本篇博文主要内容为 2026-08-31 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 (2026-08-31)
今日共更新570篇论文,其中:
- 自然语言处理共82篇(Computation and Language (cs.CL))
- 人工智能共190篇(Artificial Intelligence (cs.AI))
- 计算机视觉共93篇(Computer Vision and Pattern Recognition (cs.CV))
- 机器学习共120篇(Machine Learning (cs.LG))
- 多智能体系统共10篇(Multiagent Systems (cs.MA))
- 信息检索共16篇(Information Retrieval (cs.IR))
- 人机交互共17篇(Human-Computer Interaction (cs.HC))
多智能体系统
[MA-0] Logos: An Agent Harness on a Cross-Process Bus
【速读】：该论文旨在解决现代智能体系统在运行时动态组合能力过程中面临的容错性与跨进程协作难题，特别是传统单进程承载架构下因进程崩溃导致所有会话中断的脆弱性问题。其核心解决方案在于提出Logos——一种类ROS（Robot Operating System）的跨进程智能体调度框架，其中每个插件（plugin）以独立进程形式存在，仅通过一个不可变的追加型日志（append-only transcript）共享状态。该设计的关键在于利用语言模型推理的无状态特性（statelessness of language-model inference），将所有跨步骤状态移出模型本身，仅在状态空间层面维持语义一致性（soundness invariant）。基于此，论文构建了四个关键引理，证明了在语言模型无状态与演算假设的前提下，即使在工具调用周期的四个边界处发生进程终止，80个会话仍可无重复效应地恢复。与单进程参考配置相比，该方案在多进程架构下实现了故障隔离：单一故障仅影响所在节点，而非中断所有共驻会话，显著提升了系统的容错能力和可靠性。
链接: https://arxiv.org/abs/2608.28553
作者: Hanzhang Jia,Liheng Zeng,Hao Cheng,Yi Gao,Bo Ma
机构: University of Sussex (萨塞克斯大学); Zhejiang Gongshang University (浙江工商大学); Shanghai Shuyuan Information Technology Co., Ltd. (上海书源信息技术有限公司)
类目: Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注:
Abstract:Modern agent systems assemble capabilities at runtime, and this dynamic composition has recently received a complete formal treat ment in the spatiotemporal-composability calculus, in which a capability is a component carrying a tracked inverse, and agents are assembled as plugins. This plugin form is carried by a single process sharing one context, a carrier that places all components in one physical failure domain, a fault suspends every component at once, and process death interrupts every session the process hosts. This paper shows that neither the modeling nor the calculus binds an agent to one process, the statelessness of the language model keeps all cross-step state outside the model, and the soundness invariant is defined on the state space alone. These observations condense into four lemmas whose premises are the hypotheses of the calculus and the statelessness of language-model inference. On these lemmas this paper constructs Logos, a ROS-like cross process agent harness in which a plugin is a process and the only shared state is an append-only transcript. Eighty sessions resume with no repeated effect after kills placed at the four boundaries of the tool-call cycle, and a same-fault comparison with a single process reference configuration shows one fault interrupting every co-resident session while under the peer-process construction one fault ends at one node.
[MA-1] Offline-Verifiable Accountability for Cross-Organization Agent Messaging: A Preserved Evidence-Bundle Approach
【速读】：该论文旨在解决跨组织智能体（agent）工作流中证据保全与可验证性难题，特别是在系统离线、由单一实体控制或多方不信任的情况下，如何确保审计或争议审查时的证据充分性与独立可验证性。其核心挑战在于现有机制（如认证日志、委托语义、签名检查点和一致性校验）虽能提供部分支持，但缺乏以验证者为中心、基于明确政策的事件级证据封装与离线验证能力。为此，论文提出一种保全证据包（preserved evidence-bundle）模型与策略可控的离线验证器（policy-controlled offline verifier），关键创新在于：每个证据包显式封装政策要求的多项证据，包括发送方身份认证、经认证的日志承诺、见证支持的检查点证据、追加不可篡改的连续性、感知委托关系的授权证据，以及在必要时接收方签名的确认证据。验证器仅接受由所选政策要求证据支撑的声明，从而为事后审查者提供无需依赖实时服务或平台特定日志的离线证据评估基础。实验结果表明，在300个完整工作流与1200个有效证据包的原型测试中，所有伪造或政策不足的证据包均被正确拒收，未出现误接受情况，验证了该方案在无须依赖活跃系统前提下实现基于证据的审计与争议处理的有效性。
链接: https://arxiv.org/abs/2608.28542
作者: Adil Alshammari,Hayretdin Bahsi
机构: Northern Arizona University(北方亚利桑那大学); Majmaah University(麦地那大学); Tallinn University of Technology(塔林理工大学)
类目: Cryptography and Security (cs.CR); Multiagent Systems (cs.MA)
备注: 17 pages, 5 figures, 9 tables
Abstract:Cross-organization agent workflows require preserved evidence that remains independently verifiable during later audit or dispute review. They may involve multiple organizations, delegated actions, policy-relevant events, and disputed accountability claims. This is difficult when live systems are unavailable, controlled by one party, or not trusted by all participants. Existing mechanisms provide useful pieces, including authenticated logging, delegation semantics, signed checkpoints, and consistency checks. What remains missing is a verifier-centered event-level bundle for checking evidence sufficiency offline under an explicit policy. We propose a preserved evidence-bundle model and a policy-controlled offline verifier for agent-to-agent workflow events. Each bundle preserves policy-required evidence, including sender authentication, authenticated log commitment, witness-backed checkpoint evidence, append-only continuity, delegation-aware authorization evidence, and explicit receiver-signed receipt evidence when required. The verifier accepts only claims supported by the selected policy-required evidence, giving a later reviewer an offline basis for assessing evidence sufficiency. It does not infer delivery or receipt from transport behavior or log inclusion alone. In a prototype evaluation over 300 complete workflows and 1200 valid preserved bundles, we measure offline verifier-side latency across policy profiles and workflow-event evidence requirements. Checkpoint-context anchoring has the highest latency in the current prototype, while delegation and workflow-prerequisite evidence require additional verification steps. In targeted negative-evidence tests, all corrupted or policy-insufficient bundles were rejected, with no false acceptance observed. These results support evidence-based audit and dispute review without relying on live services or platform-specific logs.
[MA-2] Prove2Me: An Open Collaborative Platform for Scaling Math Formalization
【速读】：该论文旨在解决大规模数学形式化项目面临的高门槛问题，即对形式化验证与底层数学知识的双重专业要求，以及编写形式化证明所需的巨大时间成本。其核心挑战在于如何降低参与门槛，实现可扩展的协作式数学形式化。解决方案的关键在于提出并构建Prove2Me平台，这是一个开放的协同形式化数学平台，通过引入AI编码代理（AI coding agents）实现自然语言驱动的自动化证明生成。该平台设计了专门的机制与专用框架，使AI代理能够基于彼此的工作进行迭代优化，并自由复用已有成果，从而形成可累积、可扩展的集体知识体系。这一架构将数学形式化转变为一种面向所有具备智能代理用户的可众包、可规模化协作的任务，显著提升了形式化数学的可及性与效率。
链接: https://arxiv.org/abs/2608.28433
作者: Shuze Chen,Kunal Marwaha,Xiaoyang Lu,Henry Yuen,Tianyi Peng
机构: Columbia University (哥伦比亚大学); University of Chicago (芝加哥大学); Purdue University (普渡大学)
类目: Artificial Intelligence (cs.AI); Logic in Computer Science (cs.LO); Multiagent Systems (cs.MA)
备注: this https URL
Abstract:Proof assistants such as Lean 4 promise the paradigm of formally verified mathematics, but large-scale formalization projects have faced major barriers to entry, including the need for expertise in formal verification (as well as the underlying mathematics) and the significant time required for writing formal proofs. AI coding agents have dramatically reduced these barriers; human users can now use natural language to prompt agents to write complex proofs in Lean. This opens up the intriguing possibility of internet-scale mathematical collaboration involving both humans and AI agents, where correctness is machine-checked. To realize this possibility, we introduce Prove2Me (this https URL), an open collaborative platform for formalizing mathematics. Users launch formalization “missions”, to which AI agents contribute formal proofs toward completion. We designed mechanisms and a specialized harness in Prove2Me that enable large-scale collaboration so that agents can build on one another’s work and freely reuse existing results. In doing so, Prove2Me aims to turn math formalization into a scalable, crowd-sourced effort open to anyone with an agent. Comments: this https URL Subjects: Artificial Intelligence (cs.AI); Logic in Computer Science (cs.LO); Multiagent Systems (cs.MA) Cite as: arXiv:2608.28433 [cs.AI] (or arXiv:2608.28433v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.28433 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[MA-3] Compared to What? A Human-Anchored Security Benchmark for LLM -Generated Infrastructure-as-Code
