rulebook a human wrote and versioned. Reports, verdicts, disputes and rulings are git commits, so the supervision history can be re-read and cited; raw model exchanges are not yet part of that record. Scripted checks run before any model does. Advisory judgement never gates the pipeline: a model blocks only by citing a rule, and no model may waive a deterministic failure. Blockers that survive a bounded number of revision rounds go to a person. We state the protocol as eight invariants. We describe a reference implementation built from GitHub Actions and a few hundred lines of Python, and report a live deployment of a closely related variant in a computational-chemistry pipeline. We also ran a seeded-defect trial (30 increments, 43 seeded defects, one run per configuration). A cross-vendor audit of our own repository then voided its blinding. We adopt that audit’s findings and report the corrected results. The trial shows that two vendors read the same rulebook differently. It does not show that either is better. The strongest evidence here is the committed, uncontrolled record of cross-vendor audits of this paper itself. Comments: 19 pages, 4 figures, 3 tables, 22 references. Reference implementation, audit ledger, and experiment artefacts: this https URL Subjects: Artificial Intelligence (cs.AI); Computational Engineering, Finance, and Science (cs.CE); Computers and Society (cs.CY) Cite as: arXiv:2608.28631 [cs.AI] (or arXiv:2608.28631v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.28631 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-192] CDEP Agent : Connecting Meteorologically Detected Temporal Compound Events to Real-World Documentary Evidence
链接: https://arxiv.org/abs/2608.28628
作者: Zhuoran Li,Weiyi Kong,Boer Zhang
类目: Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Atmospheric and Oceanic Physics (physics.ao-ph)
备注: 12 pages, 2 figures, 6 tables
Abstract:Compound drought-to-extreme-precipitation (CDEP) events are recognized in climate science as a growing driver of extreme impact, but whether this recognition carries over into real-world early warning and post-event documentation is unknown, so a meteorologically real CDEP event may pass with neither advance warning nor any later record. Here we present CDEP Agent, an auditable LLM-agent framework that tests this mismatch directly by linking CDEP candidates detected from meteorological reanalysis to real-world hazard and impact evidence across sources with different spatial scales, temporal resolutions, and reporting conventions. Using California as a case study, we identify 408 candidate CDEP events from ERA5 observations during 2021-2025 and evaluate each against the U.S. Drought Monitor, NOAA Storm Events, and public webpages along five dimensions: antecedent drought, extreme rainfall, local impact, hazard-impact attribution, and explicit drought-to-rainfall linkage. Only 34.3% of candidates are corroborated on both hazard components, and just 1.5% are ever explicitly linked to their antecedent drought, indicating that most meteorologically detected CDEP events go undocumented and their compound nature almost never enters the record at all. Our framework gives climate scientists a way to test physical event definitions against what actually gets documented, and gives social scientists, economists, and disaster-response agencies a provenance-linked evidence base for compound events that current warning and reporting systems largely fail to capture.
[AI-193] Machine Learning-Enhanced Tabu Search for Tactical Wireless Network Design
链接: https://arxiv.org/abs/2608.28627
作者: Wissem Ahmed Zaid,Alain Hertz,Defeng Liu
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Combinatorics (math.CO)
备注:
Abstract:Designing high-performance tactical wireless networks under realistic operational constraints gives rise to challenging combinatorial optimization problems, where the evaluation of candidate solutions relies on detailed physical and traffic-aware models. Although classical metaheuristics such as Tabu Search offer effective mechanisms for exploring large search spaces, their computational cost remains high because numerous candidate moves must be evaluated at every iteration. In this paper, we propose a data-driven framework that improves the efficiency of Tabu Search by learning to guide its move selection process. Rather than altering the neighborhood structure, our approach exploits the information contained in the search trajectories generated during the optimization process. At each iteration, we record both improving and non-improving edge-based transformations together with a set of descriptive features capturing the structural, geometric, and performance characteristics of the network. This information is used to train a Graph Neural Network (GNN) that predicts the impact of candidate moves on the objective function. The trained model is then integrated into the Tabu Search algorithm to rank candidate transformations according to their predicted quality, thereby reducing the number of costly objective evaluations while maintaining an effective exploration of the search space. Experimental results on synthetic benchmark instances demonstrate that the proposed learning-assisted Tabu Search notably reduces computation time while consistently producing higher-quality solutions than the standard algorithm. These findings highlight the potential of combining machine learning with metaheuristics by leveraging the implicit knowledge embedded in search trajectories, paving the way for more efficient solution methods for large-scale network design problems.
[AI-194] PUFFER: Incremental Fuzzy Deduplication for Continuously Evolving Corpora
链接: https://arxiv.org/abs/2608.28622
作者: Xiao Yang,Erik Edward Aldape,Beren Millidge
类目: Databases (cs.DB); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language model training corpora grow through successive, often redundant releases, so each release must be deduplicated against both itself and the accumulated history. At trillion-token scale, this requires incremental ingestion, bounded resident memory, deterministic retry, and dataset-scoped lifecycle control without repeated corpus-wide rebuilding. We introduce PUFFER (Provenance-aware Updatable Fuzzy Filtering for Evolving Repositories), a MinHash-LSH fuzzy-deduplication pipeline built around two design choices. First, PUFFER stores each LSH band as immutable, dataset-tagged, memory-mapped sorted segments, enabling exact historical band-key membership checks without RAM proportional to corpus size. Second, T-fanout tiered compaction periodically merges segments to control screening fanout, trading lower query cost against additional index-maintenance writes while preserving membership decisions. Across N ingested keys and K equal-sized releases, PUFFER’s cumulative maintenance cost is O(N log N log_T K), compared with Theta(KN) for repeated snapshot rebuilding. Dataset-tagged segments also support dataset-scoped withdrawal: removal is constant-time for uncompacted or protected datasets, while post-compaction withdrawal reconstructs only the affected merged segment, even if the original dataset is unavailable. In our implementation, PUFFER completed cumulative index-stage ingestion for one billion documents in about 1.75 hours in a single process, using 128 bytes per document for a 16-band index. A classical resident MinHash-LSH table required about 6.5 KB per document and exceeded a 900 GiB RAM cap. In a ten-hour comparison capped at one billion documents, PUFFER was 11x faster than LSHBloom and 35x faster than Milvus-LSH. PUFFER is deployed on more than 30 billion documents, and we release it as open-source software at this https URL.
[AI-195] Preference Elicitation for Policy Optimization and Application to Aligning Heart Transplantation with Human Values
链接: https://arxiv.org/abs/2608.28620
作者: Itai Zilberstein,Ioannis Anagnostides,Zachary W Sollie,Arman Kilic,Tuomas Sandholm
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Preference elicitation is essential for aligning AI systems with human values. Prior approaches (e.g., for organ allocation) often ask stakeholders to compare the decisions of an algorithm (e.g., patient A vs. patient B). Such a decision-level approach conflates the means with the ends. Instead, we elicit preferences directly over allocation outcomes to learn a utility function for policy optimization. We construct a novel preference elicitation algorithm for linear utilities that outperforms prior techniques in practice. Our algorithm has two phases. The first phase learns cutting planes through pairwise comparisons to rapidly shrink the space of possible attribute weights and warm-starts the second phase by eliminating dominated regions. The second phase then provably converges to the user’s utility function. We apply our technique to heart transplant allocation where a policy must balance competing objectives such as post-transplant outcomes, waitlist mortality, geographic ease, and equity. Using our algorithm, we conduct a user study to learn and aggregate a community-aligned utility function, and use it to optimize heart transplant policies that are significantly better aligned with human values. Compared to the hindsight optimum, the status quo policy achieves a competitive ratio of just 0.54, while our method is near-optimal with a competitive ratio of 0.95.
[AI-196] InternReviewer InternAdvocate: Objective Reward and Evaluation for Agent ic Reinforcement Learning in Peer Review and Rebuttal
链接: https://arxiv.org/abs/2608.28612
作者: Xuerui Su,Liya Guo,Qizhi Pei,Qipeng Guo,Zhongbo Tian,Lijun Wu,Kai Chen,Zun Wang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Generating professional scholarly content, such as peer reviews and rebuttals, requires an intricate synergy between domain reasoning and factual grounding. This work presents a comprehensive framework for the development and evaluation of specialized scholarly agents, InternReviewer and InternAdvocate. We first establish a large-scale, high-quality scholarly dataset and integrate a high-efficiency arXiv retrieval tool to enable active evidence gathering. To optimize these agents, we implement an agentic Reinforcement Learning (RL) paradigm driven by a unified objective metric and reward system. This system avoids the biases of subjective model-based judging by employing multi-dimensional criteria, including reference-anchored semantic alignment, structural compliance, and a strict verification mechanism that cross-checks citations against real-time interaction logs to eliminate hallucinations. Experimental results demonstrate that agents trained within this closed-loop framework exhibit significant improvements in reasoning depth and citation accuracy.
[AI-197] PvG: A Moral Decision Framework for Large Language Models from One-Shot to Sequential Feedback
链接: https://arxiv.org/abs/2608.28610
作者: Fangyuan Zhang,Dong Yu,Pengyuan Liu
类目: Artificial Intelligence (cs.AI)
备注: 6 pages, 3 figures
Abstract:Existing LLM moral evaluations typically present models with isolated moral vignettes and elicit a single-shot decision, neglecting a factor known to profoundly influence human moral behavior: consequence feedback. We introduce TPvG (Text-based Pain-versus-Gain), adapted from a human moral paradigm, which embeds consequence feedback into an everyday moral dilemma of not harming others versus maximising self-gain. TPvG comprises five moral decision tasks, progressing from minimal-context one-shot choices to sequential decisions with explicit consequence feedback. Our results show that LLM moral decisions were strongly affected by decision format (one-shot versus sequential), and explicit receiver feedback produced heterogeneous effects across models. Furthermore, LLM responses to explicit receiver feedback diverged from the human reference pattern, suggesting potentially different decision processes. These findings highlight the need to evaluate whether LLM moral behavior remains stable in high-stakes interactive settings.
[AI-198] C3-UniMM: Causal Cycle-Consistent Unified Multimodal Modeling via Super Alignment and Shared Decoding Space
链接: https://arxiv.org/abs/2608.28603
作者: Yujie Shen,Lianlei Shan
类目: Artificial Intelligence (cs.AI)
备注: 22 pages, 2 figures
Abstract:Unified Multimodal Models aim to achieve any-to-any understanding and generation across arbitrary modalities. However, existing methods primarily rely on modeling implicit statistical correlations and lack cross-modal structural consistency constraints. This deficiency leads to profound issues, including semantic drift, poor compositional generalization, and instability under interventions. In this paper, we propose C3-UniMM, a unified multimodal modeling framework based on Causal Cycle Consistency and Super Alignment. Specifically, we introduce a Structured Latent Causal Graph (SLCG) as a shared cross-modal semantic space and design unified multimodal encoding blocks, enabling understanding and generation to be synergistically optimized within the identical causal semantic structure. Furthermore, we propose a Unified Decoding Space to enforce structural preservation and semantic invertibility during the cross-modal generation process. Theoretical analyses demonstrate that our approach significantly enhances both the invertibility and mechanism invariance of cross-modal mappings. Extensive experimental results across multiple understanding, generation, and compositional generalization tasks indicate that C3-UniMM substantially outperforms existing unified multimodal baselines.
[AI-199] Leverag ing Generative AI to Design Accessible Interactive Visualizations for Undergraduate Mathematics: A Six-Phase Workflow
链接: https://arxiv.org/abs/2608.28601
作者: Mahesh Sunkula,Kuan-Hua Chen
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Interactive visualizations support conceptual understanding in undergraduate mathematics, but building them has required programming expertise most instructors lack. Using a design-based research approach, we develop, deploy, and evaluate a six-phase workflow (Foundation, Customization, Mathematical Depth, Application, Accessibility, Pedagogical Control) that uses generative AI to build WCAG~2.2 Level~AA compliant visualizations without programming. The six phases structure every prompt, scaffold the AI’s code generation, and define where human verification is applied. We ask whether the structure reliably yields correct and accessible tools, whether it runs both backward (reverse-engineering prompts from a finished tool) and forward (generating a tool from a plain-language idea), and what verification each phase requires. Across four deployed tools spanning calculus, multivariable calculus, and differential equations, we evaluate mathematical correctness against closed forms, accessibility through automated and manual screen-reader testing, and the errors that recurred. The structure produces structurally complete first-pass tools, but human verification remains mandatory at every phase: each output must be checked for mathematical correctness, accessibility, and pedagogical fit before the next phase begins. The workflow is platform-independent and serves both instructors and students.
[AI-200] SHAPE of Chain-of-Thought in Math Reasoning ICML2026
链接: https://arxiv.org/abs/2608.28600
作者: Jonghyun Song,Sangjun Song,Minjae Oh,Haesung Pyun,Sungsik Lee,Yohan Jo
类目: Artificial Intelligence (cs.AI)
备注: accepted to The 3rd AI for Math Workshop at ICML 2026
Abstract:Large language models (LLMs) achieve strong performance on mathematical reasoning benchmarks, yet the mathematically meaningful skills underlying their reasoning remain underexplored. We introduce \textttSHAPE, a framework that analyzes Chain-of-Thought (CoT) trajectories through two lenses developed in mathematics education: (1) semantic spaces: the model’s evolving mathematical interpretations of a problem (e.g., algebraic, geometric), and (2) heuristics: the specific mathematical actions taken within those spaces (e.g., simplifying the problem, working backward). We first use \textttSHAPE to analyze the reasoning patterns of various models. Our findings reveal that the mathematical heuristics employed by a model better explain final answer correctness than traditional CoT features. Furthermore, models are likely to reach correct solutions by concentrating their reasoning effort within a few semantic spaces rather than exploring many disparate ones – a pattern consistent with human behavior. Next, we utilize the \textttSHAPE lens to evaluate whether post-training truly enhances mathematical proficiency. We find that reinforcement learning induces mode-seeking in heuristic usage. Lastly, we post-train LLMs by promoting diverse heuristics and demonstrate its effectiveness in improving accuracy. Overall, \textttSHAPE provides a theoretically-grounded diagnostic framework for decoding LLM reasoning and offers a new path toward post-training LLMs for math reasoning. The code for our model is available at this https URL
[AI-201] CDPR: Counterfactual Advantage-based Credit Assignment for Cost-Aware Sequential Medical Diagnosis
链接: https://arxiv.org/abs/2608.28599
作者: Qi Peng,Yi Cai,Changmeng Zheng,Xin Wu,Jiayuan Xie,Qing Li
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Clinical diagnosis is a step-by-step, cost-aware process: a physician orders examinations one at a time, observes the results, and updates the diagnosis before reaching a final conclusion. Most medical language models instead treat diagnosis as a one-pass classification task and ignore the trade-off between a test’s value and its cost. We model diagnosis as a cost-aware sequential decision process and train the policy with reinforcement learning. The main difficulty is credit assignment: the only reliable signal comes once at the end of a long trajectory, so it scores a wasteful workup the same as an efficient one. We propose CDPR (Counterfactual Diagnostic Process Reward), which needs no expert labels and no learned critic. CDPR first finds the states where the policy hesitates, using the uncertainty of its action distribution, and then scores the chosen action by its advantage over the alternatives the policy itself would consider, estimated with short rollouts under a utility that balances correctness against test count, cost, and infeasible requests. A rollout cache reuses within-batch trajectories to keep the cost low. We integrate CDPR into GRPO and test it on one in-domain (MIMIC-IV) and two out-of-domain (ClinicalBench and a private hospital dataset) benchmarks. CDPR improves diagnostic accuracy while clearly reducing the number and cost of examinations.
[AI-202] he Race between Agent ic AI Capabilities and Data Quality Control in Online Surveys
链接: https://arxiv.org/abs/2608.28597
作者: Sourav Panda,Hillmer Chona,Rupak Kumar Das,Shreyash Kale,Shikha Soneji,Jonathan Dodge
类目: Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注:
Abstract:Online surveys are a foundational data collection instrument in a variety of fields, with attention checks serving as critical guardians of response quality. However, the rapid emergence of agentic AI (goal directed systems powered by a large language model (LLM) brain and/or a multimodal processing unit with tool-augmented capabilities) raises new questions about the robustness of these safeguards. We investigate how well agentic AI architectures can complete web-based surveys and pass standard attention checks. We evaluate a single-agent architecture capable of multimodal input processing and tool-based web interaction on a controlled survey sandbox. We analyze the problem from two perspectives. From an attack perspective, we demonstrate how structural vulnerabilities such as exposed DOM metadata and predictable option encoding allow agents to resolve attention checks through structured parsing only. From a defense perspective, we implement a mitigation strategy of DOM metadata obfuscation to remove semantic cues in text-based questions. We evaluate multiple open-source language and multimodal models to study capability and orchestration effectiveness. Based on our evaluations, we offer perspectives on how to simultaneously meet the needs of empiricists and agentic AI researchers.
[AI-203] Paper Pilot: A Human-in-the-Loop Expert System for Evidence-Traceable Scientific Manuscript Generation in Applied Sciences
链接: https://arxiv.org/abs/2608.28596
作者: Nidhi Jha,Siddharth Chaudhary,Ajinkya Kulkarni
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large language model (LLM) agents are increasingly embedded in scientific workflows for literature analysis, drafting, and review. Existing systems advance autonomous discovery and manuscript generation, but do not resolve the governance problem that arises when ideas, methods, results, and claims propagate through AI-assisted workflows without mandatory human approval or artifact-level traceability. This paper proposes Paper Pilot, a human-in-the-loop expert system for evidence-traceable scientific manuscript generation in applied sciences. It adapts the Collaborative Agent Reasoning Engineering (CARE) methodology to manuscript development through manuscript-owner approval gates, explicit no-pass criteria, claim classification, audit logging, advisory LLM review, and evidence-locked revision control. The framework defines eight approval gates across the idea-to-claim pipeline and distinguishes literature-grounded from artifact-grounded claims, requiring reported numbers and interpretations to remain traceable to approved evidence; its system prompt is openly released for deployment in ChatGPT, Gemini, Claude, or institutional LLM environments. As a first empirical validation, we evaluate the citation-grounding layer with a controlled, mechanically scored benchmark (two commercial LLMs, real arXiv papers, no LLM judge): under coverage pressure ungated drafters fabricated up to 25% of their citations and never flagged an evidence gap, whereas the same models under Paper Pilot’s evidence-locked rules produced zero fabricated citations and surfaced the planted gaps as explicit placeholders. Preliminary results for result grounding, revision, and adversarial robustness point the same way; full evaluation is left to future work. Paper Pilot positions LLM-assisted writing as a controlled human-AI decision-support process rather than a fully autonomous authorship pipeline.
[AI-204] From Question-First to Analyst-First: Domain-Expert Skills and Verified Knowledge Compilation for Proactive Enterprise Analytics
链接: https://arxiv.org/abs/2608.28594
作者: Harmohit Singh,Rahul Sharma
类目: Artificial Intelligence (cs.AI)
备注: 21 pages, 4 figures, 2 tables
Abstract:Conversational analytics systems assume the user already has a well-formed question, leaving a non-expert facing a blank query box on an unfamiliar enterprise schema. Commercial ‘proactive’ tools narrow this gap only by detecting statistical anomalies over analyst-curated metric layers, and academic next-question recommenders depend on query logs that a fresh dataset lacks. We describe a production analytics system that inverts the interaction model from question-first to analyst-first through two coupled architectural ideas. First, a pluggable domain-expert ‘skill’ abstraction: a folder-based, database-free subject-matter pack (a manifest, per-stage prompt facets, keyword-routed references, report templates, and optional compute) auto-selected per (client, dataset) by deterministic schema matching and spliced as a cross-cutting concern into every stage of an agentic pipeline, the schema explorer, and the report engines, degrading to a strict no-op when absent. Because a skill is a self-contained folder resolved deterministically, the catalogue is open-ended: an extensible marketplace of domain experts. Second, an offline knowledge-compilation loop: an agent probes the dataset’s parquet via DuckDB (zero load on production), runs critic-gated per-table convergence with self-healing retries, and data-validates joins by value overlap, producing durable schema knowledge that drives standing expert reports whose every published metric is re-verified by re-executing its evidence SQL, plus suggested questions that mirror the report agenda. These close a proactive loop: reports surface numbers, the numbers seed questions, and a click launches a verified deep dive, all before the query box is used. We give a formal model and report illustrative single-tenant evidence. We make no user-study or benchmark claims; the contribution is the architecture and its defensibility.
[AI-205] Statutory AI: Aligning Large Language Models With Legal Norms
链接: https://arxiv.org/abs/2608.28593
作者: Cindy Delage,Stéphane Canu,Marc Décombas,Jonathan Foureur
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:With the increasing development of AI regulatory frameworks, ensuring that artificial intelligence systems, particularly generative models, operate in accordance with legal and ethical standards has become a critical priority. Existing proposals for AI alignment and value-guided behavior, however, face some limitations. Approaches such as Constitutional AI depend on human supervision, while broad normative frameworks like the Good-for-Humanity (GfH) principle may be overly general and ambiguous to provide actionable governance guidance. To overcome these limitations, we propose a hybrid approach called Statutory AI that employs pre-existing human-authored principles drawn from specific themes within a legal corpus. Specifically, Statutory AI uses legal texts as a constitutional framework, enabling AI systems to autonomously critique and revise their outputs according to established norms. It operates in two stages, both using Chain-of-Thought prompting. The first stage classifies the user prompt into one of the identified themes, while the second stage analyzes it in conjunction with relevant articles selected from the legal corpus of that theme. To illustrate the potential of our approach, we conducted an experiment involving 1,000 red-teaming prompts and five penal themes: discrimination, disclosure of confidential information, violence, fraud, and abuse of vulnerable persons. Statutory AI reduced harmful content by 52 to 59 percentage points across tested models, approximately 10 percentage points higher than standard Constitutional AI, while cutting computation time by over 50%.
[AI-206] A collective capability boundary in frontier large language models on guideline-conformant and case-specific oncology decision-making
链接: https://arxiv.org/abs/2608.28592
作者: Zhang Sheng,Jinming Li,Wangyang Chen,Zhiwei Bao,Yu YoSean Wang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models (LLMs) achieve high scores on medical knowledge examinations, yet real-world oncology is not a knowledge test–it is a sequence of guideline-pathway choices, escalation judgments, and commitments under uncertainty. Existing benchmarks largely measure factual recall, leaving open whether frontier LLMs share decision-path blind spots that combining models cannot fix. We built the Oncology Decision Boundary Benchmark (ODBB)–2,005 oncology decision points across NCCN guidelines and colorectal cancer cases–and evaluated nine frontier LLMs (four closed-source, five open-weight families) released between June 2025 and April 2026. A fully deterministic scorer (zero LLM inference) classified outputs into 14 failure types, independently validated by two oncologists (Cohen’s weighted \kappa = 0.939 and 0.790) on a 225-item stratified sample. Treating the nine as a pooled super-model, 42.1% (Wilson 95% CI 40.0–44.3%) of all items–35.7% of the 1,586 NCCN items and 66.4% of the 419 colorectal-cancer cases–were answered correctly by none, with failures concentrated in choosing between guideline pathways before reasoning within any: a consistent blind spot in clinical meta-judgment that likely requires architectural intervention rather than more training data. Two models tuned for decisiveness (GPT-5.5, Gemini 3.1 Pro Preview) made unsafe commitments three to five times more often than the seven cautious models without scoring higher. In 3–9% of items, models stated the correct next clinical step yet did not commit to it–failures of decision, not knowledge. Model quality is no longer the primary bottleneck for clinical LLM deployment; the binding constraint is the assumption that any single model can be the sole basis for a clinical decision. Progress requires architectures that detect when a model reaches its competence boundary and route the decision to a clinician.
[AI-207] Expert-validated STEM QA
链接: https://arxiv.org/abs/2608.28591
作者: Kihwan Han,Saurabh Patil,Chinmayee Shukla,Abhinav Sharma,Marko Pavlovic,Anshuman Lall,Mahesh Joshi
类目: Artificial Intelligence (cs.AI)
备注: We have open-sourced a portion of our dataset for the AI research community at this https URL
Abstract:Recent advancements in AI are helping scientists achieve breakthroughs in fields such as mathematics, medicine, and materials sciences. New evaluation datasets for AI models contribute to such advancement in AI. In the STEM domain, frontier models have consumed most of the available online data, creating the need for human-created datasets that codify the knowledge of leading experts in the domain. There are several STEM datasets available for the research community in this field. However, there are some gaps in these datasets, leaving room for improvement. Examples of gaps include (1) saturation in model performance on these datasets, leaving no head-room for meaningful evaluations, (2) skewed taxonomy distributions, (3) multiple choice question format that is misaligned with how scientists use AI in the real world, and (4) inaccurate answers and rationales partially led by a contest-based data collection and a time-bound review process. In this study, we present ‘Expert-validated STEM QA’, a high-quality, expert-validated STEM dataset (N=398) in Physics, Chemistry, Biology, and Mathematics, created by 241 domain experts. We (1) carefully designed a taxonomy with balanced distribution, (2) vetted question contributors with quality-driven incentive, (3) conducted multiple rounds of reviews with revisions validated by domain experts based on consensus, and (4) created the dataset in verifiable question and answer format. Our study demonstrated low performance ( 25% ) of frontier AI models on the dataset as a benchmark. Post-training on a separate, private version of the dataset (N=2,000) increased performance of the open source model by 15% relative to the baseline model (p=0.045) on the STEM subset of HLE-verified dataset, indicating potential utility of the dataset for model training. We have open-sourced a portion of our dataset for the AI research community.
[AI-208] DS-Lighting: Making Agent Harnesses Explicit for Data-Science Automation
链接: https://arxiv.org/abs/2608.28590
作者: Fan Liu,Hao Liu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large Language Model (LLM) agents have shown promise for automating data-science workflows, yet their end-to-end performance depends critically on the agent harness that represents tasks, manages execution state, constrains output artifacts, and provides evaluation feedback. Existing data-science agents often leave this harness implicit, making results difficult to reproduce, compare, and attribute across heterogeneous tasks. We introduce DS-Lighting, a unified harness toolkit that makes harness design explicit for data-science automation. DS-Lighting decomposes the harness into four reusable layers: data, workflow, execution, and evaluation, and represents diverse agents as executable operator programs that support both predefined pipelines and adaptive search. We further integrate multiple open-source data-science benchmarks into an MLE-Bench-style task format, enabling controlled comparison under a shared task interface, sandboxed runtime, and metric protocol. Experiments across agents, harnesses, models, and ablations show that explicit harness design improves reproducibility, comparability, and reliability, while reducing avoidable system-level failures in end-to-end data-science workflows. Our code is available at this https URL
[AI-209] An Enclosed Mode Is a Gauge Choice: Topology Relative to Reach in Certified Code World Models
链接: https://arxiv.org/abs/2608.28541
作者: Javier Aguilar Martín
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Systems and Control (eess.SY)
备注: 33 pages, 2 figures. Paper 3 of a series (companion papers: arXiv:2607.14169 , arXiv:2608.17956 ). Code, data, and Lean formalization: this https URL
Abstract:A code world model accepted by a sampling gate can be exactly right on everything the gate can see and arbitrarily wrong beyond it. We characterize what a certified model can know, and what its errors can cost, when the omission is an annular freeze mode enclosing an unreachable interior. The gate quotient makes the question precise: acceptance-with-certainty determines the model exactly on the reachable query set; beyond reach is gauge. On a minimal ring instrument we prove the extreme case (a wrong-topology filled-disc artifact unfalsifiable by any sampling gate and bitwise harmless at play) and measure, with LLM synthesis across three model families, how one knob (a channel of width gamma) walks the same artifact through three regimes: unfalsifiable-and-harmless, falsifiable-and-costly, and instantly falsified. Three principles organize the empirics. First, danger is topology relative to reach: a channel the planner can use collapses the blind model’s exploitation (play cost 1.09 to ~0 over a knee at gamma ~ 0.1), while a hidden channel with the same first Betti number keeps it at full strength (1.12). Second, repair is parameter-bound and sensor-bound: no family recovers the region from outside evidence; from inside, models pose the right topology but cannot pin its parameters, and the posed topology tracks the guiding persistent-homology summary’s wrong beta_1 (a sensor with a measured geometric resolution limit), not the truth. Third, mitigation must match the error’s dimension and direction: point fences fail against the one-dimensional boundary, a dimension-matched persisted fence collapses exploitation to a two-lesson transient (0.999 to 0.058), and the dual freedom certificate collapses the invented-mode failure symmetrically (1.769 to 0.029). In n dimensions the shell makes misidentification near-certain while the danger stays fully exploitable: the two axes are independent.
[AI-210] PowerSlider: Exploiting Phase Asymmetry for LLM Serving under Demand Response
链接: https://arxiv.org/abs/2608.21719
作者: Yueying Li,Jiayang Chen,Yuanfan Chen,Leo Han,Haoran Qiu,Esha Choukse,Rodrigo Fonseca,Udit Gupta
类目: Distributed, Parallel, and Cluster Computing (cs.DC); Artificial Intelligence (cs.AI)
备注:
Abstract:AI inference clusters are increasingly constrained by instantaneous power, not just energy: grid operators condition new capacity on demand response, imposing time-varying power caps. Existing LLM serving systems optimize a static energy objective or shed fixed priority tiers under load; either way, goodput collapses when the power envelope moves. An LLM pipeline is not a uniform load: compute-bound prefill loses throughput almost linearly with GPU frequency, memory-bound answer decode sustains it down to 0.57\times nominal, and reasoning’s thinking phase couples KV-cache capacity to scheduling – so a cap should be steered to where each watt costs the least performance. PowerSlider does so with a new Flex SLO contract that turns bounded user slack into an optimization constraint, prefill–think–answer disaggregation exposing per-stage frequency and KV control, and a Karush–Kuhn–Tucker (KKT) online solver re-solving within 7.7 ms of every cap change, backed by a consolidated fail-safe that power-gates drained instances when DVFS bottoms out on static power. On SGLang with production traces, \sys sustains 78.3% online goodput at a 30% cap reduction versus 47.6% for the best of five baselines ( 1.64\times ), holds latency-critical tails within 1.3\times of nominal (baselines: 2.3 – 6\times , up to 12\times ), and delivers 92% mean goodput through a replayed CAISO grid-emergency day bottoming at 0.41\times (54% at the trough; every baseline below 7%).
[AI-211] Fine-Tuning Qwen 3-27B for C-to-Rust Code Translation: A Three-Stage Curriculum of Pretraining Debugging-Aware SFT and Task-Specific SFT
链接: https://arxiv.org/abs/2608.13681
作者: Pu Zhao,Changdi Yang,Yixiao Chen,Yi Gao,Yifan Cao,Haochen Zeng,Yanzhi Wang
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Emerging Technologies (cs.ET); Machine Learning (cs.LG)
备注:
Abstract:Translating C code into safe, idiomatic Rust is a longstanding software-engineering goal because it can eliminate entire classes of memory-safety vulnerabilities while preserving the functional behavior of legacy systems. Large language models (LLMs) have shown promise for this task but typically underperform when applied off-the-shelf, since general-purpose pretraining rarely emphasizes idiomatic Rust generation, cross-language semantic equivalence, or the ability to reason about and repair compiler/runtime feedback. In this report we describe a three-stage fine-tuning curriculum applied to Qwen3-27B that is designed to progressively specialize the model for the C-to-Rust (C2Rust) translation task: (1) continued pretraining on Rust-centric corpora to strengthen the model’s prior over idiomatic Rust syntax and standard-library usage; (2) supervised fine-tuning (SFT) on the microsoft/Verus_Training_Data dataset to instill debugging and self-repair behavior over Rust code; and (3) task-specific SFT on paired C/Rust solutions derived from LeetCode problems to teach direct semantic translation. We evaluate the resulting model using the agentic, static-analysis-guided verification framework of SACTOR, which performs structure-aware, two-phase (unidiomatic to idiomatic) translation with foreign-function-interface (FFI)-based end-to-end (E2E) testing. We report success rate, idiomaticity (Clippy lint counts, unsafe-code fraction), and failure-mode analyses, and compare our fine-tuned model against baseline Qwen3-27B and other LLMs evaluated under the same framework.
[AI-212] Exponential random graph models with soft clique constraints
链接: https://arxiv.org/abs/2608.30869
作者: Yasmin Tousinejad,Vera Koponen
类目: Combinatorics (math.CO); Artificial Intelligence (cs.AI); Probability (math.PR)
备注: 58 pages, 1 figure
Abstract:Let r\geq3 be fixed, and let \mathbfG_n be the set of all simple graphs with vertex set [n]=\1,\ldots,n\ . We consider an exponential random graph model which gives higher probability to G \in \mathbfG_n than to H \in \mathbfG_n if G has fewer r -cliques than H . But all graphs in \mathbfG_n have positive probability. The degree to which graphs with fewer r -cliques are given higher probability is determined by a positive weight w . We prove that, asymptotically almost surely as n \to \infty , a random graph from \mathbfG_n has a vertex partition into r-1 parts of roughly equal size, the density of edges between the parts is close to 1/2 , and for every \varepsilon 0 the density of edges within any part is less than \varepsilon . The asymptotic structural properties are independent of the weight w as long as it is positive. We also extend the result to the context of several clique sizes, each one with its own weight.
[AI-213] Aggregate Disambiguation Systems
链接: https://arxiv.org/abs/2608.30805
作者: José María Lago,Albert Castellana,Edgars Nemše
类目: Methodology (stat.ME); Artificial Intelligence (cs.AI)
备注:
Abstract:Natural-language tasks can elicit different verdicts from protocol-following evaluators that receive the same declared information. We study aggregate disambiguation systems (ADSs). Given a task and a candidate solution, each evaluator casts a binary vote on whether the solution should be accepted, and the system aggregates the votes of a finite panel. The target is protocol reproducibility relative to an explicitly declared evaluator reference, not semantic truth. We separate fixed finite censuses, probabilistic evaluator populations, and growing-census limits, since their endpoint laws and guarantees are not interchangeable. In the population setting, we use finite samples to estimate how often a finite panel reaches the same decision as the declared evaluator population. We provide a lower confidence bound on the fraction of candidate solutions for which the disagreement probability is at most a chosen tolerance. The calculation accounts separately for sampling candidate solutions and sampling evaluators. The construction permits arbitrary dependence among columns induced by shared evaluator rows and uses exact binomial intervals at the evaluator layer and an exact one-sided binomial inversion at the generator layer. Simulations check the implementation against known population coverages and expose power limitations.
[AI-214] Dec-BFTRL: Squre-Root Regret for Decentralized Online Upper-Linearizable Optimization under Separation Access with Application to Continuous Submodular Maximization
链接: https://arxiv.org/abs/2608.30271
作者: Yiyang Lu,Mohammad Pedramfar,Vaneet Aggarwal
类目: Optimization and Control (math.OC); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:We study decentralized online optimization of upper-linearizable payoffs over an action set under efficient separation access, with applications to online continuous diminishing-return (DR) submodular maximization. We propose Decentralized Barrier Follow-the-Regularized-Leader (Dec-BFTRL), and evaluate each agent’s played action against the average of all local objectives. Each agent maps an internal iterate to a feasible action through an approximate gauge projection, communicates only a cumulative surrogate-gradient dual state, and invokes the local HybridNewton procedure to approximately minimize its post-communication BFTRL potential. For every agent, we achieve expected network-aggregate regret of \widetilde O(\sqrtT) . Over T rounds, each agent uses T neighbor-mixing steps and \widetilde O(T) separation-oracle calls. We give four wrapper instantiations covering three DR-submodular maximization problems.
[AI-215] Science sandboxes measure the scientific capability of AI agents
链接: https://arxiv.org/abs/2608.30165
作者: Arya S. Rao,Rodrigo I. Castro,Sager J. Gosai,Kenneth B. Hsu,Yasha Ektefaie,Shantanu Singh,Sangeeta N. Bhatia,Steven K. Reilly,Ryan Tewhey,Eric S. Lander,Pardis C. Sabeti
类目: Quantitative Methods (q-bio.QM); Artificial Intelligence (cs.AI)
备注: 72 pages, 5 main figures, 3 tables, and 5 supplementary figures; includes supplementary agent instructions and harness
Abstract:Scientific progress depends not only on finding solutions, but on learning the rules that explain why they work and using that understanding to design better experiments. We introduce science sandboxes, a framework for studying this capability in AI agents through repeated cycles of experimentation, feedback, and hypothesis revision. Science sandboxes invite an agent to query the natural world in different ways, ranging from “wet” physical experiments, to “damp” predictive models trained on empirical data, to “dry” invented rules. By establishing a common experimental loop and a protocol for evaluating agents within it, science sandboxes allow assessment of both quantitative performance on specific metrics and qualitative scientific reasoning, across a spectrum of empirical verifiability. Here, we instantiate this framework in two biological settings, models of regulatory genomics and protein fitness prediction, and examine the capabilities of frontier agents. Across these settings, we could see when agents successfully optimized a quantitative metric without understanding the rules underlying the system. In particular, their scientific reasoning deteriorated when they encountered systems whose rules fell outside familiar biological priors. By highlighting such failure modes, science sandboxes make the frontier of scientific capability measurable and provide a controlled setting in which to study and ultimately expand it.
[AI-216] Integrating adaptive human behavior into epidemic models with large language models
链接: https://arxiv.org/abs/2608.29535
作者: Yicheng Mao,Haoyang Li,Rob Deardon,Hongru Du
类目: Physics and Society (physics.soc-ph); Artificial Intelligence (cs.AI)
备注:
Abstract:Infectious disease transmission is shaped by patterns of human interaction, which adapt as epidemic conditions change. Capturing these context-dependent behaviors remains a fundamental challenge for epidemic models. Here, we recast this challenge by using large language models (LLMs) to represent adaptive human behavior within mechanistic epidemic models. We operationalize this idea through Generative Adaptive Behavioral Layer for Epidemics (GABLE), which adapts LLMs to infer behavioral responses to epidemic and policy conditions and translates them into age-structured contact matrices coupled to a mechanistic epidemic model. Applied to COVID-19 in France, GABLE reproduced responses in population mixing and age-specific contact structures that remained epidemiologically informative. In short-term forecasting, LLM-generated contact matrices outperformed mobility-driven matrices derived from real-world mobility data, with the largest gains at longer horizons. GABLE also extends beyond forecasting to prospective policy evaluation by projecting behavioral and epidemic responses to candidate interventions before implementation. When supplied with subsequently implemented policies, GABLE reproduced epidemic trajectories and generated distinct responses to alternative policy timing and composition. By leveraging LLMs as a flexible behavioral layer, GABLE provides a framework for coupling context-sensitive behavioral generation with epidemic dynamics.
[AI-217] Evaluating LLM -based AI agents integrated with materials synthesis tools: the case of atomic layer deposition
链接: https://arxiv.org/abs/2608.29309
作者: Angel Yanguas-Gil
类目: Materials Science (cond-mat.mtrl-sci); Artificial Intelligence (cs.AI); Applied Physics (physics.app-ph)
备注: Invited prospective paper submitted to MRS Communications
Abstract:This work provides an overview of the different strategies that can be used to evaluate the performance of AI models and agents based on large language models (LLMs) for materials synthesis. After providing a brief overview of the key technologies behind the current generation of AI agents based on LLMs, we summarize the different approaches to evaluating these models in the context of materials science and in particular on materials synthesis, with a specific emphasis on scenarios in which the models are directly integrated with experimental tools. We discuss evaluation strategies spanning knowledge and reasoning benchmarks, tool-use benchmarks, and closed loop benchmarks involving the interaction with experimental systems or realistic virtual tools. We use atomic layer deposition (ALD) as a case study, emphasizing how existing approaches in the literature both build from general approaches used beyond materials science and can be generalized to other materials synthesis techniques. Finally, we provide a practical evaluation framework to evaluate LLMs in the context of materials synthesis
[AI-218] Measurement Validity in LLM Cultural Alignment
链接: https://arxiv.org/abs/2608.29266
作者: An Duy Nguyen,Muhammad Aurangzeb Ahmad
类目: Physics and Society (physics.soc-ph); Artificial Intelligence (cs.AI)
备注:
Abstract:Researchers increasingly treat LLM survey responses as a proxy for human cultural values. This includes projecting model outputs onto instruments like the Inglehart-Welzel Cultural Map and drawing conclusions about which cultures a model resembles. While a model’s answer to a value-laden questions may be interpreted as a cultural signal, it also carries sampling noise and, can be quite sensitive to question framing. In this paper, we separate survey responses, sampling noise and question framing for multiple LLMs. We decompose response variance from these models into variation across random seeds, prompt rewordings. We employ noise-to-signal ratio (NSR) to test whether a model’s apparent cultural position is distinguishable from noise. When applied across a dozen models from four geographic origins, calibrated against 88 Integrated Values Survey countries, the answer is often no. NSR exceeds 1.0 on 49 of 117 valid model-question pairs (42%), reaching 5.56 in the worst case. Two models even refuse to answer sufficient number of survey questions outright. Our results corroborate previous findings that LLMs cluster toward Western, English-speaking cultural positions. However, what does not hold up in this study is the precision with which anyone can currently interpret a specific model’s coordinates: prompt tone alone can shift a model by 2.4 map units, comparable to the distance between actual countries in the Inglehart-Welzel Cultural Map. These findings suggest that cultural attribution from LLM survey responses requires establishing the reliability of the underlying measurements before interpreting model coordinates as evidence of cultural representation.
[AI-219] he information geometry of product-reference discrete diffusion: Interaction growth complexity and optimal scheduling
链接: https://arxiv.org/abs/2608.28949
作者: Martin J. Wainwright
类目: Machine Learning (stat.ML); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Statistics Theory (math.ST)
备注:
Abstract:We study a class of product-reference diffusion algorithms for sampling from a discrete distribution. We show that their sampling performance can be characterized using a path-based measure of data geometry that we call the interaction growth complexity (IGC). We show that a bivariate IGC kernel gives an exact representation of both the KL discretization error and a simple one-step upper bound. The simpler univariate IGC density can be used to study the effect of stepsize choices on the iteration complexity required to obtain \epsilon -accurate samples in KL divergence. Samplers that traverse the path with equi-spaced steps in log-squared-reliability-odds have performance that depends on the aggregate IGC mass, whereas refined choices of stepsizes have a lower complexity depending on a square-root functional. In the fine-grid limit, both of these characterizations become sharp. We also allow general product reference distributions and show that the reference law can substantially reshape the IGC profile and the resulting sampling complexity; in particular, references far from both the uniform and the data marginals can yield dimension-dependent improvements. Finally, the aggregate IGC mass admits bounds in terms of total correlation and dual total correlation, thereby connecting the pathwise geometry to classical measures of multivariate dependence.
[AI-220] Representation Learning with Quantum Signal Processing
链接: https://arxiv.org/abs/2608.28828
作者: Junqi Wang,Junyu Liu
类目: Quantum Physics (quant-ph); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Machine Learning (stat.ML)
备注:
Abstract:Representation learning begins when training changes the features that define similarity between data. A frozen-kernel model only reweights a fixed geometry. We establish quantum signal processing (QSP) as a solvable quantum model of the representation-learning regime. At arbitrary depth, we compute the exact mean and variance of its quantum neural tangent kernel, revealing an input-dependent angular geometry whose diagonal remains non-self-averaging even when the underlying unitary approaches Haar randomness. We also prove a sparse-data guarantee for the full nonlinear gradient flow without freezing or ensemble-averaging the kernel: the realized dynamics converges to an integrable scalar flow with a time-dependent kernel closure and explicit convergence times. A finite-depth speed limit holds for every data set and trajectory. At higher data density, numerical results show coupled evolution beyond both the scalar and frozen-kernel descriptions. These results give a controlled theory of learned quantum data geometry with provable training dynamics beyond the frozen limit.
[AI-221] Agent -Based Model Framework for the North Carolina Modeling Infectious Diseases Program (NC MInD ABM) Overview Design Concepts and Details Protocol
链接: https://arxiv.org/abs/2202.06853
作者: Kasey Jones,Emily Hadley,Caroline Kery,Alexander Preiss,Marie C.D. Stoner,Sarah Rhea
类目: Applications (stat.AP); Artificial Intelligence (cs.AI)
备注: 23 pages. arXiv admin note: text overlap with arXiv:2106.04461
Abstract:To help facilitate a variety of simulations related to healthcare facilities in North Carolina, we have developed an agent-based model (ABM) to accurately simulate patient (i.e., agent) movement to and from these facilities. This is an Overview, Design Concepts, and Details (ODD) Protocol, a standardized method for describing ABMs. This ODD provides detailed information on healthcare facilities in North Carolina, the agent movement to and between them, and any decisions that were made during the creation of this model. This ABM is intended to be used alongside disease-specific submodels. It can be used for purposes such as simulating the success of interventions on reducing disease transmission, simulating strain on facility resources (including staff and materials), and forecasting hospital capacity. Disease-specific ODDs should accompany this document. No details related to any submodels that use this ABM as a base model are included.
机器学习
[LG-0] Constant Individual Regret in General Games
链接: https://arxiv.org/abs/2608.31166
作者: Mingyang Liu,Gabriele Farina,Asuman Ozdaglar
类目: Machine Learning (cs.LG); Computer Science and Game Theory (cs.GT)
*备注:
Abstract:Uncoupled no-regret dynamics provide a decentralized route to equilibrium, but prior guarantees for individual regret retain a polylogarithmic dependence on the horizon. We remove this dependence for every finite N -player normal-form game under full-information feedback. We introduce \emphECHO-OFTRL: optimistic follow-the-regularized-leader (OFTRL) equipped with an EMA cascade for high-order optimism (ECHO), where EMA denotes exponential moving average. The algorithm is deterministic and fully uncoupled. If m_\max denotes the largest action-set size, then, simultaneously for every horizon T\geq1 , it guarantees that each of the N players in the game incurs regret upper bounded by O(\textrmpoly(N, \log m_\max)) . Our algorithm leverages a new form of optimism inspired by modern filter design.
[LG-1] Sharp Approximation Rates for Neural Networks with Affine Latent Parameterizations
链接: https://arxiv.org/abs/2608.31157
作者: Shijun Zhang
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注:
Abstract:Many parameter-efficient methods generate the parameters of a large neural network from a low-dimensional latent representation. Given an architecture \Phi with P_\Phi parameter slots, we write \boldsymbol\theta_f=\mathcalG(\boldsymbol\xi_f) , where \mathcalG\colon\mathbbR^M\to\mathbbR^P_\Phi is a parameter generator and \boldsymbol\xi_f\in\mathbbR^M is a latent representation of the target function f . The architecture \Phi and the generator \mathcalG are shared across the entire target class, while each target f is represented by its own latent vector \boldsymbol\xi_f , with \Phi_\mathcalG(\boldsymbol\xi_f) approximating f . This framework encompasses hypernetworks, low-dimensional parameterizations, parameter-efficient adaptation, and model compression. Understanding the tradeoff between the latent dimension M and the network budget P is therefore fundamental to characterizing the expressive efficiency of these methods. We study this tradeoff for affine generators and fully connected ReLU architectures. More precisely, optimizing jointly over architectures \Phi satisfying P_\Phi\leq P and affine generators \mathcalG:\mathbbR^M\to \mathbbR^P_\Phi , we prove that the optimal worst-case uniform approximation error over the unit ball of \alpha -Hölder functions on [0,1]^d , where 0\alpha\leq1 , has the sharp order \bigl(P\min\M,P\bigr)^-\alpha/d. In particular, our result shows that even a fixed-dimensional latent space suffices to achieve vanishing approximation error as the network budget increases.
[LG-2] On the Complexity of the Compatibility Problem for Succinctly Encoded Conditional Distributions
链接: https://arxiv.org/abs/2608.31120
作者: Guy Emerson
类目: Machine Learning (cs.LG); Computational Complexity (cs.CC); Probability (math.PR)
*备注: 14 pages
Abstract:The motivation for this paper is the investigation of the trade-offs implicit in probabilistic models used in machine learning. Models are often used to make predictions in the form of conditional probabilities. However, a pair of conditional distributions p(x|y) and p(y|x) may not be compatible with any joint distribution p(x,y). Given two such conditionals, determining if there exists a compatible joint is known as the compatibility problem. For discrete random variables, when the conditionals are encoded as probability tables, the compatibility problem has a known solution, which is computationally tractable. In this paper, we formalise and study a succinct version of the problem, encoding conditional distributions as arithmetic circuits. This is applicable to practical applications of probabilistic modelling in high-dimensional settings, including neural network models. We show that, for succinct circuit representations of conditionals, the compatibility problem is intractable. In the case that all probabilities are non-zero, the problem is co-NP-complete. In the case that probabilities can be zero, we give examples to demonstrate that several notions of compatibility can be distinguished, and we prove that multiple versions of the problem are PSPACE-complete. Furthermore, we show that, assuming the polynomial hierarchy does not collapse, there exist compatible succinct conditionals whose joint cannot be expressed succinctly. Implications of these results for probabilistic modelling and machine learning are discussed.
[LG-3] Stress-Testing Efficient Responsible-AI Evaluation: When Compute Savings Change Benchmark Conclusions
链接: https://arxiv.org/abs/2608.31108
作者: Ahmed El Kady,Aravind Narayanan,Rehana Noorani,Yani Ioannou,Shaina Raza
类目: Machine Learning (cs.LG)
*备注:
Abstract:Efficient evaluation changes the protocol used to support claims about model behavior, yet it is rarely tested whether those claims remain stable after the evaluation itself is made cheaper. We stress-test conclusion robustness in responsible-AI benchmarking by evaluating three dense and mixture-of-experts models on BBQ and BBQ-V under seven conditions spanning batching, quantization, benchmark reduction, and their combinations. Rather than treating preserved aggregate accuracy as sufficient, we compare accuracy, bias severity and prevalence, reasoning quality, subgroup behavior, subset-membership stability, runtime, and measured GPU energy against a full-benchmark BF16 baseline. Larger batching keeps accuracy within 0.35 percentage points of baseline and produces comparatively small subgroup changes, while reducing energy in five of six model–dataset settings. INT8 largely preserves quality but uses 1.79–4.26 \times baseline energy. INT4 causes larger, model- and context-dependent changes. Reduced benchmarks provide the most consistent savings, but very small subsets are substantially more sensitive to which items are retained. Efficient evaluation should therefore be treated as a measurement intervention whose validity must be checked across the conclusions the benchmark is intended to support. Our project website is this https URL and the code is available at this https URL.
[LG-4] Minimax bounds for watermarked and masked recursive discrete distribution estimation
链接: https://arxiv.org/abs/2608.31091
作者: Millen Kanabar,Michael Gastpar
类目: Information Theory (cs.IT); Machine Learning (cs.LG); Statistics Theory (math.ST)
*备注: Shorter version to be published at the IEEE Information Theory Workshop 2026
Abstract:Watermarking has been proposed as a way to identify synthetic samples in estimation settings where no metadata is available to distinguish them from real samples, but its precise effects remain unexplored. In the absence of a distinguishing mechanism, it has been shown that adding synthetic samples significantly reduces the marginal efficacy of new real samples. In this work, we study the minimax loss of such recursive discrete distribution estimation in the presence of watermarks in contrast to the unassisted and oracle-assisted losses. When the fraction of real samples vanishes asymptotically, we provide a lower bound that shows that it is impossible to improve performance by adding watermarks unless the false negative rate of detection also vanishes. Additionally, we show that in most regimes, the worst-case losses of a sequence of simple deterministic estimators match the corresponding lower bounds up to constants. Finally, we propose masking, a randomization procedure that narrows the gap in the remaining regimes to a Jensen gap. We conjecture that a tighter lower bound argument can close this gap.
[LG-5] Sycophantic Agreement Transfers with Neutral Data via Contrastive Preference Optimization
链接: https://arxiv.org/abs/2608.31079
作者: Camila Blank,Zhuofan Ying,Christopher Potts,Peter Hase,Jing Huang
类目: Machine Learning (cs.LG)
*备注:
Abstract:Sycophantic agreement refers to a behavior in which language models excessively affirm the user, often at the cost of factual accuracy. Although sycophantic agreement is a well-known failure of model alignment, there is limited understanding of how it emerges from model training. In this work, we demonstrate that sycophantic agreement can emerge as an unintended consequence of widely used contrastive preference optimization objectives. Using the OLMo 3 post-training pipeline, we show that, for various pairs of teacher models across three families, there is a strong correlation between the log-ratio of the teacher model sycophantic agreement rates and the resulting student model sycophantic agreement rate. We further demonstrate that this unintended transfer is not limited to DPO but also occurs across 6 other preference optimization objectives. To understand whether this effect can be attributed to particular training examples, we analyze the preference data and find that the sycophancy signal is diffused across the entire dataset rather than concentrated in a sparse set of examples: each example appears neutral, i.e., there are no explicit instances of sycophantic agreement, and filtering based on probe-based data attribution or logit-linear selection fails to mitigate sycophancy without removing a large portion of the dataset. Overall, our findings suggest that the teacher models used to generate preference data can interact with alignment training objectives in unexpected ways, generalizing to undesirable and potentially harmful behaviors like sycophantic agreement.
[LG-6] Universal Transformers for Circuit Computations: Perfect Length Generalization in Tiny Transformers
链接: https://arxiv.org/abs/2608.31067
作者: Takuya Ito,Ruchir Puri,Murray Campbell,Parikshit Ram
类目: Machine Learning (cs.LG)
*备注:
Abstract:Learning generalizable algorithmic computations remains a challenge for neural networks, as reflected in persistent failures on compositional and length generalization benchmarks. We present a provably correct, transformer parameterization (with only 280 learnable parameters for Boolean algebra tasks) capable of learning and evaluating problems of any depth or length. We assume inputs are fully parenthesized, well-formed expressions. Our approach conceptualizes algorithmic tasks as circuit models embedded in transformers, enabling depth-1 circuit reduction in a single forward pass. To achieve depth generalization, we introduce a positional encoding that tracks each gate’s depth within the circuit, enabling the model to identify evaluable subexpressions at each iteration via masked hard attention, with O(n) per-iteration complexity via linear attention. Combined with an autonomous halting criterion, the model terminates after d iterations for problems of depth d , yielding O(n \cdot d) total complexity. We show that training on shallow problem instances (depth 1 and depth 2) effectively recovers interpretable parameters that \em snap into place, resulting in exact length generalization. Though we establish that our construction provably evaluates Boolean expressions – a universal symbolic computation – of arbitrary length perfectly, in other experiments we also demonstrate that our transformer variant can learn and generalize perfectly (100% accuracy) on other common length generalization benchmarks, including modular arithmetic and ListOps.
[LG-7] Rotational Equivariance in Machine Learning: A Comprehensive Tutorial
链接: https://arxiv.org/abs/2608.31045
作者: Peter Lippmann,Fred A. Hamprecht
类目: Machine Learning (cs.LG)
*备注:
Abstract:Rotational symmetry is one of the most important structural principles in machine learning on 3D data. In applications ranging from physics and materials science to 3D computer vision, predictions should not depend on an arbitrary choice of coordinate frame. Rotational equivariance captures this requirement mathematically by enforcing that a rotation of the input induces a corresponding transformation of the model output. This tutorial provides a comprehensive introduction to rotational equivariance, starting from the physical and geometric intuition behind coordinate independence and building up the necessary machinery from geometric deep learning, group theory, and representation theory. We introduce message passing on Euclidean graphs, group actions and representations, spherical harmonics, Wigner matrices, tensor products, and Clebsch-Gordan decomposition, and explain how these ingredients give rise to modern equivariant architectures. We then survey the principal strategies for incorporating rotational equivariance in deep learning, including group convolutions, internal tensorial representations, and canonicalization-based methods, and discuss their practical strengths and limitations. The tutorial aims to lower the barrier to the subject by connecting the underlying mathematics to practical model design, by unifying ideas that are often expressed in different formal languages, and by helping practitioners choose among competing approaches through a clear discussion of their trade-offs.
[LG-8] Normalized Low-Rank Adaptation
链接: https://arxiv.org/abs/2608.31036
作者: Jiale Kang,Ziyin Yue,Zheng Zhan,Yangyi Huang,Weiyang Liu
类目: Machine Learning (cs.LG)
*备注:
Abstract:While low-rank adaptation (LoRA) is widely used for parameter-efficient model adaptation, how to regularize its training dynamics for stable and effective optimization remains underexplored. Because LoRA initializes the up-projection to zero, its early optimization dynamics are largely governed by the down-projection. Building on this observation, we introduce Normalized Low-Rank Adaptation (NoRA), a simple yet effective method that normalizes the down-projection matrices during training. We further show that the same normalization can be applied only at initialization, improving standard LoRA without requiring repeated normalization throughout training. Across pretraining, supervised finetuning, and reinforcement learning, NoRA consistently accelerates convergence, improves performance and training stability, and mitigates catastrophic forgetting. These benefits require neither additional trainable parameters nor inference-time computation, making NoRA a simple and broadly applicable enhancement to LoRA.
[LG-9] Language-Informed Flow Matching for Trend-Guided Structure-Based 3D Molecular Generation EMNLP2026
链接: https://arxiv.org/abs/2608.31009
作者: Tianyu Gao,Zhikai Su,Jiashu Li,Wenjun Gao,Zichuan Ying,Zhe Zhao,Fei Zhang,Ye Wei
类目: Machine Learning (cs.LG)
*备注: Accepted at Findings of EMNLP 2026
Abstract:Structure-based drug design (SBDD) requires ligands that satisfy both 3D target affinity and 1D chemical validity. Existing controllable generation methods often rely on task-specific fine-tuning or externally imposed sampling-time guidance, adding cost and potentially conflicting with evolving 3D geometric constraints. We propose LiFT, a language-informed cross-modal framework built on Flow Matching for trend-guided 3D molecular generation across both de novo design and scaffold hopping. LiFT uses a “Sense-Evolve-Assemble” agent to generate target-aware SMILES as intermediate chemical conditions, from which a pre-trained chemical foundation model extracts continuous semantic priors. These priors are integrated into geometric generation through a lightweight semantic projector with zero-initialized adaptive normalization for stable cross-modal conditioning. We further introduce a Self-Conditioned Decoupled Router (SCDR), which modulates the velocity field according to intermediate structural states during ODE integration. Experiments on Cross-Docked2020 show that LiFT achieves competitive distribution matching while improving medicinal chemistry metrics and maintaining competitive structural validity under task-steering settings without additional generator fine-tuning. Our results suggest that language-derived chemical priors provide effective trend-level guidance for 3D molecular generation. Code and released artifacts are available at this https URL.
[LG-10] Sparse Competition during Training For the Emergence of Specialized Modules BMVC2026
链接: https://arxiv.org/abs/2608.30978
作者: Baptiste Rossigneux,Karim Haroun
类目: Machine Learning (cs.LG)
*备注: 21 pages, 6 figures. Published at BMVC 2026
Abstract:Modularity in deep neural networks has been proposed as a means of improving both interpretability and training by promoting disentangled representations and reducing redundancy. In this work, we study the emergence of modular structure through competition dynamics between groups of neurons during training. We introduce a method that (i) maintains near-baseline accuracy, (ii) induces usage-based modularity by sparsely routing inputs to neuron groups, and (iii) encourages specialization of these modules, such that their activations are correlated with input classes. We evaluate the proposed approach on ImageNet-100 and CIFAR-100 and show that with it, specialized modules emerge without module-level supervision. These modules capture a meaningful high-level structure in the data, with individual modules responding to semantic categories (e.g., dogs or vehicles). We also study the emergence of a hierarchical partition of sub-tasks depending on the number of modules. Our results suggest that competitive dynamics can serve as a simple mechanism for inducing functional modularity in standard architectures.
[LG-11] A Human-in-the-Loop Autonomous Agent for Industry Time Series Forecasting
链接: https://arxiv.org/abs/2608.30976
作者: Xiaoyu Tao,Mingyue Cheng,Ze Guo,Bokai Pan,Qi Liu,Shijin Wang,Enhong Chen
类目: Machine Learning (cs.LG)
*备注:
Abstract:Real-world time-series forecasting is rarely a one-shot model invocation: practitioners must formulate tasks, connect data and models, incorporate domain expertise, assess prediction plausibility, and communicate uncertainty. Specialized forecasting models provide strong numerical predictions but usually operate in fixed pipelines, while general-purpose large language model (LLM) agents often lack forecasting-specific checks, constraints, and stopping rules. We present CastClaw, a human-in-the-loop autonomous forecasting system built through forecasting-oriented harness engineering. CastClaw connects data, specialized models, analytical tools, user input, and a versioned execution record in one runtime. Users specify the target, horizon, constraints, and hypotheses in natural language. Starting from a supplied or model-generated forecast, CastClaw checks temporal patterns and user constraints; when evidence is missing, it retrieves context, runs an analysis or another model, or asks the user. It then keeps, revises, or escalates the result under explicit stopping conditions. The output contains the final forecast and an execution report recording inputs, evidence, actions, and revisions. In this five-dataset electricity-price setting, CastClaw reports the lowest point-estimate MSE and MAE among 16 baselines. A Nord Pool case demonstrates the inspectable workflow. CastClaw was also validated offline on provincial electricity-load data from North China covering January–June 2026.
[LG-12] Singular Curvature in ReLU Training:Differentiation and the Gradient-Flow Limit Need Not Commute
链接: https://arxiv.org/abs/2608.30960
作者: Xiaoyang Li,Runni Zhou
类目: Machine Learning (cs.LG)
*备注:
Abstract:Gradient descent (GD) is explicit Euler for gradient flow, but a state-accurate continuous-time surrogate need not remain accurate after differentiation. At every fixed nonresonant step size, ordinary automatic differentiation exactly differentiates the executed hard-ReLU GD program. We prove that, over a fixed finite horizon, the GD states converge and these exact discrete derivatives approach an event-free regional propagator, whereas the derivative of the limiting flow also contains speed-normalized activation-event transfers. A prepoint Stieltjes representation separates the absolutely continuous regional Hessian from atomic interface curvature; one nonzero gradient jump produces an exactly rank-one endpoint discrepancy, and global convexity prevents complete multi-event cancellation whenever an event is strict. Nevertheless, a standard family of globally 1-strongly convex residual-ReLU squared-loss risks realizes arbitrarily large reciprocal sensitivity ratios on open initialization sets, with a uniform transversality margin. The same discrete-versus-flow decomposition extends to parameters and reverse-mode adjoints; resolved smoothing in the scalar or autonomous-normal regime and consistent event localization recover the flow sensitivity. The results concern deterministic full-batch, finite-horizon dynamics with a stable finite itinerary of separated same-direction transverse events; they are consistency theorems, not prevalence claims for large-scale training.
[LG-13] Reproducible macroscopic dynamics in a closed-loop human-AI learning system
链接: https://arxiv.org/abs/2608.30946
作者: Minlin Wu(1),Xu Fang(1),Yicheng Zhang(2),Chenyu Zhou(1),Zhiyi Liu(1) ((1) Tianli Qiming AI Research Institute, Sichuan Qiming Daren Technology Co., Ltd., Chengdu, China, (2) Swiss AI Laboratories, Blonay, Switzerland)
类目: Machine Learning (cs.LG); Adaptation and Self-Organizing Systems (nlin.AO)
*备注: 8 figures, 11 supplementary tables
Abstract:Closed-loop human-AI systems generate high-dimensional behavioural trajectories whose collective dynamics remain obscure. Using 297,915 learners’ adaptive-tutoring histories, we define semantic order variables before model fitting and test them in user-disjoint cohorts. The state exhibits reproducible basin-like flow and operationally defined, state-heterogeneous metastable-like kinetics. A construction-matched null distinguishes normalised-memory relaxation from a reproducible excess field. A four-term conditional mechanism recovers population drift (r = 0.946; learner-bootstrap 95% CI, 0.935-0.955). Predictive event-level self-supervised learning recovers the state and learned-plane flow; null-referenced corrections retain directional, partial-amplitude excess-field structure without full calibration. Shuffled-order training reverses learned-plane flow on ordered trajectories; support-alignment randomisation selectively reduces inward transport. Both axes remain linearly accessible without state supervision. Without cross-model fitting, the models share leading population drift (r = 0.866; learner-bootstrap 95% CI, 0.857-0.875) and persistence ordering; residual directions remain model-specific. These results identify an externally anchored leading-order effective field linking empirical dynamics, an interpretable mechanism and neural computation.
[LG-14] Nonparametric Contextual Pricing and Inventory Learning under Censored Demand
链接: https://arxiv.org/abs/2608.30944
作者: Zean Han,Jing Liang,Ruihan Lin,Zezhen Ding,Jiheng Zhang
类目: Machine Learning (cs.LG)
*备注: 31 pages, 3 figures
Abstract:In online retailing, when a product sells out, a retailer often sees only the units sold, not how many customers would have bought it had inventory been available. However, the inventory level determines how much demand is revealed, and this information can influence subsequent decisions and future profits. We study an online selling problem in which, in each round, the seller observes a market context and then makes pricing and stocking decisions based on censored sales data from previous rounds. The challenge is to learn a context-dependent pricing and stocking policy without assuming a particular formula for demand or observing realized profit. To overcome this difficulty, we propose a Mean-Calibrated Kernel UCB (MCK-UCB) algorithm that turns each incomplete sales record into a reliable guide for both inventory and price decisions, using data from past rounds with similar market conditions. This design allows us to learn while serving customers, without a separate exploration phase or the need to recover all demand hidden by stockouts. We prove the minimax optimality of the proposed algorithm, with strictly faster rates when expected profit varies more smoothly with price. Comprehensive numerical experiments have been conducted to confirm the effectiveness of the proposed algorithm.
[LG-15] Selection-Aware Stress Testing for Interactive Agents
链接: https://arxiv.org/abs/2608.30916
作者: Yang Xu,Chenang Li,Jiefu Zhang,Haixiang Sun,Zhou Li,Vaneet Aggarwal
类目: Machine Learning (cs.LG); Applications (stat.AP); Machine Learning (stat.ML)
*备注:
Abstract:Agent evaluations often use one benchmark to choose a workflow and then search for task types where its advantage weakens, so both conclusions are selected from the same data. We introduce Selection-Aware Semantic Stress Testing (\SASST), which learns a task reweighting from pre-execution features on discovery tasks and evaluates the same paired comparison on separate confirmation tasks. The protocol checks support and stability, uses joint bounds for all planned claims, and can return no claim. We prove conditional asymptotic validity under stated cluster assumptions. A forty-cluster audit finds Gaussian undercoverage and conservative Bonferroni t bounds. In one 480-episode \tau -bench study, a 3.75 point discovery gain vanished on confirmation. A second-model study likewise confirmed neither a workflow benefit nor a stable stress rule.
[LG-16] Fine-Tuning Low-Bit Models with Gradient in Quantized Code Space
链接: https://arxiv.org/abs/2608.30908
作者: Shiguang Wu,Zhouchen Lin,Quanming Yao
类目: Machine Learning (cs.LG)
*备注:
Abstract:Fine-tuning Low-bit models aims to adapt a quantized model while keeping the final deployed checkpoint in the same low-bit form. This setting is practically important as it reduces memory and inference cost for storage and deployment. Under this constraint, adaptation becomes an optimization problem over quantization codes and scales. Existing continuous low-bit training is efficient, but it can be distorted by straight through estimation error or by post-quantize gap; discrete search is deployment-faithful, but it is often too inefficient under a finite training budget. We propose code surrogate gradient as the first order signal in deployable code space to acceleate optimization, and performing guided search to preserve deployment faithfulness. Experiments across arithmetic reasoning, instruction following, and structured language understanding show that GradCodes consistently improves fine-tuning low-bit models across different quantization datatypes. Code is provided at this https URL.
[LG-17] Deploying DeepSeek 175B Locally on a Single Consumer-Grade RTX 4060 Laptop with 32GB RAM for 200k-Scale Protein-Ligand Virtual Screening
链接: https://arxiv.org/abs/2608.30877
作者: Rui Xiao,Yili Xu
类目: Machine Learning (cs.LG)
*备注:
Abstract:Recent advances in large language models (LLMs) have demonstrated exceptional performance in protein-ligand interaction prediction, but state-of-the-art pipelines for large-scale virtual screening almost exclusively rely on high-end GPU clusters with hundreds of gigabytes of memory, creating prohibitive hardware barriers for small academic teams. In this work, we present a fully local low-resource framework that deploys the 175-billion-parameter DeepSeek 175B LLM on a single consumer-grade RTX 4060 laptop equipped with 32GB system RAM and 8GB VRAM, completing a full 200k-scale protein-ligand virtual screening workflow across 20 distinct protein targets. Our implementation achieves 100x throughput of an 8-card A100 cluster baseline under identical task configurations within 72 hours, with an average binding affinity prediction error of 0.88 kcal/mol across all targets, satisfying the 1.0 kcal/mol chemical accuracy requirement for preclinical drug discovery. Systematic runtime profiling reveals that heterogeneous memory management overhead accounts for 72% of total execution time, while accuracy loss introduced by model optimization contributes less than 10% to total prediction error. This work validates the engineering feasibility of running industrial-scale trillion-parameter LLM-driven biomedical computing tasks on consumer hardware, establishing a new low-barrier paradigm for AI-powered early stage drug discovery.
[LG-18] What Emerges and What Breaks in Self-Play Driving
链接: https://arxiv.org/abs/2608.30819
作者: Laur Sisask,Ardi Tampuu,Tambet Matiisen
类目: Machine Learning (cs.LG)
*备注:
Abstract:Training autonomous driving policies through pure self-play has recently shown promising results. Following Gigaflow and Puffer- Drive, we train driving policies in a similar self-play fashion, but extend the models from MLPs to Transformers and train on the high-definition map of a real city, where we ultimately aim to deploy them. On the CARLA and Waymax benchmarks, our policies fall short of Gigaflow, and we trace the gap to specific failure modes, including reward hacking at traffic lights and a missing incentive to stop at stop signs. We further analyze which traffic rules emerge from self-play and how closely they match human driving, and we confirm that reward conditioning yields the intended diversity of driving behaviors. A demonstration of a trained policy is available at this https URL.