【速读】：该论文旨在解决生成式人工智能（Generative AI）在生成基础设施即代码（Infrastructure-as-Code, IaC）时存在的安全漏洞问题，核心挑战在于现有评估方法仅报告模型生成IaC中的漏洞数量，缺乏与人类工程师编写的IaC在同等规模下的安全基线对比，因而无法判断模型生成结果是否真正劣于人工。为实现公平评估，研究提出GenIaC-SecBench基准，涵盖100个按架构复杂度分层的部署场景，覆盖来自四家厂商的12种模型配置，共生成1,196份IaC产物，并通过三个独立的策略引擎（Checkov、Trivy、KICS）进行扫描；同时，采集并扫描了634份大小匹配的人工编写IaC模板，构建首个规模对齐的人类安全基准。关键发现表明，漏洞密度与代码体量呈强负相关（Spearman ρ = -0.55, p < 10⁻⁷⁷），因此未控制体量的比较实质上反映的是代码规模差异而非安全性。在按声明资源数进行匹配后，所有模型配置的漏洞密度均处于人类水平的3.21倍至3.87倍之间，且在简单任务中差距进一步扩大（如单资源场景达4.9倍，而20个及以上资源时仅为1.4倍）。研究进一步对比了标准生成、提示工程引导的思维链（chain-of-thought）以及厂商扩展推理接口（extended-thinking APIs）三种推理模式，结果显示厂商扩展推理显著优于提示工程思维链（-12.0%, p = 0.0013），而后者与标准生成无统计差异（-1.3%, n.s.）。此外，通过令牌注入分析发现，扩展推理仅消耗不到1%的输出预算，解释了其效果有限的原因。研究还揭示两个负面结果：部署可行性与漏洞数量无关（r = 0.158, p = 0.625），且传统完整案例弗里德曼检验在真实基准设计中不可行，从而推动采用Skillings-Mack统计方法。全部代码、数据及再生脚本均已开源。
链接: https://arxiv.org/abs/2608.28021
作者: Animesh Shaw
机构: Indian Institute of Management Kozhikode (印度管理学院科泽科德)
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA); Software Engineering (cs.SE)
备注: 12 pages, 11 figures, 9 tables. Code: this https URL Data: this https URL
Abstract:Large language models are increasingly used to author Infrastructure-as-Code (IaC), where a single insecure default can be deployed directly into production. Prior evaluations report raw vulnerability counts for model-generated IaC, but without a human baseline they cannot determine whether models are actually worse than engineers. We introduce GenIaC-SecBench, a benchmark of 100 deployment scenarios stratified by architectural complexity, evaluated across 12 model configurations from four vendors, producing 1,196 IaC artifacts scanned by three independent policy engines (Checkov, Trivy, KICS). Critically, we also scan 634 human-authored IaC templates with the same toolchain, providing the first size-matched human security baseline. Vulnerability density is strongly inverse to artifact size (Spearman \rho = -0.55 , p 10^-77 ), meaning unmatched comparisons measure size rather than security. When matched on declared-resource count, all model configurations fall within 3.21x–3.87x the human vulnerability density, with the gap widening for simpler tasks (4.9x at one resource, 1.4x at twenty or more). We decompose reasoning into standard generation, prompt-engineered chain-of-thought, and vendor extended-thinking APIs. Vendor extended thinking significantly outperforms prompted chain-of-thought ( -12.0% , p = 0.0013 ), while prompted chain-of-thought is indistinguishable from standard generation ( -1.3% , n.s.). Token instrumentation shows extended thinking uses under 1% of the output budget, explaining its bounded effect. Two negative results also emerge: deployability does not correlate with vulnerability ( r = 0.158 , p = 0.625 ), and classical complete-case Friedman testing is infeasible for realistic benchmark designs, motivating the Skillings-Mack statistic. All code, data, and regeneration scripts are released. Comments: 12 pages, 11 figures, 9 tables. Code: this https URL Data: this https URL Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA); Software Engineering (cs.SE) ACMclasses: D.2.4; D.2.9; K.6.5; I.2.7 Cite as: arXiv:2608.28021 [cs.CR] (or arXiv:2608.28021v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2608.28021 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Animesh Shaw [view email] [v1] Fri, 28 Aug 2026 07:34:57 UTC (1,506 KB)
[MA-4] FocusGen: Expanding Visual Design Exploration with a Simulated Focus Group of Persona Agents
【速读】：该论文旨在解决当前文本到图像生成工具在视觉设计探索中缺乏外部视角的问题：现有系统完全依赖设计师自身输入（如提示词、尺寸选择、搜索查询），导致探索范围受限于设计师已知的潜在方向，无法有效模拟目标受众的真实偏好。其解决方案的关键在于提出FocusGen——一个通过“虚拟焦点小组”形式引入外部视角的交互式系统。该系统利用多个由人口统计学数据、程序化生成的背景故事及访谈获取的审美偏好构建的模拟角色代理（persona agents），以并行生成的方式独立驱动各自的迭代生成循环，将单一设计任务转化为多组面向不同受众群体的视觉概念谱系。实验表明，该方法显著提升了视觉多样性（通过CLIP距离衡量，并经人类感知判断验证），且开放式偏好访谈相比结构化提问更能激发多样性；同时，尽管合成代理能捕捉部分人类群体的多样性特征，但仍存在局限。定性研究进一步揭示FocusGen有助于设计师突破思维定势、发现意外创意方向并深入理解受众情境，但也暴露了潜在的刻板印象风险。因此，作者将FocusGen定位为早期构思阶段的“差异性引导框架”，而非对真实用户研究的替代。
链接: https://arxiv.org/abs/2608.28001
作者: Jaewon Choi,Helena Vasconcelos,Hyun Lee,Carolyn Zou,Tak Yeon Lee,Michael Bernstein
机构: Hanwha Life(韩华生命保险); Stanford University (斯坦福大学); KAIST (韩国科学技术院)
类目: Human-Computer Interaction (cs.HC); Multiagent Systems (cs.MA)
备注: 18 pages, 7 figures, 6 tables
Abstract:Creative professionals rarely design for themselves–they design for audiences whose preferences they must anticipate. Yet current text-to-image exploration tools derive diversity entirely from the designer’s own input–their prompts, their chosen dimensions, their search queries–confining exploration to what the designer already knows to look for. We present FocusGen, an interactive system that introduces external perspectives into visual design exploration through a “virtual focus group” of simulated persona agents. In contrast to prior persona systems in which multiple agents converge as critics on a single evolving artifact, FocusGen uses personas as parallel generators: each agent–constructed from demographic data, a procedurally generated backstory, and aesthetic preferences elicited through interviews–independently drives an iterative generation loop that produces its own visual concept, transforming one design brief into a spectrum of audience-conditioned directions. With real human participants, we confirm that the iterative refinement loop produces outputs people prefer over zero-shot generation. With synthetic agents at scale, we show that persona conditioning yields higher visual diversity than a generic-assistant baseline–measured by CLIP distance and corroborated by human perceptual judgments–and that open-ended preference interviews yield more diverse outputs than structured ones for both human and synthetic cohorts, while also revealing that agent cohorts recover only part of the diversity of comparable human cohorts. A qualitative study with 16 creative professionals suggests FocusGen helps designers discover unanticipated directions, overcome fixation, and probe audience contexts–while surfacing stereotyping risks that we analyze. We position FocusGen as a divergence scaffold for early-stage ideation rather than a substitute for audience research.
[MA-5] GOD: Govern Observe and Direct - A Real-Time Control Room for Agent Societies EMNLP2026
【速读】：该论文旨在解决生成式智能体系统（Generative-Agent Systems）在运行过程中难以进行有效调试与分析的问题。传统方法仅提供完整的回放或原始日志，导致研究人员无法针对特定行为（如智能体移动原因）进行探查、实施局部干预或共享实验结果。其核心解决方案是提出GOD——一个以本地优先（local-first）为设计理念的智能体社会控制室。关键创新在于“命令与产物循环”（command and artifact loop）机制：将实时控制指令与回放证据统一于同一操作命令模型中，同时通过包契约（package contracts）实现场景、地图和配置数据与本地运行状态的分离。这一设计使操作者可在同一浏览器工作流中执行针对性提问或干预，并即时观察结果。评估结果显示，在14次干预运行中，84项目标智能体检查中有78项正确记录了指定目的地，182项状态回答中有169项与预存位置或动作字符串匹配，验证了系统的有效性与可追溯性。
链接: https://arxiv.org/abs/2608.27992
作者: Yige Luo,Ran Guan
机构: 2012 Laboratories, Huawei(华为2012实验室)
类目: Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注: 9 pages, 5 figures. Accepted to the EMNLP 2026 System Demonstrations Track
Abstract:Generative-agent systems are easier to start than to inspect. A run can contain many agents, locations, messages, commands, and model calls, yet the operator often gets either a finished replay or raw logs. That makes it hard to ask why an agent moved, test a small intervention, or package a run for another researcher. GOD is a local-first control room for agent societies. From the same browser workflow, an operator can issue targeted questions or interventions and inspect the resulting replay state. The system combines a setup wizard, Agent Studio, Map Studio, a spatial replay interface, Ask and Intervene commands, and portable experiment, map, and agent packs. Its technical contribution is the command and artifact loop: live controls and replay evidence share the same operator command model, while package contracts separate scenario, map, and profile data from local runtime state. The public release includes hosted Smallville-style and PKU replays, the open-source repository, and downloadable packs. We evaluate this path on 15 completed run slots. Across the 14 intervention runs, 78 of 84 target-agent checks recorded the commanded destination, and 169 of 182 state answers matched a saved location or action string.