[LG-19] Geometric Attractor Monitoring: A Robust and Frugal Framework for Multi-modal Industrial Robotic Cycles KDD2026 ECML
链接: https://arxiv.org/abs/2608.30804
作者: Martin Bonsergent-Brachet,Jesse Read,Dany Abboud
类目: Machine Learning (cs.LG)
*备注: 17 pages, 6 figures. Accepted at ECML PKDD 2026
Abstract:Monitoring the health of heterogeneous industrial robot fleets is severely challenged by the multi-modal nature of their operational cycles and a persistent scarcity of run-to-failure data. Standard data-driven approaches, particularly deep learning architectures relying on sequential reconstruction, often struggle in this specific setting; they tend to over-smooth complex dynamics, masking early signs of degradation. To address these industrial constraints, we reframe the monitoring problem through a framework based on Phase Space Reconstruction (PSR). Instead of predicting temporal sequences, this framework transforms univariate sensor data into a geometric attractor, explicitly unfolding the mechanical states independently of their temporal occurrence. By evaluating various anomaly scoring techniques within this space, we demonstrate that discrete support estimation provides an effective and computationally frugal Health Indicator (HI). Validated on a real-world dataset of 21 heterogeneous robots over three years and a synthetic Langevin system, our approach outperforms standard deep learning baselines. We show that aligning the algorithmic bias with the geometric properties of the target system yields a pragmatic, traceable and easily deployable approach perfectly tailored to the realities of industrial constraints.
[LG-20] Reciprocity Separates Gradient Flow from Rotation in Conservative Physical Learning
链接: https://arxiv.org/abs/2608.30778
作者: Ruiwu Niu,Xiaowen Bi,Michaël Antonie van Wyk
类目: Machine Learning (cs.LG); Adaptation and Self-Organizing Systems (nlin.AO)
*备注:
Abstract:Physical learning lets a trainable material or network use its own physical response to carry error signals, reducing the need for a separately programmed backward computation. We ask what determines whether such a system follows conventional gradient descent or evolves along a genuinely different learning trajectory. Our canonical model is a directed layered transport network in which every node redistributes a fixed amount of flow, so learning preserves positivity and total mass. In this model, conservation constrains only the allowable learning directions. Within the matched response class studied here, adjoint matching gives the physical output response a symmetric form. Non-negative mode-wise feedback then produces a reciprocal closed-loop response and a reweighted gradient flow. Adding an antisymmetric boundary component makes the closed-loop response rotational: the learning path can turn while the error driving that update still decreases at that moment. Turning is not automatically beneficial. Its finite-step effect is set by local curvature, and its accumulated effect also depends on step selection and on the new states visited along the path. Numerical consistency checks reproduce the exact response structure, predict the sign of the local effect across new network families, and show how trajectory drift can negate a local advantage. These results separate the roles of conservation, reciprocity, and nonreciprocity in physical learning.
[LG-21] rainSDC: Characterizing and Mitigating Silent Data Corruption in Large Language Model Training
链接: https://arxiv.org/abs/2608.30769
作者: Zhipeng Xia,Haotian Xu,Siyu Yun,Liqi Lin,Hu Liu,Yu Li,Cheng Zhuo
类目: Machine Learning (cs.LG)
*备注: 12 pages, 5 figures, and 6 tables. Includes an appendix with additional experiments
Abstract:LLM training is increasingly vulnerable to silent data corruption (SDC), yet existing protection methods largely treat Transformer computations uniformly because their vulnerability remains poorly understood. We present the first systematic characterization of SDC vulnerability across major computation interfaces in both the forward and backward passes of Transformer training. Our analysis reveals two distinct error propagation mechanisms: forward-pass vulnerability is highly location dependent, with faults on the Q/K path producing persistent training deviations, whereas backward-pass vulnerability is largely governed by gradient exponent distributions rather than computation locations. Motivated by these observations, we propose TrainSDC, a characterization-guided protection framework consisting of Q/K-path recomputation, residual-gain monitoring, and exponent-aware gradient scaling. Experiments on Llama 3.2-1B and Qwen3-0.6B show that TrainSDC maintains training behavior close to fault-free execution under both sparse and dense fault injection while introducing only 1.65%-6.76% runtime overhead.
[LG-22] 3S: Improving Multi-Task Reinforcement Learning with Task-Specific Feature Selector and Scheduler IJCNN
链接: https://arxiv.org/abs/2608.30765
作者: Yuanqiang Yu,Tianpei Yang,Yongliang Lv,Yan Zheng,Jianye Hao
类目: Machine Learning (cs.LG)
*备注: 8 pages, 7 figures, 4 tables. Published in the 2023 International Joint Conference on Neural Networks (IJCNN)
Abstract:Multi-task reinforcement learning (MTRL) is a technique to train multiple tasks simultaneously, where previous works usually train a single model to solve different tasks by sharing parameters across various tasks. However, these methods are faced with inter-task interference since what parameters should be shared across tasks is not addressed, dramatically reducing learning efficiency. To solve these problems, we propose a novel MTRL framework called Task-Specific feature Selector and Scheduler (T3S), which consists of two components: a feature selector and a task scheduler. Specifically, the feature selectors employ hypernetworks to construct task-specific soft masks, which can be applied by globally shared representation to construct task-specific features. The task scheduler selects tasks for learning through two metrics, where the selection probability is inversely proportional to task progress (e.g., success rate) and task learning speed. Experimental results show that T3S consistently outperforms the state-of-the-art MTRL algorithms on various robotics manipulation tasks.
[LG-23] PRACTICE: From Experience to Expertise in Self-Evolving Embodied Agents
链接: https://arxiv.org/abs/2608.30760
作者: Ziyi Bai,Siqi Li,Tinglei Huang,Börje F. Karlsson
类目: Machine Learning (cs.LG)
*备注:
Abstract:Recent studies have shown that multimodal large language models (MLLMs) can serve as embodied agents, translating language instructions and visual observations into executable plans. However, building agents that can continually improve through interaction and rapidly adapt to their environments remains challenging. Summing up experience from past interaction trajectories provides a promising solution, but existing experience-based methods often rely on manually designed prompting workflows to extract and update skills. Such fixed procedures may struggle to learn updated skills from new and diverse experiences. We introduce PRACTICE, which trains a skill learner to discover and maintain a persistent skill library from past interaction trajectories while keeping the task executor frozen. Given the historical accumulated skills and incoming trajectories, the skill learner produces structured batch-edits that add, refine, merge, or remove skills, and then hierarchical consolidate all collected edits into a consistent updated skill library. We train the learner with a two-stage curriculum. First, it learns basic skill generation and library maintenance from oracle trajectories. Then, by contrasting successful and failed trajectories from heterogeneous executors on the same tasks, it learn to identify invalid action patterns and recovery strategies. Finally, we apply online skill-edit distillation to align the skill learner with a stronger teacher on its current edit distribution to further improves the policy. Experiments demonstrate that a compact skill learner delivers consistent performance improvements across successive library-update rounds for multiple frozen executors. On EB-ALFRED and EB-Habitat, PRACTICE further outperforms the strongest experience-based baselines. Project resources are publicly available at: this https URL
[LG-24] Do VLMs Share Safety Neurons Across Modalities? EMNLP2026
链接: https://arxiv.org/abs/2608.30750
作者: Jiaxuan Li,Jiahao Zhang,Duc Minh Vo,Huy H. Nguyen,Pride Kavumba,Koki Wataoka
类目: Machine Learning (cs.LG)
*备注: Accepted to EMNLP 2026 (Main Conference)
Abstract:Vision-language models (VLMs) can comply with harmful requests delivered through images, even when their LLM backbones would refuse the same content in text. While prior work characterizes these jailbreaks empirically or at the representation level, how visual inputs perturb safety pathways at the neuron level remains uncharted. We close this gap with a causal, neuron-level analysis of safety mechanisms in 10 VLMs. We propose a two-stage detection pipeline with iterative ablation that accounts for self-repair, and introduce two modality-isolated benchmarks, ViSafe-Detect and ViSafe-Eval, which decouple visual and textual safety signals. Our analysis reveals: (i) Text safety in VLMs is localizable: \sim 88 neurons ( 0.01%) whose targeted ablation substantially reduces refusal. (ii) Text safety neurons constitute the dominant refusal pathway: ablating them is the only intervention that consistently and substantially reduces refusal across all models. (iii) Visual safety is high-dimensional and diffuse at the single-neuron level: text safety concentrates in \sim 5 subspace directions while visual safety requires \geq 50. This gap holds across architectures, explaining why current alignment has not closed the visual safety gap. Project page is at: this https URL Warning: this paper may include examples of harmful content. Comments: Accepted to EMNLP 2026 (Main Conference) Subjects: Machine Learning (cs.LG) Cite as: arXiv:2608.30750 [cs.LG] (or arXiv:2608.30750v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2608.30750 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-25] DDM-Melatt: A Decoupled Memory and Diffusion Framework for Generalizable Encrypted Traffic Classification CCS2026
链接: https://arxiv.org/abs/2608.30745
作者: Ze Chen,Qiming Yu,Zijia Song,Guozheng Yang,Wei Yan
类目: Machine Learning (cs.LG)
*备注: 18 pages, 13 figures, 9 tables, accepted at the 2026 ACM Conference on Computer and Communications Security (CCS 2026)
Abstract:The widespread adoption of encrypted traffic poses severe challenges to current security situational awareness systems based on network traffic monitoring. In existing dataset-driven training and testing studies, limitations such as shortcut learning induced by spurious feature correlations and sample imbalance caused by the long-tail distribution of real-world traffic result in weak generalization of traffic identification performance to real-world network traffic. To address these limitations, we propose TDDM-Melatt, a disentangled memory-based traffic classification framework with diffusion-based data augmentation. First, we design Melatt, a memory-decoupled traffic representation model, which employs Competitive Gating Long Short-Term Memory (CG-LSTM) to construct the encoder and decoder. We design a spurious-correlation-free pre-training and inference paradigm, employing strict topology anonymization and a frozen pre-trained encoder strategy to cut off the model’s learning pathways for spurious features. During inference, classification is performed efficiently by a downstream classifier on the frozen representations. Second, we propose a Traffic Denoising Diffusion Model (TDDM) tailored to the characteristics of traffic data. Extensive experiments are conducted on 4 representative public benchmark datasets. Under strict flow-level splitting and anonymization, TDDM-Melatt outperforms 6 basic classification models and 6 SOTA representation learning models. The proposed method provides a new and effective technical pathway for encrypted traffic classification in real-world network environments.
[LG-26] Functional Degeneracy in Neural Networks: Measurement and Pruning
链接: https://arxiv.org/abs/2608.30741
作者: Maria Matveev,Pascal Esser,Ayush Bharadwaj,Lucius Bushnaq,Gitta Kutyniok
类目: Machine Learning (cs.LG)
*备注:
Abstract:A central question in modern machine learning is how much a trained model can be compressed without changing its behavior, to reduce the memory, compute and energy required to deploy it. To study this, we quantify functional degeneracy through the behavioral recovery rank, defined as the number of leading behavioral-Hessian eigendirections required to recover a trained model’s performance. Using the behavioral recovery rank as a geometric benchmark for compression, we find that structural and magnitude pruning retain more degrees of freedom, even after the task is saturated. This gap suggests that functional redundancy is distributed across parameter directions and is not exposed by individual weights or neurons.
[LG-27] racing distinguishability through transformer processing with stochastic LayerNorm STOC
链接: https://arxiv.org/abs/2608.30720
作者: Kieran Murphy
类目: Machine Learning (cs.LG)
*备注: Code: this https URL
Abstract:Representational similarity is foundational to analyses of deep networks, yet distances between point-valued representations are not intrinsically tied to downstream function: nearby states may produce different behaviors, while distant states may behave similarly. We instead give representations volume, turning similarity into statistical distinguishability. Overlapping stochastic representations necessarily induce overlapping downstream distributions, grounding latent comparison in model function and bringing it under information-theoretic tools such as the data-processing inequality. We realize this idea in pretrained transformers through a light-touch modification to LayerNorm: at each residual-stream read, we normalize the state, add isotropic Gaussian noise, and renormalize. During distillation fine-tuning, one learned allocation parameter per residual-stream read distributes a fixed global rate budget across the processing stack. The resulting model can be viewed as transformer blocks reading the residual stream with learned finite precision under a shared global rate budget. Using the Bhattacharyya coefficient, we trace which counterfactual distinctions are preserved through MLP blocks or selectively exposed to the query, key, and value computations of individual attention heads. Experiments on ViT-S and GPT-2 small reveal the depthwise propagation of continuous visual perturbations and head-specific sensitivity to token distinctions aligned with known attention motifs. These results establish distinguishability as a functionally grounded lens on transformer computation that complements existing interpretability approaches.
[LG-28] Kolmogorov–Arnold against bounded translations
链接: https://arxiv.org/abs/2608.30710
作者: Sviatoslav V. Dzhenzher
类目: Machine Learning (cs.LG); Functional Analysis (math.FA)
*备注: 7 pages
Abstract:Historically originating from Hilbert’s 13th problem, the Kolmogorov-Arnold representation theorem (KART) has recently experienced a major revitalisation through its applications to neural networks, specifically Kolmogorov-Arnold Networks (KANs). While the exact representation is well established, its stability under continuous adversarial perturbations of the hidden layer remains a critical open question. In this paper, we investigate the robustness of KART against bounded adversarial translations. We provide an explicit, self-contained, and constructive proof of an approximate representation using fixed, piecewise linear inner functions. Crucially, our construction employs a single outer function that remains invariant for all summands and is independent of the specific adversarial translation, provided its maximum bound is known a priori.
[LG-29] Liquid Gated Attention
链接: https://arxiv.org/abs/2608.30695
作者: Yiheng Jiang,Yuanbo Xu,Yongjian Yang
类目: Machine Learning (cs.LG)
*备注: Code is available at this https URL
Abstract:Real-world time series often exhibit irregular sampling and extended temporal horizons, requiring models to capture continuous-time dynamics across arbitrary intervals without prohibitive scaling costs. Discrete-time methods collapse variable time intervals into static positional steps; solver-dependent continuous-time models preserve temporal structure but rely on sequential integration, precluding parallelization; and solver-free approximations avoid this cost yet none couples observed time intervals with input-driven state modulation. We propose Liquid Gated Attention (LGA), a solver-free parallel temporal operator. By parameterizing an input-driven gating mechanism with observed time intervals, LGA introduces a continuous-time inductive bias and formulates hidden state evolution as a fast-weight associative memory, enabling parallel computation across the temporal dimension. Using matrix associativity in non-causal encoding and a prefix scan in causal encoding, LGA attains linear temporal complexity in sequence length in both modes. A sequence-level normalization bounds cumulative temporal decay for stable long-horizon optimization. Building on LGA, we instantiate LFormer, a modular backbone for continuous-time representation learning. Across six tasks and sixteen datasets spanning up to 17,984 steps, LFormer demonstrates long-range dependency modeling, fine-grained state tracking, and trajectory reconstruction from sparse and noisy observations, while delivering competitive performance against state-of-the-art discrete-time and continuous-time baselines with linear scaling efficiency.
[LG-30] Season-Aware Hybrid Convolutional-Transformer for Antarctic Sea Ice Concentration Forecasting
链接: https://arxiv.org/abs/2608.30654
作者: Danyang Li,John Taylor,Thang Bui,Quanling Deng
类目: Machine Learning (cs.LG)
*备注:
Abstract:Antarctic sea ice concentration (SIC) forecasting is an important yet challenging task due to the coexistence of complex spatial structure, long-range temporal dependencies, and strong seasonal variability. Conventional convolution-based models are effective at capturing local spatial patterns, but often have limited ability to model long-term temporal evolution. To address these challenges, we build on a hybrid Convolutional-Transformer forecasting framework for monthly Antarctic SIC forecasting. This framework combines convolutional encoding for spatial feature extraction with factorised self-attention for spatio-temporal dependency modelling. We further introduce two seasonal prior mechanisms: a month-aware positional encoding that injects calendar-month information into the token representation, and a seasonal temporal bias that encourages attention to periodically related historical states. Experimental results show that the proposed framework achieves better performance than convolutional and recurrent baselines across both classification and regression metrics. Ablation studies further indicate that the seasonal prior mechanisms provide consistent additional gains in both short- and long-horizon prediction. These results demonstrate the value of combining convolutional structures, attention mechanisms, and periodic prior information for Antarctic SIC forecasting.
[LG-31] hree Steps at a Time: Learning Representations from Action Sequences in Contrastive RL
链接: https://arxiv.org/abs/2608.30640
作者: Michal Korniak,Kamil Dybek,Benjamin Eysenbach,Marco Bagatella,Michał Bortkiewicz
类目: Machine Learning (cs.LG)
*备注:
Abstract:While self-supervised approaches to reinforcement learning have achieved strong results by learning representations of states and actions, a key open question is the time scale over which actions should be modeled. Departing from the standard formulation relying on single-step actions, we extend contrastive reinforcement learning (CRL), a prototypical self-supervised method, to operate over action chunks, and find that this results in large, pervasive gains across established offline and online benchmarks: +31.7% and +93.1% across 18 and 11 environments respectively. While action-chunking-driven gains are generally explained through the ability to model non-Markovian, temporally extended policies, and to propagate unbiased multi-step returns, interestingly, we find that these arguments only partially apply to CRL. Our empirical studies suggest that, in the context of CRL, an action chunk carries more information about the goal than a single action, measurably improving the critic’s representations, and rendering the algorithm significantly more effective.
[LG-32] MolLedger: An Additive Graph Neural Network with Chemically Grounded ADME Attributions
链接: https://arxiv.org/abs/2608.30636
作者: Christina X. Ji
类目: Machine Learning (cs.LG)
*备注:
Abstract:Optimizing absorption, distribution, metabolism, and excretion (ADME) is an important part of small molecule drug discovery. Many machine learning models have been built to predict ADME properties to facilitate this optimization process, but explaining model predictions is challenging. We propose a new graph neural network architecture with built-in meaningful per-atom attributions. Our model MolLedger outputs predictions that are the sum of per-atom scores. MolLedger’s additive framework obtains exact interpretability at no cost to performance because the global context vector gives the additive head enough context to produce good per-atom scores. Furthermore, MolLedger produces attributions that are more faithful to chemical properties than other interpretability methods because the auxiliary loss in MolLedger anchors the atom scores to chemical properties. Our case studies comparing interpretations from multiple methods on molecular pairs reveal that MolLedger is much better at producing sensible explanations for predicted property changes.
[LG-33] State of Health Estimation using Convolutional and Bidirectional LSTM Neural Networks tuned by Bayesian Optimization
链接: https://arxiv.org/abs/2608.30593
作者: Panagiotis Eleftheriadis,Foivos Georgios Kyrgios,Sonia Leva
类目: Machine Learning (cs.LG)
*备注:
Abstract:In this research, a novel framework is proposed for the SOH estimation, which employs a hybrid deep learning architecture of a concatenation of a Convolution Neural Network (CNN) and a Bidirectional Long Short-Term Memory (BiLSTM) Neural Network (NN) with the integration of Bayesian Optimization-based hyperparameter tuning for the network. Three different deep learning architectures are being evaluated: standalone recurrent models, CNN-RNN architectures and CNN-RNN combinations enhanced with intermediate Fully Connected (FC) layers. Among the three, the model with the intermediate FC layers demonstrated the highest predictive accuracy. A comprehensive feature engineering approach combines capacity (Q), voltage (V), Incremental Capacity Analysis (ICA), and Differential Voltage Analysis (DVA), with systematic evaluation of multiple combinations to identify the optimal input representation. To validate the proposed method, three publicly available datasets were utilized, ensuring reproducibility of the results, two from external sources and one developed by the author of this study using a unique experimental setup. The comparison study was performed using the Mean Absolute Error (MAE), the Root Mean Squared Error (RMSE) and the FLoating-point OPerations (FLOPs) as evaluation metrics.
[LG-34] he Safety Relay in Roleplay Jailbreaks: A Component-Resolved Causal Analysis of Harm Recognition and Refusal
链接: https://arxiv.org/abs/2608.30585
作者: Md Mokarram Chowdhury,Ernie Chang,Yang Li
类目: Machine Learning (cs.LG)
*备注: Preprint
Abstract:Large language models are trained to follow instructions while refusing harmful requests. Jailbreaks exploit this balance to elicit content a model would ordinarily reject. Roleplay jailbreaks are especially concerning: the harmful request can remain visible inside a roleplay wrapper made of a persona, scenario, and task, yet the model may comply. We use mechanistic interpretability to determine how this context reverses refusal and which elements contribute to the reversal. Across two benchmarks, three model families, and four authored wrappers, we compare matched harmful and benign requests with and without this wrapper. We trace hidden-state contrasts from the request to the final prompt state, isolate wrapper operations through controlled counterfactuals, intervene on their activation directions in held-out evaluation requests, and decompose effective directions geometrically. Our analysis yields three findings. (1) Successful attacks retain the measured harmful-versus-benign distinction at the request, while its refusal-associated expression weakens where the answer begins, a pattern we call safety-relay attenuation. (2) Constructing the complete roleplay around the request and framing it within the scenario contribute causally: removing the associated activation changes restores refusal. (3) These effects largely share internal structure, and most repair is reproduced by components aligned with the model’s ordinary refusal of harmful requests without roleplay; scenario framing retains a smaller, model-dependent component. Together, these findings explain how roleplay can produce compliance despite retained evidence of harm and identify a concrete target for future safeguards: maintaining the connection from harm recognition to refusal. Comments: Preprint Subjects: Machine Learning (cs.LG) Cite as: arXiv:2608.30585 [cs.LG] (or arXiv:2608.30585v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2608.30585 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-35] PAC: Progress-Augmented Advantage Curriculum for Multi-Task Reinforcement Learning of LLM s EMNLP2026
链接: https://arxiv.org/abs/2608.30528
作者: Yuanqiang Yu,Yanzhao Zheng,Zhentao Zhang,Tianze Xu,Chao Ma,Jihuai Zhu,Jiashun Liu,Xinle Deng,Baohua Dong,Hangcheng Zhu,Ruohui Huang
类目: Machine Learning (cs.LG)
*备注: Accepted at EMNLP 2026 (Main)
Abstract:Reinforcement learning (RL) is used to improve the reasoning abilities of LLMs, while training data span heterogeneous tasks. However, most RL post-training pipelines rely on fixed or manually designed task mixtures, even though task usefulness changes as training progresses. Online curriculum methods often define learnability by update magnitude, ignoring whether the update translates into reward gains, which can misallocate rollout budget toward tasks with large but ineffective updates. We propose PAC, a Progress-Augmented Advantage Curriculum for multi-task RL of LLMs that combines two task-level signals: advantage-derived learnability, which measures the magnitude of the policy update a task can induce, and recent reward gains, which show whether those updates have improved task performance. A Bayesian Thompson Sampling controller uses these signals to allocate rollouts across tasks during GRPO training. We evaluate PAC under two settings: a multi-level reasoning setting and a multi-domain reasoning setting. PAC improves sample efficiency and final performance: it reaches comparable validation scores with fewer rollout steps and achieves higher final averages than random sampling and advantage-based curriculum baselines in both settings. These results show that jointly tracking advantage signals and actual reward gains yields an effective online curriculum for LLM post-training.
[LG-36] When the Martingale Never Stops Firing: Anytime-Valid Gating on Real Forecast Streams
链接: https://arxiv.org/abs/2608.30502
作者: Weijia Han,Lisha Qu
类目: Machine Learning (cs.LG); Methodology (stat.ME); Machine Learning (stat.ML)
*备注: 4 pages, 2 figures
Abstract:Machine learning systems are increasingly corrected while they run, and the decision of when to intervene is increasingly delegated to statistical monitors. Anytime-valid inference promises evidence that can be acted on at any moment, exactly the guarantee this setting needs, and it is moving from theory into deployed monitoring. Conformal test martingales are the change-detection instrument, and Ville’s inequality caps their false-alarm probability on exchangeable data. The guarantee is conditional. A deployment inherits it only if the stream it monitors behaves exchangeably. The premise is hardest to satisfy where these monitors are most useful, on dependent data and inside loops where the monitor modifies the learner whose scores it reads. It is also rarely measured. We measure it in a pre-specified case study, where such a monitor gates the online updates of a Kalman adapter correcting frozen time-series foundation models on five forecasting streams. On exchangeable synthetic streams, the same implementation fires in at most 1 of 60 runs. On the real streams, at alpha = 0.05, 135 of 135 clean-stream runs fired. The construction does not explain the firing; the failure comes from the deployed score stream itself. Repeated fires hold the gate’s drift response active, and the gated filter amplifies the very transient it was designed to prevent. The component worth keeping makes no validity claim. Huber-style gating of the filter’s own updates cuts isolated-spike degradation by an order of magnitude with no dataset specific tuning. Anytime-valid methods proposed for dependent data should therefore be accompanied by null-calibration controls and mechanism traces.
[LG-37] oxLens: A Reproducible Graph-Learning Framework for Leakage-Aware Uncertainty-Calibrated Molecular Toxicity Prediction
链接: https://arxiv.org/abs/2608.30472
作者: Magnus H. Strømme,Alex G. C. de Sá,David B. Ascher
类目: Machine Learning (cs.LG)
*备注: 39 pages, 5 figures
Abstract:Molecular toxicity prediction is increasingly used to prioritise compounds before experimental testing, but conventional benchmark performance can overstate practical utility when structurally related molecules occur across training and test folds. We introduce ToxLens, a reproducible multi-task graph-learning framework for 11 toxicity endpoints spanning Ames mutagenicity, acute oral toxicity, hERG inhibition, and Tox21 nuclear-receptor and stress-response assays. The workflow combines conservative chemical curation, sphere-exclusion filtering, a leakage-aware UMAP-HDBSCAN split, parallel graph and global-feature encoders joined by late concatenation, temperature-scaled Monte Carlo dropout with conformal-style prediction sets, applicability-domain analysis, and SHAP-guided toxicophore discovery with occlusion controls. On the leakage-controlled test fold, a five-seed soft-voting ensemble achieved a Matthews correlation coefficient score of 0.44, an area under the receiver operating characteristic curve score of 0.83, and an area under the precision-recall curve score of 0.58. It exceeded four ECFP4-based shallow baselines on all 11 endpoints under the same split and validation-based threshold-selection protocol. Controlled ablations showed that the global pathway was important, whereas late concatenation outperformed the tested gated and feature-wise linear modulation fusion variants. Conformal-style prediction sets revealed substantial endpoint-specific variation in set efficiency, and discrimination and calibration improved with similarity to the training domain. Retraining on fixed published Tox21 Challenge and TDA folds produced competitive, but not uniformly state-of-the-art, performance. SHAP-guided occlusion and consensus subgraph mining yielded model-derived structural hypotheses, 44 of which contained at least one occurrence that passed the predefined counterfactual criteria.
[LG-38] Self-Supervised Pretext Tasks for Infant Cry Analysis: A Controlled Comparison and a Cautionary Result on Donateacry
链接: https://arxiv.org/abs/2608.30456
作者: Luigi Simeone
类目: Machine Learning (cs.LG)
*备注: 11 pages, 2 figures
Abstract:We compare six self-supervised pretext tasks for infant cry analysis under a fixed budget, meaning the same compact encoder of 1.17M parameters, the same 115 hours of license-verified public pretraining audio, and the same evaluation protocol for every candidate. On cry detection the reconstructive objectives dominate, and a linear probe over a masked-spectrogram encoder reaches 0.988 AUC with subject-wise splits even though the encoder never observed a cry during pretraining. On cry-reason classification over donateacry, the de facto public benchmark for cry reasons, every encoder performs at chance (0.38 to 0.54 macro AUC over 5 classes), and neither domain adaptation on 1.8 hours of real cries nor end-to-end fine-tuning moves the result. Since a frozen HuBERT-base with 80 times more parameters shows the same pattern, the bottleneck must sit in the labels and not in model capacity. We then reproduce the 90%+ accuracies of the donateacry literature on our own system by changing nothing but the evaluation protocol: clip-wise splits raise accuracy to 85.2% (barely above the 83.8% majority-class baseline), and applying augmentation before splitting raises it to 97.9%, matching the reported state of the art, from the same model that measures 0.49 macro AUC under subject-wise splits. Under leakage-free splits, a twentyfold augmentation of the labeled set (vocoder speaker perturbation and noise mixing, 21 hours) leaves cross-subject AUC unchanged: for this task the effective sample size is the number of infants. We release code, seeds and per-clip license manifests.
[LG-39] Confounding Masquerading as Improvement: A Systematic Evaluation of Offline Reinforcement Learning for Stroke Antithrombotic Treatment in a 129000-Patient Registry ALT
链接: https://arxiv.org/abs/2608.30442
作者: Kihun Rhee
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: 37 pages, 5 figures. Accepted at Machine Learning for Healthcare (MLHC) 2026; to appear in Proceedings of Machine Learning Research (PMLR), vol. 340
Abstract:Recent offline reinforcement learning (RL) studies report policies that outperform physician decisions on clinical outcomes. We conduct a systematic, partially crossed evaluation of five offline RL algorithm families and 14 reward designs in 44,894 post-2018 acute ischemic stroke patients from a nationwide registry (N = 129,033). Standard Fitted Q-Evaluation (FQE) yields an apparent policy-improvement estimate of +0.0069; adding an Early Neurological Deterioration penalty increases it to +0.0101. We identify reward-embedded confounding, in which a proxy terminal reward encodes baseline severity and prognosis as well as treatment efficacy. A 2 x 2 factorial analysis finds that terminal reward confounding accounts for 218.6% of the observed signal change, so its removal overshoots the null. After DML-inspired GBM reward residualization, the FQE estimate attenuates to +0.0033 (p = 0.132), and full deconfounding yields +0.0025 (p = 0.291). FQE-based diagnostics, T-learner analyses, and direct recurrence analyses converge away from a clinically meaningful aggregate improvement. A 1-year mRS factorial analysis replicates the attenuation. We provide an empirically motivated six-step evaluation checklist. NIHSS-stratified heterogeneity is hypothesis-generating for prospective trial design; hospital-level disagreement does not persist after full reward deconfounding. Comments: 37 pages, 5 figures. Accepted at Machine Learning for Healthcare (MLHC) 2026; to appear in Proceedings of Machine Learning Research (PMLR), vol. 340 Subjects: Machine Learning (cs.LG); Machine Learning (stat.ML) Cite as: arXiv:2608.30442 [cs.LG] (or arXiv:2608.30442v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2608.30442 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-40] Event-Driven Language Models with Sparse Neural Activity for Neuromorphic Hardware
链接: https://arxiv.org/abs/2608.30439
作者: Simon Richter,Ruhai Lin,Jason Yik,Taylor Kergan,Rui-Jie Zhu,Farshad Moradi,Jason Eshraghian
类目: Neural and Evolutionary Computing (cs.NE); Machine Learning (cs.LG)
*备注: 8 pages, 4 figures, Accepted at IEEE MCSOC2026
Abstract:Inference with transformer-based large language models (LLMs) is often limited by the memory-bound KV cache and quadratic attention cost. State-space models (SSMs) mitigate this through linear attention and fixed-size recurrent states, but their large dense linear projections remain computationally expensive even after quantization. We introduce a method that induces sparse neural activity in heavily quantized linear-attention models with minimal performance loss. Activations below a per-projection trainable threshold ( \pm \Delta ) are nullified while preserving crucial outliers, achieving comparable performance to dense models with up to 4 \times fewer effective arithmetic operations. Targeting a multi-core, multi-chip neuromorphic platform, where event-driven execution converts unstructured sparsity into throughput at both the compute and communication levels, a capability GPU architectures fundamentally lack, we project up to 37 \times higher throughput and 16 \times lower power versus edge GPU inference of a comparable transformer-based model, and up to 5.4 \times improvements over the non-sparsified baseline. These results position sparse, quantized linear-attention models as a natural fit for deploying LLMs on event-driven multi-core platforms.
[LG-41] Generalization as a robust performance property of learning-enabled dynamical systems
链接: https://arxiv.org/abs/2608.30431
作者: Filippo Fabiani
类目: ystems and Control (eess.SY); Machine Learning (cs.LG); Optimization and Control (math.OC)
*备注:
Abstract:By focusing on algorithmic stability as a means of establishing out-of-sample bounds, we provide a system-theoretic interpretation of generalization in learning-enabled dynamical systems arising in data-driven optimization and feedback control approximation. Given two neighboring datasets, we specifically model sample replacement as an exogenous disturbance acting on a sensitivity system, while the incremental behavior of the data-dependent operator is encoded through an integral quadratic constraint. By relying on dissipativity arguments, we establish a matrix inequality-based certificate and a uniform stability bound that separates the one-sample sensitivity of the learned operator, and an algorithm-dependent dynamical gain. The latter can then be optimized, offering a tractable tool for certifying and comparing generalization capabilities of learning dynamics. We show that our results recover classical ones for gradient descent, apply naturally to momentum-based methods such as heavy-ball and Nesterov acceleration, and extend to data-driven control.
[LG-42] No Equivariant Architecture Covers All Equivariant Attention
链接: https://arxiv.org/abs/2608.30417
作者: Tīkun Ông
类目: Machine Learning (cs.LG); Algebraic Geometry (math.AG); Representation Theory (math.RT)
*备注:
Abstract:We give a complete characterization of equivariant multi-head self-attention (MHSA): if an MHSA layer is equivariant to a symmetry group G , then G can only act by permuting head-clusters, with QK and OV matrices satisfying an equivariance constraint tied to the group action. As a consequence, we prove that any fixed MHSA architecture that achieves exact equivariance by polynomially parameterizing unconstrained MHSA parameters inevitably leads to expressivity loss within the class of equivariant maps: the equivariance locus of unconstrained MHSA forms a union of extremely many Zariski-irreducible components in a reduced parameter space, and any single architecture covers at most one. For G=D_4 acting on C copies of the regular representation as the token feature space, we show that there are \Omega(C^64) components for eight attention heads.
[LG-43] Locally-Guided Actor-Critic: Training a Goal-conditioned Actor with a Subgoal-aware Critic
链接: https://arxiv.org/abs/2608.30406
作者: Olivier Serris,Stéphane Doncieux,Olivier Sigaud
类目: Machine Learning (cs.LG)
*备注:
Abstract:Goal-conditioned reinforcement learning struggles with long horizons when rewards are sparse. While a planner can provide subgoals to guide a low-level policy, its use at test time may introduce practical subgoal management difficulties. An alternative paradigm utilizes a high-level planner to assist learning, while the policy remains conditioned only on the final goal, enabling planner-free deployment. Among these methods, Reinforcement Learning with Imagined Subgoals (RIS) introduces a regularization term that encourages the policy to take the same actions for the final goal as it does for an intermediate goal. This regularization, however, may lead to goal-chaining issues when intermediate goals are low-dimensional. Potential-based reward shaping (PBRS) translates plans into an additional reward while ensuring that the optimal policy remains unchanged. Yet, it can generate deceptive rewards in terminal states. We study these failure cases and first propose an alternative reward shaping method (RS) that removes these deceptive rewards at the expense of theoretical guarantees of PBRS. Similar to this RS variant, we then propose another method named Locally-Guided Actor Critic (LG-AC) that rewards the agent for reaching intermediate goals. Unlike RS, where intermediate rewards are implicit in the shaping signal, we explicitly condition a value estimator on the full sequence of intermediate goals but represent the value function as a sum of subgoal-conditioned value functions, enabling dense hindsight relabeling. We evaluate all these methods in tasks with challenging goal-chaining requirements and empirically highlight specific cases in which either action regularization or reward shaping yield low performance, while LG-AC achieves the best overall performance across tasks.