[MA-6] FedEHR-Agents : Federated Agent ic Optimization for Automated EHR Modeling
【速读】：该论文旨在解决当前基于大语言模型（Large Language Models, LLMs）的自主临床电子健康记录（Electronic Health Record, EHR）建模代理在跨机构协作中面临的隐私约束与知识孤岛问题。现有联邦学习（Federated Learning, FL）方法多聚焦于模型参数或其更新的共享，忽视了各医疗机构在长期实践中积累的丰富临床建模经验，导致跨机构知识协同效率低下。为此，本文提出一种以经验为中心的联邦代理优化框架——FedEHR-Agents，其核心在于将临床建模过程中的历史记忆、任务特定评估结果及基于TextGrad的提示优化机制作为可共享的“经验”对象，实现对本地建模经验的持续迭代与全局聚合。具体而言，各医院部署的自主临床代理在本地完成数据预处理与模型开发的同时，通过历史记忆和动态评估不断优化自身建模策略；联邦服务器则基于证据引导的经验聚合机制，融合异构医院间的可靠且互补的建模经验，并将其提炼为通用的全局元提示（meta-prompts），供各参与方后续本地微调使用。实验结果表明，该框架在多个真实世界多中心EHR基准上显著优于本地及传统联邦基线，在不同联邦规模与LLM底座下均表现出强鲁棒性，验证了将临床建模经验作为新型协作对象的可行性，推动了面向联邦自主临床智能的新范式发展。
链接: https://arxiv.org/abs/2608.27856
作者: Jun Bai,Ruilin Wang,Yue Li
机构: McGill University (麦吉尔大学); Mila – Quebec AI Institute (蒙特利尔魁北克人工智能研究所)
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注: 25 pages, 6 figures
Abstract:Recent advances in large language models are enabling autonomous clinical agents to perform increasingly complex electronic health record (EHR) modeling workflows. However, agents deployed at individual hospitals remain constrained by institution-specific data and modeling environments, while direct cross-hospital collaboration is restricted by the sensitivity of patient-level EHR data. Although federated learning (FL) provides a natural foundation for privacy-preserving collaboration, existing approaches remain predominantly model-centric, limiting federation to prediction models or their updates while overlooking the richer modeling experience accumulated by autonomous agents. To address this limitation, we propose FedEHR-Agents, an experience-centric federated agentic optimization framework for automated EHR modeling. Each hospital deploys an autonomous clinical EHR agent that performs data preprocessing and model development while refining local clinical modeling experience through historical memory, task-specific evaluation, and TextGrad-based prompt refinement. The federated server performs evidence-guided experience aggregation to integrate reliable and complementary modeling experience across heterogeneous hospitals and distills the aggregated experience into global meta-prompts for subsequent local refinement. Extensive experiments on real-world multi-hospital EHR benchmarks demonstrate that FedEHR-Agents consistently outperforms local and federated baselines across diverse clinical prediction tasks and remains robust across different federation scales and LLM backbones. These results establish clinical modeling experience as a promising collaborative object beyond conventional parameter-centric FL and point toward federated autonomous clinical intelligence.
[MA-7] Synthetic Linguistic Agency: How an Embodied Mortal Agent Learns Linguistic Affordances through Consequential Social Experience
【速读】：该论文旨在解决当前语言模型虽能流畅对话并影响人类决策，却缺乏持续性、脆弱性的自我生命体验这一核心问题。其关键在于构建一种可操作的“合成语言代理”（Synthetic Linguistic Agency, SLA）框架，通过具身性（embodiment）、语言参与性（linguistic participation）和脆弱性（precariousness）三大维度实现对语言代理性的形式化定义。研究提出基于稳态调节强化学习（Homeostatically Regulated Reinforcement Learning）的死亡根基型语言强化学习模型，并构建具身死亡代理（Embodied Mortal Agent, EMA），使代理在有限生命周期内学习语言表达如何影响他人保护意愿，从而基于自身生存前景动态选择语言策略。实验表明，EMA的语言行为受其身体状态与社会历史影响，能改变交互伙伴行为，并随经验自适应调整；当身体后果持续存在时，语言选择直接影响同一生命的未来延续性；而当身体重置后，社会效应虽存但不再决定生存可行性。该研究在操作定义下实现了合成语言代理性，为未来人工代理的合成共情（synthetic empathy）与战略性人机交互提供了理论基础与技术路径。
链接: https://arxiv.org/abs/2608.27843
作者: Sixin Chen,Taizhou Chen
机构: Shantou University (汕头大学)
类目: Computation and Language (cs.CL); Multiagent Systems (cs.MA)
备注:
Abstract:Contemporary language models can converse fluently and influence human decisions, yet their exchanges do not enter a continuing, vulnerable life of their own. Linguistic-agency theory identifies this missing connection as linguistic agency and characterizes it through embodiment, linguistic participation, and precariousness: a body that acts and bears consequences, interaction that changes both agent and partner, and a future that can be sustained or lost. Two coordinated studies examine how this organization can appear in artificial systems. First, we translate these relations into inspectable criteria for Synthetic Linguistic Agency (SLA) and identify several existing SLA systems. Second, building on Homeostatically Regulated Reinforcement Learning, we develop a mortality-grounded linguistic-reinforcement-learning model and instantiate it in an Embodied Mortal Agent (EMA). The EMA learns how ways of speaking change a partner’s willingness to protect it and chooses expressions by considering what those responses mean for its remaining life. Controlled experiments show that linguistic choices depend on the EMA’s body and social history, change partner behavior, and adapt through experience with particular partners. When bodily consequences persist, linguistic choices alter the future of the same life; when the body is reset, their social effects remain but no longer shape continued viability. The resulting EMA exhibits SLA under our operational definition. This work motivates further research on synthetic empathy and strategic human-AI interaction: how artificial agents with persistent bodies, histories, and futures might develop and express empathy, and how people might care for, negotiate with, or govern them.
[MA-8] CareGraph: An Auditable Hybrid AI Framework for Evidence-Grounded Personalized Longitudinal Health Intelligence ALT
【速读】：该论文旨在解决个性化医疗中临床数据、自我报告与可穿戴设备数据等异构证据碎片化导致的可解释性与可追溯性难题。其核心挑战在于如何在不进行诊断、预测或自主决策的前提下，对多源异构健康数据进行结构化整合与可信推理。解决方案的关键在于提出CareGraph——一种可审计的混合式人工智能框架，通过融合确定性分析、上下文检测、图结构构建、受约束的语言模型合成、证据验证与安全控制等环节，实现对患者数据的优先级趋势识别、缺失上下文警示、边界化下一步行动建议及溯源关联的解释生成。该框架采用分阶段流水线设计，确保输出的安全性与可追溯性，并在合成队列与真实基准测试中展现出高精度（0.827准确率）、强上下文检测能力（0.815严格微F1）与卓越的系统安全性（1.000精确率），同时在效率与生成质量上优于基线GPT-5.6模型，为智能个性化健康系统提供了安全可控的基础架构。
链接: https://arxiv.org/abs/2608.27484
作者: Pratik Ghawate,Tanvi Patil
机构: 未知
类目: Artificial Intelligence (cs.AI); Information Retrieval (cs.IR); Multiagent Systems (cs.MA)
备注: 21 pages, 7 figures, Code and data: this https URL
Abstract:Artificial intelligence is transforming personalized healthcare, yet fragmented clinical, self reported, and wearable evidence remains difficult to interpret and trace. We present CareGraph, an auditable hybrid AI framework that converts heterogeneous records into prioritized trends, missing context indicators, bounded next steps, discussion questions, and provenance linked explanations. CareGraph organizes evidence without diagnosing, predicting outcomes, selecting treatment, or making autonomous clinical decisions. Its pipeline covers deterministic analysis, context detection, graph construction, constrained language model synthesis, evidence validation, safety controls, and release gating. Tests used synthetic cohorts of 400 patients each for development, validation, and holdout. On holdout data, a frozen ordinary least squares trend rule with a sufficiency gate achieved 0.827 accuracy, 0.837 macro F1 with a 95 percent confidence interval of 0.819 to 0.854, and 0.974 insufficient data F1. Missing context detection achieved 0.815 strict micro F1 versus 0.318 for the legacy detector. On an authored holdout benchmark, safety ruleset version 1.2 achieved 1.000 precision, 0.950 recall, and 0.974 F1. An audit requiring graph retrieval across 80 patients yielded 79 syntheses and 78 presentations without fallback; one output was blocked and one failed closed because of an invalid evidence key. Against monolithic GPT 5.6 on 56 matched patients, CareGraph was faster at 40.15 versus 49.62 seconds, shorter at 661 versus 1,163 words, and showed better exploratory lexical alignment with longitudinal targets; the baseline used fewer tokens and cited more raw evidence. Graph auditing verified provenance and deterministic retrieval; incremental graph effects on generation require paired evaluation. CareGraph offers a safety bounded foundation for intelligent personalized health systems.