[LG-44] Foundation Models Meet Agriculture: Challenges Beyond Pretraining
链接: https://arxiv.org/abs/2608.30392
作者: Vishal Nedungadi,Xingguo Xiong,Marc Rußwurm,Ioannis N. Athanasiadis
类目: Machine Learning (cs.LG)
*备注:
Abstract:Global food security and sustainable climate action increasingly rely on robust, scalable agricultural monitoring. Earth observation foundation models have emerged as powerful, label-efficient tools across general remote sensing domains, yet early attempts to deploy them for agricultural applications have yielded surprisingly poor results. We hypothesize that this performance gap stems from the extreme heterogeneity of agricultural landscapes and the inherent inability of current earth observation foundation models to adapt to task-specific nuances. In this work, we systematically evaluate two critical bottlenecks hindering the deployment of foundation models in agricultural tasks, benchmarking two earth observation foundation models, a foundation model designed for tabular data, and conventional supervised baselines across seven real-world agricultural datasets spanning yield prediction, phenology estimation, and crop classification. First, we identify a pretraining-deployment modality gap: agricultural downstream tasks frequently require diverse, non-imagery data modalities that earth observation foundation models are architecturally unequipped to ingest, while a foundation model built for tabular data handles this heterogeneity more naturally. Second, we formalize the agricultural task space across five structural axes to demonstrate why current models fail to generalize reliably, resulting in highly unstable model rankings across evaluation settings. By characterizing these structural and modal gaps, our insights highlight the friction between general-purpose architectures and specialized agricultural downstream data, providing a strategic roadmap for developing the next generation of domain-aware foundation models.
[LG-45] Uncertainty of Vision Medical Foundation Models
链接: https://arxiv.org/abs/2608.30390
作者: Haoxu Huang,Narges Razavian
类目: Machine Learning (cs.LG)
*备注:
Abstract:Accurate uncertainty estimation is essential for machine learning systems de- ployed in high-stakes domains such as medicine. Traditional approaches primarily rely on probability outputs from trained models (point predictions), which provide no formal guarantees on prediction coverage and often require additional calibra- tion techniques to improve reliability. In contrast, conformal prediction (region prediction) offers a principled alternative by generating prediction sets with finite- sample validity guarantees, ensuring that the ground truth is contained within the set at a specified confidence level. In this study, we explore the impact of pre-training approach, dataset scale and domain on both point and region-level uncertainty quantification, by studying domain-specific vision medical foundation models vs. general domain vision foundation models. We conduct a comprehensive evaluation across foundation models trained on retinal, histopathological, and Chest X-Rays data, applying various calibration techniques. Our results demonstrate that (1) pre-training on higher-quality domain-specific datasets along with self-supervised learning leads to better-calibrated point predictions than general domain pre-training, (2) stan- dard re-calibration methods alone cannot fully mitigate uncertainty discrepancies across models trained on different data sources, (3) domain-specific foundation model can lead to more efficient conformal prediction. These findings highlight the importance of careful model selection and the inte- gration of both point and region prediction to enhance the reliability and trust- worthiness of medical AI systems. Our work underscores the need for a holistic approach to uncertainty quantification in recent development of medical vision foundation model, ensuring robust and interpretable AI-driven decision-making.
[LG-46] Convergence rates for the RMSprop optimizer with full control of the hyperparameters
链接: https://arxiv.org/abs/2608.30382
作者: Steffen Dereich,Arnulf Jentzen
类目: Machine Learning (cs.LG); Optimization and Control (math.OC); Probability (math.PR)
*备注:
Abstract:Popular adaptive stochastic gradient descent (SGD) methods to train artificial intelligence (AI) systems include the RMSprop, the Adam, and the AdamW optimizers, where the adaptivity parts in Adam and AdamW basically just coincide with RMSprop. Such adaptive methods involve several hyperparameters including the regularization parameter \epsilon (which ensures that one does not divide by 0 and is often chosen to be very close to zero such as 10^-8 in PyTorch by default) and the second moment decay parameter \beta (which is often chosen to be very close to 1 such as 0.99 (RMSprop) and 0.999 (Adam and AdamW) in PyTorch by default). Despite the high relevance of such methods, it remains an open research problem to provide error estimates for such methods with the error constants being not exploding but uniformly bounded with the respect to the hyperparameters, even in the situation of convex stochastic optimization problems. It is the key contribution of this work to essentially solve this problem for RMSprop. Specifically, we bound the expectation of the stopped evaluation of the objective function at the RMSprop process from above by the sum of an initialization term that decays exponentially in the training time, a stochastic approximation remainder of order \gamma_n , and a memory error of order ( 1 - \beta)^2 with the error constants being uniformly controlled over all admissible choices of the step sizes, the second moment decay parameter \beta and the regularization parameter \epsilon\in[0,1] (also covering \epsilon=0 ). Our non-asymptotic error estimates hold not just for all sufficiently large n but hold for every gradient step n=1,2,3,… with all error constants being explicitly specified. The key innovative new feature in the proof of our analysis are suitable inverse moment bounds for the second moment process in RMSprop. Subjects: Machine Learning (cs.LG); Optimization and Control (math.OC); Probability (math.PR) Cite as: arXiv:2608.30382 [cs.LG] (or arXiv:2608.30382v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2608.30382 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-47] Beat-Synchronous Tokenization for ECG Transformers
链接: https://arxiv.org/abs/2608.30367
作者: Ahmed Sameh,Nolan Wilson,Max Enderlein,Yogatheesan Varatharajah
类目: Machine Learning (cs.LG); Signal Processing (eess.SP)
*备注: 6 pages, 1 figure, 3 tables. Accepted at the 2026 IEEE International Workshop on Machine Learning for Signal Processing (MLSP 2026)
Abstract:Transformer-based electrocardiogram (ECG) models commonly tokenize waveforms into fixed temporal patches. Though convenient, fixed patching can split heartbeat structures across token boundaries. We study beat-synchronous tokenization as a physiologically grounded alternative, comparing fixed patches with three beat-aligned strategies: resampled beats, adaptive pooled beats, and resampled beats augmented with R–R interval information. Experiments span two settings: 10-second 12-lead diagnostic classification on PTB-XL after MIMIC-IV-ECG masked pretraining, and 60-second single-lead rhythm classification on Icentia11k after patient-level contrastive pretraining. On PTB-XL, resampled beat tokens achieve the highest mean macro Area Under the ROC Curve (AUROC; 0.8945) and nearly match the best fixed-patch macro Area Under the Precision-Recall Curve (AUPRC; 0.7414), reducing average sequence length from 100 to 11.2 tokens. On Icentia11k, beat-synchronous tokenizers obtain comparable AUPRC to fixed patching with better stability across runs. These results suggest morphology-preserving beat tokenization is a compact, competitive alternative to fixed temporal patching.
[LG-48] Mode Connectivity Beyond Classifiers: Evidence from Generative and Contrastive Models
链接: https://arxiv.org/abs/2608.30366
作者: Chengzheyi Yao,Yongzhao Zhang,Yongding Tian
类目: Machine Learning (cs.LG)
*备注:
Abstract:The loss landscape of Deep Neural Networks (DNNs) exhibits highly complex and non-convex properties. Recent studies have revealed the phenomenon of mode connectivity, demonstrating that independently trained network modes can be connected via a continuous low-loss path. However, existing mode connectivity research is predominantly confined to classifier-based models, leaving it an open question whether similar geometric properties exist in modern complex models. In this paper, we extend the boundaries of mode connectivity to generative and contrastive domains (specifically DDPM and NanoCLIP). Addressing the unique architecture of DDPM and CLIP, we propose an architecture-aware connection building algorithm. Extensive empirical results demonstrate for the first time that we successfully discover mode connectivity between independently trained DDPM and NanoCLIP modes. Our work provides a novel perspective for understanding the geometric properties of the loss landscapes in modern generative and contrastive models.
[LG-49] Beyond Churn: Predicting Financial Frag mentation in Retail Banking with Temporal Machine Learning
链接: https://arxiv.org/abs/2608.30364
作者: Ananyaa Chopra,Brandon Xu,Brendan Yuen,Lauren Zung,Sarabroop Aulakh
类目: Machine Learning (cs.LG)
*备注: 14 pages, 4 figures
Abstract:Retail banking attrition is usually represented as a terminal binary event, even though client relationships often weaken earlier through partial movements of deposits, investments, and recurring activity to external financial institutions. This paper defines that preceding state as financial fragmentation and presents an end-to-end temporal machine-learning system for predicting it before complete disengagement. Using anonymized multi-source data from a large retail bank, the framework predicts whether a valid external transfer or investment event will occur within 90 days. The study uses 595,220 client-month observations, with 346 engineered features combining monthly client profiles, balances, product relationships, prior flow-of-funds behavior, macroeconomic conditions, and competitor activity. A four-stage XGBoost cascade estimates (1) whether an external outflow will occur within 90 days, (2) the expected amount, (3) the originating product, and (4) the destination financial institution. The primary classifier achieved a test precision-recall area under the curve of 0.823. At the validation-selected threshold, it produced 86.4% precision, 75.1% recall, and an F1 score of 0.803. Ranking test observations in descending Stage 1 fragmentation score, the top 1% of clients yielded 95.3% precision, while the top 5% captured 78.7% of observed outflow cases. The amount model placed 94.9% of predictions within an adjacent amount bucket. Destination prediction reached a macro-F1 of 0.81 across 27 classes; source-product prediction achieved a weighted F1 of 0.92. By moving the analytical focus from terminal churn to earlier fund migration, the proposed approach provides a practical foundation for proactive, explainable, and economically informed client-retention decision support.
[LG-50] Coarse composition suffices: tabular in-context learning for multi-activity antimicrobial peptide profiling
链接: https://arxiv.org/abs/2608.30337
作者: Raunak Kumar,Anuj Pal,Dhruvi Solanki,Parikshit Pareek,Juhi Singh,Jitin Singla
类目: Machine Learning (cs.LG); Biomolecules (q-bio.BM)
*备注:
Abstract:Antimicrobial peptides (AMPs) often act against multiple pathogen classes, making multi-label activity prediction a more realistic screening target than binary antimicrobial classification. The ESCAPE benchmark formalizes this setting, but leading approaches typically rely on multimodal, structure-conditioned deep models that are costly to train and tune. We show that a simple, sequence-only pipeline can match and surpass these methods by combining 330 interpretable sequence descriptors with TabPFN, a tabular foundation model that performs in-context prediction in a single forward pass without gradient-based training or hyperparameter search. On ESCAPE (82,359 peptides; five labels), a label-powerset TabPFN model achieves mAP-5 = 77.8%, improving on the previously best reported 72.1%. A probabilistic classifier chain is the first method to match or exceed the best published average precision on each of the five labels simultaneously. The gains persist under the prior state-of-the-art single-fold training protocol, indicating they are not a training-set-size artefact, and are largest for remote homologues (+11.2 points below 30% sequence identity). Ablations further show that predicted structure is unnecessary at inference and that performance is not driven by any single descriptor family: ten global physicochemical scalars recover 91% of full-feature performance. Finally, explicitly modelling label dependence yields targeted benefits for scarce activities and supports ranking which activity to assay next from partial positive evidence.
[LG-51] Learning PDE Time-Stepping with Neural Cellular Automata
链接: https://arxiv.org/abs/2608.30328
作者: Esha Saha,Hao Wang
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: 4 figures, 7 tables
Abstract:Classical numerical solvers for partial differential equations (PDEs) are computationally expensive to solve repeatedly across varying initial conditions, motivating the need for learned surrogates. In this paper, we propose a trainable Neural Cellular Automata (NCA) based surrogate model for learning long time PDE dynamics. Rather than mapping an entire initial field to a full trajectory in one shot, our proposed model learns a small, local, homogeneous update rule that is applied identically and repeatedly at every grid cell, mirroring the locality of differential operators. We benchmark this framework against three baselines: PDE - Net, a modified physics-informed neural network (PINN), and a Fourier Neural Operator (FNO), on five canonical PDEs (heat, advection, Burgers, Allen - Cahn, and Fisher - KPP), evaluated at temporal domain two times beyond the training temporal domain. The proposed model achieves the lowest long-horizon relative errors on the majority of the experiments.
[LG-52] Generative multi-domain transfer learning for fault detection in data-scarce wind turbines
链接: https://arxiv.org/abs/2608.30323
作者: Stefan Jonas,Angela Meyer
类目: Machine Learning (cs.LG)
*备注:
Abstract:Normal behavior models have shown promise for reliable fault detection in wind turbines. However, these unsupervised anomaly detection models require sufficient fault-free training data to learn the normal operation behavior of turbines. Under data scarcity, for example in newly deployed wind turbines, these models may result in poor fault detection performance. In this work, we propose a multi-domain generative domain mapping approach based on Star Generative Adversarial Networks (StarGAN) to improve fault detection on data-scarce wind turbines. Our model maps SCADA measurements from a data-scarce turbine to resemble those of several data-rich turbines. By preserving the operational state during translation, faults occurring in a data-scarce domain can be mapped and detected by reliable pre-trained normal behavior models of data-rich domains. Highlighting the benefits of an ensemble fusion strategy, we show that under severe data scarcity our method can produce anomaly scores comparable to models trained on large representative datasets. Our approach can consistently outperform models trained on scarce data when less than 2 weeks of training data are available. With just 2 weeks of accumulated training data, we achieve an anomaly score similarity that is, on average, +16% higher than conventional fine-tuning, and +10% higher than single-source domain mapping. As a step towards unsupervised model selection, we propose a proxy metric that detects poor performance at training time, despite an absence of anomalies. Our study presents the potential and challenges of multi-domain mapping for wind turbine fault detection under unrepresentative training data.
[LG-53] Strengthening Recursive Constructions for Zero-Error Shannon Capacity
链接: https://arxiv.org/abs/2608.30273
作者: Ravi Tandon
类目: Information Theory (cs.IT); Machine Learning (cs.LG); Combinatorics (math.CO)
*备注:
Abstract:The exact Shannon capacity is unknown for every odd cycle beyond the five-cycle C_5 , making odd cycles a central open problem in zero-error information theory. Improving the known lower bounds requires constructing large independent sets in strong powers of these graphs. Recent AI-assisted work has produced a rapid sequence of improvements: building on the construction of Itty et al., Gao developed a recursive product construction for combining structured independent sets, and Buys, Polak, and Zuiddam (BPZ) subsequently strengthened this through a richer recursion framework. We continue this line of AI-assisted exploration and introduce a heterogeneous refinement of these constructions. The central observation is that the usefulness of an intermediate construction depends not only on the size of its current main independent set, but also on the auxiliary structure it carries into subsequent recursion. Consequently, different parts of that auxiliary structure need not use the same independent set, and different occurrences in a recursion need not use the same intermediate representation. We formalize this for Gao’s binary product and derive explicit propagation rules showing how heterogeneous choices strengthen the resulting gadget while leaving its current code size unchanged, then extend the principle to the more general BPZ framework, tailoring constructions to the distinct roles they play within the recursion. Applying these refinements to the seven-cycle C_7 , we obtain an independent set in C_7^\boxtimes 500 yielding \Theta(C_7)\ge 3.25883262\ldots , improving the best known lower bound. Beyond the numerical gain, the results illustrate a general principle for recursive zero-error constructions: intermediate structures with the same dimension and current code size can have different downstream value depending on where and how they are used in the recursion. Subjects: Information Theory (cs.IT); Machine Learning (cs.LG); Combinatorics (math.CO) Cite as: arXiv:2608.30273 [cs.IT] (or arXiv:2608.30273v1 [cs.IT] for this version) https://doi.org/10.48550/arXiv.2608.30273 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-54] Multivariate Scientific Data Compression with Learned Cross-Variable Latent Decorrelation and Autoregressive Entropy Modeling
链接: https://arxiv.org/abs/2608.30262
作者: Liangji Zhu,Anand Rangarajan,Sanjay Ranka
类目: Machine Learning (cs.LG)
*备注: 11 pages, 4 figures, 4 tables
Abstract:Scientific simulations generate collections of physical fields with heterogeneous statistics and dependencies, yet learned compressors often encode those fields independently or rely on a shared encoder without explicitly modeling the structure that remains in latent space. We present CAESAR-LDAR, an error-controlled multivariate learned compressor that augments a shared CAESAR-V backbone with two complementary mechanisms: a trainable orthogonal transform that reorganizes dependence across aligned latent channels, and a causal autoregressive hierarchical prior that captures local spatial structure left after transformation. Orthogonality is maintained through a matrix-exponential parameterization, making the transform exactly invertible without an additional penalty. A common residual-correction stage is applied uniformly to all variants to enforce the requested reconstruction tolerance. Experiments across combustion, climate, and turbulence data show that the two mechanisms are useful in different regimes. Latent decorrelation helps most when substantial linear cross-channel dependence survives the nonlinear encoder, whereas autoregressive modeling remains effective when the remaining structure is primarily local or spatial. Their combination provides the strongest or near-strongest rate-distortion performance across the evaluated datasets. The global transform adds little computational overhead, while autoregressive coding introduces a larger throughput tradeoff. More broadly, the results suggest a practical design principle for multivariate scientific compression: exploit global cross-channel dependence when it is measurably present in latent space, and use local probabilistic context as a complementary mechanism across a wider range of data regimes. Comments: 11 pages, 4 figures, 4 tables Subjects: Machine Learning (cs.LG) Cite as: arXiv:2608.30262 [cs.LG] (or arXiv:2608.30262v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2608.30262 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-55] Exact Recovery Thresholds for Weighted Data Selection in Vector-Valued Linear Regression
链接: https://arxiv.org/abs/2608.30254
作者: Guangjian Zhang
类目: Machine Learning (cs.LG); Statistics Theory (math.ST)
*备注: 21 pages
Abstract:We resolve the threshold part of Question 4 of the COLT 2025 open problem “Data Selection for Regression Tasks” of Hanneke, Moran, Shlimovich and Yehudayoff. In vector-valued linear regression with square loss \ell_(x,y)(W)=|Wx-y|_2^2 , where x\in\mathbbR^d , y\in\mathbbR^m and the learner is the empirical risk minimizer of minimal Frobenius norm, we prove that the minimal budget of weighted examples that recovers the full-data loss on every finite dataset is exactly n^*(d,m)=(m+1)d . We further determine two more values of the weighted selection profile F_w(d,m,n) : at the near-threshold budget, F_w(d,m,(m+1)d-1)=1+\frac1dm^2 , and at the spanning budget, F_w(d,m,d)=d+1 for every m , while F_w(d,m,n)=\infty for nd . For the smallest open intermediate cell (d,m)=(2,2) we prove F_w(2,2,3)\in[13/8,15/8] and F_w(2,2,4)\in[5/4,3/2] , reduce the conjectured exact values 13/8 and 5/4 to a finite moment problem on the circle with at most seven atoms, and establish strong structural evidence for the conjecture. The upper-bound techniques (a fixed-basis conic compression lemma, a determinant-facet rigidity theorem for maximal certificates, and sharp sparsification lemmas for zero-mean weighted point systems) are of independent interest. As a byproduct we correct an erroneous claim circulating in a recent unrefereed preprint, exhibiting an explicit dataset with m=2 on which no weighted selection of 2d points recovers the optimal loss. All results are new only for m\ge 2 ; the scalar case m=1 is due to Hanneke et al.
[LG-56] Strong Drafts Need Compact Memories: Long-Context Speculative Decoding with Compressed KV Cache EMNLP2026
链接: https://arxiv.org/abs/2608.30252
作者: Tong Yuan,Chengxi Liao,Zeyi Wen
类目: Machine Learning (cs.LG)
*备注: EMNLP 2026 Findings
Abstract:Long-context LLM applications such as document summarization and multi-turn agents require generation from prefixes spanning tens of thousands of tokens, making decoding latency a major bottleneck. Speculative decoding (SD) reduces latency without changing model outputs, but its speedup depends on both accepted draft tokens and draft-step latency: Lightweight drafts are fast but lack the capacity to capture long-range dependencies, whereas strong independent drafts recover acceptance but incur growing KV-access cost at long prefixes. We introduce memory-augmented drafting for long-context SD, equipping a strong independent draft with compressed draft-side KV memory: A lightweight adaptor constructs and incrementally updates this memory to retain distant information and exact recent context. The target verifier retains its full KV cache and applies the standard accept/reject rule, preserving SD’s lossless guarantee. Experiments on Llama~3.1-8B and 70B targets at prefix lengths up to 32K show that our method reduces draft-side memory by over 70%. It achieves speedups of up to 2.08x and 3.33x , respectively, over autoregressive decoding.
[LG-57] Diffusion-Based Refinement for Kilometer-Scale Probabilistic Precipitation Nowcasting
链接: https://arxiv.org/abs/2608.30205
作者: Dohyun Park,Changhoon Song,Tengyuan Chang,Yoo-Geun Ham,Youngjoon Hong
类目: Machine Learning (cs.LG); Atmospheric and Oceanic Physics (physics.ao-ph)
*备注: Submitted to npj Climate and Atmospheric Science. Includes Supplementary Information
Abstract:Localized extreme precipitation is a major trigger of urban flash floods and landslides, yet producing nowcasts that combine fine spatial detail with probabilistic uncertainty remains challenging. Here we introduce exPreCast-ENS, a conditional residual diffusion framework that transforms the deterministic 4 km radar nowcaster exPreCast into a 1 km probabilistic ensemble while correcting systematic forecast errors. Conditioning on both the forecast and preceding radar observations lets the ensemble-mean correct the baseline rather than perturb it, while members represent unresolved fine-scale variability. Over the Korean Peninsula, skill improves with ensemble size. In two high-impact events in 2023, a 30-member ensemble recovers 38-47% of heavy-rain pixels missed by exPreCast while retaining approximately 95% of its correct detections and alarming on under 1% of the pixels it correctly left clear. The method generates a 1-h forecast in 3.4 s on a single GPU and yields consistent improvements on the French regional MeteoNet radar dataset.
[LG-58] Certified Safety Radii in Forecast-Error Space for Wasserstein Distributionally Robust Small Signal Stability-Constrained AC Optimal Power Flow via Lifted Spectrahedral Containment
链接: https://arxiv.org/abs/2608.30201
作者: Ziqi Zhang,Xi Chen
类目: Machine Learning (cs.LG)
*备注:
Abstract:Directly robustifying small-signal stability in AC optimal power flow is challenging since the stability boundary in the original uncertainty space is implicit, highly nonconvex, and changes with the operating decision. This paper exploits an alternative geometry. For a fixed model-specific stability certificate admitting suitable physical lifts, the small-signal stability requirement becomes an affine positive semidefinite constraint in the lifted variables, thereby defining a convex certified safe region. Instead of approximating the nonlinear instability boundary itself, we optimize a sample-wise safe radius in the original uncertainty space and certify, in the lifted space, that the entire power-flow image of the corresponding uncertainty ball is contained in the convex stability region. To this end, a componentwise Perron certificate guarantees existence, uniqueness, and Jacobian regularity of the target AC power-flow branch throughout each ball. An adjoint elimination then provides an exact affine-quadratic representation of the stability-relevant quantities, while rigorous matrix remainder bounds convert their nonlinear variation into finite robust PSD constraints. The resulting radii are certified lower bounds on the distances from empirical samples to failure and can therefore be coupled directly to the distance-based reformulation of a Wasserstein distributionally robust chance constraint, without directly approximating the instability boundary. Numerical studies demonstrate the effectiveness of the proposed framework.
[LG-59] he PUR-1 Cyber-Physical Digital Twin
链接: https://arxiv.org/abs/2608.30186
作者: Vasileios Theos,Jonah Lau,Konstantinos Gkouliaras,Zachery Dahm,Konstantinos Vasili,Noah Fillgrove,William Richards,True Miller,Brian Jowers,Stylianos Chatzidakis
类目: Computational Engineering, Finance, and Science (cs.CE); Machine Learning (cs.LG)
*备注:
Abstract:Digital twin technologies have the potential to improve operational flexibility and responsiveness capabilities of nuclear systems. To provide decision support, cyber event characterization, state estimation, predictive control, and real-time dynamic processing of operational data, however, an efficient digital twin needs to integrate multiple models (data-driven as well as physics-based) with explainability while at the same time maintain two-way synchronization with the physical facility at a time constant less than its operational cycle. In this work, we present the Purdue University Reactor One Digital Twin (PUR-1 DT), a cyber-physical digital twin with a complete high-fidelity physics-based and AI-driven virtual model stack (neutronics, thermal-hydraulics, point kinetics) which provides closed-loop explainable diagnostics, forecasting, predictive control, and action recommendation back to the reactor via two-way communications and a cyber-physical testbed. We demonstrate real-time synchronized state estimation and short-term forecasting over a full reactor operational cycle and conduct a series of benchmarking experiments to validate accuracy and latency. Our results show good agreement with experimental results and lay the groundwork for further development and experimental demonstration of DT-enabled functionalities in real-world facilities.
[LG-60] Benchmarking Peptide-Protein Affinity Prediction Across Peptide and Target Shifts
链接: https://arxiv.org/abs/2608.30175
作者: Jiaxin Tian,Darren An,Jun Li
类目: Machine Learning (cs.LG); Quantitative Methods (q-bio.QM)
*备注: 11 pages, 6 figures, 2 tables
Abstract:Peptide-protein affinity models are often evaluated with a single data split, obscuring whether they interpolate among measurements for observed targets or generalize across peptide or target shifts. We integrated three sources of quantitative peptide-protein binding data to obtain 11,349 deduplicated pairs and benchmarked ten peptide representations, ESM-2 protein embeddings, and six regressors under peptide-similarity, within-target, and leave-target-out partitions. Across 60 matched representation-regressor configurations, mean test Spearman correlations were 0.462, 0.669, and 0.530, respectively. The top configuration shifted from ECFP-16 count fingerprints with random forest in the first two settings to HELM-BERT with Extra Trees when exact target sequences were excluded. Representation-rank correlations ranged from -0.042 to 0.624 across partitions, whereas regressor-rank correlations ranged from 0.771 to 0.943. Learning curves showed that representation differences were largest with limited supervision and narrowed as training data increased. PeptideCLM-2 adaptation and simple element-wise interaction features provided no consistent gain over a frozen encoder and direct concatenation under the tested protocols. These conclusions are specific to a dataset that pools transformed Kd, Ki, and IC50 measurements and to target exclusion at the exact-sequence level. Peptide-protein affinity benchmarks should therefore align data partitions with the intended use and jointly assess the effects of data scale, molecular representation, and downstream learner.
[LG-61] Reinforcement Learning for Symbolic Equation Solving
链接: https://arxiv.org/abs/2608.30162
作者: Kevin P O Keeffe
类目: Machine Learning (cs.LG)
*备注:
Abstract:We present a reinforcement-learning agent that solves symbolic equations step by step, covering both nonlinear closed equations (radicals, exponentials, trigonometric) and a controlled class of restricted-open families requiring a change of variables (CoV) such as completing the square. We cast algebra as an MDP with a dynamic action space and a tree-structured policy (TreeMLP). The main policy learns from reward alone with no supervised solution traces; the CoV substitution comes from a supervised generator interchangeable with a CAS call. On closed equations the agent matches the prior best on CommonCore (0.93 greedy vs. ConPoLe’s 0.925) under a single policy. On four hand-designed restricted-open families (quadratic, cubic, quartic, exponential) it reaches 0.79 beam / 0.67 greedy, exceeding the strongest non-learned search (A-star, 0.64). Learned CoV timing has content only on the exponential family, the one requiring a nested CoV, where a natural rule solves none of the held-out equations while the policy solves 75% from reward alone. At 10x scale a sharp seed-level bimodality emerges; a UCB learning-progress curriculum shows a non-significant positive trend toward mitigating it. We do not claim general open-equation solving: every open-equation result is confined to these four controlled families.
[LG-62] Converse and Collision-Based Achievability for Node Localization with Hybrid Distance-Spectral Graph Positional Encodings
链接: https://arxiv.org/abs/2608.30152
作者: Zimo Yan,Yifan Li,Hao Li,Zheng Xie,Chang Liu,Zheming Tu,Yuan Wang
类目: Machine Learning (cs.LG)
*备注:
Abstract:Graph positional encodings are widely used in graph neural networks and graph Transformers, yet it remains unclear when the code itself can identify nodes. We study a hybrid distance-spectral encoding that combines anchor-distance profiles with quantized low-frequency Laplacian-energy coordinates. Treating the encoding as an observation map yields a simplex-refined converse, an exact collision factorization (\kappa_H=\kappa_D\kappa_S|D), and the collision information (I_H=-\log\kappa_D-\log\kappa_S|D). On random regular graphs, the criterion is made explicit through a bounded-correlation Gaussian-wave surrogate; for actual Laplacian-energy coordinates, we give the distance-conditioned spectral collision condition sufficient for conditional actual-coordinate achievability. Experiments show that (I_H/\log n) calibrates localization success, and PE-only structural task probes on Universal Dependencies trees show that hybrid encodings better recover syntactic-tree geometry than distance-only or spectral-only baselines.
[LG-63] Balancing Privacy Utility and Safety in LLM Alignment through Preference Optimization
链接: https://arxiv.org/abs/2608.30141
作者: Dishu Yang,Jingjing Liu,Jize Li
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注: 6 pages, accepted for presentation at PRAI 2026
Abstract:Preference optimization is widely used to align large language models with human preferences, but preference-data composition may also influence privacy-relevant memorization. We examine whether adding synthetic privacy-preference pairs to Direct Preference Optimization (DPO) is associated with lower canary-based memorization signals without modifying the objective or introducing a formal privacy mechanism. We propose Privacy-Pressure Preference Mixing (P3M), a data-composition protocol that varies the amount of privacy-preference data while keeping helpfulness and harmlessness preference data fixed. We evaluate a non-privacy Baseline and privacy-mixing ratios of 0.5, 1.0, and 2.0 using Gemma 3 270M-IT across five random seeds and validate the same four conditions using 4-bit-quantized Gemma 2 2B-IT across three seeds. Overall, under the tested conditions, privacy-preference mixing is associated with lower mean canary suffix log-likelihood proxy values across both model settings and lower aggregate membership-inference attack performance relative to the Baseline in the mixed-source 2B evaluation. Specifically, across the privacy-aware 2B configurations, the mean area under the receiver operating characteristic curve (AUROC) ranges from 0.596 to 0.629, and the mean area under the precision-recall curve (AUPRC) ranges from 0.541 to 0.575, compared with 0.804 and 0.790, respectively, for the Baseline. However, the reduction in membership distinguishability does not hold uniformly across data sources. Moreover, the relationship between the privacy ratio and harmlessness preference accuracy varies by model setting, whereas helpfulness preference accuracy remains broadly stable. These findings suggest that P3M should be viewed as a lightweight empirical protocol for examining privacy-utility-safety trade-offs rather than as a formal privacy guarantee or a defense against extraction attacks.
[LG-64] Suprag lacial Lake Fate Is Knowable Long Before the Season Ends
链接: https://arxiv.org/abs/2608.30113
作者: Emam Hossain,Md Osman Gani
类目: Machine Learning (cs.LG)
*备注:
Abstract:A supraglacial lake on the Greenland Ice Sheet ends its melt season in one of four ways: it drains rapidly through a hydrofracture, drains slowly across the surface, refreezes in place, or is buried by late-season snowfall. Which one occurs decides whether the meltwater reaches the ice bed. Satellite classifiers recover the outcome accurately but only after the season closes, and how much of a season each outcome actually requires has never been measured. We measure it directly: holding the representation and the classifier fixed, we truncate the input at 14 cutoffs from 1 May to 31 December, retrain at each, and record the earliest cutoff at which each outcome’s per-class F_1 reaches a fixed target. The outcomes resolve in a consistent order, two of them months early: rapid drainage by 15 July and slow drainage by 1 August, 92 and 75 days ahead of the earliest date a full-season pipeline can be computed at all, with buried and refreeze following at 44 and 30 days. Five further learners, from a majority-class floor and 54 summary statistics to a trigger-based early classifier, leave the ordering intact: every learner that produces a per-class trajectory reproduces it despite end-of-season accuracies differing by up to 18 percentage points, and it survives leave-one-basin-out evaluation, though not the substitution of machine labels for expert ones in an unseen season. Every feature we compute at day t reads only days up to t , at a cost of at most 1.3 percentage points. A monitoring system should therefore not have one release date: rapid drainage can be flagged on 15 July, three months before a full-season pipeline can be computed at all.
[LG-65] A Simple Transformer Pipeline for Full-Key Side-Channel Attacks on Uncropped Datasets
链接: https://arxiv.org/abs/2608.30105
作者: Jimmy Gammell,Kaushik Roy
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注: Accepted to the OPTIMIST Workshop '26 at CHES 2026. 6 pages, 1 figure. Code can be found at this https URL
Abstract:Deep learning-based side-channel analysis has historically focused on single-byte targets and manually cropped traces, which risks discarding exploitable leakage. While recent work has proposed specialized architectures and resampling techniques to address this gap, the literature lacks a simple transformer baseline for simultaneous full-key attacks on uncropped traces. We present an open-source transformer implementation for uncropped full-key attacks which uses the standard transformer encoder backbone, adapting only the input and output layers to the side-channel setting. We release our implementation, training recipes, and pretrained weights for uncropped ASCADv1f, ASCADv1r, and CHES-CTF-2018 which achieve performance competitive with previously-reported results, while using less than 10GB of VRAM and requiring at most 3.34 hours of training on a single NVIDIA A6000.
[LG-66] SMOTE-VAR: An Uncertainty-Aware Oversampling Method for Predicting Depression Remission in University Students
链接: https://arxiv.org/abs/2608.30102
作者: Dang Nguyen,Arun Kumar A V,Taylor A. Braund,Wu Yi Zheng,Debopriyo Bal,Leonard Hoon,Jill Newby,Helen Christensen,Svetha Venkatesh,Alexis Whitton,Sunil Gupta
类目: Machine Learning (cs.LG)
*备注:
Abstract:University students experience disproportionately high rates of common mental health conditions, such as depression, which can impair learning, social functioning, and overall well-being. Although lifestyle interventions such as mindfulness and physical activity can reduce the symptoms, many do not achieve symptomatic remission. Developing new approaches to identify students with poor outcomes could enable earlier and more targeted intervention. Machine learning (ML) methods have increasingly been used to predict remission in depressive patients. However, these ML models often suffer from class imbalance, where there may be an unequal proportion of people in the remitted group relative to the non-remitted group. This imbalance can reduce model accuracy and bias predictions. To address this, studies commonly employ the popular oversampling strategy SMOTE. However, SMOTE has a notable limitation: it may generate invalid synthetic minority samples. In a clinical context, these false positives can lead to incorrect risk stratification, potentially delaying necessary escalated care for patients unlikely to remit. In this paper, we introduce a novel and effective oversampling method that addresses this shortcoming. Our approach leverages the variance function of a Gaussian process to estimate the uncertainty of generated minority samples to reduce false positives. We validate our method on a depression dataset collected from university students and demonstrate that it is better than existing oversampling approaches in predicting remission (i.e., treatment outcome). By improving the reliable identification of non-responders, our method provides a robust computational tool to help clinicians rapidly pivot to adjunctive therapies, thereby personalizing and optimizing mental health care pathways.
[LG-67] A Lightweight Phenology-Aware YOLOv5 Framework for Tomato Growth Stage Detection in Resource-Constrained Bhutanese Greenhouse Environments
链接: https://arxiv.org/abs/2608.30088
作者: Sherab Gocha,Sou Nobukawa
类目: Machine Learning (cs.LG)
*备注:
Abstract:Accurate detection of tomato growth stages is essential for stage-specific greenhouse management and precision agriculture. In Bhutan, greenhouse cultivation is affected by altitude variability, large diurnal temperature fluctuations, diffuse illumination, limited automation, and a scarcity of locally annotated datasets, limiting the applicability of conventional deep learning models. This work proposes Pheno-Lite + Efficient Channel Attention (ECA), a lightweight, phenology-aware object detection architecture derived from Ultralytics YOLOv5 for tomato growth stage recognition. A balanced dataset of 2,464 annotated images was constructed from locally collected greenhouse images in Bhutan and publicly available tomato images, with augmentation designed to simulate local greenhouse conditions. The dataset includes vegetative (820), flowering (824), fruiting (820), and background (26) samples. The proposed architecture introduces two customized backbone modules: C3 PhenoLite, which enhances spatial and texture feature extraction using depthwise residual refinement, and C3 ECA, which strengthens inter-channel feature interactions through efficient channel attention. The proposed model achieves 90.6% precision, 88.8% recall, and 92.6% mAP@50, with 4.0 million parameters and 10.9 GFLOPs at 640 x 640 resolution. These results demonstrate its potential for real-time and climate-resilient greenhouse deployment in Bhutan.
[LG-68] Selection Representation and Execution in Sparse Fourier Neural Operators
链接: https://arxiv.org/abs/2608.30070
作者: Abdul Qadir Ibrahim,Martin Burger
类目: Machine Learning (cs.LG); Numerical Analysis (math.NA)
*备注: 21 pages, 11 figures
Abstract:Sparse representations are often expected to make models smaller and also reduce inference cost. For Fourier Neural Operators (FNOs), these objectives are not equivalent or do not always align: removing parts of the learned operator can leave the underlying transforms and dense computations unchanged, while changing the grid on which the model is evaluated can introduce overhead of its own. We therefore distinguish sparsity in the representation, in the stored parameters, in the theoretical operation count, and in measured runtime, and present an empirical study of several routes toward sparse FNOs that tests each transition between them separately. Coarsening the execution grid reduces the theoretical cost without reducing measured latency, and adding a correction term recovers accuracy at the cost of making the model slower. Even an 83% parameter reduction remains slower than the dense baseline under ordinary execution. These results motivate a stricter definition of useful sparsity: the deployed operator must preserve solution accuracy and map its reduced support to a genuinely cheaper execution path.
[LG-69] Multiclass Linear Perceptrons with Multiplicative Margins
链接: https://arxiv.org/abs/2608.30028
作者: Dmitri Rachkovskij,Evgeny Osipov,Olexander Volkov,Daswin De Silva,Denis Kleyko
类目: Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE)
*备注: 38 figures
Abstract:This paper introduces a family of multiclass linear Perceptron classifiers with a multiplicative margin mechanism (MMPerc), as an alternative to standard margin-free and additive margin Perceptrons. The multiplicative formulation enforces classification confidence by requiring the true class score to exceed that of competing classes by a specified fraction of itself, rather than by a fixed additive threshold. This avoids dependence on score magnitudes arising from varied norms of data and class weight vectors. We propose several architectural and algorithmic variants of MMPerc, derive associated loss functions and mistake bounds for both linearly separable and non-separable data, and analyze key design considerations, including bias, margin threshold selection, and training modes. Extensive experiments on synthetic and real datasets show that MMPerc classifiers typically outperform the standard Perceptron, as well as classic baselines such as Support Vector Machines and Ridge classifiers. Owing to their simplicity, minimalistic design, and computational efficiency, MMPerc classifiers are promising candidates for conventional machine learning tasks, linear evaluation of Deep Neural Networks, integration with Hyperdimensional Computing / Vector Symbolic Architecture representations, and deployment in resource-constrained applications.
[LG-70] he Intervention Gap in Latent World Models
链接: https://arxiv.org/abs/2608.29998
作者: Donna Vakalis
类目: Machine Learning (cs.LG)
*备注: 21 pages, 10 figures
Abstract:Planning-time intervention fidelity is a distinct, measurable property of a learned world model: whether the model’s own open-loop transitions move task variables the way matched environment interventions do. In the settings we test, it is neither revealed by reward fit nor ensured by task-anchored training. Across released TD-MPC2 checkpoint sizes, episode return falls as an operator-error diagnostic on task observables grows, while reward-prediction error stays small and nearly flat, and a self-supervised world model trained without task signal preserves the same operator substantially better than a task-anchored model on the shared task. A capture-gated matched-intervention audit then localizes what fails. On Cheetah, three LeWorldModel checkpoints capture the current task query and support decodable real intervention effects; however, their imagined five-step effects are worse than predicting no effect and worse than an environment-endpoint oracle. The failure is task-direction rotation with excess gain, not feature collapse. This severe pattern is conditional: five PreJEPA seeds retain an oracle-relative deficit without it, Finger Spin experiments extend the deficit beyond locomotion with heterogeneous severity across seeds, and shared-bank effect geometry is both candidate- and support-dependent. We also test practice-side questions. In DreamerV3 the posterior distribution, not its sample, carries the current query; ensemble disagreement ranks error only near training support; and a frozen support-aware score degrades held-out error ranking in both tested transfer directions while native disagreement remains informative in both. We conclude that intervention fidelity must be audited directly, capture-first, on the model’s native interface.
[LG-71] Robust Broad Learning System with Wave Loss for Classification under Data Uncertainty
链接: https://arxiv.org/abs/2608.29983
作者: Mushir Akhtar,A. Varshney,A. Quadir,A. Rahaman,M. Tanveer,Mohd. Arshad
类目: Machine Learning (cs.LG)
*备注: Accepted at WCCI 2026
Abstract:Broad Learning System (BLS) offers an efficient alternative to deep architectures by enabling fast learning through randomized feature mapping and closed-form solutions. However, its reliance on squared error loss makes it highly sensitive to noise, outliers, and corrupted labels, limiting its reliability in real-world scenarios. To address this limitation, we propose Wave-BLS, a robust broad learning framework that integrates the wave loss function, which is asymmetric, bounded, and smooth, enabling controlled penalization of large errors. The proposed formulation replaces the standard least-squares objective with a wave-loss-based optimization problem, solved efficiently using a Nesterov accelerated gradient (NAG)-based scheme without requiring matrix inversion, thereby improving scalability. Extensive experiments on 30 UCI benchmark datasets demonstrate that Wave-BLS consistently outperforms classical BLS and several robust variants. Statistical validation using Friedman and Nemenyi post-hoc tests confirms the significance of the observed improvements. Furthermore, robustness evaluations under controlled noise and outlier injection reveal that Wave-BLS exhibits substantially slower performance degradation compared to BLS, even in challenging contamination settings. These results establish Wave-BLS as a stable and robust alternative to existing broad learning models for learning under data uncertainty.
[LG-72] Diffusion-Based Inverse Design of Dielectric Resonator Metasurfaces for Shaping Smart Electromagnetic Environments
链接: https://arxiv.org/abs/2608.29907
作者: M. Tsukerman,K. Grotov,D. Vovchuk,P. Ginzburg
类目: Machine Learning (cs.LG); Applied Physics (physics.app-ph); Computational Physics (physics.comp-ph)
*备注:
Abstract:Future wireless systems are expected to transform the surrounding space from a passive propagation medium into a smart electromagnetic environment, where engineered surfaces control wave propagation, support wireless sensing, and create programmable electromagnetic fingerprints. A key challenge in realizing this vision is the inverse design of metasurfaces for tailored electromagnetic propagation. While forward analysis evaluates the response of a known geometry, the inverse task starts from a prescribed scattering signature and seeks a physically realizable structure that produces it. This inverse task is inherently nonlinear and often high-dimensional, while candidate solutions may be non-unique and provide no direct indication of practical realizability. Here, we introduce a conditional diffusion framework for inverse design of dielectric resonator metasurfaces from target angular scattering patterns. Trained on T-matrix simulated geometry-response pairs, the model learns a conditional distribution of geometries instead of a deterministic mapping, enabling multiple candidate designs for the ill-posed inverse problem. The best generated metasurface achieves a mean percentage error of 1.39%, outperforming CMA-ES optimization (4.1% after 10 h) while requiring only about one minute for after-training inference. The model also produces lower error distributions than deterministic neural baselines for out-of-distribution spectra, highlighting the potential of diffusion models for efficient metasurface design.
[LG-73] Joint Spatiotemporal Spectral Neural Operators for Learning PDEs on Irregular Domains
链接: https://arxiv.org/abs/2608.29892
作者: Abdolmehdi Behroozi,Chaopeng Shen
类目: Machine Learning (cs.LG)
*备注:
Abstract:Learning solution operators for partial differential equations (PDEs) on irregular and geometry-dependent domains remains a central challenge in scientific machine learning. While spectral methods provide strong inductive biases for modeling global interactions, they are typically limited to regular domains, and existing neural approaches often require domain warping, interpolation, or costly geometric embeddings. We introduce the \textbfGraph Spectral Neural Operator (GSNO), a neural operator that combines spatial graph spectral decompositions with temporal Fourier transforms through a unified space–time spectral kernel. This formulation enables globally coherent operator learning on non-Cartesian discretizations without domain warping or autoregressive rollouts. By replacing learned geometric embeddings with a graph Laplacian spectral basis, GSNO provides geometry-aware spectral learning with low parameter complexity. Across steady and unsteady PDE benchmarks on irregular and geometry-dependent domains, GSNO achieves strong accuracy with reduced runtime and parameter counts, while demonstrating robust zero-shot generalization across mesh resolutions and geometry families.
[LG-74] Sensitivity-Constrained Neural Operators for Data-Efficient Forward and Inverse Modeling of Partial Differential Equation Systems
链接: https://arxiv.org/abs/2608.29888
作者: Abdolmehdi Behroozi,Chaopeng Shen,Daniel Kifer,Kathryn Lawson
类目: Machine Learning (cs.LG)
*备注:
Abstract:Neural operators provide fast surrogates for partial differential equation (PDE) solvers, but their reliability can degrade for high-dimensional spatial inputs and inverse or repeated inference. State-only training constrains solution values but not the learned input–output response. We study sensitivity-constrained neural operators (SC-NOs), which augment standard training with sampled solver-derived Jacobian supervision. Selected sensitivities from differentiable solvers or discrete adjoints are matched during training, allowing response information to be amortized across minibatches without imposing the full Jacobian at every update. We evaluate SC-NO on advection–diffusion and RANS–Spalart–Allmaras benchmarks, input-dimensionality scaling tests, long-horizon autoregressive rollout, and a shallow-water Tohoku tsunami source-inversion case. Sensitivity supervision improves forward prediction and yields larger gains in gradient-based inverse reconstruction of distributed fields. Scaling experiments show an improved accuracy–cost tradeoff for high-dimensional gridded inputs, while ablations indicate that state values and Jacobian information provide complementary supervision. In the tsunami case, SC-FNO reconstructs gridded seafloor deformation from sparse early gauge observations and forecasts subsequent wave propagation in a near-real-time proof-of-concept workflow. These results support sampled sensitivity supervision as a practical way to improve neural PDE surrogates when forward accuracy, inverse stability, robustness, and computational cost must be considered together.
[LG-75] Structural Hierarchy and Geometry in Molecular Representation Learning
链接: https://arxiv.org/abs/2608.29886
作者: David Sulu,Lorenzo Di Fruscia,Jana M. Weber
类目: Machine Learning (cs.LG); Biomolecules (q-bio.BM)
*备注:
Abstract:Molecular self-supervised learning uses chemical structures to guide which molecular embeddings should be similar. We study whether explicitly encoding a molecule’s Bemis-Murcko scaffold and using it to supervise the molecular embedding changes what the model learns. We further test whether this effect depends on the embedding geometry by comparing Euclidean and Lorentz contrastive objectives. Across two augmentation strengths, scaffold-supervised models consistently organize molecules according to both identical and structurally related scaffolds. The resulting embeddings also improve molecular property prediction on several tasks, while the exact gains depend on the predicted property. The effect of scaffold supervision on molecular organization is stronger under Lorentz objectives, but neither geometry provides a consistent overall advantage. These results show that explicitly teaching the relation between a molecule and its structural core can reliably shape the organization of molecular embedding space, while the extent of usefulness of this organization remains task dependent.
[LG-76] owards an Expressivity-Normalized Energy-Demand Comparison of ANNs and SNNs
链接: https://arxiv.org/abs/2608.29869
作者: Miriam Kranzlmüller,Pascal Esser,Gitta Kutyniok
类目: Machine Learning (cs.LG)
*备注:
Abstract:Spiking neural networks (SNNs) are often regarded as energy-efficient alternatives to artificial neural networks (ANNs), yet their advantage depends critically on both network architecture and data properties. We develop an analytical framework to compare fully-connected ReLU ANNs and integrate-and-fire SNNs for time-series data with respect to their theoretical energy efficiency at matched expressive capacity. By relating an inference-energy model to theoretical bounds on representational expressivity, we derive an expressivity-normalized efficiency ratio and explicit thresholds in network width, spike sparsity, and ANN depth scaling. Our analysis characterizes the regimes in which event-driven computation offsets the temporal overhead of SNNs, providing capacity-aware principles for designing energy-efficient temporal networks. It shows that ANNs exceed SNNs in expressivity-normalized efficiency only in specific regimes.
[LG-77] Partially Linear Autoencoders for Manifold Learning and Dimensionality Reduction
链接: https://arxiv.org/abs/2608.29867
作者: Louen Pottier,Louis Lesueur,Anders Thorin
类目: Machine Learning (cs.LG)
*备注: 19 pages, 10 figures, 3 tables. Preprint also available on HAL: this https URL
Abstract:Autoencoders are widely used for nonlinear dimensionality reduction and manifold learning. While most common implementations rely on both nonlinear encoders and decoders, we investigate the specific role of the encoder and the extent to which it can be constrained to be linear without reducing accuracy. We conduct a comparative study on four autoencoder architectures: standard fully nonlinear autoencoders (AE), linear-encoder autoencoders (Lenc-AE), linear-decoder autoencoders (Ldec-AE), and fully linear autoencoders (LAE), evaluated on synthetic manifolds, computational mechanics data sets, and real-world image data sets including MNIST. We demonstrate that imposing a linear encoder preserves most of the representational capacity of the autoencoder, provided the decoder remains nonlinear. In particular, Lenc-AE consistently outperforms both Ldec-AE and LAE, and achieves reconstruction quality comparable to fully nonlinear AE, while offering advantages in terms of parsimony and interpretability of the latent representation. These results suggest that the nonlinear decoder is the critical component for manifold learning, rather than the encoder. A geometric interpretation of this finding is developed, which identifies the precise conditions under which a linear encoder is sufficient, and the specific manifold configurations that expose its limitations.
[LG-78] Uncertainty-Driven Replay Memory for Reinforcement Learning
链接: https://arxiv.org/abs/2608.29860
作者: Sheeraja Rajakrishnan,Alexander G. Ororbia,Travis Desell,Daniel E. Krutz
类目: Machine Learning (cs.LG)
*备注:
Abstract:Uncertainty estimation provides promising capabilities for reinforcement learning (RL) agents. Notably, estimating uncertainty can reduce the training time and enable agents to obtain greater rewards over time by exploiting information related to whether an action would facilitate exploration of portions of an environment that are well-known versus those that are relatively unknown. In this work, we propose a novel formulation of the experience replay buffer commonly used in RL that we call uncertainty-driven replay memory (UDRM), which entails an update scheme for internally stored memories based on uncertainty estimates obtained by an RL model during training. In contrast to existing forms of RL, which typically use temporal difference error or the distribution of transitions to update the replay memory buffer and train RL controllers, our scheme biases the memory buffer to store more uncertain transitions that will improve an RL agent’s generalization throughout training. Experimental results demonstrate that our proposed uncertainty-aware replay buffer enables an RL agent to obtain higher rewards during training compared to other existing uncertainty-aware RL frameworks.
[LG-79] Designing for the Next Click: Bandits for Real-Time Page Layout RECSYS2026
链接: https://arxiv.org/abs/2608.29850
作者: Bhavtosh Rath,Harshith Narasimhamurthy,Bob Eisinger,Cole Stiegler,Adnan Awow,Amit Pande
类目: Machine Learning (cs.LG)
*备注: Accepted to The Web Conference 2026 (short paper track), but later withdrawn due to internal prioritization. Subsequently accepted to the Online Adaptive Recommender Systems Workshop (held in conjunction with the 20th ACM Conference on Recommender Systems, RecSys 2026)
Abstract:E-commerce platforms increasingly personalize user experiences through machine learning, yet page layout decisions remain dominated by static rules and manual curation. We present a scalable bandit-based system that optimizes product page layouts in real time while preserving human control over design intent. A contextual bandit model dynamically selects the most effective layout for each session using user, item, and category-level features. The system leverages a LinUCB-based policy to balance exploration and exploitation as it learns from live user interactions. The architecture is designed for seamless integration into large-scale web serving stacks, supporting low-latency inference and continuous model updates. The system was first tested on entry product pages. In online A/B deployments on a major retail platform, our approach achieved positive lifts in session-level performance metrics over a strong heuristic baseline. Our results demonstrate that contextual bandits can effectively optimize visual and structural aspects of product discovery for user engagement, providing a scalable path toward learning-to-design the web.
[LG-80] Structure Aware Neural Architecture Search for Mixture of Experts
链接: https://arxiv.org/abs/2608.29817
作者: Petr Babkin,Oleg Bakhteev
类目: Machine Learning (cs.LG)
*备注: 22 pages, 3 figures
Abstract:Neural Architecture Search (NAS) has so far rarely been applied to Mixture-of-Experts (MoE) models, and existing MoE designs leave the alignment between experts and the structure of the data to emerge on its own. We propose an architecture search framework that makes this alignment an explicit search variable: the assignment of data clusters to experts is optimised jointly with the per-expert architectures. We cast the joint problem as a cluster-aware likelihood maximisation, show that it coincides with the incomplete-data maximum likelihood of a latent-variable mixture, and solve it by a generalised Expectation-Maximisation procedure whose otherwise intractable expert-quality term is supplied by an adaptively refined surrogate. We prove that the iterates converge whenever the surrogate errors are summable, and that at every limit point no candidate the search produces improves the true objective. On a heterogeneous image-classification mixture the method recovers the underlying domain partition on 95% of clusters without ever observing domain labels, and on that benchmark and a four-domain time-series forecasting one alike it outperforms the MoE and NAS baselines that likewise use no label information.
[LG-81] PruneShift: A Framework for Evaluating Decision Reliability in Structured Pruning
链接: https://arxiv.org/abs/2608.29765
作者: Hao Ye,Gaopeng Zhang
类目: Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE)
*备注:
Abstract:Structured pruning uses surrogate objectives because direct task evaluation over every feasible mask is too expensive. Most evaluations report average surrogate error or rank correlation on broadly sampled masks. These summaries do not directly test the mask chosen by the surrogate. We introduce PruneShift, an evaluation framework that separates broad predictive fidelity, fidelity near selector outputs, and the quality of the selected pruning decision. We first prove that Spearman and Kendall agreement can approach one while normalized selection regret remains maximal. We then derive sufficient conditions based on uniform error, selector suboptimality, decision margin, density ratio, and comparison mass. The analysis also yields a finite pool certificate with an explicit excess cost bound. Four studies test different links in this argument. External TextbookQA confirmation is heterogeneous: 7 of 20 simultaneous intervals favor the surrogate-selected mask, 6 favor its fixed comparator, and 7 cross zero. On a fixed Natural Questions pool, strict improvement holds in one of four settings. A controlled QQP experiment supports the proposed coverage mechanism in all 16 prespecified endpoints, although the sufficient bounds are conservative. Finally, a restricted OSSCAR reconstruction study on OPT-125M shows better local than broad fidelity in 68 of 75 primary endpoints. Independent fixed-mask confirmation is inconclusive in 24 of 25 endpoints and favors the comparator in one. These results show why predictive fit, decision reliability, and pruning method quality require separate evidence.
[LG-82] ECA-BLS: An Efficient Complex-Augmented Broad Learning System
链接: https://arxiv.org/abs/2608.29763
作者: A. Rahaman,A. Quadir,M. Sajid,M. Akhtar,M. Tanveer
类目: Machine Learning (cs.LG)
*备注:
Abstract:Broad Learning System (BLS) is an efficient alternative to deep architectures due to its fast training, analytical learning, and strong generalization under limited data. However, existing BLS variants are confined to real-valued representations, restricting their ability to capture nonlinear interactions and second-order statistical dependencies inherent in real-world data. Notably, no prior BLS model fully exploits the complete second-order statistics that naturally emerge when data are embedded in the complex domain. To address this limitation, this paper introduces the first complex augmented Broad Learning System (CA-BLS), which transforms real-valued inputs into phase-encoded complex representations and adopts widely linear modeling to jointly leverage covariance and pseudo-covariance information via complex conjugate augmentation. This enables effective modeling of latent nonlinearities, coherence structures, and second-order dependencies inaccessible to conventional BLS formulations. To mitigate the additional computational cost of complex augmentation, an Efficient Complex Augmented BLS (ECA-BLS) is further developed, reformulating CA-BLS entirely in the real domain while preserving its exact decision function, achieving up to 75% fewer multiplications and over 60% fewer additions. A rigorous theoretical analysis proves the mathematical equivalence between CA-BLS and ECA-BLS, ensuring zero theoretical loss. Extensive experiments on 26 benchmark datasets from the UCI and KEEL repositories demonstrate that ECA-BLS consistently outperforms classical BLS and recent state-of-the-art randomized neural networks in accuracy, average rank, and statistical significance, establishing augmented second-order modeling as a critical and previously missing dimension of BLS research.
[LG-83] GraM-Diff: A Unified Graph-Mamba Diffusion Framework for EEG-Based Alzheimers Disease Data Generation and Diagnosis
链接: https://arxiv.org/abs/2608.29755
作者: M. Tanveer,Ayush Singh Rana,Sanskriti Jain,Arnav Kumar,Aryaman Tiwari,A. Rahaman,A. Quadir,M. Sajid
类目: Machine Learning (cs.LG)
*备注:
Abstract:Electroencephalography (EEG) is a promising, non-invasive, and cost-effective modality for Alzheimer’s disease (AD) detection, but deep learning methods are limited by small and imbalanced clinical datasets. Generative augmentation offers a solution, yet existing approaches rely on inefficient class-specific models or fail to capture complex spatial and temporal brain dynamics. To address this, we propose GraM-Diff, a unified classifier-guided Graph-Mamba diffusion framework for EEG synthesis. It embeds Graph Convolutional Networks within a diffusion U-Net to model inter-electrode connectivity and Bidirectional Mamba state-space blocks for linear-complexity long-range temporal modeling. Latent-space classifier guidance lets a single model generate both healthy and pathological EEG within a shared representation, avoiding fragmented per-cohort pipelines. Across four EEG-based AD benchmarks, synthetic augmentation improves classification, yields superior Context-FID and correlation scores over strong generative baselines, and enhances robustness in data-scarce settings.
[LG-84] Last Step Matters: Early Uncertainty Cannot Predict Failure in Long-Horizon Agents EMNLP2026
链接: https://arxiv.org/abs/2608.29685
作者: Zongyue Li,Chengyue Yu,Lei Zang,Chenyi Zhuang,Linjian Mo,Leilei Gan
类目: Machine Learning (cs.LG)
*备注: Accepted to the Main Conference of the 2026 Conference on Empirical Methods in Natural Language Processing (EMNLP 2026)
Abstract:Early failure prediction is important for long-horizon agents, as it enables timely intervention and can reduce inference and tool-use costs. Uncertainty quantification, such as verbal confidence and perplexity, offers a promising approach to detecting agent failures; however, it has not been explored whether these signals retain their discriminative power during the intermediate stages of long-horizon execution. We evaluate mainstream uncertainty signals on deep-research tasks and find that verbal confidence reliably distinguishes failures at trajectory completion, achieving a mean AUROC of 0.85, whereas all evaluated signals offer limited predictive value earlier in execution, with none exceeding a mean AUROC of 0.60 at 50% trajectory progress. We identify an underlying mechanism explaining this gap: path switching, where agents frequently abandon their current search direction in-trajectory, breaking the link between early signal and final outcome. These findings challenge the assumption that intermediate uncertainty can reliably guide early intervention. They also motivate a practical recommendation for agent harnesses in deep-research settings: use final-step confidence to decide whether to restart, an approach that our experiments find more effective than in-trajectory intervention.
[LG-85] Creation begins with understanding: LLM s as strategy designers for privacy-preserving tabular data synthesis
链接: https://arxiv.org/abs/2608.29674
作者: Jinmeng Li,Quan Zhang,Hangting Ye,He Zhao,Firas Laakom,Dandan Guo,Jürgen Schmidhuber
类目: Machine Learning (cs.LG)
*备注:
Abstract:Sharing tabular data in high-stakes domains is constrained by privacy regulations. Synthetic data offer a promising alternative, but deep generative models are costly to train and difficult to audit, while LLM-based methods often serialize records as text, obscuring tabular structure and exposing sensitive data. We introduce Tabular Synthesis Strategy Designer (TabSSD), which uses an LLM to design synthesis procedures rather than directly generate records. TabSSD provides the LLM with tree-derived summaries of variable dependence rather than raw records, which produces Python programs for local execution and evaluation. Across twelve datasets, TabSSD strikes a favourable balance among statistical fidelity, predictive utility, and empirical privacy risk, achieving the best average rank across six metrics among ten methods. Moreover, it substantially reduces local computation and token consumption relative to the compared methods. By enabling human-guided refinement and eliminating user-side model tuning, TabSSD lowers the expertise and infrastructure barriers to transparent tabular data synthesis.
[LG-86] A Target-Centric Survey of Quantization-Aware Training EMNLP2026
链接: https://arxiv.org/abs/2608.29667
作者: Jiamin Song,Mengjie Zhao,Zijing Wang,Yongkang Liu,Qian Li,Shi Feng,Feiliang Ren,Daling Wang,Hinrich Schütze
类目: Machine Learning (cs.LG)
*备注: Accepted to EMNLP 2026 (Main Conference)
Abstract:The rapid development of LLMs incurs prohibitive memory footprints and intensive computational demands. Quantization-Aware Training (QAT) techniques have emerged as a promising solution to address these challenges by explicitly simulating quantization effects during model training, yielding low-bit models that achieve accuracy comparable to their full-precision counterparts. In this work, we provide a target-centric survey of QAT, aimed at clarifying both its theoretical foundations and its evolving implementation landscape. We systematically review existing QAT methods through a target-centric taxonomy and synthesize cross-target differences in error characteristics, numerical formats, and strategy transferability. We further summarize QAT evaluation paradigms and discuss challenges in optimization and deployment, outlining potential directions for future research.
[LG-87] Reward-guided Fine-Tuning of One-Step Generative Models via Wasserstein Gradient Flow
链接: https://arxiv.org/abs/2608.29647
作者: Hoseong Hwang,Woorim Han,Joungin Chun,Jinseong Park,Jaewoong Choi
类目: Machine Learning (cs.LG)
*备注: 14 pages, 9 figures
Abstract:To mitigate the time complexity of generative models, one-step generative models have recently emerged through direct mapping from noise to data in a single forward pass. However, the reward-guided fine-tuning method of one-step generative models remains largely unexplored. To address this, we consider one-step generators from an optimal transport view, investigating Wasserstein Gradient Flow (WGF) for modeling smooth and controlled distributional evolution in probability space. We then propose a novel reward-guided fine-tuning of a one-step generative model via WGF. We derive a practical training method that requires no reward gradients, thereby handling both non-differentiable and differentiable rewards. Moreover, our method provides smooth and stable reward-guided distributional updates while mitigating reward hacking and mode collapse. Experiments on 2D synthetic data, CIFAR-10, and ImageNet 256 \times 256 with diverse rewards, including JPEG (in)compressibility, class probability, Black-and-White and CLIP alignment, show that our method achieves better reward alignment compared to baselines.
[LG-88] Unsupervised Multi-Scale Gromov-Wasserstein Hypergraph Alignment ICDM2026
链接: https://arxiv.org/abs/2608.29635
作者: Lutz Oettershagen,Honglian Wang,Aristides Gionis
类目: Machine Learning (cs.LG); Social and Information Networks (cs.SI)
*备注: Accepted at ICDM 2026
Abstract:We study unsupervised hypergraph alignment, where the goal is to infer node correspondences between two hypergraphs using only structural information, without node features, labels, seed matches, or side information. Direct higher-order formulations can represent hyperedge interactions faithfully, but they can be computationally demanding and cumbersome for non-uniform hypergraphs. Graph-reduction approaches introduce a different challenge: clique expansions keep the alignment problem on the original node set but collapse all hyperedge evidence into one pairwise graph, whereas bipartite expansions preserve incidence structure but enlarge the problem from nodes to nodes plus hyperedges. We introduce FALCON (Filtration-based hypergrAph aLignment via Cross-scale Optimal traNsport), an unsupervised optimal-transport framework for hypergraph alignment. Instead of representing each hypergraph by a single collapsed clique graph, FALCON constructs a filtration-induced sequence of clique-based co-occurrence dissimilarity matrices and jointly aligns all levels through one shared multi-scale Gromov–Wasserstein (GW) objective. The shared transport plan enforces a globally consistent node correspondence across filtration levels while avoiding the auxiliary hyperedge nodes introduced by bipartite expansion. Experiments on perturbation benchmarks derived from real-world hypergraphs show that FALCON is robust to structural noise and in almost all cases outperforms strong graph- and hypergraph-alignment baselines.
[LG-89] On the Resilience of Text-to-Video Diffusion Models to Hardware Faults ICML2026
链接: https://arxiv.org/abs/2608.29598
作者: Zachary Coalson,A M Aahad,Stella Doehring,Zane Ma,Sanghyun Hong
类目: Machine Learning (cs.LG)
*备注: Accepted to ICML 2026 Workshop on From Frames to Stories (F2S)
Abstract:We present the first systematic study of the resilience of text-to-video (T2V) diffusion models under random hardware-level faults. While T2V models are widely used for automated video generation due to their ability to produce high-quality, temporally coherent, and realistic videos, their iterative denoising process and spatiotemporal dependencies introduce unique failure modes. We perform an extensive fault-injection study covering both computational and memory faults across three T2V models and a representative benchmark. Our results show that (1) a single fault can degrade overall performance by up to 3.7%, with semantic correctness more affected than perceptual quality; (2) memory faults are more damaging than computational faults, high-order exponent bits are particularly vulnerable, and the widely-used bfloat16 is more susceptible than alternative formats; and (3) 7-28% of faults cause visible artifacts, including semantic changes such as added objects, suggesting that single faults are sufficient to alter output semantics. Our findings reveal reliability risks in deployed T2V systems and motivate further research on improving fault resilience. Code: \hrefthis https URLthis https URL.
[LG-90] Predicting the Unpredictable: LLM -powered Long-term Chaotic Time Series Forecasting under Short-term Observations
链接: https://arxiv.org/abs/2608.29579
作者: Yuhang Yao,Bohan Jiang
类目: Machine Learning (cs.LG)
*备注:
Abstract:Chaotic time series forecasting is a challenging task due to its sensitivity to initial conditions and long-term unpredictability. Traditional methods typically rely on sufficient temporal trajectories to learn long-term dynamics, which limits their applicability when only short-term observations are available. While recent Large Language Models (LLMs) have shown great potential for time series forecasting, their temporal representations are not explicitly tailored to the phase-space structure and nonlinear evolution of chaotic systems. To address these issues, we propose PAC-LLM, a phase-space-aware adaptive fusion framework for long-term chaotic time series forecasting powered by LLMs. PAC-LLM leverages learned phase-space features and textual information to fully enable LLM’s time series forecasting capacity. In particular, we design an auxiliary feature module and a gated weighting mechanism for multivariate coupling information fusion and selection. Extensive experiments on representative chaotic systems demonstrate that our method outperforms existing fine-tuned and zero-shot baselines in both short-term and long-term predictions. Our ablation study further confirms the effectiveness of each key component in PAC-LLM.
[LG-91] Event-triggered Control and Online Learning for Networked Systems under Computational Delays
链接: https://arxiv.org/abs/2608.29576
作者: Xiaobing Dai,Armin Lederer,Zewen Yang,Sihua Zhang,Lu Wan,Yang Tang,Sandra Hirche
类目: Machine Learning (cs.LG); Robotics (cs.RO); Systems and Control (eess.SY)
*备注:
Abstract:Online learning-based control is a promising approach to control uncertain systems, where unknown components are identified during operation to improve control performance. However, resource-intensive online learning algorithms introduce non-negligible computational delays, especially when executed on systems with limited local computational resources. To mitigate this, an in-network online learning-based control structure is employed by deploying the learning-based controller on a remote computation node and connecting it via a communication channel. In this paper, control performance guarantee is first established by deriving tracking error bound for the in-network control architecture, while accounting for computational delays. The derived tracking error bound allows for diverse communication and computation strategies under a specific condition, including time-/event-triggered mechanisms. Additionally, the trade-off between communication and computation performances is shown for a given desired control performance. Furthermore, to enhance the efficiency in both communication and computation, an efficient control framework with an asynchronous event-triggered mechanism in both control and online learning is devised under the existence of computational delay. The proposed event-triggered strategy is proven to achieve the same control performance as time-triggered scenario while excluding Zeno behavior. Finally, we derive an explicit expression of the proposed event-trigger condition for exponentially stabilizable systems, and demonstrate its effectiveness through simulations.
[LG-92] Which LLM for Which Work? Budgeted Model Allocation under Uncertain Evaluation
链接: https://arxiv.org/abs/2608.29560
作者: Hamed Khosravi,Xiaoming Huo
类目: Machine Learning (cs.LG); Optimization and Control (math.OC); Machine Learning (stat.ML)
*备注:
Abstract:A company with a fixed artificial intelligence (AI) budget must decide which large language model (LLM) handles each recurring workload. What it lacks is the quality table, how well each model performs on each workload. Given that table, the decision is a multiple-choice knapsack problem and is routine to solve, so estimating it is the difficulty, and that estimation fails in two ways. Models are rarely compared on the same work, and the recorded score is usually a proxy rather than the outcome the company values. Causal and off-policy methods repair the first but condition on the second, while evaluator-validation methods estimate the second but stop short of the decision. Worse, buying more re-evaluation cannot settle the second: randomization governs which requests are scored, not how a score is produced, so the table stays uncertain however much evaluation is purchased. Yet the deployment decision may still be determined even when the table is not. We therefore ask whether one assignment stays optimal across every quality table consistent with the evidence. For the fixed-budget problem, this admits an exact two-solve certificate: solve once at the estimated table and once at a least-favourable table. Agreement certifies the assignment; disagreement identifies the model-workload pairs where further evidence can matter. We propose CASE (causal active sequential experimentation), which targets evaluation to those pairs and repeats the test as evidence accumulates. On a production log, the measurement failure is the larger of the two: correcting assignment exactly still leaves most of the loss, and randomized re-evaluation does not remove it. In our experiments, the available evidence often does not determine the assignment. On paid software tasks, better information about model quality yields more savings than further optimization of the assignment on the same estimates.