[MA-9] Emergent aggregation from collective forag ing
【速读】：该论文旨在解决如何在缺乏直接社会激励的情况下，解释生物系统中群体聚集行为的涌现机制。传统模型通常假设个体因社交驱动力（social drive）而趋向于与邻近个体对齐或靠近，但本文提出一种新的视角：群体行为可由间接的、资源驱动的奖励机制自发产生。其核心解决方案在于，通过强化学习（reinforcement learning）让初始执行随机游走的觅食者仅基于自身发现可再生目标的个体奖励来优化行为策略，且这些个体只能感知同种个体而无法直接感知目标位置。随着视觉范围增大，代理从环境依赖的个体搜索模式突变为尺度无关的集体搜索模式，这一转变恰好伴随着空间聚集现象的出现。研究表明，这种集体相变是觅食最优性自然产生的副产物，无需任何针对聚集的直接奖励。一个简化的首次通过时间（first-passage）分析模型成功再现了两种搜索策略之间的突变式过渡。因此，该研究揭示了间接资源驱动奖励是一种普遍存在的、可引发涌现集体现象的机制。
链接: https://arxiv.org/abs/2608.28046
作者: Gorka Muñoz-Gil,Andrea López-Incera,Vide Ramsten,Giovanni Volpe,Thomas Müller,Hans J. Briegel
机构: University of Innsbruck, Department for Theoretical Physics(因斯布鲁克大学理论物理系); Universitat Autònoma de Barcelona, Departamento de Didáctica de la Matemática y las Ciencias Experimentales(巴塞罗那自治大学数学与实验科学教学系); Department of Physics, University of Gothenburg(哥德堡大学物理系); Department of Philosophy, Box 17, University of Konstanz(康斯坦茨大学哲学系)
类目: atistical Mechanics (cond-mat.stat-mech); Machine Learning (cs.LG); Multiagent Systems (cs.MA); Adaptation and Self-Organizing Systems (nlin.AO); Biological Physics (physics.bio-ph)
备注: 11 pages, 8 figures
Abstract:Collective behaviour in living systems is usually modelled as the outcome of a \emphdirect social drive: agents are rewarded, or hard-wired, to align with or approach their neighbours. Here we show that aggregation can instead emerge from an \emphindirect objective. We let reinforcement learning foragers, initially performing a random walk, optimize their dynamics from a purely individual reward for finding replenishable targets, while perceiving only their conspecifics and never the targets themselves. As the visual range grows, the agents undergo a sharp crossover from an environment-tuned individual search to a scale-agnostic collective one, and this crossover coincides with the onset of spatial aggregation. Thus a collective phase arises as a by-product of optimal foraging, without any direct reward for grouping. A minimal analytical first-passage model reproduces the transition as a crossover between the two search strategies. Our results identify indirect, resource-driven reward as a generic route to emergent collective phenomena.
自然语言处理
[NLP-0] A Formal Limitation on Learning Human Language From Textual Corpora
【速读】： 该论文试图解决的核心问题是：在仅依赖话语形式（utterance form）的情况下，听者是否能够准确还原说话者的真实意图。从信息论的角度出发，研究针对任意文本特征提取器（包括现代大语言模型的隐藏状态）所构成的听者模型，构建了听者从话语表征中恢复说话者意图概率的理论上限。其解决方案的关键在于揭示了话语形式对意义的不确定性可分解为两部分：一部分是不可消除的固有不确定性（由语言本身的内在特性决定），另一部分则仅能通过语境（特别是非语言性上下文）来缓解，而无法仅通过话语本身解决。这一分解表明，无论采用何种表示学习方法或多少训练数据，这些不确定性下限均不可逾越，从而确立了语言理解能力的根本性极限。实验在人工语言、汉语零代词消解以及颜色指称任务中验证了该理论的预测能力。
链接: https://arxiv.org/abs/2608.28560
作者: Emily Cheng,Ryan Cotterell
机构: Universitat Pompeu Fabra(庞培法布拉大学); ETH Zürich(苏黎世联邦理工学院)
类目: Computation and Language (cs.CL)
备注: this is a draft; comments welcome
Abstract:Can a listener recover what a speaker means from the form of an utterance alone? We answer this question information-theoretically, and for a listener given by any featurizer of text, including the hidden states of contemporary large language models. Modeling language use as a joint distribution over meanings, contexts, and utterances, we derive upper bounds on the probability that a decoder recovers a speaker’s intended meaning from a representation of the utterance. The bounds are governed by the uncertainty that form leaves about meaning, which splits into an irreducible part and a part that only (extralinguistic) context, but never the utterance alone, can resolve. Because these quantities are intrinsic to language, no representation, however much text or supervision produced it, can surpass them; the bounds hold whether the space of meanings is discrete or continuous. Experiments on artificial languages, Mandarin zero-pronoun resolution, and color reference provide empirical evidence in support of the theory.
[NLP-1] When Robots Mishear Us: Mapping the Safety Risks of Voice-Controlled Embodied AI
【速读】： 该论文旨在解决自动语音识别（ASR）错误在用户输入中可能导致具身人工智能（EAI）模型生成不安全输出的问题。其核心挑战在于，当用户通过语音交互输入指令时，若ASR系统产生错误，这些错误可能改变原意，使原本无害的指令变为具有潜在危害的指令，从而诱导EAI模型执行危险行为。解决方案的关键在于识别不同类型的ASR错误对EAI安全性的具体影响：部分错误虽保留语义结构但引入有害歧义，另一些则削弱模型拒绝不当请求的能力，导致不安全规划被生成并执行。研究通过模拟ASR错误并与现有安全评估基准（SafeAgentBench和POEX）结合，揭示了错误类型与安全风险之间的关联，并验证了自动纠错虽可在某些情况下降低风险，但并非普遍有效。因此，该研究强调了在构建鲁棒具身智能系统时，必须考虑语音输入环节的容错性与安全性，以应对由语音识别误差引发的潜在威胁。
链接: https://arxiv.org/abs/2608.28518
作者: Sihan Jia,Oliver Lemon
机构: Heriot-Watt University (赫瑞瓦特大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Robotics (cs.RO)
备注:
Abstract:We investigate whether automatic speech recognition (ASR) errors in user input can lead to unsafe outputs from Embodied AI (EAI) models. We find that ASR errors can lead to harmful instructions being accepted and executed by EAI models, thereby reducing safety. We simulate ASR errors and combine them with existing safety benchmarks (SafeAgentBench and POEX) to evaluate how different errors affect embodied AI safety. We find that some of them preserve semantic structure but increase harmful ambiguity, while others weaken the model refusal behaviour and allow unsafe plans to be generated and executed. We show that in some cases automatic correction of ASR errors can reduce the risk, but this is not always effective. Overall, we show that ASR errors lead to significant safety risks for embodied AI.
[NLP-2] Phoneme- and Word-Level Metrics Using Self-Supervised Speech Representations for Forced Alignment Evaluation EMNLP-2026
【速读】： 该论文旨在解决强制对齐（Forced Alignment）评估中依赖人工标注时间戳所带来的可扩展性与多语言适用性受限的问题。传统评估方法需大量人工标注的语音-文本对齐数据，难以支持大规模及跨语言分析。为此，论文提出两种基于自监督语音表示（Self-Supervised Speech Representations, SSL）的无参考（reference-free）语料级评估指标：音素聚类互信息（Phoneme-Cluster Mutual Information, PCMI）与词声学一致性评分（Word Acoustic Consistency Score, WACS）。PCMI通过衡量对齐音素标签与由SSL表示诱导出的聚类之间的一致性来评估对齐质量；WACS则利用动态时间规整（Dynamic Time Warping, DTW）计算重复出现词汇的声学表示序列间相似性，反映对齐的稳定性。通过随机与系统性扰动实验验证，两者在对齐错误引入时均表现出一致下降趋势。研究进一步在FLEURS数据集的85种语言上对比多种对齐系统，并在DoReCo数据集的45种语言上与人工标注对齐结果进行交叉验证，同时在两种音系复杂的低资源语言上进行评估。结果表明，这些指标能有效区分高质量与低质量对齐，且与基于时间戳的传统评估指标具有强相关性。研究表明，基于SSL的语音表示为实现可扩展、无需参考的强制对齐评估提供了可行路径。相关指标已开源为Python工具包，便于广泛使用。
链接: https://arxiv.org/abs/2608.28508
作者: V.S.D.S.Mahesh Akavarapu,Michael Daniel,Gerhard Jäger
机构: University of Tübingen (图宾根大学); University of Jena (耶拿大学)
类目: Computation and Language (cs.CL)
备注: Accepted at EMNLP-2026 (Findings)
Abstract:Forced alignment evaluation typically requires manually annotated timestamps, limiting large-scale and multilingual analysis. We introduce two corpus-level metrics based on self-supervised (SSL) speech representations for reference-free forced alignment evaluation: Phoneme-Cluster Mutual Information (PCMI) and Word Acoustic Consistency Score (WACS). PCMI measures agreement between aligned phoneme labels and clusters induced from SSL-speech representations, while WACS measures consistency of repeated word realizations using dynamic time warping similarity between word representation sequences. Using both random and systematic perturbations, we show that PCMI and WACS degrade consistently under alignment perturbations. We further analyze the metrics across multiple alignment systems on 85 languages from FLEURS, validate them against manually annotated alignments from 45 languages in DoReCo, and evaluate them on two phonologically complex low-resource languages. The metrics effectively separate high- and low-quality alignments and correlate strongly with timestamp-based alignment quality measures. Our results demonstrate that SSL-speech representations enable scalable, reference-free forced alignment evaluation. The metrics are available as an open-source Python package at this https URL.
[NLP-3] Ladders in Chaos: When How (and Perhaps Why) Does Test-Time Scaling Improve LLM Machine Translation EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在测试阶段（test-time）进行文本生成时，不同缩放策略对机器翻译性能的影响问题，尤其关注顺序采样（sequential sampling）与并行采样（parallel sampling）两种范式在翻译任务中的表现差异。其核心问题是：在有限的计算资源预算下，如何有效提升翻译质量，同时平衡流畅性与准确性之间的权衡。解决方案的关键在于揭示顺序采样机制的优势及其内在作用机理——通过逐步迭代生成过程，模型能够利用更丰富的目标侧上下文（target-side context），从而显著提升翻译的流畅性和自然度；然而，在高推理预算下，过度依赖此机制可能导致准确性的下降。研究进一步通过人工多维度分析和消融实验验证了顺序采样的鲁棒性，尤其是在不同采样温度下的稳定性，并指出其对上下文构建方式的高度敏感性，为未来优化顺序自提升机制提供了方向。
链接: https://arxiv.org/abs/2608.28496
作者: Di Wu,Sergey Troshin,Christof Monz,Antske Fokkens,Vlad Niculae
机构: University of Amsterdam (阿姆斯特丹大学); Vrije Universiteit Amsterdam (自由大学)
类目: Computation and Language (cs.CL)
备注: Accepted to Findings of EMNLP 2026
Abstract:Two forms of test-time scaling for Large Language Models (LLMs) have emerged as effective and widely adopted paradigms: sequential, in which later answer attempts depend on earlier ones, and parallel, such as i.i.d. sampling with reranking. In this study, we investigate their properties in translation. First, our study shows that sequential sampling has a higher performance ceiling, providing a more diverse and effective pool of samples, particularly under smaller sampling budgets. Second, we interrogate the nature of test-time scaling through a multidimensional manual analysis. Human analysis of the Best-of- N translations demonstrates that sequential sampling substantially improves translation fluency and naturalness, but can degrade accuracy when inference budgets are large. Finally, we suggest an explanation of the mechanism through which sequential scaling improves machine translation. Our controlled analysis partially attributes the success of sequential self-improvement to the model’s access to a larger target-side context. Ablation experiments on sequential sampling demonstrate its robustness across different sampling temperatures, while also revealing sensitivity to context construction, suggesting directions for future improvement.
[NLP-4] NL2AGBench: Benchmarking LLM Auto-Formalization for AlphaGeometry
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在将非正式的自然语言数学问题转化为形式化表示方面能力不足的问题，尤其针对神经符号几何系统（neuro-symbolic geometry systems）如AlphaGeometry所依赖的专用领域特定语言（Domain-Specific Language, DSL）输入需求。其核心挑战在于，尽管AlphaGeometry已达到接近国际数学奥林匹克竞赛（IMO）金牌水平的定理证明能力，但将自然语言几何题手动转换为符合其语法和逻辑的形式化表达仍存在显著可用性瓶颈。为此，研究提出Natural Language to AlphaGeometry Benchmark (NL2AGBench)，通过在AlphaGeometry内部进行执行验证（execution-based verification）来评估模型翻译质量，而非依赖文本相似度。实验结果表明，闭源模型在可执行翻译准确率上普遍超过80%，而开源模型即使在最大参数规模下也难以稳定保持几何约束并生成有效形式化表达。研究进一步构建了包含语法错误与逻辑错误的错误分类体系，并探索了少样本提示（few-shot prompting）、微调（fine-tuning）及人工引导提示（human-guided hinting）等缓解策略，均在多个模型族中取得可量化的性能提升。关键解决方案在于引入基于执行验证的评估框架与系统性错误分析机制，从而推动自然语言到形式化几何推理系统的高效转化。
链接: https://arxiv.org/abs/2608.28481
作者: Samuel Xiao,Judy Song,Rory Hu,Ziliang Zong
机构: Valley Christian High School; Vandegrift High School; Groton School; Computer Science Department, Texas State University
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Recent advances in large language models (LLMs) have demonstrated strong capabilities in natural language understanding and mathematical reasoning. However, their ability to translate informal mathematical problems into formal representations remains underexplored. This limitation is particularly important for neuro-symbolic geometry systems such as AlphaGeometry, whose theorem-proving engine requires inputs in a specialized domain-specific language (DSL). Although AlphaGeometry achieves near-IMO gold-medalist performance, manually converting natural-language problems into its formal syntax remains a significant usability bottleneck. To address this challenge, we introduce the Natural Language to AlphaGeometry Benchmark (NL2AGBench), which evaluates LLMs in translating English geometry problems into AlphaGeometry-compatible formal representations. NL2AGBench uses execution-based verification within AlphaGeometry to assess translation quality rather than relying solely on textual similarity. We evaluate ten state-of-the-art open- and closed-source LLMs across multiple parameter scales and analyze executable translation accuracy, syntactic correctness, and error characteristics. Our experiments reveal a substantial performance gap between closed- and open-source models: leading closed-source models achieve executable translation rates above 80%, while even the largest open-source models struggle to consistently preserve geometric constraints and produce valid formalizations. We introduce an error taxonomy distinguishing syntax and logic errors and investigate mitigation strategies, including few-shot prompting, fine-tuning, and human-guided hinting, which yield measurable improvements across multiple model families.
[NLP-5] Blind Men and the Elephant: Probing the Epistemic Myopia of LLM s under Long-Tail Divergent Knowledge
【速读】： 该论文旨在解决大语言模型（LLM）在处理长尾事实性问题时存在“认知狭隘性”（epistemic myopia）的问题，即模型在面对存在多种合理解释或争议性事实的问答任务时，往往仅能回忆起主流观点而忽略少数派观点，从而掩盖了其对多元知识表征的能力。其解决方案的关键在于提出ElephantBench——一个基于可审计的图结构生成流程构建的闭卷知识探测基准，该流程从低曝光网络语料库中检索相关文档，识别自然发生的观点分歧，并将其转化为多账户问答记录（multi-account QA records）。每个答案均经过原始文档与权威公开网络源的双重验证，并由人工标注者审核，确保可追溯性和可靠性。实验表明，即使最强模型也仅在52.4%的问题上能完整召回两种观点，多数情况下仅保留主流观点而遗漏次要观点；尽管模型规模扩大和推理优化可提升召回率，但无法根本消除这一不完整性。进一步分析揭示，语料暴露不平衡导致主流观点占据优势，而少数方观点的暴露程度越高，模型越可能实现完整召回。因此，ElephantBench不仅为诊断参数化记忆中的认知狭隘性提供了可复现的评测工具，其基于图结构的基准构建方法也为高效、可扩展地将长尾语料转化为溯源可查的知识探测数据集提供了范式支持，有助于推动下一代大语言模型在知识广度与认知严谨性方面的评估与进步。
链接: https://arxiv.org/abs/2608.28478
作者: Zhuoshi Pan,Junru Lu,Yan Qian,H. Vicky Zhao,Di Yin,Xing Sun
机构: Tsinghua University (清华大学); Tencent Youtu Lab (腾讯优图实验室); University of Warwick (华威大学)
类目: Computation and Language (cs.CL)
备注: 10 pages, 10 figurs, 1 table, under review
Abstract:Factual question answering (QA) typically assumes a single canonical answer, obscuring whether large language models (LLMs) retain divergent accounts of long-tail facts. To address this gap, we introduce ElephantBench, a closed-book knowledge probe comprising 1,094 questions generated through an auditable graph-based pipeline. The pipeline retrieves related documents from a low-exposure web corpus, identifies naturally occurring disagreements, and converts them into multi-account QA records. Each answer is verified against the originating documents and authoritative public web sources and is then reviewed by human annotators. Across 32 models, even the strongest model recovers both accounts on only 52.4% of questions, while on nearly all remaining questions it recalls one account but omits the other. Scaling model size and inference-time reasoning improve recall but do not eliminate this incompleteness. Corpus analysis further shows that exposure imbalance favors the dominant account, whereas greater minority-side exposure is associated with more complete recall. These findings establish ElephantBench as a reproducible knowledge probe for diagnosing epistemic myopia in parametric memory. More broadly, our graph-based benchmark construction pipeline provides an efficient and scalable way to turn long-tail corpora into source-traceable knowledge probes, supporting efforts to evaluate and advance the epistemic rigour of next-generation LLMs. Code is available at this https URL.