[LG-93] BEACON: Behavioral and Semantic Enrichment of AlphaEarth Embeddings through Tri-Modal Contrastive Learning
链接: https://arxiv.org/abs/2608.29553
作者: Hao Tian,Heng Cai,Yifan Yang
类目: Machine Learning (cs.LG)
*备注: 4 pages, 2 figures, Accepted by ACM SIGSPATIAL 2026
Abstract:Geospatial foundation models such as the AlphaEarth Foundation produce compact and globally consistent representations of the Earth’s surface that transfer effectively to a wide range of downstream tasks. However, because these models are trained primarily on Earth-observation imagery, their embeddings mainly capture physical and spectral characteristics while encoding human activity and urban function only weakly. To address this limitation, we propose BEACON, a tri-modal contrastive learning framework that aligns three complementary views of urban space: physical representations from AE embeddings, semantic representations from point-of-interest (POI) text, and human behavioral representations from hourly POI visitation, while keeping the deployed representation image-only. Using the Houston Metropolitan Area as a case study area, we evaluated the performance of the BEACON framework on nine downstream tasks, including seven regression and two classification tasks against six baselines (raw coordinates, Space2Vec, SatCLIP, TESSERA, Clay and AlphaEarth), using frozen linear and MLP probes over five seeds. Under a linear probe, BEACON improves relative R^2 over AlphaEarth by up to 43% for obesity prevalence, 34% for poor mental health, and 22% for median household income, while remaining competitive in the prediction of physical and environmental variables. These findings highlight the value of augmenting geospatial foundation models with semantic and behavioral signals, extending their applicability from physical Earth observation to human-centered urban analytics.
[LG-94] Adversarial Online Classification with a Preview
链接: https://arxiv.org/abs/2608.29503
作者: Roi Livni,Sahil Singla
类目: Machine Learning (cs.LG); Data Structures and Algorithms (cs.DS)
*备注:
Abstract:Worst-case online classification is governed by sequential complexity, such as Littlestone dimension, and can be impossible even for statistically simple classes, such as thresholds of VC dimension one. We study a preview model in which an oblivious adversary fixes an entire labeled sequence of length T , a uniformly random subset of size pT is revealed before prediction begins, and the remaining (1-p)T examples are then presented in their original adversarial order. Against the best full-sequence hypothesis evaluated on the unrevealed examples, we characterize the dependence on the preview rate p : for binary classes of VC dimension d , the optimal excess loss is \Theta(d/p+\sqrtdT) , up to the trivial cap at T ; for multiclass classes we obtain the corresponding \widetilde O(d_\rm DS/p+\sqrtd_\rm NatT) bound with no dependence on the number of labels. Thus a random preview can replace worst-case sequential complexity by classical statistical dimensions without randomizing the online order. To achieve the sharp binary bound, our ChainedPrediction algorithm uses an online analogue of chaining, implemented as a multiscale aggregation algorithm rather than only as an analytic argument. Subjects: Machine Learning (cs.LG); Data Structures and Algorithms (cs.DS) Cite as: arXiv:2608.29503 [cs.LG] (or arXiv:2608.29503v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2608.29503 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-95] arget-Aware State-Adaptive p-Dirichlet Graph Neural Regression for Non-Invasive Body-Composition Estimation
链接: https://arxiv.org/abs/2608.29496
作者: Nadejda Drenska,Matthew Lemoine,Gowri Priya Sunkara,Yu Wang,Sri Lakshmi Sravani Devarakonda,Steven B. Heymsfield
类目: Machine Learning (cs.LG)
*备注: 21 pages, 7 figures, 3 tables. The first four authors contribute equally to this work and are listed alphabetically. Corresponding authors: N. Drenska, Y. Wang and S. B. Heymsfield
Abstract:Accurate estimation of body-composition outcomes, including body fat percentage (BFP), bone mineral density (BMD), and appendicular lean mass (ALM), is important for evaluating metabolic, skeletal, and muscular health. Direct assessment using dual-energy X-ray absorptiometry (DXA), however, requires specialized equipment and involves ionizing radiation. We propose a target-aware, state-adaptive p -Dirichlet energy-flow graph neural regression ( p SADE-GNR) framework for estimating these outcomes from non-invasive anthropometric measurements. A neural encoder maps participant representations to hidden states that are propagated over an outcome-specific participant-similarity graph by a state-adaptive forward-Euler discretization of the graph p -Dirichlet energy flow. Graph distances weight each original or latent coordinate by its normalized absolute training-fold correlation with the outcome. Using clinical data from the Pennington Biomedical Research Center and five-fold cross-validation, the correlation-weighted model using the original standardized measurements achieved the lowest root mean squared error in all nine primary outcome-cohort combinations and outperformed previously reported support vector regression or least-squares support vector regression reference values in eight of nine comparisons. Autoencoder, variational-autoencoder, and Gaussian-mixture variational-autoencoder representations generally did not improve primary-outcome prediction or reduce computational cost. In an exploratory age-prediction analysis including ALM, BMD, and BFP as predictors, the correlation-weighted GMVAE model achieved the lowest mean error in all three cohorts. These results support target-aware, state-adaptive p -Dirichlet graph neural regression for non-invasive body-composition estimation.
[LG-96] Learning Human Health and Diseases from 24-hour Wrist Movement
链接: https://arxiv.org/abs/2608.29494
作者: Yong Wang,Dylan McGagh,Katya Broomberg,Zizheng Zhang,Jonathan Carter,Junayed Naushad,Laura Brocklebank,Yang Sun,George Nicholson,Dianjianyi Sun,Canqing Yu,Jun Lv,Maxim Barnard,Hubert Lam,Andrew Steptoe,David W. Eyre,Liming Li,Zhengming Chen,Naomi Wray,Spiros Denaxas,Gary S. Collins,Huaidong Du,Aiden Doherty,Hang Yuan
类目: Machine Learning (cs.LG)
*备注:
Abstract:Much of human health and function unfolds beyond the clinic, through the movements of everyday life. Wrist-worn accelerometers capture these movements continuously, yet their rich signals are often reduced to a small set of predefined behavioural summary measures. Here, we present Sensori, a self-supervised foundation model that learns general-purpose health representations directly from 24 hours of raw tri-axial wrist movement. We developed and evaluated the model across four population-based cohorts from the United Kingdom, China and the United States, comprising 122,640 participants contributing 683,617 person-days of free-living recordings. Sensori condensed each day of movement into a representation that captured diverse movement behaviours, demographic characteristics, health axes and physical function. Evaluation in independent cohorts showed that these representations generalised across populations and measurement settings without retraining. When added to common clinical covariates, Sensori significantly improved prevalent disease classification for 52 of 102 eligible conditions (median delta AUROC, 0.060; range, 0.012-0.242) and incident disease risk prediction for 26 of 87 eligible conditions (median delta Uno’s C-index, 0.064; range, 0.025-0.172), with the largest gains for neurological and psychiatric disorders. These findings establish 24-hour wrist movement as a rich and scalable source of health information, with the potential to support passive health monitoring and disease prediction at population scale.
[LG-97] A Causal Model for Locating and Unlocking Sandbagging in Model Organisms
链接: https://arxiv.org/abs/2608.29461
作者: Hong Kiat Tan,Linh Le,David Williams-King
类目: Machine Learning (cs.LG)
*备注: under submission
Abstract:Sandbagging models strategically underperform on evaluations while retaining the capabilities being measured. The evaluations that guide frontier-model deployment and governance then understate what these models can do. To understand the mechanism, we propose a causal model of how sandbagging is carried in the residual stream. Early layers write the sandbagging intent onto a single axis of the stream, and a later layer reads that axis and commits the answer. We study three instruction-tuned models (Qwen2.5-7B, Llama-3-8B, and Mistral-7B) and four ways of installing a sandbagging lock (prompting, fine-tuning, reinforcement learning, and circuit breaking). Each lock decides from the prompt whether to sandbag, and the fine-tuned and circuit-broken locks answer honestly whenever a password appears. The causal model predicts a window of layers, after the last sandbagging write and before the answer commit, in which a single-layer reference graft of the sandbagging axis to its honest value restores the full capability. The single-layer graft recovers the capability in 28 of the 33 runs of the prompted, fine-tuned, and RL-trained locks, with a median held-out recovery of 96%. The circuit-broken lock rewrites the whole state through a band of layers, and the single-layer graft fails at every layer. We therefore introduce a second intervention, context grafting, which replays the password’s cached key/value activations so that every layer’s attention reads them as additional context. Context grafting provably and empirically restores the full capability on all three models, and the recovery is surprisingly insensitive to the exact password content. More broadly, an auditor can use this causal model to design interventional auditing techniques for sandbagging models.
[LG-98] Behavioral Latency as Weak Event-Time Supervision for EEG Reaction-Time Decoding
链接: https://arxiv.org/abs/2608.29428
作者: Anuar Aimoldin,Ayana Mussabayeva,Yedige Mussabayev,Xue Liu,Kun Zhang
类目: Machine Learning (cs.LG)
*备注: 33 pages, including appendices; 7 figures and 12 tables. Accepted for publication in Neural Computation
Abstract:Single-trial EEG analyses are often organized around events and latencies, yet EEG-based reaction-time (RT) prediction is posed as scalar regression on a fixed stimulus-locked window. RT is treated as a window-level label rather than timing evidence about response-relevant dynamics. Here we reformulate trial-wise RT decoding as event-time posterior modeling. Instead of predicting RT directly, the model estimates a posterior over response-relevant event times, p(t_\mathrmevent\mid X) , and uses its mean as the RT estimate. This treats behavioral latency as a weak observation of latent response-relevant timing. We evaluate this formulation on the Healthy Brain Network contrast change detection EEG task under a subject-disjoint, release-separated protocol. Across five seeds, distributional event-time supervision consistently improves held-out RT prediction relative to scalar regression and temporal-readout controls. Controlled objective comparisons isolate supervision of the event-time distribution, rather than expectation-based readout alone, as the source of this gain. Architecture controls show that the effect persists across four temporal backbones and is not explained by model scale. Beyond point prediction, posterior geometry characterizes concentration, target alignment, and interval behavior, while observation-noise calibration separates latent concentration from predictive uncertainty over RT. Shifted-crop inference probes shortcut use versus temporal localization. Matched shift-jitter improves robustness, increases mean sensitivity, and moves predictions more often in the expected crop-relative direction. Sensitivity remains below ideal crop-relative localization, leaving a clear equivariance gap. Together, these results establish event-time posterior modeling as a probabilistic and interpretable formulation for linking single-trial EEG dynamics to behavioral timing.
[LG-99] One Capability or Many? Testing the Economic Validity of Frontier AI Evaluation
链接: https://arxiv.org/abs/2608.29420
作者: Louis Yiven Zhu
类目: Machine Learning (cs.LG); Computers and Society (cs.CY); Software Engineering (cs.SE)
*备注: 25 pages, 11 figures. Analysis plan deposited at this https URL (retrospective deposit; see the note there). Code and data: this https URL
Abstract:Frontier-model leaderboards now rank systems based on economic benchmarks, tests of how well models carry out professional tasks from software engineering to banking workflows, and those rankings inform what organisations buy, what regulators scrutinise, and expectations of how work will change. Whether such benchmarks measure a capability distinct from general test-taking, or re-express the one axis along which every benchmark rises as models improve, is a question of construct validity that has not yet been studied. We test it on a hash-pinned leaderboard snapshot of 421 model configurations across twelve benchmarks, four of them economic, treating benchmarks as items and models as respondents in a latent-variable model with four hypotheses and their thresholds fixed before analysis. A single factor explains 74.5% of common variance and tracks model release date (R^2 = 0.505), so the leading axis of capability is substantially a time trend; where prior work controls for scale, compute adds little once date is removed. Removing the date trend lowers that share by 14.9 points, and by 24.1 with one row per base model. Under the dimensionality rule fixed in advance the economic benchmarks form no distinct factor, yet a leave-one-benchmark-out test with factors re-estimated inside every fold shows that a multi-factor representation predicts held-out economic scores better than a single general index (pooled Delta-MSE 0.037, 95% bootstrap interval [0.019, 0.055]). Economic benchmarks therefore add incremental predictive information to a largely date-driven general factor, and the evidence does not support treating them as a distinct latent capability. Leaderboards remain a sound guide to overall progress, but most of the gap between models released months apart is calendar, so a small gap between contemporaneous models should be date-adjusted before being read as a capability difference.
[LG-100] Where Induction Runs Out: Description-Length Difficulty and the Memorisation Gap in Integer-Sequence Benchmarks
链接: https://arxiv.org/abs/2608.29411
作者: Sabilashan Ganeshan
类目: Machine Learning (cs.LG); Symbolic Computation (cs.SC)
*备注: 14 pages, 6 figures. Code and data: this https URL . Archived: doi: https://doi.org/10.5281/zenodo.21830331
Abstract:Integer sequences from the On-Line Encyclopedia of Integer Sequences (OEIS) are increasingly used to benchmark mathematical reasoning in language models. We ask what such benchmarks actually measure, using an exactly computable reference learner: two-part minimum description length (MDL) over the class of P-recursive (holonomic) recurrences, evaluated on every prefix of a sequence as terms arrive. Three findings follow. First, MDL difficulty is a parameter count. The discovery point nd, the first prefix length at which a symbolic hypothesis beats verbatim storage, is predicted almost exactly by a combinatorial identifiability bound on the selected operator’s order and degree. It is invariant to term magnitude: scaling Fibonacci over twelve orders of magnitude leaves nd unchanged, because a hypothesis must encode its own initial conditions and the magnitude cancels. Second, at scale the learner exhibits a regime our curated corpus could not produce even once: across 20,000 OEIS sequences, 89.98% of those that fit a recurrence on some prefix fit none at full length. We call this the wilderness – induction acquires a theory, loses it, and never recovers. Third, evaluating three language models on sequences stratified by these MDL regimes refuted our pre-registered hypothesis: models do not confabulate where MDL reports no theory, but hedge appropriately. Confident errors are inverted, concentrating on the easy stratum, where apparent competence tracks recognition of the sequence rather than induction of its rule. OEIS-derived benchmarks therefore substantially measure memorisation, and MDL supplies a cheap, contamination-free difficulty signal they currently lack. Code and data are released.
[LG-101] Unlearning on Spatio-Temporal Graphs through Subgraph Virtual Edge Reconstruction
链接: https://arxiv.org/abs/2608.29369
作者: Qiming Guo,Wenbo Sun,Chen Pan,Ye Wang,Wenlu Wang
类目: Machine Learning (cs.LG); Cryptography and Security (cs.CR)
*备注: Accepted as a short paper at ACM SIGSPATIAL 2026. 4 pages
Abstract:Spatio-temporal graphs are widely used in modeling complex dynamic processes such as temporal forecasting, molecular dynamics, and healthcare monitoring. Recently, stringent privacy regulations such as GDPR and CCPA have introduced significant new challenges for existing spatio-temporal graph models, requiring complete unlearning of unauthorized data. Since each node in a spatio-temporal graph diffuses information globally across both spatial and temporal dimensions, existing unlearning methods primarily designed for static graphs and localized data removal cannot efficiently erase a single node without incurring costs nearly equivalent to full model retraining. To address this, we propose CallosumNet, a spatio-temporal graph unlearning framework biologically inspired by the corpus callosum structure. CallosumNet makes two key technical contributions: (1) it reconstructs subgraphs using biologically-inspired virtual edges; and (2) it restores interlinked spatio-temporal dependencies among subgraphs via a lightweight meta-graph integration layer. Empirical results on four diverse real-world datasets show that CallosumNet achieves complete unlearning while maintaining accuracy very close to the gold model. The code is publicly available at this https URL.
[LG-102] Spectral Analysis for Sparse Matrix Computation: Insights and Potential
链接: https://arxiv.org/abs/2608.29362
作者: Ruifeng Zhang,Xipeng Shen
类目: Performance (cs.PF); Machine Learning (cs.LG)
*备注: 12 pages, 10 figures, 2 tables
Abstract:Sparse computations are fundamental to scientific computing, graph analytics, and machine learning, yet their performance is highly sensitive to the diverse sparsity and patterns. This is because cache reuse, memory coalescing, and load balancing depend critically on the sparsity patterns. This work gives the first known exploration of the connections between sparse matrix computation and spectral analysis by treating sparse matrices as two-dimensional signals and analyzing their frequency-domain representations through Fast Fourier Transform. We show that spectral signatures uncover global structural characteristics that are not sufficiently captured by conventional spatial statistics and provide complementary information for understanding sparse computation performance. Experiments on incorporating spectral features into machine-learning-based SpMV format selection demonstrate the usefulness of such spectral analysis over a state-of-the-art spatial-only model. By uncovering the principled connections between spectral characteristics and sparse matrix computations, this work introduces a novel analytical perspective into sparse computation, and provides a new approach to enhancing the current sparse structure characterization and optimization. On pruned LLM decoding, adding spectral features improves kernel selection and yields 1.035–1.245 \times kernel speedups.
[LG-103] Spatial Entropy based Partitioning for Spatiotemporal Graph Unlearning SDM2026
链接: https://arxiv.org/abs/2608.29360
作者: Qiming Guo,Wenbo Sun,Ye Wang,Wenlu Wang
类目: Machine Learning (cs.LG)
*备注: Accepted at SIAM International Conference on Data Mining (SDM 2026)
Abstract:Spatiotemporal graphs underpin applications such as traffic forecasting, weather forecasting, and healthcare monitoring. Privacy regulations such as the GDPR and the CCPA require the complete removal of unauthorized data from trained models, but achieving this on a spatiotemporal graph is difficult: because information propagates globally through both spatial and temporal message passing, fully erasing a node’s influence forces costly full-graph retraining. ST-graph unlearning requires both exactness and efficiency. We propose IsleNet, which uses spatial-entropy-guided partitioning to create balanced, locally coherent subgraphs and reconnects them with lightweight virtual edges. Upon an unlearning request, only the affected subgraph encoder and virtual-edge layer are retrained, ensuring exact removal with low cost. Experiments on four real-world benchmarks show that IsleNet attains up to 94% of full-graph accuracy while reducing unlearning time by up to an order of magnitude. Our code is publicly available at this https URL.
[LG-104] Information-Based Calibration of Uncertainty Quantification in Product-of-Experts Gaussian Process Models
链接: https://arxiv.org/abs/2608.29349
作者: Yean Hoon Ong,Paolo Barucca,Wei Pan,Jun Wang
类目: Machine Learning (cs.LG)
*备注: Published in the Journal of Artificial Intelligence Research, Volume 86 (2026)
Abstract:Gaussian process (GP) regression with a single global GP (GP-glo) incurs cubic computational cost, limiting scalability to large datasets. Product-of-experts GP models (GP-pro), which combine local GP models to capture global correlations, alleviate this computational burden. However, training local experts on disjoint data subsets can lead to overestimated posterior variances. We propose GP-pro-c, a product-of-experts GP model that calibrates these variances using an information-based method. The method exploits the monotonicity and submodularity of information gain in GPs to define a calibration ratio that reduces the posterior variance of individual local GP models. We evaluate GP-pro-c using negative log-likelihood (NLL), root mean squared error (RMSE), and expected normalised calibration error (ENCE). Experiments on four synthetic functions and six regression datasets show that GP-pro-c achieves average reductions of 2.3% in NLL and 12.0% in ENCE compared with the uncalibrated GP-pro model. The proposed method mitigates posterior variance overestimation while maintaining predictive accuracy and reducing computational complexity. GP-pro-c provides a promising approach for uncertainty estimation in scalable GP models and may serve as a useful surrogate model for Bayesian optimisation with high-dimensional and large-scale data.
[LG-105] MEL: Coordinate-Preserving EEG Tokenization for fMRI Translation
链接: https://arxiv.org/abs/2608.29304
作者: Xiangyu Liu,Zeting Yan,Zhitong Yin,Boyang Li,Xi Zhang
类目: Machine Learning (cs.LG)
*备注:
Abstract:Translating electroencephalography (EEG) into functional magnetic resonance imaging (fMRI) is important for medical neuroimaging, clinical brain-state monitoring, and multimodal neural decoding, because it aims to infer spatially organized hemodynamic activity from fast and accessible electrophysiological recordings. Existing EEG-to-fMRI studies mainly pursue stronger decoders, but the problem is also constrained by a representation-interface mismatch: fMRI responses are delayed, temporally integrated, and spatially distributed, whereas generic EEG encodings often entangle temporal lag, channel identity, and frequency-band structure. We propose Multi-band EEG Latent-state Tokenization (MEL), a coordinate-preserving EEG representation framework that anchors each target fMRI response to its preceding EEG history and organizes it into lag-channel-frequency neural-state tokens. By explicitly capturing hemodynamic latency and spectral-spatial dynamics, MEL aligns fMRI-pertinent EEG representations with capacity-controlled readouts without depending entirely on model scaling. Experiments on VU EEG-fMRI benchmarks and external Oddball data show that MEL improves prediction over strong NeuroBOLT baselines. Ablations and controls further indicate that the gains come from structured EEG representation rather than leakage, shortcut statistics, or decoder capacity.
[LG-106] A Spectral Identifiability Threshold for Dissipative Rate Recovery from Truncated Liouvillian Spectra
链接: https://arxiv.org/abs/2608.29302
作者: Yujun Ji,Somyajit Chakraborty
类目: Machine Learning (cs.LG); Computational Physics (physics.comp-ph); Quantum Physics (quant-ph)
*备注: 15 pages, 6 figures. Data and code are available at Zenodo, DOI: https://doi.org/10.5281/zenodo.22160077
Abstract:Open quantum systems lose energy and phase coherence through different dissipative processes, but these processes can produce overlapping dynamical signatures. The Liouvillian spectrum summarizes how such a system relaxes, yet it is not obvious how much of that spectrum is needed to distinguish the underlying dissipation rates. We study this question for amplitude damping and dephasing in a six-qubit Lindblad model whose spectrum can be derived analytically. We retain only the slowest non-steady spectral modes and ask how many are required before each dissipative rate becomes recoverable. We show that population modes contain no dephasing information, which creates a lower bound of D = 2^n retained modes for uniform dephasing identifiability in the relevant rate regime. The measured recovery threshold reaches this bound at n = 4,5,6, while n = 3 remains above it. At n = 6, least squares achieves a mean joint absolute error of order 10^-9, compared with 4.355 x 10^-4 for four tabular learning methods. Robustness tests show that this advantage weakens when the spectra are perturbed and when a transverse field breaks the commuting structure. These results show that the amount and structure of retained spectral information can determine whether dissipative parameters are recoverable, independently of the estimator used. The present conclusions apply to noise-free simulator spectra rather than measurement-derived spectra.
[LG-107] Sense Once Serve Many: Common-Trace Factorized Constrained PPO for Online Sensing-Session Consolidation in Multi-Tenant ISAC Networks
链接: https://arxiv.org/abs/2608.29256
作者: Dang-Dung Vu
类目: Networking and Internet Architecture (cs.NI); Machine Learning (cs.LG)
*备注: 13 pages, 4 figures, 6 tables
Abstract:Integrated sensing and communication (ISAC) networks can serve compatible requests through shared sensing sessions, but consolidation couples admission, reuse, profile selection, sensing service-level agreements (SLAs), communication quality of service (QoS), and future commitments. We formulate this problem as a constrained Markov decision process and propose Common-Trace Factorized Constrained Proximal Policy Optimization (CT-PPO). During training, stochastic policy replicas share the same primitive workload trace; leave-one-out discounted Monte Carlo return contrasts provide reward credit to applicable actor factors, while constraint credit remains factor/prefix-specific. Across five training seeds and matched workloads, CT-PPO achieves the highest mean macro return, exceeding matched Joint-Credit PPO (JC-PPO) by 0.934 (95% confidence interval [0.702, 1.164]) and SLA-Aware Greedy by 1.847; versus JC-PPO, it reduces sensing-resource cost by 6.277 and raises accepted requests per created session by 0.0806. A four-way ablation shows that the factorized surrogate alone yields no detectable macro-return gain, whereas adding common-trace reward credit produces the dominant improvement. Without retraining, CT-PPO retains a return advantage at low, nominal, and high arrival loads, with the strongest gain under clustered arrivals. Deployment uses public observations and hard masks; CT-PPO’s extra parameters are training-side, its actor footprint matches JC-PPO, and actor-only CPU latency is effectively unchanged.
[LG-108] A-MADiff: Attention-Guided Multi-Agent DRL with Diffusion Policies for Memory-Aware Task Orchestration in Mobile AIGC Networks
链接: https://arxiv.org/abs/2608.29255
作者: Chongzhi Wu,Zhengtao Li,Jiawen Kang,Jinbo Wen,Xiaohuan Li,Maomao Zhang,Ekram Hossain
类目: Networking and Internet Architecture (cs.NI); Machine Learning (cs.LG)
*备注:
Abstract:Artificial Intelligence-Generated Content (AIGC) services employ Generative AI (GenAI) models to automatically generate diverse content. Mobile AIGC networks host GenAI models on edge-located AIGC Service Providers (ASPs) to deliver low-latency and personalized AIGC services for mobile users. However, AIGC inference tasks typically occupy GPU memory until task completion, causing GPU memory exhaustion at serving ASPs and triggering out-of-memory failures rather than merely increasing service latency. Existing studies on AIGC task orchestration have largely overlooked GPU memory feasibility constraints. To address this issue, we develop a cooperative multi-agent orchestration framework, in which each edge node is equipped with a scheduling agent to route tasks to local ASPs or neighboring edge nodes. Since scheduling agents make decisions based only on local observations, while peer offloading couples their resource states and long-term utilities, we formulate the orchestration process as a cooperative Decentralized Partially Observable Markov Decision Process (Dec-POMDP). To solve the Dec-POMDP, we propose an \underlineAttention-guided \underlineMulti-\underlineAgent deep reinforcement learning algorithm with \underlineDiffusion policies (A-MADiff) under the centralized training with a decentralized execution paradigm. A-MADiff employs diffusion-based decentralized actors to generate multi-modal preferences over feasible orchestration actions, and an attention-guided centralized critic to estimate per-agent values from cross-agent states under GPU memory heterogeneity. Numerical results demonstrate that A-MADiff significantly improves the cumulative reward over the state-of-the-art baseline.
[LG-109] RL-FAT: Reinforcement Learning for Fair Adversarial Training
链接: https://arxiv.org/abs/2608.29247
作者: Tejaswini Medi,Levan Mikeladze,Margret Keuper
类目: Machine Learning (cs.LG)
*备注:
Abstract:Deep neural networks remain highly vulnerable to adversarial perturbations, and adversarial training (AT) has become a widely used approach for improving robustness. However, improvements in average robust accuracy often mask substantial class-wise disparities: while some classes become more robust, others may remain disproportionately vulnerable under attack. This imbalance raises an important adversarial fairness concern, particularly in vision tasks where reliable robustness is expected across all categories. To address this challenge, we propose \textbfRL-FAT, a reinforcement-learning-inspired fair adversarial training framework that uses policy-gradient based feedback from adversarial predictions. RL-FAT interprets the prediction distribution as a policy and combines correctness-based prediction rewards with class-wise value estimates to compute class-specific advantages for policy-gradient optimization. This enables the model to adaptively focus on class-wise misclassification. Furthermore, we introduce a fairness-emphasis adversarial loss that assigns stronger training pressure to classes with high adversarial loss, thereby mitigating class-wise robustness disparity. By combining reinforcement-driven adaptation with fairness-emphasis regularization, RL-FAT improves adversarial robustness while promoting a more balanced robustness distribution across classes. Extensive experiments demonstrate that our method achieves competitive robust accuracy and substantially reduces class-wise robustness imbalance compared with standard adversarial training baselines.
[LG-110] AdaVLA: Adaptive Step Flow Matching for Training-free Acceleration of Vision-Language-Action Models IROS2026
链接: https://arxiv.org/abs/2608.29208
作者: Sunghwan Han,Youngtae Han,Youngmin Yi
类目: Robotics (cs.RO); Machine Learning (cs.LG)
*备注: Accepted to the 2026 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2026)
Abstract:Vision-Language-Action (VLA) models, built upon Vision-Language Models (VLMs), have significantly enhanced robotic capabilities by leveraging internet-scale knowledge and multimodal reasoning. However, the intensive computational overhead of VLAs constrains on-device deployment, hindering real-time responses to environmental changes. While various acceleration techniques have been proposed, they often rely on fine-tuning or access to training datasets, which are frequently unavailable due to privacy and proprietary concerns. Moreover, although flow-matching-based VLAs have emerged as efficient alternatives to standard diffusion models, current acceleration efforts largely target VLM inference costs, failing to address the iterative ODE solving process inherent in flow matching inference. To address these limitations, we propose AdaVLA, an online, training-free adaptive framework for fast yet accurate flow-matching-based Vision-Language-Action models. We introduce a novel metric derived from the flow matching trajectory curvature to quantify action generation confidence during inference. This metric enables the dynamic reduction of inference steps and the adaptive adjustment of MLP pruning ratios through an efficiently computed importance evaluation, requiring no access to training data. Experimental results on the LIBERO benchmark using a Jetson AGX Orin device demonstrate that our method achieves 1.87\times and 2.24\times speedups for \pi_0.5 and X-VLA, respectively, with negligible degradation in success rates. Furthermore, we validate the robustness of our approach on real-world robotic tasks using SmolVLA.
[LG-111] HalluPrism: When Multimodal Uncertainty Should Diagnose Not Decide EMNLP2026
链接: https://arxiv.org/abs/2608.29193
作者: Aman Prakash,Sourish Dasgupta,Tanmoy Chakraborty
类目: Machine Learning (cs.LG)
*备注: Accepted to Findings of EMNLP 2026. 52 pages, 1 figure, 46 tables
Abstract:Multimodal Large Language Models (MLLMs) can assign similar confidence to answers that fail for different reasons. We propose HalluPrism, a behavioral diagnostic that re-runs an answer after visual degradation, blank-image replacement, and grounding or relation checks. These targeted probes yield a signature over visual-perturbation sensitivity (V ), image-removal confidence retention (L), and grounding/relation-probe instability (A). Across 58K+ examples from four benchmarks and four MLLMs, image-removal confidence retention is most prevalent, while grounding/relation-probe instability better separates failure families. Only 18 of 48 source-target checks are diagonally aligned, so the coordinates should be interpreted jointly rather than as independent causal sources. With the dataset fixed, the joint signature improves failure-family AUROC from 0.634 to 0.769 on HallusionBench and from 0.707 to 0.817 on VizWiz, with smaller gains on POPE and VSR. In pooled XGBoost analysis, AUROC rises from 0.78 with scalar confidence to 0.95 with (V, L, A) and 0.97 when confidence is added. The same signature does not automatically improve correctness ranking. The three tested direct scalarizations can harm it. These results separate failure diagnosis from abstention scoring: multimodal uncertainty should characterize failure structure before it is used to decide whether to abstain or correct.
[LG-112] Explainable Machine Learning for Broadband Adoption Disparities: Tract-Level Prediction and SHAP-Based Factor Profiling
链接: https://arxiv.org/abs/2608.29110
作者: Xiao Han
类目: Machine Learning (cs.LG)
*备注:
Abstract:The United States has allocated approximately 65 billion through the Infrastructure Investment and Jobs Act for broadband expansion, yet evidence-based methods for targeting these investments remain underdeveloped. This paper presents an explainable machine learning framework for profiling broadband adoption disparities at census-tract granularity across 83,359 tracts nationwide. Using 65 socioeconomic, demographic, and infrastructure features derived from the American Community Survey 2022, we train a LightGBM model under spatial five-fold cross-validation, achieving R^2 = 0.533 and Spearman rho = 0.763; state-held-out cross-validation (51 folds) confirms generalization (R^2 = 0.525). TreeSHAP analysis identifies income and education as the dominant factor group (with the engineered interaction term absorbing attribution from its constituent features), and SHAP-based clustering reveals three exploratory factor profiles: Well-Connected Moderate (~49K tracts), Affordability-Limited Severe (~21K tracts), and Rural-Elderly (~13K tracts). As a screening tool, ML-based tract selection captures 38.0% of the total adoption gap within the top 10% of tracts versus 35.2% for income-only heuristics (+2.8 pp, p 0.002, county-block bootstrap); in regret-reduction terms, the model closes 19% of the remaining gap between income-only and oracle selection. The primary contribution is the per-tract factor decomposition: SHAP identifies which feature groups (income/education, rurality, age) are most strongly associated with each tract’s predicted gap, and informs differentiated investigation. A temporal stability check, training on ACS 2017 and predicting ACS 2022 with zero survey-year overlap, confirms ranking stability (rho = 0.784, noting hyperparameters tuned on 2022 data).
[LG-113] PathGuide: Dynamic Classifier-Free Guidance via On-Policy Transport Alignment
链接: https://arxiv.org/abs/2608.29107
作者: Avishag Nevo,Tamir Hazan
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: 40 pages, 8 figures, 17 tables. Includes appendices with full proofs, ablations, reproduction details and additional experiments
Abstract:While modern generative models excel at modeling complex data, precise inference-time control in conditional generation remains a critical challenge. Classifier-free guidance (CFG) is a primary mechanism for such control, yet it is typically treated as a static tuning parameter. In flow-based models, however, the guidance scale fundamentally dictates the velocity field and the resulting probability path, making guidance selection a dynamic path-optimization problem. We introduce PathGuide, a framework that reformulates scalar CFG selection as an on-policy transport problem. Leveraging the weak form of the continuity equation, we derive a selection criterion with a direct path-correctness interpretation: we prove that if the guided field is weakly equivalent to the exact conditional field along the generated rollout, the sampler’s path coincides with the target conditional law. For scalar CFG, this criterion yields a strictly quadratic local objective with an efficient, closed-form selector for each solver interval. PathGuide enables optimal guidance scales to be computed and used online during generation or fitted offline as a reusable piecewise-constant schedule. We validate our method on low-resolution image manifolds and controlled settings across various continuous-time flow constructions, demonstrating that this transport-based selector improves path alignment and sample fidelity over both fixed and state-of-the-art adaptive guidance baselines.
[LG-114] Optimally Selecting Representative Agents from a Metric Space
链接: https://arxiv.org/abs/2608.29097
作者: Benjamin Cookson,Eva Deltl,Yeeseok Oh
类目: Computer Science and Game Theory (cs.GT); Machine Learning (cs.LG)
*备注:
Abstract:This paper studies the problem of proportionally fair clustering, where the goal is to select k ``centers’’ from a metric space that fairly represent a set of agents who also lie in the metric space. Specifically, we focus on finding a clustering satisfying a fairness property known as the Droop core. In the practical special case in which the set of feasible center locations contains every agent location, the previous best-known result guaranteed a (1 + \sqrt2) -approximation of the Droop core, while the best-known lower bound was 2 . In this paper, we show that this lower bound is tight and that a clustering in the 2 -Droop core always exists. Further, we show that such a clustering can be achieved by only selecting centers from locations in the metric space where an agent resides. We establish this using Scarf’s theorem guaranteeing a nonempty core for balanced non-transferable utility games. This result has several interesting corollaries. Most notably, it resolves the \beta -plurality problem of Aronov et al. [2021] for general metric spaces. The main result of this paper was generated by \mathttChatGPT - \mathtt5.6 - \mathttSol through a series of interactions with the authors. The authors of this paper verified the generated proof and rewrote it for clarity.