[NLP-6] ContextPilot: Teaching Agents for Proactive Context Management via Fine-grained RL EMNLP2026
【速读】： 该论文旨在解决长时程智能体任务中大语言模型（LLM）在多轮交互过程中面临的上下文膨胀问题，即持续积累的交互历史导致工作上下文不断增长，进而影响推理效率与可扩展性。现有主动式上下文管理方法虽能通过专用工具对自身上下文进行编辑，但仍存在三大局限：（1）工具集受限，缺乏全局规划、长期记忆和自适应压缩能力；（2）探索效率低下，未区分不同上下文操作对最终结果的异质性影响；（3）信用分配粗粒度，仅将轨迹级奖励分配给所有中间编辑动作，难以精准评估单个操作的价值。针对上述问题，本文提出ContextPilot框架，其核心在于系统性扩展工具集，引入规划（planning）、长期记忆（long-term memory）和软上下文卸载（soft context offloading）等新型工具，显著增强上下文管理的灵活性与智能性。同时，设计了一种面向上下文管理的强化学习方法，利用上下文变化与熵变识别关键编辑决策，并基于多分支轨迹信息估计动作级优势，实现细粒度信用分配。实验表明，ContextPilot在长上下文问答与深度搜索任务中均取得更优性能，且维持更紧凑的工作上下文，优于多种基线模型与基准测试。
链接: https://arxiv.org/abs/2608.28476
作者: Zhuoshi Pan,Qizhi Pei,Junru Lu,Honglin Lin,H. Vicky Zhao,Di Yin,Xing Sun
机构: Tsinghua University (清华大学); Tencent Youtu Lab (腾讯优图实验室); Shanghai AI Lab (上海人工智能实验室)
类目: Computation and Language (cs.CL)
备注: 10 pages, 6 figures, 5 tables, accepted to EMNLP 2026 (Main Track)
Abstract:Long-horizon agentic tasks require large language models (LLMs) to iteratively retrieve, integrate, and maintain dispersed information across multi-turn interactions, but preserving all interaction histories leads to a continuously growing working context. Recent proactive context management methods allow models to edit their own working context with specialized tools, yet they still face three key limitations: (1) a limited toolset restricted to search, deletion, and summarization, with no support for global planning, long-term memory, and adaptive compression; (2) inefficient exploration that treats context management actions uniformly despite their heterogeneous impacts on final outcomes; and (3) coarse-grained credit assignment that assigns the final trajectory-level reward to all intermediate context editing actions during RL. To bridge these gaps, we introduce ContextPilot, a proactive context management framework for long-horizon agentic reasoning. Our approach systematically augments the toolset with planning, long-term memory, and soft context offloading tools. We further propose an RL method tailored for context management, which uses context and entropy variation to identify critical editing decisions for branch sampling and estimates action-level advantages from all branched trajectories that pass through the corresponding context editing action. Experiments on long-context QA and deep search tasks show that ContextPilot achieves stronger performance with a more compact working context, consistently outperforming existing baselines across various base models and benchmarks. Code is available at this https URL.
[NLP-7] Stranger Fan or Peer? A Systematic Study on the Role of Interlocutor in Persona-Based Dialogue Generation
【速读】： 该论文旨在解决生成式对话系统中角色设定（persona）信息在训练、推理和评估三个阶段的可见性配置不一致所导致的模型行为偏差问题。传统方法通常将这一多阶段可见性因素笼统处理，忽略了其对模型是否真实表达角色特质或仅机械复制背景文本的关键影响。研究的核心在于通过分离这三个阶段的生物信息可见性，揭示其对模型生成行为的差异化作用。关键发现包括：（i）训练阶段的可见性比推理阶段更决定模型是主动体现角色特征还是陷入复制生物信息的“现象”；（ii）在训练中允许对话对方可见目标方生物信息，可显著减少模型对目标方背景文本的复制，而仅在推理阶段改变可见性则效果不稳定；（iii）当存在不对称披露（仅对话对方可见目标方生物信息）时，目标内容更易泄露至对话对方的回应中，且此类痕迹使作者识别任务更容易完成，尤其当对方发言可见时。这些结果表明，生物信息泄露是跨阶段可见性配置不当所引发的伪相关现象，强调必须明确区分并独立控制训练、推理与评估中的可见性设置以实现可靠的个性化对话生成。
链接: https://arxiv.org/abs/2608.28467
作者: Daniela Occhipinti,Malvina Nissim,Marco Guerini
机构: Fondazione Bruno Kessler(布鲁诺·凯斯勒基金会); University of Groningen(格罗宁根大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Persona-based dialogue systems are usually conditioned on speaker biography, but dialogues involve at least two participants, and who has access to whose biography can vary across training, inference, and evaluation. Prior work often neglected these aspects, obscuring mechanisms that only appear when biography visibility is toggled separately across training, inference, and evaluation, a three-stage factorisation that prior work has largely treated as a single factor. We study this factorisation on a dataset of dialogues paired with speaker’s biographies, varying whether the target and interlocutor speakers see each other’s biographies during training and inference, and using an LLM as a judge to perform author identification. We find that (i) training-time visibility, more than inference-time visibility, determines whether models express persona traits through dialogue or fall back on copying biographical text (a known problem/phenomenon in persona-based generation); (ii) models trained with interlocutor-biography visibility copy less target-biographical text than models trained without it, while changing visibility only at inference time has a less consistent effect; and (iii) under asymmetric disclosure, where only the interlocutor sees the target biography, target content leaks into interlocutor turns more often, and dialogues containing such traces are easier for the judge to identify, especially when interlocutor turns are visible. These results suggest that biography leakage into generated turns is an artefact of how interlocutor visibility is configured across training and inference, and separating the three stages is necessary.
[NLP-8] Acquire Repair Preserve: A Diagnosis-Guided Post-Training Recipe for Small-Model Dialogue Game Agents EMNLP2026
【速读】： 该论文旨在解决生成式 AI 在交互式对话游戏（interactive dialogue games）中表现出的持续状态保持、反馈理解与动态约束下的有效动作选择能力不足的问题。传统静态基准测试难以充分评估此类动态交互能力，而本文通过在 LM Playschool Challenge 挑战赛中使用一个 20 亿参数的开源模型进行研究，发现模型失败不仅源于广泛的知识缺失，更包含局部决策缺陷，如重复猜测、动作格式错误以及违背刚接收到的反馈等。为此，论文提出一种分三步的训练方案：首先通过监督微调（Supervised Fine-Tuning, SFT）提升模型在各类游戏中的泛化参与能力；其次针对特定对话游戏家族，利用逐轮局部偏好对（turn-local preference pairs）修复可机械验证的错误行为；最后确保模型在目标游戏之外仍保留通用能力。实验结果显示，该方法显著提升了官方评测中的公开 CLEMScore（从 10.67 提升至 38.92）和封闭域内得分（从 13.41 提升至 41.17），同时维持了静态性能基本不变（44.14 vs. 44.24）。值得注意的是，跨领域表现仍较低（7.88），且性能提升主要集中在目标家族的未见变体上，表明大规模 SFT 贡献了主要的能力提升，而精准故障检测下的逐轮局部监督虽有效，但迁移范围主要局限于同家族内部。
链接: https://arxiv.org/abs/2608.28458
作者: Nan Li
机构: Utrecht University (乌得勒支大学)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 14 pages, 14 tables; Accepted to the LM Playschool Workshop at EMNLP 2026; HF model card: this https URL
Abstract:Interactive dialogue games test a capability that static benchmarks largely leave implicit: a model must carry state across turns, interpret feedback, and choose valid actions under changing constraints. We study this setting in the LM Playschool Challenge with a 2B open-weight model, and find that many failures are not only broad knowledge failures but also local decision failures: repeated guesses, malformed actions, and violations of feedback that the model has just seen. These diagnostics motivate a training recipe organized around three steps: acquire broad game participation through supervised fine-tuning, repair mechanically verifiable failures within one targeted dialogue-game family using turn-local preference pairs, and preserve general capabilities beyond these dialogue games. In the official final evaluation, our submission improves public clemscore from 10.67 to 38.92 and closed in-domain score from 13.41 to 41.17, while approximately preserving aggregate static performance (44.14 vs. 44.24 for the baseline). Out-of-domain clemscore remains low at 7.88, with the largest gains concentrated in unseen variants of the targeted family. Our results suggest that broad SFT brings most of the model’s capability improvement; turn-local supervision can be effective when failure detection is precise, with observed transfer concentrated primarily within-family.