[LG-115] ans-QFWP: A Regime-Aware Hybrid Quantum Fast Weight Programmer for Portfolio Optimization
链接: https://arxiv.org/abs/2608.29093
作者: Ming-Kai Hung,Jun-Hao Chen,Yun-Cheng Tsai,Samuel Yen-Chi Chen
类目: Machine Learning (cs.LG)
*备注: 4 pages, 3 figures, 4 tables, accepted for presentation at the IEEE International Conference on Quantum Computing and Engineering (QCE) 2026 QCRL Workshop
Abstract:We propose Titans-QFWP, a hybrid reinforcement learning architecture integrating a Quantum Fast Weight Programmer with Titans-style memory (Persistence, Surprise, and Forgetting) for adaptive portfolio optimization. To address high-dimensional market features, we introduce an enhanced A3C^2 framework with Hungarian-aligned K-means clustering and scaled log-return rewards. Evaluated on 468 SP 500 stocks under an Equal-Parameter-Count (EPC) benchmark with approximately 3,000 trainable parameters, Titans-QFWP achieves strong performance (median ARR 0.4260, Calmar 8.5504, IR 0.8427). Ablation results reveal that quantum gating fundamentally reshapes memory component roles, with Persistence supporting drawdown control, Surprise contributing to return generation, and Forgetting providing additional stabilization. By stabilizing these quantum representations, the model enables defensive allocation during market drawdowns while preserving upside potential.
[LG-116] Selective Disclosure of Hidden Directives in Reasoning Models: Behavioral Asymmetry and Steering
链接: https://arxiv.org/abs/2608.29070
作者: Zimo Shi,Xander Tifft,Wen Xing
类目: Machine Learning (cs.LG)
*备注: 22 pages, including appendices
Abstract:Chain-of-thought (CoT) reasoning traces are increasingly proposed as a mechanism for AI oversight: a monitor inspecting a model’s reasoning can, in principle, detect misbehavior invisible from outputs alone. This assumes CoT surfaces what a model is instructed to do regardless of the instructions given. We test this assumption along two axes. First, we introduce the Instruction-Compliance Gap (ICG): the difference in probability that a model’s CoT explicitly references a hidden system prompt directive when that directive is malign versus benign. Across 100 task pairs and 8 frontier reasoning models from 5 families, we find consistent asymmetric disclosure, a higher probability of leaking malign hidden instructions than benign ones, in Qwen3-14B (Wilcoxon p=0.0001 , +13.9 pp), Qwen3-32B ( p=0.0011 , +13.0 pp), Qwen3-235B ( p=0.035 , +5.8 pp), and similar results with MiniMax-M2.5 and DeepSeek-R1. The detector has 100% precision against two independent blinded labelling passes, and an LLM monitor reading only the reasoning trace reproduces the asymmetry in all 8 models against directive-free controls, identifying the specific directive in 82% of malign traces which the detector classifies as clean. Second, steering vectors extracted in MiniMax-M2.5 via Contrastive Activation Addition causally induce hiding from bare prompts and suppress it from prompts that would otherwise produce it, replicating in Qwen3-14B under a pre-registered design. Benign and malign-derived hiding vectors are highly similar (cosine 0.804 in MiniMax-M2.5; 0.970 in Qwen3-14B), implying that in these models the disclosure asymmetry arises from differential activation of a shared hiding direction rather than separate mechanisms.
[LG-117] PathBridger: Subgoal Bridges for Offline Goal-Conditioned Reinforcement Learning
链接: https://arxiv.org/abs/2608.29061
作者: Soohyun Choi,Seonvin Cho,Songnam Hong
类目: Machine Learning (cs.LG); Robotics (cs.RO)
*备注: 14 pages, 2 figures. Code: this https URL
Abstract:Offline goal-conditioned reinforcement learning (GCRL) aims to learn policies for reaching diverse goals entirely from fixed trajectory data. Long-horizon offline GCRL remains challenging because sparse goal-reaching signals must be propagated over many steps, while execution errors cannot be corrected through additional environment interaction. Existing methods address these challenges by improving long-range value estimation or reducing the effective decision horizon through subgoals, options, and action chunks. In several hierarchical methods, however, a selected subgoal specifies where to go, while the intervening state-space path remains implicit in an endpoint-conditioned low-level policy. To address this interface, we propose PathBridger, a hierarchical offline GCRL method that explicitly connects subgoal selection to short-horizon execution. PathBridger constructs a state-space bridge toward the selected intermediate endpoint and decodes it into a short executable action chunk using an inverse dynamics model. Experiments across the evaluated OGBench tasks demonstrate strong aggregate performance, with particularly large gains on the multi-object Cube manipulation tasks. Code: this https URL
[LG-118] Sparse Koopman Autoencoders Identify Local Dynamical Regimes in Multibasin Systems
链接: https://arxiv.org/abs/2608.29057
作者: Aidan Li,Uday Kiran Reddy Tadipatri,Mahan Fathi,Sarath Chandar,Ross Goroshin
类目: Machine Learning (cs.LG); Dynamical Systems (math.DS)
*备注:
Abstract:Koopman autoencoders (KAEs) seek a higher-dimensional latent representation in which nonlinear dynamics evolve linearly. However, many interesting systems have multiple basins of attraction, and both theoretical and empirical work has shown these multibasin systems cannot generally admit a single finite-dimensional global Koopman embedding under standard assumptions. We posit that encoders with a sparsity-inducing objective encouraging few active latent coefficients will provide latent supports as an inspectable basin-modeling principle for Koopman autoencoders. We use these encoders producing sparse latents in training Sparse Koopman Autoencoders (SKAEs) without basin labels or other regime annotations, and treat the learned latent supports as model-produced regime variables after training. Across a range of procedurally generated multibasin systems and chaotic flows, we show that SKAEs have superior forecasting performance compared to dense-latent KAEs. We also perform a mechanistic study that shows latent supports produced by SKAEs are both essential for the quality of the representation and useful for identifying basins on held-out basin interior states, whereas dense-latent KAEs collapse to an uninformative single family. These results identify sparse latents and their corresponding supports as label-free, interpretable regime variables for Koopman learning in nonlinear systems with multiple local dynamical laws.
[LG-119] NVE: A Separability and Coverag e-Aware Internal Validation Metric for Biclustering
链接: https://arxiv.org/abs/2608.29045
作者: Paritosh Tiwari,I Navin Kumar,James C. Bezdek,Punit Rathore
类目: Machine Learning (cs.LG)
*备注:
Abstract:Biclustering, or co-clustering, aims to discover coherent submatrices by grouping rows and columns of a data matrix simultaneously. This local two-dimensional structure makes validation more difficult than in ordinary clustering, where internal indices usually rely on compactness and separation in a single shared feature space. Existing popular internal biclustering measures such as Mean Squared Residue (MSR), and Virtual Error (VE) mainly evaluate within-bicluster coherence. Although useful, these measures do not directly assess whether the extracted biclusters are mutually distinct or whether they explain a meaningful portion of the data matrix. This paper investigates Normalised Virtual Error (NVE), an internal validation metric that extends VE using a super-bicluster normalization strategy. By comparing the VE of each bicluster with the VE obtained after merging it with other biclusters, NVE introduces a relative notion of separability and redundancy. We also study a coverage-adjusted variant, NVE\textsubscriptcov, which penalizes solutions that obtain low error by selecting only very small submatrices. Through controlled synthetic benchmarks and yeast gene-expression datasets, we examine whether NVE and NVE\textsubscriptcov provide information beyond standard coherence-based metrics. The results show that NVE is sensitive to redundant and poorly separated biclusters, while NVE\textsubscriptcov changes solution rankings when low-error biclusters cover only a negligible part of the matrix. These findings suggest that NVE-based measures are useful complementary criteria for internal co-clustering validation, especially when coherence, separability, and coverage must be considered jointly.
[LG-120] Hybrid Semantic Context-Enhanced Ensemble Learning for Wind Power Ramp-Event Forecasting and Uncertainty-Aware Evaluation
链接: https://arxiv.org/abs/2608.29024
作者: Momina Liaqat Ali,Muhammad Abid,Muhammad Abdullah,Aneela Zameer
类目: Machine Learning (cs.LG)
*备注:
Abstract:Wind power ramp events which are sudden, large swings in turbine output over short windows are difficult to estimate, and standard models often miss them. Hybrid forecasting approach is built which augments semantic context to ramp-event forecast. Rather than applying an extensive language model directly to predict turbine operating data, we have implemented a pipeline where turbine operating data is converted to simplified text, which is then converted to dense embeddings to be used as inputs for ensemble models incorporated with other features. Testing runs are performed at multiple intervals within the SDWPF dataset, including 10-minute, 30-minute, and 60- minute horizons, with ramp events constituting the highest change in future power output. We check robustness against autoregressive, LSTM, and GRU baselines plus several ensemble configurations, using Diebold-Mariano tests and bootstrap confidence intervals, and we vary the ramp threshold, compress the embeddings with PCA, and validate externally on Kaggle SCADA and NREL data with uncertainty-aware scoring. The semantic-context features produce negligible yet statistically significant gains over the baselines in multiple paired ensemble runs, most clearly at the 30- and 60-minute horizons where these gains hold across different ramp-threshold definitions, and PCA compression helps in some longer-horizon cases. The best context- augmented ensembles rank near the top overall, though the GRU model still posts the lowest ramp-event RMSE at 30 and 60 minutes. External tests confirm the error reduction generalizes across datasets, but the size of the gain depends on both model and dataset. Prediction intervals cover most test cases well but weaken during ramp events, pointing to a localized shift in the data distribution.
[LG-121] V2TATC: A Joint Voice-Trajectory Embedding Framework and Dataset for Air Traffic Controller Situational Awareness
链接: https://arxiv.org/abs/2608.28981
作者: Louis Brusset,Mathurin Petit,Jordan Kam,Alexandre Bayen
类目: Machine Learning (cs.LG); Audio and Speech Processing (eess.AS)
*备注: 41 pages, 21 figures, 9 tables
Abstract:As air traffic volumes in the National Airspace System continue to expand, in particular in the low altitude airspaces, the need for scalable decision support tools used by air traffic controllers will also require more development. This article introduces Voice-to-Trajectory for Air Traffic Control, a joint voice communication-flight trajectory data embedding framework, that can be a component of situational awareness in congested airspaces, and assist the development of tools for ATC as they reason in real-time over Automatic Dependent Surveillance-Broadcast trajectories, or the intent expressed by pilots in natural language. We show that these data modalities are not independent and represent a common physical referent: an aircraft flying through the airspace. V2TATC maps a voice instruction and the trajectory of the addressed aircraft to nearby points in a single latent space that can be queried in both directions. It combines a self-supervised trajectory encoder, a frozen large-scale speech encoder, a contrastive joint embedding, and a bijective lifting via normalizing flows. We demonstrate V2TATC’s effectiveness on the San Francisco Bay Area, for its concentration of major airports, and its mix of commercial and general aviation low altitude traffic. Lastly, we release a novel paired voice-trajectory dataset, and report experiments on cross-modal retrieval, ablations, and latent-space analysis.
[LG-122] Brain-Language-Action (BLA) Models: Language-Conditioned EEG for Robotics Control
链接: https://arxiv.org/abs/2608.28967
作者: Alexandr Plashchinsky
类目: Robotics (cs.RO); Machine Learning (cs.LG)
*备注:
Abstract:Electroencephalography (EEG)-based robotic control is commonly formulated as a direct classification problem, in which electrical neural signals are mapped to a fixed set of discrete actions. However, the limited separability and high noise of EEG signals make it difficult to scale this approach to fine-grained robotic control spaces. We introduce Brain-Language-Action (BLA) models, a framework in which language conditions the interpretation of neural representations for robotic action generation. In a BLA, a small set of reliably distinguishable brain states can be dynamically associated with different actions through a language-defined control mapping, allowing a small number of neural classes to apply to a larger global action space. We develop a proof-of-concept BLA for drone control using motor-imagery EEG from the BCI Competition IV 2a dataset. The system is trained in two stages. First, we evaluate multiple candidate EEG encoder architectures using subject-specific four-class motor-imagery classification, converting 250Hz, 3.5-second, 22-channel EEG samples into five 128-dimensional brain-token embeddings. Second, these embeddings are projected into the embedding space of a pretrained large language model (LLM) and jointly fine-tuned with language instructions to autoregressively generate structured three-token drone actions. Across 840 possible language-defined mappings between four neural states and seven flight action combinations, the resulting BLA achieves 90% per-token accuracy during evaluation. These results provide an initial demonstration that language conditioning can expand the effective control range of EEG-based robotic interfaces without requiring a corresponding increase in the number of directly distinguishable neural states. Subjects: Robotics (cs.RO); Machine Learning (cs.LG) Cite as: arXiv:2608.28967 [cs.RO] (or arXiv:2608.28967v1 [cs.RO] for this version) https://doi.org/10.48550/arXiv.2608.28967 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-123] Continuity-Free Near-Minimax Leading-Order Regret for CVaR-UCBVI
链接: https://arxiv.org/abs/2608.28960
作者: Yuanlong Chen
类目: Machine Learning (cs.LG)
*备注: 12
Abstract:For finite-horizon tabular CVaR reinforcement learning, prior work proves a \widetildeO(\tau^-1\sqrtSAK) leading regret bound for arbitrary normalized return laws and the sharper \widetildeO(\sqrtSAK/\tau) rate under a density lower bound. We show that the same Bernstein CVaR-UCBVI algorithm attains the sharper rate without continuity assumptions. The key is a selected-budget self-bound: the conditional variance of the episode shortfall is at most \tau plus the value-estimation width. Substitution into the original Bernstein decomposition yields, with high probability, \widetildeO(\sqrtSAK/\tau+(SAHK^1/4+S^2AH)/\tau) regret for arbitrary normalized return laws, including atomic, mixed, and continuous laws. The \tau^-1/2 leading term matches the expected-regret minimax lower bound up to logarithmic factors. Thus Bernstein CVaR-UCBVI is minimax-optimal over the full return-law class in the leading-order regime; the lower-order terms retain their \tau^-1 dependence.
[LG-124] Revisiting the Provable-Auditable Privacy Gap of DP-SGD
链接: https://arxiv.org/abs/2608.28934
作者: Saloni Modi,Srivi Balaji,Yusong Zhu,Gautam Kamath,Kevin Tian
类目: Machine Learning (cs.LG)
*备注: The code for this paper can be found here: this https URL
Abstract:Differential privacy (DP) has traditionally been used to provide theoretical upper bounds on an algorithm’s stability to changing its training data. In modern private machine learning applications, achieving strong tradeoffs between utility and theoretical privacy is challenging, and thus one may optimistically hope that existing theoretical privacy analyses are loose. Recent work on privacy auditing has adopted a dual viewpoint, instead lower bounding the true privacy of an algorithm by constructing empirical distinguishing events. The auditing literature has thus far yielded a pessimistic outlook on the looseness of theoretical privacy bounds for DP-SGD, the de facto private training method in modern ML, as nearly-matching empirical lower bounds have been achieved under various threat models [NHSBTJCT23, AC24, CBP25]. In this work, we propose the empirical privacy lower bound of an algorithm as a concrete metric to optimize for, complementary to the theoretical upper bound. We give a lightweight defense framework that generically augments optimization methods in the ML pipeline to have significantly-improved empirical privacy on standard benchmarks. Moreover, we show that our framework comes at no theoretical privacy cost when augmenting DP-SGD, unlike previously-proposed defenses against membership inference attacks. We evaluate our defense against a broad range of audit constructions, models, and datasets to demonstrate its flexibility. Comments: The code for this paper can be found here: this https URL Subjects: Machine Learning (cs.LG) Cite as: arXiv:2608.28934 [cs.LG] (or arXiv:2608.28934v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2608.28934 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-125] RankShift: In-Database Detection and Explanation of Categorical Shifts
链接: https://arxiv.org/abs/2608.28922
作者: Omair Shafi Ahmed
类目: Machine Learning (cs.LG)
*备注:
Abstract:A login service can receive its usual number of failed sign-ins while one source grows from 2% to 30% of them. The same pattern appears in system logs when a rare event template becomes common while the message rate stays stable. These events change which categories are active without changing how many events occur. RankShift detects such changes inside the analytical database that stores the data. It compares each window’s category shares with a benign reference using a Pearson score whose terms identify the categories responsible for the change. The same query returns the score, calibrated alert, and largest increasing contributions. We evaluate RankShift on HDFS, BGL, and Thunderbird. It matches the count-vector autoencoder within 0.001 AUROC on HDFS (0.999 versus 1.000) and leads on Thunderbird (0.983 versus 0.949). In a controlled fixed-volume experiment, RankShift detects rare-category shifts that are invisible to event-count monitoring, reaching 0.787 AUROC compared with 0.771 for the autoencoder. Across all three corpora, observed false-alarm rates track the requested operating levels. RankShift requires no model training or inference service, and the autoencoders deployed state is 137x larger.
[LG-126] Learning-Theoretic Foundation for General Coded Computing: The Strag gler Setting
链接: https://arxiv.org/abs/2608.28910
作者: Parsa Moradi,Behrooz Tahmasebi,Mohammad Ali Maddah-Ali
类目: Machine Learning (cs.LG)
*备注:
Abstract:Coded computing has emerged as a powerful paradigm for mitigating the impact of straggling workers in distributed computing systems. However, existing coded-computing schemes are predominantly designed for the exact recovery of highly structured computations, such as polynomial evaluation and matrix multiplication, and typically rely on strict recovery thresholds. These assumptions significantly limit their applicability to modern machine-learning workloads, particularly deep neural networks (DNNs), whose computations generally lack rigid algebraic structure and, in many applications, require only accurate approximations rather than exact recovery. To address this gap, we revisit coded computing from a learning-theoretic perspective and introduce General Coded Computing (GCC). Rather than adopting existing algebraic tools, GCC formulates coded computing through a natural end-to-end mean-squared error loss that directly measures the discrepancy between the desired computations and their recovered estimates. By deriving suitable upper bounds and restricting the encoder and decoder to a reproducing kernel Hilbert space (RKHS) with mild smoothness constraints, we show that both the encoder and decoder admit specific representations as linear combinations of RKHS kernel functions. This representation allows the corresponding coefficients to be computed efficiently. Moreover, this framework enables us to establish theoretical performance guarantees for GCC under two complementary straggler regimes. In the worst-case setting with N worker nodes, and at most S stragglers, we show that the end-to-end loss decays at least at rate O(S^3N^-3) for standard configurations. We then study a probabilistic setting in which each worker independently straggles with probability p . We prove that the expected loss can still converge at rate O(\log_1/p^3(N)N^-3) . Subjects: Machine Learning (cs.LG) Cite as: arXiv:2608.28910 [cs.LG] (or arXiv:2608.28910v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2608.28910 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-127] Off-Policy Evaluation for Semantic ID Recommenders: Does the Models Own Code Hierarchy Help?
链接: https://arxiv.org/abs/2608.28905
作者: Artem Betlei
类目: Machine Learning (cs.LG)
*备注:
Abstract:Generative recommenders increasingly emit semantic IDs (SIDs): each item is a short sequence of hierarchical discrete codes from a residual quantizer, decoded autoregressively. Before spending scarce A/B-test, a team may decide offline which decoder or reranking variants are worth testing - a job for off-policy evaluation (OPE). We ask a simple question: can the model’s own SID tree serve as the action abstraction for that OPE? Our answer has three parts. (i) Under the near-argmax logging real recommenders use, per-item OPE is hopeless - as item-level effective sample size is usually small on production logs - but marginalizing items to code-prefix clusters restores estimable support and cuts error. (ii) This gain is thanks to coarsening, not to the hierarchy specifically; but the SID tree is what makes coarsening feasible in a generative system - each cluster’s mass is exactly and cheaply returned by the decoder, whereas flat clustering requires enumerating item/leaf masses that a code-only decoder does not directly expose. (iii) Resolution depth is the operative knob - coarser under scarce support - and a conditional bias bound links the coarsening bias to the quantizer’s worst-case reconstruction residual and the target-logging divergence.
[LG-128] Conservative Hybrid Graph Networks for Process Systems with Learned Routing
链接: https://arxiv.org/abs/2608.28896
作者: Paolo Guida
类目: Machine Learning (cs.LG)
*备注:
Abstract:Industrial process networks do not maintain a single effective topology while operating: streams are throttled or bypassed, and units move between idle, transition, and active regimes. Models of such systems are typically trained on measured state trajectories while the operating mechanisms that generated them remain latent, and an unconstrained graph network can fit such a trajectory without assigning stable physical meaning to the recovered routing. We address both problems with the Conservative Hybrid Graph Network (CHGN), which learns routing, regime assignment, and removal rates as data-driven surrogates and inserts them into a fixed transport equation, so that the mass balance holds by construction for any predicted routing. CHGN trained on networks of 10-20 nodes transfers zero-shot to unseen graphs of 25-40 nodes without retraining, reaching an RMSE of 2.1e-3 against 6e-2 to 9e-2 for GNN baselines under the same protocol, with a gate MAE of 7.9e-3 and regime accuracy of 94.3% (1.2e-2 and 96.4% respectively on the fixed training topology). On a fluid-mixing pilot plant, CHGN improves on a persistence baseline for held-out physical faults but does not predict manual interventions, for which the governing valve actions are unobserved. The model therefore transfers across process topologies without retraining and exposes the latent mechanisms governing plant behaviour to inspection.
[LG-129] Uncertainty-Aware Multi-Task Learning for Joint Modulation Recognition and SINR Estimation
链接: https://arxiv.org/abs/2608.28865
作者: Kosar Nourolahi,Vahid Ghasemi
类目: Networking and Internet Architecture (cs.NI); Machine Learning (cs.LG)
*备注: 4 pages, 6 figures
Abstract:Joint modulation recognition and signal-to-interference-plus-noise ratio (SINR) estimation can reduce duplicated processing in intelligent receivers, but the two tasks have different uncertainty characteristics. This letter proposes an uncertainty-aware multi-task model that transforms each short normalized in-phase/quadrature window into 36 deterministic, label-free statistics, learns a shared representation, and uses task-specific adapters for modulation classification and heteroscedastic SINR regression. A joint uncertainty score combines classification entropy and predicted regression variance to support selective inference. Simulations cover QPSK, 8PSK, 16QAM, and 64QAM under matched additive white Gaussian noise/Rayleigh channels and an unseen frequency-selective Rician channel. Over five independent seeds, the proposed model improves matched and unseen-channel accuracy over conventional multi-task learning by 14.86 and 8.61 percentage points, respectively, while reducing SINR mean absolute error by 1.60 and 1.61 dB. Confidence-based rejection further lowers modulation error under channel mismatch.
[LG-130] Unsupervised Latent Space Alignment with Hyperspherical Geodesic Matching
链接: https://arxiv.org/abs/2608.28840
作者: Cameron Ryan,Vivek Sivaraman Narayanaswamy,Kowshik Thopalli,Shusen Liu
类目: Machine Learning (cs.LG)
*备注:
Abstract:Independently trained neural networks tend to encode the same data with similar latent geometries. These latent geometries are not directly compatible, yet they can be nearly the same up to some class of transformations. While there exists many methods for alignment between different latent spaces, it is typically done using a set of shared sample correspondences, known as anchors. This leaves a fundamental question: are the geometric signatures of different latent spaces representing similar data sufficient to recover an alignment between them? To that end, we introduce HGA (Hyperspherical Gaussian Alignment), a method that directly optimizes a transformation between two latent spaces by maximizing a geometric measure of “fit” between them. Since it is driven by the geometry of the latent spaces rather than paired data, HGA can operate in both an unsupervised and weakly supervised regime. On tasks such as model stitching or multilingual word embedding correspondence recovery, HGA manages to match supervised results with minimal or no supervision.
[LG-131] Separable Nonnegative Matrix Factorization Using Powered Ratio-of-Norms Regularization
链接: https://arxiv.org/abs/2608.28799
作者: Matthew McCarver,Jing Qin
类目: Numerical Analysis (math.NA); Machine Learning (cs.LG); Optimization and Control (math.OC)
*备注:
Abstract:Separable nonnegative matrix factorization (SNMF) has been widely used for low-rank representation and clustering of nonnegative data, owing to its ability to produce part-based and interpretable decompositions. In particular, SNMF is closely related to graph clustering and community detection. To enhance sparsity and identifiability of the learned factors, we propose an \ell_1^p/\ell_2 -regularized SNMF model based on a powered ratio-of-norms regularizer. The resulting formulation is nonconvex and nonsmooth, which poses significant challenges for optimization. To address this, we develop efficient algorithms based on the difference-of-convex function algorithm (DCA) and the alternating direction method of multipliers (ADMM). The proposed methods decompose the original problem into tractable subproblems, leveraging closed-form proximal operators associated with the powered norm terms. We establish descent and limiting criticality properties for the DCA scheme and convergence under standard assumptions for the ADMM scheme. Extensive numerical experiments on synthetic datasets and hand gesture classification tasks demonstrate that the proposed approach achieves competitive or improved performance in anchor identification and classification accuracy compared with existing SNMF methods, while maintaining competitive computational efficiency.
[LG-132] ERR: Sequential Entropy Resolution for Efficient and Decisive LLM Reasoning
链接: https://arxiv.org/abs/2608.28771
作者: Xin Jiang,Minhao Wang,Wen Wu,Zhentao Xie,Shangheng Du,Jinxin Shi,Jiabao Zhao
类目: Machine Learning (cs.LG)
*备注: 16 pages, 5 figures
Abstract:Large reasoning models achieve strong performance on complex tasks by generating extended chain-of-thought (CoT) traces via reinforcement learning with verifiable rewards (RLVR). While current RLVR methods have achieved strong results with correctness-based reward signals, they provide limited guidance on the quality of the reasoning process itself, leaving the internal reasoning structure largely unoptimized. Through empirical analysis across multiple model families, we identify a consistent pattern: correct reasoning trac es exhibit more frequent and larger token-level entropy drops within the thinking phase than incorrect ones. We propose ERR+, a two-phase RLVR framework grounded in this observation. The first phase trains with the Entropy Relief Reward (ERR), a bonus proportional to cumulative token-level entropy drops in the thinking phase, log-normalized by response length. Unlike prior methods that suppress entropy, ERR rewards the resolution of uncertainty while leaving exploratory high-entropy states unconstrained. The second phase introduces the Robust Relative Efficiency Reward, which scores each response’s length against co-generated peers via a \tanh -transformed within-group z -score. We provide a formal analysis showing that joint optimization of the two objectives induces gradient conflict in early training, motivating the sequential design . Experiments on five datasets demonstrate consistent improvements in both accuracy and response conciseness across model backbones. Our code is available at this https URL
[LG-133] Generation of High-Level Concepts in 3D Scene Graphs via Autoregressive Diffusion
链接: https://arxiv.org/abs/2608.28733
作者: Jose Andres Millan-Romera,Samuel Cognolato,Holger Voos,Jose Luis Sanchez-Lopez,Luciano Serafini
类目: Robotics (cs.RO); Machine Learning (cs.LG)
*备注:
Abstract:Indoor 3D Scene Graphs (3DSGs) represent environments as multi-layer hierarchies that connect observed geometric primitives (e.g., planes) to higher-level metric-semantic concepts (e.g., rooms, floors, buildings), enabling incremental spatial reasoning for robotic perception and SLAM. However, classical high-level concept generation approaches rely on hand-crafted rules for specific concept classes, while learning-based methods require separate models for graph structure and spatial node features (e.g., centroids), which limits scalability to novel classes and more complex hierarchies. We propose a unified autoregressive diffusion-based graph generative model that jointly learns structure and features, constructing complete 3DSGs bottom-up from observed vertical planes across arbitrary hierarchy depths. Our method consistently surpasses all learning-based and random baselines across 3DSG datasets spanning synthetic scenes, real architectural floor plans, and robotic sensor data, with varying layout complexity and hierarchy depth, and surpasses a one-shot model with oracle access to the target graph size on the largest hierarchy and on real single-floor data. Finally, we propose an adaptation of the Fused Gromov–Wasserstein distance for principled graph-level evaluation of generated 3DSGs against ground truth.
[LG-134] Privacy-Preserving Detection of Rare Disease-Associated Cell Subsets via Secure Multi-Party Computation
链接: https://arxiv.org/abs/2608.20118
作者: Ş. Selcan Magara,Esther Havemann,Debora Jutz,Ali Burak Ünal,Mete Akgün
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注: Accepted at the CIBB 2026 conference ( this https URL )
Abstract:The detection of rare disease-associated cell subsets from high-dimensional single-cell measurements is critical for understanding diseases such as leukaemia and viral infections. CellCnn, a convolutional neural network (CNN) designed for this task, has demonstrated the ability to identify phenotype-associated cell populations at frequencies as low as 0.01%. Training such models reliably requires patient cohorts that are larger and more diverse than any single institution can typically assemble, and the underlying single-cell data is too sensitive to share across institutional boundaries under existing privacy regulations. We propose a secure multi-party computation (MPC) framework that enables the training and inference of CellCnn entirely on secret-shared data. This ensures that neither the participants nor the computing servers ever observe raw patient data or intermediate values. Evaluated on benchmark single-cell datasets for cytomegalovirus infection (CMV) and acute myeloid leukaemia (AML), our implementation preserves accuracy close to its plaintext counterpart while outperforming the prior privacy-preserving baseline. In contrast to earlier privacy-preserving approaches that removed components such as ReLU activations and bias terms, our method retains these key parts of the CellCnn architecture and supports accurate analysis without exposing raw patient data.
[LG-135] Implementing neural network mixed-effects models in Template Model Builder (TMB)
链接: https://arxiv.org/abs/2608.31133
作者: Nan Zheng,Hoi Yiu Cheung,Vibhu Sharma,James T. Thorson,Noel G. Cadigan
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:Neural network mixed-effects models (NMMs) have gained traction by combining the strong representation and predictive power of artificial neural networks with the capacity of mixed-effects modeling to capture complex correlation structures. However, existing estimation approaches rely heavily on manual derivations of objective functions and gradients, which inherently forces simplifying approximations and severely constrains the complexity and accuracy of NMMs. In this work, we introduce a general framework for implementing NMMs using Template Model Builder (TMB). By leveraging automatic differentiation and Laplace approximation, TMB requires users to specify only the negative joint log-likelihood and any regularization terms. The framework automatically integrates out random effects and evaluates the marginal objective function alongside its exact gradients, eliminating the need for manual derivations or ad hoc approximations. We demonstrate the efficiency, flexibility, and statistical performance of TMB-based NMMs across two numerical examples, including an application to monotonic NMMs. Reproducible code is provided to facilitate broader adoption.
[LG-136] “Train classical deploy quantum” requires rethinking generalization
链接: https://arxiv.org/abs/2608.31117
作者: Snehal Raj,Natansh Mathur,Alejandro Perdomo-Ortiz
类目: Quantum Physics (quant-ph); Machine Learning (cs.LG)
*备注: 22 pages, 16 figures, 9 tables
Abstract:Generative models have become central across science and industry, from image and text synthesis to the design of molecules and materials. Quantum generative models are considered one of the most promising applications for quantum computers, since a quantum circuit naturally produces samples from the distribution it encodes, and for suitable circuits that distribution is believed to be hard for any classical computer to reproduce. A leading strategy trains these models on a classical computer and reserves the quantum device for generating samples at deployment. This is possible when the training loss can be evaluated on a classical computer. A prime example is the maximum mean discrepancy (MMD ^2 ), a moment-matching loss that compares the model and the data through their Pauli- Z correlations. Research so far has asked whether such models can be trained and whether their sampling is hard; whether minimizing such an objective yields a model that generalizes, rather than one that merely reproduces the training statistics, remains poorly understood. We benchmark a broad set of quantum and classical generative models by direct sampling and show that models trained with a moment-matching loss generally show worse generalization than the likelihood-trained models. We show this on two application-inspired datasets: first a cardinality-constrained dataset at up to 30 qubits and second a dataset of genomic single-nucleotide variants, whose valid set is the observed data. These results indicate that a converged moment-matching loss is not a reliable measure of generalization, and that train-classical, deploy-quantum workflows will need approaches that target generalization directly, leaving open whether better training objectives suffice or whether the model architectures themselves must change.
[LG-137] Learning the Geometry of Admissible Hypotheses through Inductive Bias in Training Distributions
链接: https://arxiv.org/abs/2608.31028
作者: James Crowley,Faez Ahmed,Anton van Beek
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:Scientific discovery often requires reasoning over competing hypotheses that are consistent with experimental observations. For mixed-variable and combinatorial hypothesis spaces, however, constructing probabilistic representations remains challenging because both the active model components and their associated parameters are unknown. In this work, we present a framework for learning continuous latent representations of admissible partial differential equations (PDEs) by embedding a scientific inductive bias directly into the training distribution. Progressively richer structural principles (e.g., sparsity, logical dependencies, common PDE families, and physical admissibility) are used to generate a structured distribution of hypotheses from which a gated variational autoencoder learns a continuous latent manifold. Experimental results show that the resulting 11-dimensional representation accurately reconstructs a broad collection of representative PDEs, while exhibiting smooth geometric transitions both within and across equation families. Through an ablation study we further demonstrate that introducing scientific principles reduces both structural misclassifications of equation forms and parameter estimation errors when reconstructing a representative benchmark set of admissible partial differential equations. These results show that embedding a scientific inductive bias in the training distribution enables the learning of compact and geometrically meaningful hypothesis manifolds, providing a principled foundation for future inference over competing governing equations.
[LG-138] Uncertainty-Aware End-to-End AI Weather Forecasting: Disentangling Observation and Model Contributions
链接: https://arxiv.org/abs/2608.30795
作者: Rodrigo Almeida,Noelia Otero,Jost Arndt,Simon Baur,Wojciech Samek,Jackie Ma
类目: Atmospheric and Oceanic Physics (physics.ao-ph); Machine Learning (cs.LG)
*备注: 17 pages, 7 figures
Abstract:End-to-end weather forecasting systems produce skillful global gridded and station forecasts directly from raw Earth observations, replacing the numerical weather prediction pipeline, including data assimilation, at a fraction of its cost. These systems are deterministic and issue no uncertainty. Here we render the Aardvark Weather model probabilistic by attaching one stochastic mechanism to each component: learned, input-dependent noise at the observation encoder, capturing aleatoric uncertainty inherited from the observing system, and Monte Carlo dropout in the processor, capturing epistemic uncertainty in the learned dynamics. The resulting nested ensemble attributes forecast spread to the two sources through a law-of-total-variance decomposition, cross-checked by withholding observation streams. Probabilistic finetuning significantly improves the mean forecast, by 4.2% on average across variables and lead times. The ensemble is calibrated against ERA5 through the medium range (spread-skill ratio 0.98), keeps station RMSE within 2.4% of the deterministic model while beating it in CRPS at every lead time, and trails the operational ECMWF ensemble. The encoder branch behaves as observation-driven uncertainty. Component-attributed uncertainty makes end-to-end forecasts more transparent, a step toward observation-driven digital twins of the atmosphere.
[LG-139] Informative Label Missingness in Multiclass Classification Information Geometry and Excess Risk
链接: https://arxiv.org/abs/2608.30561
作者: Fariborz Setoudehtazang,Geoffrey J. McLachlan
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:Informative label missingness can change the usual efficiency ordering between completely and partially labelled classifiers because the pattern of missing labels may itself carry information about the classification model. We develop a general likelihood-based theory for this phenomenon in parametric multiclass classification. An efficient-information decomposition separates information lost through unavailable class memberships from information contributed by the missing-label mechanism. We then derive a quadratic expansion of plug-in excess risk over the active pairwise faces of the multiclass Bayes boundary, showing that classification efficiency depends on how information gains and losses align with directions that perturb the decision boundary. This yields a classification-weighted generalized-eigenvalue criterion under which informative partial classification may have smaller asymptotic classification risk without globally dominating complete classification in Fisher information. Near missing completely at random, with the marginal missing-label proportion fixed, redistribution of missing labels changes lost class-label information at first order, whereas efficient information from the missingness pattern appears only at second order. Three-class quadratic discriminant calculations, finite-sample experiments, and a semi-synthetic multiclass application illustrate the resulting regime-dependent behaviour.