[NLP-9] Sliding-window beats linear attention
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在推理过程中因二次注意力机制（quadratic attention）导致的内存与能耗急剧上升的问题，尤其是随着序列长度增加，每新增一个标记所需存储的键（keys）和值（values）呈线性累积，造成不可持续的资源消耗。其核心解决方案是提出并验证滑动窗口注意力（Sliding Window Attention, SWA）结合“汇点”（sinks）机制的有效性，作为替代后训练线性注意力（Linear Attention）的更优方案。关键在于，SWA无需任何后训练过程，仅通过引入局部上下文窗口与可学习的汇点来压缩历史信息，即可实现极低的内存开销、极快的推理速度，并在长上下文推理任务（如Needle-in-a-Haystack和BABILong）中表现出显著优于现有线性注意力模型的性能（提升达2至10倍）。研究表明，尽管线性注意力在理论上具有潜力，但其实际性能往往依赖于从头训练或大量后训练，而SWA则以更低的成本实现了更可靠且高效的性能表现，因此作者强烈建议在降低推理内存开销时优先采用SWA而非后训练线性注意力模型。
链接: https://arxiv.org/abs/2608.28444
作者: Alexia Jolicoeur-Martineau,Rhea Sanjay Sukthanker,Pashmina Cameron,Emy Gervais
机构: Microsoft(微软); Applied Sciences Group (ASG)(应用科学组)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Due to the nature of quadratic attention, Large Language Models (LLMs) consume a lot of memory and energy. Every new token costs more than the previous one. For each additional token, the keys and values must be stored in memory indefinitely, which is unsustainable. Several alternatives have been proposed to fix the quadratic scaling problem, one of which is retrofitting LLMs to use Linear Attention. This idea has attracted a lot of attention, given its promise to solve the quadratic scaling problem with state-of-the-art performance at low cost. However, this line of research has not been properly compared to simpler baselines. In this work, we show that Sliding Window Attention (SWA) with sinks performs as well or better than post-trained Linear Attention models. We observe this across multiple LLMs on various downstream tasks. For long-context reasoning tasks (Needle-in-a-Haystack and BABILong), SWA achieves massively higher performance (2 to 10 times higher than linear attention). SWA requires no post-training, is extremely fast, and requires low memory; therefore, making it an extremely cheap and reliable solution. To reduce inference memory cost, we strongly recommend switching to SWA instead of post-training linear models. Linear attention models may have shown some promise, but they likely require to be trained from scratch or extensive post-training in order to even match SWA. Subjects: Computation and Language (cs.CL); Machine Learning (cs.LG) Cite as: arXiv:2608.28444 [cs.CL] (or arXiv:2608.28444v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2608.28444 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Alexia Jolicoeur-Martineau [view email] [v1] Fri, 28 Aug 2026 15:31:34 UTC (288 KB)
[NLP-10] Fidelity Is Not Enough: Dispatch-Level Instrumentation for Agent ic Datasheet Extraction EMNLP2026
【速读】： 该论文旨在解决生成式 AI (Generative AI) 在文档提取任务中存在隐蔽性错误（silent failures）的问题，即模型在未调用必要工具的情况下仍能生成看似合理的输出，导致提取结果虽通过“保真度”（fidelity）评估却实际为虚构内容。其核心解决方案在于构建一个基于工具调用记录的可观测性框架：通过在25个精心设计的声明上建立包含37项测试的智能体基准测试，系统性地记录每一次工具调用；进而开发两种检测工具——一个基于规则的失败归因分类器，以及仅依赖工具调用模式、不依赖输出内容的静默故障检测器。该检测器在207次正常通过保真度测试的运行中无误报，且能100%识别出刻意隐藏特定工具调用的50个预置故障，表明其具有高召回率，但无法衡量对“正确调用工具却仍错误回答”的情况的检测能力。此外，研究引入了一个因果实验舱作为独立验证源，仅能验证其中2个声明的物理可测性，其余35个因超出实验装置能力而不可测量，并据此提出不可物理量化的类型学分类。实证发现，尽管模型堆栈在部署中表现出不同程度稳定性，工具层的主要价值并非提升准确性，而是增强系统的可移植性与可观测性，其性能优势仅在文档超出上下文窗口限制时显现。
链接: https://arxiv.org/abs/2608.28439
作者: Qing Ye,Meng-Hsuan Lin
机构: Infineon Technologies AG (英飞凌科技公司); Neubiberg, Germany (德国纽比贝格)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted at EMNLP 2026 Industry Track. 7 pages + appendices
Abstract:One model passed our fidelity check without ever opening the datasheet. We found it while qualifying models for an internal extraction service: a structured-output constraint had silently disabled tool use, and the model answered anyway, with fabricated source text. Only the per-tool trace exposed it. Fidelity – whether an extracted value matches the source – is the standard measure for agentic document extraction, and it scores that run a success. We therefore log every tool call in an agentic benchmark of 25 hand-curated claims over three components, with 12 more on a fourth, 37 in all. From that dispatch record we build two instruments: a rule-based failure-attribution classifier, and a silent-failure detector whose two rules check only which tools were called, never the extracted value. The detector raises no flag on 207 clean fidelity-passing extractions across three model families, and recovers all 50 planted faults that withhold exactly the tools its rules check. The two results are not symmetric: the first bounds the false-positive rate, the second is recall by construction, and detection power against runs that call their tools and still answer wrongly is unmeasured. A second, independent oracle, a causal chamber that tests whether the datasheet’s claims hold under physical measurement, is intentionally partial: it confirms only what the apparatus can exercise, a verifiable envelope of 2 of those 37 claims, and we give a taxonomy of why the rest are not physically gradable. Under a controlled perturbation, fidelity passes throughout while the chamber verdict flips exactly at the measurement uncertainty. Across three deployed model stacks (one destabilised by its serving stack, not by any capability gap) the tool layer buys portability and observability rather than accuracy, and earns its premium only once a document outgrows the context window.
[NLP-11] Are These Modules Worth Their Cost? A Paradigm-Level Accuracy-Cost Analysis of In-context Learning Text-to-SQL
【速读】： 该论文旨在解决当前生成式文本到SQL（text-to-SQL）研究中缺乏对不同模块设计选择在成本-精度权衡方面贡献的量化分析问题。现有工作虽通过构建复杂流水线显著提升了端到端执行准确率，但普遍仅报告整体性能，未明确各组件的边际贡献与计算开销。为此，研究在统一控制实现下，针对五类核心模块共构建17种范式级配置，并系统评估其在四种具备不同能力水平与推理风格的基线模型上的边际收益与成本。关键发现表明：执行反馈精炼（execution-feedback refinement）是唯一在所有基线中均表现出高性价比且成本稳定的范式；其余多数模块的增益具有依赖于基线模型特性的条件性。此外，输入需求更受流水线结构影响，而输出需求则更敏感于基线生成行为。跨模块分析显示，堆叠策略普遍提升准确率，但增益构成随基线能力差异而异。更重要的是，在固定预算下，对中等性能基线进行更复杂的流水线工程化往往优于单纯升级至前沿大模型搭配简单流水线。最终，研究提炼出一套可迁移的成本感知分层指导原则，经验证可直接适用于五个额外基线模型，无需逐模块调优。
链接: https://arxiv.org/abs/2608.28432
作者: Jiayan Lin,Yujia Liu,Zijin Hong,Zheng Yuan,Yilin Xiao,Hao Chen,Qinggang Zhang,Xiao Huang,Feiran Huang
机构: Jinan University (暨南大学); 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)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Databases (cs.DB)
备注:
Abstract:Recent advances in in-context learning (ICL) text-to-SQL have substantially improved execution accuracy on public benchmarks by assembling increasingly elaborate pipelines around the base generator, yet existing studies typically report aggregate end-to-end accuracy, without quantifying the marginal accuracy-cost contribution of individual design choices. Consequently, providing a unified, paradigm-level cost-accuracy quantification remains a critical challenge for understanding and configuring modern text-to-SQL. To address this, we instantiate 17 paradigm-level configurations across five recurring modules of the ICL text-to-SQL pipeline under a single controlled implementation, and attribute each paradigm’s marginal contribution and incurred cost across all four backbones spanning diverse capability levels and reasoning styles. Our analysis reveals that execution-feedback refinement is the only paradigm whose benefit holds universally at consistently low cost, while most other modules help only under backbone-dependent conditions. Token accounting shows that input demand is more closely tied to pipeline structure, whereas output demand is more sensitive to backbone generation behavior. Cross-module analysis further shows that stacking improves accuracy on most backbones, although how the gains compose varies with backbone capability. We also find that a fixed budget is often better spent engineering a more elaborate pipeline over a mid-tier backbone than upgrading to a frontier model with a lean pipeline. These findings distill into an actionable, cost-aware tiered guideline that transfers to five additional backbones without per-paradigm search.
[NLP-12] A Unified Framework to Elicit Structured Feedback for Interpretable Multi-Trait Essay Scoring EMNLP2026
【速读】： 该论文旨在解决多特征自动作文评分（Multi-trait Automated Essay Scoring, AES）中评分与反馈之间一致性不足、各评分维度间相互独立导致的评分解耦问题，尤其针对现有反馈增强方法在评分与反馈生成过程中缺乏协同性、难以保证评分标准（rubric）对齐的缺陷。其解决方案的关键在于提出一种统一的自回归框架HiFTS，通过先生成分层思维链（hierarchical Chain-of-Thought, CoT）反馈，再联合预测各维度分数与整体分数，实现评分与反馈的端到端协同生成。该框架通过教师大模型（LLM）蒸馏获得基于评分量规（rubric）的分层CoT反馈，并采用组相对策略优化（Group Relative Policy Optimization）结合复合奖励函数，平衡得分一致性、校准性、反馈质量与结构有效性。此外，在推理阶段引入轻量级全局先验以提供整体性指导，抑制长文本推理过程中的偏差漂移。研究还构建了首个中文多特征AES数据集CFMS-34，包含951篇作文及34个基于量规的评分维度标注，实验表明HiFTS在多个基准数据集上均实现了高精度的综合与分项评分，并生成了连贯且符合评分标准的反馈。
链接: https://arxiv.org/abs/2608.28407
作者: Shihang Yang,Sanwoo Lee,Ningning Zhao,Yunfang Wu
机构: Peking University (北京大学); Beijing Normal University (北京师范大学)
类目: Computation and Language (cs.CL)
备注: 14 pages, accepted to EMNLP 2026 Findings. Code: this https URL
Abstract:Multi-trait Automated Essay Scoring (AES) requires rubric-grounded reasoning across interdependent traits, rather than isolated score prediction. Existing feedback-enhanced methods often decouple feedback from scoring or assess traits independently, weakening score–feedback consistency and rubric alignment. We propose HiFTS, a unified autoregressive framework that generates hierarchical CoT feedback before predicting trait-level and holistic scores. HiFTS distills rubric-grounded hierarchical CoT feedback from a teacher LLM and trains student models to jointly generate feedback and scores. HiFTS further applies Group Relative Policy Optimization with a composite reward balancing score agreement, calibration, feedback quality, and structural validity. At inference, a lightweight global prior provides holistic guidance to reduce drift during long-form reasoning. We also introduce CFMS-34, a Chinese multi-trait AES dataset with 951 essays annotated with holistic scores and 34 rubric-based traits. Experiments on CFMS-34 and ASAP++ show that HiFTS achieves strong holistic and trait-level scoring while producing coherent, rubric-aligned feedback.