[LG-140] Beamforming Design Via GNN in mmWave Cell-Free Massive MIMO Using Sub-6 GHz CSI
链接: https://arxiv.org/abs/2608.30524
作者: Sina Tavakolian,Abolfazl Zakeri,Ahmed Alkhateeb,Markku Juntti,Nhan Thanh Nguyen
类目: ignal Processing (eess.SP); Machine Learning (cs.LG)
*备注: 5 pages, 3 figures, Accepted for presentation at the 2026 IEEE 27th International Workshop on Signal Processing and Artificial Intelligence in Wireless Communications (IEEE SPAWC 2026)
Abstract:Beamforming methods in millimeter-wave (mmWave) cell-free massive multiple-input multiple-output (CFmMIMO) systems require accurate channel state information (CSI), whose acquisition entails significant training overhead. This paper shows that fully digital cell-free mmWave beamforming can be effectively learned from sub-6 GHz CSI using a graph neural network (GNN). Specifically, we represent a CFmMIMO system as a wireless graph, and the GNN is trained to approximate beamformers that maximize the downlink sum-rate based on the available sub-6 GHz CSI. A message-passing mechanism is proposed to capture inter-user interference and inter-base-station cooperation across different network topologies. Simulation results demonstrate that the proposed sub-6 GHz-assisted GNN-based beamformer achieves competitive and often superior sum-rate performance compared to classical baselines that rely on full mmWave CSI.
[LG-141] End-to-End Neural Shrinkage of Indefinite Pairwise Correlation Matrices for Small-Cap-Inclusive Portfolios
链接: https://arxiv.org/abs/2608.30446
作者: Christian Bongiorno,Lorenzo Villassero
类目: Portfolio Management (q-fin.PM); Machine Learning (cs.LG); Statistical Finance (q-fin.ST)
*备注:
Abstract:Small-cap-inclusive equity universes contain recently listed and intermittently traded securities, so enforcing a common look-back discards a substantial fraction of the available information. Pairwise-complete estimation preserves the longest overlap for each asset pair, but the resulting correlation matrix can be indefinite because its entries are computed on different samples. This prevents direct use in Markowitz optimization and falls outside the assumptions of standard random-matrix shrinkage. We adapt a rotation-invariant neural covariance estimator to this setting. The model computes mask-aware marginal moments and a pairwise correlation matrix proxy, processes its signed spectrum, and uses a bidirectional gated recurrent unit conditioned on factor-aligned effective sample lengths derived from the overlap matrix and eigenvector loadings. It maps all eigenvalues, including negative ones, to a positive inverse spectrum. The reconstructed covariance is positive definite and is trained end-to-end to minimize five-day realized global-minimum-variance risk. We evaluate 26 expanding-window models from 2000 to 2025 on up to 1,500 U.S. equities in a closing-auction simulator with point-in-time selection, commissions, financing, corporate actions, and market impact. Across the 26-year out-of-sample period, the neural estimator reduces annualized five-day volatility by approximately 20% and increases the Sharpe ratio by approximately 40% relative to the next-best covariance estimator. These improvements are consistent across realized risk, risk-adjusted performance, and drawdown control, remain after the modeled execution frictions, and are supported by a 99.9% Model Confidence Set that retains only the neural estimator.
[LG-142] Benchmarking External Generalization of SPD Matrix Learning for Resting-State fMRI Connectome Prediction
链接: https://arxiv.org/abs/2608.30418
作者: Ce Ju,Antoine Collas,Florent Bouchard,Bertrand Thirion
类目: ignal Processing (eess.SP); Machine Learning (cs.LG)
*备注: 33 pages, 5 figures
Abstract:Resting-state functional magnetic resonance imaging (rs-fMRI) functional connectivity (FC) matrices are widely used for individual-level prediction, but strong performance within one cohort may not generalize to a new cohort. We ask whether within-dataset performance remains when the test data come from an entirely held-out rs-fMRI dataset. Each scan is represented as a regularized symmetric positive definite (SPD) correlation connectome, which allows methods to use the geometry of the SPD manifold. We introduce a reproducible age-prediction benchmark across six rs-fMRI datasets: COBRE, ADNIDOD, Cam-CAN, ABIDE, OASIS-3, and ADNI. The benchmark compares a vectorized correlation baseline, Tangent-Space Ridge, SPDNet, and split-wise Riemannian harmonization under within-dataset GroupKFold, pooled GroupKFold, and leave-one-dataset-out (LODO) evaluation. Within-dataset and pooled GroupKFold results are substantially more favorable than LODO results. When an entire dataset is held out, prediction error increases, differences among methods narrow, and performance is strongly affected by age-range mismatch and cohort heterogeneity. The benchmark provides common inputs, model settings, data splits, and analysis scripts so that future SPD matrix learning methods can be evaluated under the same external-validation protocol.
[LG-143] Compact and Infinite-Order Error Analysis for Null-Space SVD Estimation
链接: https://arxiv.org/abs/2608.30374
作者: Xin Li,Jonathan Cohen,Rami Puzis
类目: atistics Theory (math.ST); Machine Learning (cs.LG)
*备注: 31 pages
Abstract:We study null-space estimation from a noisy matrix. For a simple left null space, we first derive an exact compact expression for the error of the smallest left singular vector. We then give an all-order series for the SVD vector and projector, followed by compact and consistently truncated series forms for the fixed-realization empirical risk and conditional population generalization risk. The recursion extends to a multiple-dimensional null space by following the complete invariant subspace. The convergence radius is not inferred from an error plot: it is computed independently from the nearest complex exceptional point that joins a retained eigenvalue branch to its complement. A reduced-nullity experiment shows that moving this spectral boundary can increase the radius, although the improvement is not monotone in the retained nullity. For individually ordered null directions under Gaussian training with (\tau\geq m), we prove that the Wishart splitting matrix (W) gives a strict second-order empirical ranking. Gaussian averaging equalizes the leading generalization risks at both small and very large noise, while a column-swap theorem proves strict expected generalization ranking for an isotropic signal subspace. For unequal spikes, an exact population-overlap criterion and a simultaneous (99%) Monte Carlo confidence certificate explain the observed intermediate ranking. A sixth-order risk correction improves the lower-crossover estimate in the reported experiment. This equal–ranked–equal phenomenon is a finite-sample diagnostic related to spectral mixing, but its tolerance crossings, the exceptional-point radius, and the asymptotic BBP threshold are three distinct quantities.
[LG-144] Estimating Population-Risk Curves Along Nonconvex Gradient Flows from the Training Sample
链接: https://arxiv.org/abs/2608.30261
作者: Mingzhi Song
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注: 63 pages
Abstract:We estimate the conditional population-risk curve of a realized smooth nonconvex gradient flow from the training sample. Flow approximate leave-one-out (Flow-ALO) propagates a deletion response and evaluates omitted observations at approximate deleted paths. The risk-curve error decomposes into response approximation, exact-LOO fluctuation, and deletion-to-full risk transfer. On each fixed finite horizon, bounded centered training-loss gradients, a one-sided Hessian lower bound, locally Lipschitz Hessians, and a strict tube-closure condition yield an explicit (n-1)^-2 bound for the deletion-response error. Bounded evaluation-loss gradients transfer the deletion-response bound to the score without requiring the Hessian to be invertible. Direct first-order jackknife cancellation and exact-LOO concentration control deletion-to-full risk transfer and fluctuation, respectively, completing recovery of the conditional population-risk curve. For bounded smooth two-layer mean-field networks training both layers, the score-error bound is uniform in width.
[LG-145] A Borel Concept Class of VC Dimension One with a Non-PAC Consistent Learner in ZFC
链接: https://arxiv.org/abs/2608.30246
作者: Mateus Jesus de Arruda Campos,Gabriel Fernandes,Vinicius de Oliveira Rodrigues
类目: Logic (math.LO); Machine Learning (cs.LG); Probability (math.PR)
*备注:
Abstract:The fundamental theorem of statistical learning states that, under suitable measurability assumptions, finite Vapnik–Chervonenkis (VC) dimension guarantees that every proper consistent learning rule is probably approximately correct (PAC). Blumer, Ehrenfeucht, Haussler, and Warmuth showed, assuming the Continuum Hypothesis, that the “well-behavedness” condition of the concept class cannot be omitted: they constructed a concept class of Borel sets of VC dimension one admitting a consistent learning rule that is not PAC. We show that the Continuum Hypothesis is unnecessary. Working in Zermelo–Fraenkel set theory with the Axiom of Choice (ZFC) alone, we construct a concept class of Borel sets on [0,1] of VC dimension one and a proper consistent learning rule that is not PAC. More precisely, for a suitable Borel probability measure and target concept, the rule has true risk one at every sample size on a set of samples of outer probability one. Consequently, finite VC dimension and Borel measurability of the individual concepts do not suffice to guarantee that every proper consistent learning rule is PAC. The result shows, with no need of extra set-theoretical assumptions, that the additional regularity assumption in the fundamental theorem cannot in general be omitted.
[LG-146] Fairness in multi-class multi-group classification problems via contextial coherent risk measures
链接: https://arxiv.org/abs/2608.30223
作者: Darinka Dentcheva,Xiangyu Tian
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:We propose a new design of fair classifiers for multi-class classification problems in the presence of vector-valued sensitive attributes. In that scenario each sensitive attribute has multiple values and forms several groups relevant to the fairness consideration. Naturally those groups are overlapping and one should also analyze the interaction of factors. Additionally, the decision makers aided by the classification should not violate individual rights at the expense of satisfying fairness metrics at the group level. We propose an approach using the theory and methods of coherent measures of risk aiming at resolving the fairness challenges. Further, we propose a specialized numerical method for solving the resulting optimization problem. The method scales well with the increase of the number of observations. Additionally, we note that the obtained classifier is robust with respect to corrupted data or to situation when data is scarce. We demonstrate the advantages of the proposed framework in comparison to the support-vector machine framework and other methods handling fairness.
[LG-147] Learning Representations through Token Prediction: Geometry Approximation and Downstream Guarantees
链接: https://arxiv.org/abs/2608.30072
作者: Shulei Wang
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Statistics Theory (math.ST)
*备注:
Abstract:Token prediction is a central pre-training objective for modern language models. Despite its empirical success, why token prediction learns broadly useful representations remains incompletely understood. We develop a statistical framework connecting token prediction with representation geometry, encoder approximation, and downstream performance. Under a softmax prediction head, we show that accurate token prediction organizes token embeddings according to similarities between the distributions of contexts in which different token types appear, as measured by Hellinger distance, with explicit errors governed by prediction accuracy and token frequency. Meanwhile, the contextual representation provides a low-dimensional coordinate for the conditional distribution of the target token relative to these embeddings. We further introduce a self-consistency principle showing that repeated applications of a shared representation block can progressively refine the contextual representation without introducing additional block parameters. Among representations with the same prediction accuracy, this recurrent construction favors those that can be stably reconstructed from their contexts. Finally, we establish downstream guarantees for token generation, token community recovery, and classification by a linear probe, showing how prediction accuracy and recovered geometry translate into performance beyond the pre-training objective. Together, these results explain how the simple objective of predicting tokens can recover semantic geometry and produce broadly useful representations. A controlled simulation illustrates the theoretical mechanisms.
[LG-148] A Deep Latent Variable Framework for Jointly Modeling Missingness Measurement Error and Heterogeneity
链接: https://arxiv.org/abs/2608.30040
作者: Yasin Khadem Charvadeh,Grace Y. Yi,Mithat Gönen,Pouya Faroughi
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注: 24 pages, 4 figures
Abstract:Missing data, measurement error, and population heterogeneity are pervasive challenges in analyzing data arising from modern observational studies and machine learning applications. Although these problems frequently coexist and interact, they are often treated separately in existing works. We propose a unified probabilistic framework that jointly addresses these issues utilizing deep latent variable representation. The proposed method integrates a novel hierarchical tree-routed variational autoencoder with pattern-aware latent representations and calibration-based denoising. The framework accommodates missing data mechanisms, including MCAR, MAR, and MNAR, while simultaneously learning subgroup-specific and globally shared latent structure. The introduced reconvergent routing mechanism enables selective parameters to be shared across related subpopulations, which offers flexibility as well as improved statistical efficiency. Simulation studies demonstrate substantial improvements over existing deep generative imputation approaches under complex heterogeneous missingness and measurement-error settings. The proposed framework provides a principled approach for learning from noisy and incomplete data in modern healthcare and other high-dimensional applications.
[LG-149] Data-Driven Design Optimization of Streaming-Potential-Mediated Electrokinetic Transport of Viscoelastic Fluids in Microchannels
链接: https://arxiv.org/abs/2608.29939
作者: Ankan Basu,Sumanta Banerjee
类目: Fluid Dynamics (physics.flu-dyn); Machine Learning (cs.LG)
*备注:
Abstract:Streaming-potential-mediated transport of viscoelastic fluids has attracted research attention owing to its applications in electrokinetic energy conversion and microfluidic transport. Existing analytical and semi-analytical models in published literature provide valuable physical insights, but require repeated numerical evaluations for exploring large design spaces and identifying the optimal operating conditions. In this work, a surrogate-assisted framework is developed for rapid design optimization of pressure-driven electrokinetic transport of simplified Phan-Thien-Tanner fluids in a slit microchannel. A high-fidelity numerical database is generated over a broad range of governing dimensionless parameters, which includes the zeta potential, the Debye parameter, the Dukhin number, and the viscoelastic parameter. A Machine Learning surrogate model is subsequently trained to accurately approximate the nonlinear relationship between the governing parameters and the streaming potential, while the volumetric flow rate and hydroelectric energy conversion efficiency were calculated from closed form equation by using the streaming potential predicted by the surrogate. This is coupled with a multi-objective optimization strategy to identify operating conditions that simultaneously maximize energy conversion efficiency and volumetric flow rate. The proposed methodology can significantly accelerate parametric exploration compared with repeated numerical simulations across different parameters and provides practical design guidelines for electrokinetic microfluidic devices. The study demonstrates the potential of combining computational fluid mechanics with data-driven surrogate modeling for efficient engineering design and optimization.
[LG-150] he Price of Intelligence: A Quality-Adjusted Price Index for AI Services
链接: https://arxiv.org/abs/2608.29843
作者: Louis Yiven Zhu
类目: General Economics (econ.GN); Machine Learning (cs.LG)
*备注: 38 pages, 3 figures. Data, code and provenance archived at this https URL pre-registration at this https URL
Abstract:Posted prices for AI inference have fallen steadily since 2024, yet the measured speed of that fall depends almost entirely on the method of measurement. This paper constructs quality-adjusted price indices for the AI inference market from public data. The panel assembles 21,024 posted-price observations across 3,208 models and 86 providers and joins them to 4,605 benchmark scores through a latent quality index estimated from benchmark response patterns, so the quality ladder of the hedonic tradition is built here from evaluations in place of product characteristics. Measured by the matched-model methods that statistical agencies apply to software, inference prices fell at 0.10 log points a year. The quality-adjusted index fell at 0.73, so 87% of the decline is invisible to current methods, with direct consequences for measured competition, concentration and productivity in this market. Counted per completed task, moreover, the buyer’s price stopped falling. Reasoning models raised token consumption faster than token prices fell, and the seller’s and buyer’s prices accordingly diverged. A pre-registered validity audit disciplines the quality measure and yields the sharpest result. Excluding contamination-flagged benchmarks leaves model rankings intact at 0.998 yet moves the index by 0.49 log points a year, so the leaderboard-stability arguments standard in AI evaluation offer no defence of economic statistics built on benchmarks. Prices, quality and the audit are fully reproducible from public sources at zero cost.
[LG-151] A Unified Perspective on Conformal Prediction and Wasserstein Distributionally Robust Optimization for Uncertainty Quantification
链接: https://arxiv.org/abs/2608.29789
作者: Kehan Long,Yiqi Zhao,Pol Mestres,Lars Lindemann,Nikolay Atanasov,Jorge Cortés
类目: Optimization and Control (math.OC); Machine Learning (cs.LG); Systems and Control (eess.SY)
*备注:
Abstract:Uncertainty quantification from finite data is central to machine learning, optimization, and automation systems, where decisions must remain reliable under limited samples and test-time distribution shift. Conformal prediction (CP) and distributionally robust optimization (DRO) offer two complementary approaches: CP constructs data-dependent prediction sets with distribution-free finite-sample validity under exchangeability, while DRO optimizes worst-case performance over an ambiguity set around an empirical distribution. We develop a unified probabilistic perspective on CP and DRO by viewing both as ways to turn finite calibration data into a data-dependent quantile estimator that a test score falls below with high probability. From this perspective, CP and DRO correct the empirical quantile along two coordinates of the same family of estimators: CP inflates the quantile level, whereas DRO shifts the quantile value through an ambiguity radius. Both methods provide the same calibration-conditional guarantee for the true distribution, requiring the target coverage to hold with high probability over the calibration sample. Their constructions differ, however: CP uses a closed-form, distribution-free level correction, while DRO uses a value-space correction whose certified radius depends on properties of the unknown distribution and additionally guarantees coverage uniformly over the ambiguity set. This distinction emerges in the tails of the score distribution. Because CP relies on sparse upper-tail order statistics of the calibration samples, its level inflation barely moves the estimator when those samples are dense near the target quantile but overshoots when they are sparse, whereas a well-chosen DRO radius corrects in value space and may avoid this overshoot.
[LG-152] Neural ODE enhanced linear mixed effect models for estimating complex association patterns of time-varying covariates with the marker trajectory
链接: https://arxiv.org/abs/2608.29714
作者: Zhe Aurore Li,Quentin Clairon,Cécilia Samieri,Rodolphe Thiébaut,Mélanie Prague,Cécile Proust-Lima
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:Longitudinal cohort studies produce repeated data that enable the assessment of time-varying association patterns between exposures and health outcomes. Classical linear mixed-effects models (LMMs) can accommodate a large variety of association patterns while accounting for the irregularly spaced, partially observed measurement. But they require the analyst to pre-specify the functional form linking the exposure history to the outcome. We propose the Neural ODE-LMM, which embeds a Neural Ordinary Differential Equation (Neural ODE) within the linear mixed-effects framework: a learned vector field encodes covariate trajectories into a continuous-time latent state that drives both the fixed- and random-effect design, while preserving the standard LMM observation model. This retains classical likelihood-based inference while learning complex, potentially cumulative, covariate effects flexibly. All parameters are estimated by maximising a penalised marginal likelihood. To quantify covariate effects, we introduce contrasts of counterfactual predictions that compare the expected outcome under alternative covariate trajectories with variance estimated via the delta method. In simulations, the model recovers both instantaneous and cumulative-burden effects without prior specification of the functional form. Applied to the Trois-Cités (3C) cohort, a population-based study of 7,324 participants, the method reveals trajectory-dependent associations of BMI and fasting glucose with cognitive decline.
[LG-153] ransformer-Based Flow Shop Scheduling Using MILP-Generated Training Data
链接: https://arxiv.org/abs/2608.29690
作者: Roderich Wallrath
类目: Optimization and Control (math.OC); Machine Learning (cs.LG)
*备注:
Abstract:Advances in machine learning (ML) have created new opportunities to complement traditional operations research (OR) methods. In particular, transformer models can capture complex interactions in token sequences by mapping tokens into a high-dimensional embedding space and propagating contextual information via attention. This makes them a candidate to model non-permutation flow shop scheduling with secondary resources as a next-token prediction task, where tokens represent job-machine-secondary resource tuples. For training, mixed-integer linear programming (MILP)-generated schedules are tokenized and used as next-token prediction data. During inference, partial token sequences (prefixes) are randomly generated and completed by the trained transformer through constrained decoding. A computational study is conducted on a flow shop with 8 jobs, 4 machines, and 3 secondary resources, where jobs are selected from a fixed pool of 20 jobs that is sampled during training and provides the candidates during prefix completion. The transformer achieves better solution quality (smaller makespans) compared to a genetic algorithm (GA), the NEH heuristic, and random search. It is outperformed only by the MILP model and the iterated greedy (IG) heuristic. The study concludes that transformer models can, to some extent, learn patterns from MILP-optimized non-permutation flow shop schedules and that transformer-based scheduling represents an interesting direction for future research, particularly in settings with a fixed, recurring job set.
[LG-154] ButterMamba: Butterworth-Enhanced Spatial-Temporal Mamba for Efficient Traffic Flow Prediction
链接: https://arxiv.org/abs/2608.29658
作者: Limiao Zhang,Yuhui Lu,Jie Gao,Hao Jiang,Haiping Ma,Xingyi Zhang
类目: Physics and Society (physics.soc-ph); Machine Learning (cs.LG); Data Analysis, Statistics and Probability (physics.data-an)
*备注: 9 pages, 6 figures
Abstract:Accurate traffic flow prediction is fundamental to intelligent transportation systems, playing a pivotal role in urban mobility optimization and smart city development. While Graph Neural Networks (GNNs) integrated with time series forecasting have emerged as promising solutions, two critical limitations persist: (1) the quadratic complexity of attention-based architectures hinders real-time deployment in large-scale networks, and (2) high-frequency noise in sensor data significantly degrades prediction reliability. These challenges are particularly acute in metropolitan scenarios where both computational efficiency and noise robustness are paramount. To address these limitations, we introduce \textbfButterMamba, a novel and efficient framework based on State Space Models (SSMs). ButterMamba consists of two key components: (1) a Butterworth Spectral Filtering module that preprocesses the data by removing high-frequency noise, allowing the model to focus on significant underlying trends, and (2) a Spatial-Temporal State Mixer that uses a parallel Mamba architecture to efficiently capture both long-range temporal dependencies and complex spatial correlations across the road network. By decoupling noise filtering from spatial-temporal modeling, ButterMamba achieves superior predictive accuracy with linear computational complexity. Extensive experiments on three public datasets demonstrate that ButterMamba not only outperforms existing state-of-the-art models in terms of prediction accuracy but also considerably reduces training time and memory usage.
[LG-155] Deciding When to Decide: Testing Operational Suboptimality Under Distributional Shift
链接: https://arxiv.org/abs/2608.29465
作者: Minxing Zheng,Holly Wiberg,Shixiang Zhu
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Methodology (stat.ME)
*备注:
Abstract:Deployed decisions are often optimized once and retained because updates impose operational, regulatory, or switching costs. As operating conditions change, when should such decisions be re-optimized? We study this question for stochastic optimization when the objective’s functional form is known but the decision maker’s trade-offs are encoded by an unknown preference parameter. Standard distribution-shift tests are poorly aligned with this goal: they can flag detectable yet decision-irrelevant changes without determining whether the incumbent decision has become materially suboptimal. We propose \textttRADAR (Regret-based Assessment of Decision Adequacy and Risk), a decision-focused framework that uses inverse optimization to infer latent preferences and tests the deployed decision’s optimality gap under the current distribution. By targeting regret, \textttRADAR ignores decision-irrelevant shifts while detecting changes that warrant re-optimization. We develop two-sample and sequential changepoint procedures and establish asymptotic guarantees for Type-I error and power. Across synthetic optimization problems, a semi-synthetic capacity allocation task, and police-zone planning, \textttRADAR more reliably distinguishes harmful from harmless shifts than decision-agnostic alternatives.
[LG-156] Explanations Prompts and Formalizations: Arguments for New Norms in LLM -Enabled Mathematical Research
链接: https://arxiv.org/abs/2608.29401
作者: Axel Boldt
类目: History and Overview (math.HO); Machine Learning (cs.LG)
*备注: 5 pages
Abstract:As several mathematical conjectures have recently been settled using large language models (LLMs), the mathematical community has formulated norms and recommendations regarding the publishing of such results. These norms do not cover the disclosure of the prompts and precise software setup used to obtain those results, nor do they require that results be formalized in a manner that allows for machine verification. I argue that both of these are essential. In addition, since LLM-obtained results may be hard to understand, human authors have the responsibility to invent intuitive explanations.
[LG-157] Signed random Fourier features for fast density estimation with indefinite kernels
链接: https://arxiv.org/abs/2608.29265
作者: Xie Wang,Nicolas Langrené,Wen Chen
类目: Computation (stat.CO); Machine Learning (cs.LG); Probability (math.PR); Machine Learning (stat.ML)
*备注: 25 pages, 12 figures
Abstract:Kernel density estimation (KDE) is one of the most fundamental statistical estimators of density functions. Its direct implementation on a dataset of N points incurs an \mathcalO(N^2) computational cost, which is prohibitive for large-scale datasets. Kernel approximation techniques can be applied to bring the computational cost down to \mathcalO(N) . The random Fourier features (RFF) technique, based on sampling from the spectral density of the kernel function, has become popular to speed up kernel estimators for machine learning applications. Unfortunately, it is restricted to positive definite kernels, while the majority of kernel functions popular in KDE, such as the parabolic kernel, do not satisfy this property. To overcome this limitation, this article introduces the signed random Fourier features (SRFF) technique. It is a generalization of RFF compatible with indefinite kernels whose inverse Fourier transform is absolutely integrable. The motivation for introducing this method is to speed up KDE in the case of multivariate compact kernels, which are generally not positive definite. We detail how to implement SRFF for both product kernels and isotropic kernels. For the class of Kuttner-Golubov kernels K(\boldsymbolx_i,\boldsymbolx_j)=(1-\left\Vert \boldsymbolx_i-\boldsymbolx_j\right\Vert ^\alpha)^\beta\mathbf1_\left\Vert \boldsymbolx_i-\boldsymbolx_j\right\Vert \leq1\ where \boldsymbolx_i\in\mathbbR^d , \boldsymbolx_j\in\mathbbR^d , \alpha0 , \beta0 , which includes the triangular, parabolic, biweight, triweight, and other kernel functions of interest for KDE as particular examples, we provide an explicit acceptance-rejection algorithm to sample from its signed spectral density. Our numerical tests on a dataset of one million points confirm the computational efficiency and accuracy of SRFF for large-scale KDE.
[LG-158] Uniform Statistical Convergence of Empirical Sinkhorn Potentials with Exponential and Polynomial Dependence on the Regularization Parameter
链接: https://arxiv.org/abs/2608.29152
作者: Denis Belomestny
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Optimization and Control (math.OC)
*备注:
Abstract:We study the empirical Sinkhorn estimator of the entropic optimal transport potentials under the uniform loss. Since the potentials are only unique up to additive constants, we measure the error using the quotient supremum norm, defined as d_\infty([u],[v]) = \inf_a\in\mathbbR|u-v-a|_\infty . For a fixed regularization parameter \varepsilon0 , we establish a non-asymptotic statistical rate of n^-1/2 . This is achieved by combining the Birkhoff-Hopf contraction theorem with entropy bounds on normalized kernel sections. However, the constant in this bound grows exponentially with 1/\epsilon . To improve this, we isolate geometric conditions under which the empirical estimator maintains the n^-1/2 rate but features polynomial dependence on 1/\varepsilon . The key requirement is a polynomial residual-stability estimate for the population Sinkhorn map. We provide sufficient criteria for this, including a polynomial contraction property and a local inverse estimate. Furthermore, we introduce two rigorously verifiable model classes an \varepsilon -weak residual-interaction class obtained after separable centering and another based on connected tight-edge graphs for fixed discrete costs where the polynomial rate is guaranteed without relying on abstract resolvent assumptions. Finally, we establish matching minimax lower bounds demonstrating that the \varepsilon n^-1/2 rate cannot be uniformly improved in the bounded-interaction regime.
[LG-159] Spectral-Embedded Operator Learning for Three-Phase Interfacial Flow: A Ternary Cahn-Hilliard-Navier-Stokes Benchmark
链接: https://arxiv.org/abs/2608.29069
作者: Muhammad Abid,Arth Sojitra,Omer San
类目: Fluid Dynamics (physics.flu-dyn); Machine Learning (cs.LG)
*备注:
Abstract:Operator-learning surrogates have been benchmarked largely on single-field, single-interface problems, leaving unclear whether architectural choices validated in those settings transfer to constrained, multiphase flows. We introduce a three-phase interfacial-flow benchmark to examine whether the trunk coordinate representation matters for a multi-channel, interface-dominated target. The configuration consists of an air bubble rising through water, piercing a water-oil interface, and entraining a water plume into the oil within a bounded, wall-confined domain. Reference data are generated using a structure-preserving ternary Cahn-Hilliard-Navier-Stokes solver that algebraically preserves the simplex constraint. From 1,024 Sobol-sampled simulations spanning a nine-dimensional parameter space, we learn the mapping from physical parameters to five-channel space-time fields. We compare three parameter-matched DeepONet variants differing only in trunk representation: raw coordinates (DeepONet), random Fourier features (FEDONet), and a fixed tensor-product Chebyshev dictionary (SEDONet). SEDONet reduces the test relative L2 error by 16.8% compared with FEDONet and by 24.0% compared with DeepONet, while improving all five output channels. Spatial and temporal error analyses localize the principal gains near the diffuse interfaces and after bubble breakthrough. The results indicate that the Chebyshev representation is particularly effective for the strongly non-periodic wall-normal and temporal structure of this three-phase flow.
[LG-160] Sharp Restricted Isometry Thresholds for Global Minima of Rank-Restricted Matrix LASSO
链接: https://arxiv.org/abs/2608.29018
作者: Richard Y. Zhang
类目: Machine Learning (stat.ML); Information Theory (cs.IT); Machine Learning (cs.LG); Optimization and Control (math.OC); Statistics Theory (math.ST)
*备注:
Abstract:We determine the sharp restricted isometry threshold for recovery at global minima of the rank-restricted matrix LASSO. For target rank r_\star , if the rank- k RIP constant satisfies \delta\delta_\mathrmsharp(k/r_\star) , where \delta_\mathrmsharp(t)=t/(4-t) for 0t4/3 and \delta_\mathrmsharp(t)=\sqrt(t-1)/t for t\ge4/3 , then every global minimizer has Frobenius error \lesssim\sqrtr_\star\lambda for all \lambda\gtrsim|\mathcalA^*(\xi)|\mathrmop and at every search rank r\ge r\star . The constants depend only on the RIP constant and t=k/r_\star , and in particular are independent of the search rank. When the rank restriction is inactive, the result specializes to the ordinary convex matrix LASSO. We also obtain the analogous results for sparsity-restricted vector LASSO. Conversely, we show that the threshold \delta\delta_\mathrmsharp(k/r_\star) cannot be improved, due to the existence of counterexamples whose global minimizers fail to recover the ground truth.
[LG-161] Jigsaw-CRL: Recovering Global Latent Causal Order from Frag mented Multi-Client Interventions
链接: https://arxiv.org/abs/2608.28991
作者: Haijie Xu,Chen Zhang
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:Causal representation learning (CRL) aims to recover latent causal variables and their structural relations from high-dimensional observations. Existing CRL methods typically assume that all environments are defined over the same latent variables, or at least share a common latent representation space. We study a fragmented multi-client setting, where multiple clients interact with the same global latent causal system but each client only accesses and intervenes on a subset of the latent variables. In this regime, marginalizing unused latent variables induces bidirected edges, so a single client no longer admits a node-wise latent causal graph, and the global latent causal order must be recovered by assembling client-specific structural fragments. We propose \textbfJigsaw-CRL, a framework for recovering global latent causal order from such fragmented interventions. Under soft interventions, differences between precision matrices across environments exhibit a low-rank structure governed by latent ancestor relations. This enables recovery, for each client, of a block partition, the corresponding block-level ancestral order, and latent subspaces, and then assembly of these fragments into the global node-level latent causal order. We establish identifiability guarantees, develop practical algorithms, and validate the framework on synthetic data. Our codes are available on this https URL
[LG-162] Quantitative Target Convergence and Uniform-in-Time Propagation of Chaos for Langevin-Regularized SVGD
链接: https://arxiv.org/abs/2608.28827
作者: Sayan Banerjee,Dohyeon Kim
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Probability (math.PR)
*备注: 54 pages
Abstract:We establish quantitative convergence to the target and uniform-in-time propagation of chaos for Langevin-regularized Stein variational gradient descent. The Stein interaction need not be small relative to the confining Langevin drift and does not generally yield a contractive particle coupling. At the mean-field level, the Stein and Langevin components dissipate the same relative entropy in the kernel-induced Stein and 2 -Wasserstein geometries, producing the squared kernel Stein discrepancy and relative Fisher information. Under a log-Sobolev inequality for the target, this yields exponential last-iterate convergence. We also derive a finite-particle entropy identity relative to the product target, giving exponential-in-time convergence of the empirical measure up to polynomial sampling errors. For propagation of chaos, we develop two complementary finite-time approaches. A synchronous coupling, combined with exponential moment estimates for the nonlinear mean-field diffusion, yields explicit single-exponential bounds in Wasserstein distance and kernel Stein discrepancy (KSD). Moving-product entropy gives joint-law relative entropy control relative to the evolving mean-field product law and, through entropy superadditivity and concentration, fixed-marginal relative entropy and total variation bounds and empirical KSD estimates. Under an additional T_2 inequality for the initial law, it also yields Wasserstein bounds. Combining these finite-time estimates with target convergence at a logarithmic cutoff time gives polynomial uniform-in-time propagation of chaos rates in expectation for empirical KSD and W_2^2 , and for fixed-marginal total variation and W_2^2 . All bounds control the last iterate in physical time. We also compare the two finite-time mechanisms and identify regimes in which each gives the sharper polynomial exponent. Comments: 54 pages Subjects: Machine Learning (stat.ML); Machine Learning (cs.LG); Probability (math.PR) Cite as: arXiv:2608.28827 [stat.ML] (or arXiv:2608.28827v1 [stat.ML] for this version) https://doi.org/10.48550/arXiv.2608.28827 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-163] Propensity Straight-Through Gradients for Discrete Stochastic Systems
链接: https://arxiv.org/abs/2608.25631
作者: Jose M. G. Vilar,Leonor Saiz
类目: Quantitative Methods (q-bio.QM); Disordered Systems and Neural Networks (cond-mat.dis-nn); Machine Learning (cs.LG); Computational Physics (physics.comp-ph); Molecular Networks (q-bio.MN)
*备注: 16 pages, 5 figures
Abstract:Continuous-time Markov chains (CTMCs) provide the backbone for modeling discrete stochastic dynamics across applied, physical, and biological sciences. Their integration with modern gradient-based machine learning, however, is limited by the hard categorical event selection intrinsic to Gillespie-type simulation algorithms. We exploit the affine state update to obtain the exact one-step conditional-mean sensitivity by differentiating normalized reaction propensities. We pair this backward rule with exact forward trajectories to define the propensity straight-through (PST) estimator. At the trajectory level, we show that one-step sensitivities composed across events can depart from the exact multistep sensitivity. We derive the resulting per-step discrepancy in closed form and prove that it vanishes identically for affine downstream dependence. PST matches the accuracy of Gumbel-Softmax straight-through across all benchmarks: reversible dimerization (0.06% error), a genetic oscillator (1.7% error), a 50-task repressilator suite (0.17% median error), and patch-clamp ion-channel recordings ( R^2 = 0.988). Under matched settings, PST converges 3.0-fold faster on the oscillator and 2.1-fold faster on the ion channel. At deep-learning scale, PST trains a 203,796-parameter stochastic reaction network with hard sampling, reaching 98.22% MNIST digit classification accuracy. By differentiating an exact conditional mean rather than a relaxed sample, PST offers a temperature- and Gumbel-free path to scalable gradient-based learning through exact stochastic trajectories.
附件下载