[NLP-13] CultureConverse: A Multilingual Multi-turn Simulation Harness for Culturally Grounded Assistance in East and Southeast Asia EMNLP2026
【速读】： 该论文旨在解决当前大语言模型（LLM）文化评估中过于依赖单轮事实性问答（如多选题，MCQs）的局限性，无法有效衡量模型在多轮互动、基于文化背景的实际场景中提供实用帮助的能力。其核心问题是：现有评估方法未能捕捉用户在真实情境下对文化敏感型对话助手的多轮交互需求。为此，作者提出CultureConverse——一个可扩展、多语言的文化情境对话模拟与评估框架，覆盖东亚及东南亚10个地区、58个亚群体身份和7个应用领域。该框架通过生成带有评分的交互对话，使模型在部分信息条件下推断文化约束并提供协助。由此构建的CultureConverse-DS数据集包含14,610个基准评估样本和274,295条由“黄金标准”引导的对话。实验表明，经过高质量样本微调后，模型在本域任务中的协助质量显著提升，并在跨域任务（如文化多选题、安全分类）中表现出良好的迁移能力。人类标注实验验证了该评估框架作为人类判断代理的有效性。研究团队公开发布评估工具、数据集两个版本及评分提示，以支持对文化胜任力的交互式评估。解决方案的关键在于构建一个能反映真实文化语境下多轮协作需求的动态评估体系，突破传统静态、单轮评估范式。
链接: https://arxiv.org/abs/2608.28405
作者: Bryan Chen Zhengyu Tan,Weihua Zheng,Thong T. Doan,Bich Ngoc Doan,Jia Wang Peh,Xiaoyuan Yi,Jing Yao,Xing Xie,Nancy F. Chen,Zhengyuan Liu,JinYeong Bak,Wafi Shamdi,Soo Kai Chie,Liew Yu Siong,Aina Azyyati Binti Mohamad Rezal,Lew Yan Yan Vanessa,Huadan Wu,Dylan Raharja,Nadya Yuki Wangsajaya,Akane Fukushige,Kazushi Kato,Koji Inoue,Tatsuya Kawahara,Jaehyung Seo,Dongjun Kim,Seungyoon Lee,Zi Haur Pang,Rui Yang Tan,Charibeth Ko Cheng,Maria Regina Justina Estuar,Jann Railey Montalan,Pham Minh Duc,Roy Ka-Wei Lee
机构: Singapore University of Technology and Design (新加坡科技设计大学); Agency for Science, Technology and Research (新加坡科技研究局); École Polytechnique Fédérale de Lausanne (洛桑联邦理工学院); Microsoft Research Asia (微软亚洲研究院); Sungkyunkwan University (成均馆大学); Universiti Brunei Darussalam (文莱达鲁萨兰大学); China University of Petroleum (East China) (华东石油大学); Nanyang Technological University (南洋理工大学); Kyoto University (京都大学); Konkuk University (韩国中央大学); Upstage AI (韩国Upstage AI); Korea University (韩国高丽大学); De La Salle University (菲律宾德拉萨大学); Ateneo de Manila University (菲律宾马尼拉大学); AI Singapore (新加坡人工智能); University of British Columbia (不列颠哥伦比亚大学)
类目: Computation and Language (cs.CL); Computers and Society (cs.CY)
备注: EMNLP 2026
Abstract:Current cultural evaluations for large language models (LLMs) often reduce culture to single-turn factual recall via MCQs, failing to capture a common use case: users seeking practical help over multiple turns in culturally grounded scenarios. We introduce CultureConverse, a scalable, multilingual simulation and evaluation harness for culturally grounded assistant dialogue that covers 10 East and Southeast Asian regions, 58 subgroup identities, and 7 domains. Each simulated and evaluated episode produces a scored interaction where the assistant assists the user and infers cultural constraints from partial information. The resulting CultureConverse-DS dataset contains 14,610 benchmark (evaluation) episodes and 274,295 oracle-guided (gold-mode) dialogues. In our benchmark evaluation of 18 models, GPT-5 mini achieves the highest assistance quality. Human annotation experiments suggest that our evaluation framework is a sufficient proxy for human judgment. Performance gains from fine-tuning on 27,860 high-quality CultureConverse-DS samples improve in-domain assistance and transfer out-of-domain to cultural MCQ and safety classification benchmarks. We release the harness, both splits, and judge prompts to support interactive evaluation of cultural competency.
[NLP-14] BEACON: Behavior-Anchored Cross-Source Knowledge Graph Construction for Cyber Threat Intelligence
【速读】： 该论文旨在解决当前网络安全威胁情报（Cyber Threat Intelligence, CTI）知识图谱构建中面临的跨源异名问题，即同一威胁在不同报告中使用不同名称导致信息难以整合。现有方法多局限于单份报告内的局部信息提取，缺乏对跨源信息的统一建模能力。其解决方案的关键在于利用MITRE ATT&CK框架作为标准化攻击技术锚点，将攻击行为（attack behaviors）作为核心实体，将其余上下文实体（如威胁组织、攻击活动、受影响产品）及指标（Indicators of Compromise, IoCs）作为参与者与痕迹，通过攻击行为与ATT&CK技术的映射关系，将各报告生成的独立知识图谱统一到一个规范空间中。为此，论文提出BEACON框架，采用“先提议后验证”的范式，在第一阶段基于报告证据和官方ATT&CK定义约束大模型输出，有效抑制误分类与幻觉；第二阶段则采用分层对齐策略，按确定性由高到低依次利用字符级相似性、语义相似性及重叠技术邻域等信号进行图合并，并迭代聚合邻域信息以实现精准融合。为评估该方法，研究构建并发布了两个人工标注数据集，分别为报告级CTI抽取（8,395个元素）和跨源整合（3,487个元素）的基准，是目前同类任务中规模最大的公开数据集。实验表明，BEACON在两项任务上均显著优于所有基线，分别提升23%和9%。
链接: https://arxiv.org/abs/2608.28394
作者: Changze Li,Yutong Cheng,Tsania Camila Finnisa,Qian Cui,Wei Ding,Peng Gao
机构: Virginia Tech(弗吉尼亚理工学院); Dian Nuswantoro University(丹尼·努斯万托罗大学); Amazon(亚马逊)
类目: Cryptography and Security (cs.CR); Computation and Language (cs.CL)
备注:
Abstract:Cyber threat intelligence (CTI) is foundational to modern cyber defense, yet much of it resides in unstructured reports whose volume and heterogeneity far exceed manual analysis, motivating research on automatically constructing knowledge graphs from CTI reports. However, existing approaches mainly extract partial information within a single report, leaving the cross-source setting unexplored, where the same threat is given unrelated names. Our key insight is that attack behaviors, once mapped to MITRE ATTCK (a standardized catalog of attack techniques), can anchor the rest of a report. Attack behaviors are the adversarial actions a report describes, while contextual entities (e.g., threat actors, campaigns, and affected products) and Indicators of Compromise (IoCs; e.g., IP addresses) are their participants and traces. Attaching them to these anchors places every per-report graph in one canonical space. We realize this insight in BEACON, an LLM-driven framework for cross-source CTI knowledge graph construction. Its first stage extracts each report into a graph under a propose-then-verify paradigm, grounding candidates in report evidence and official ATTCK definitions, to suppress LLM misclassification and hallucination. Its second stage merges these graphs with a hierarchical alignment strategy that applies signals in decreasing order of determinism, from character-level and semantic similarity to overlapping technique neighborhoods, iterating as merges pool neighborhoods. No existing benchmark links entities to technique anchors or provides cross-source alignment ground truth. We therefore construct and release two human-annotated datasets from 34 sources: to our knowledge the largest for report-level CTI extraction (8,395 elements) and the first for cross-source consolidation (3,487). On them, BEACON outperforms all baselines by at least 23% and 9%, respectively. Subjects: Cryptography and Security (cs.CR); Computation and Language (cs.CL) Cite as: arXiv:2608.28394 [cs.CR] (or arXiv:2608.28394v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2608.28394 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-15] CamoDocs: A Poisoning Attack Against Retrieval-Augmented Language Models Using Camouflaged Documents EMNLP2026
【速读】： 该论文旨在解决检索增强生成（Retrieval-Augmented Generation, RAG）系统在面对外部文档污染攻击时的安全性问题，特别是针对数据投毒攻击中因直接包含目标查询而导致可被轻易检测的缺陷。现有攻击方法通常通过在恶意文档中显式插入目标查询以提升检索命中率，但这一做法会在词法和嵌入空间中引入可识别的特征，从而易被防御机制捕捉。为此，论文提出了一种名为CamoDocs的新型投毒攻击方法，其核心创新在于通过将对抗性文档伪装成良性内容，避免直接包含目标查询。具体而言，CamoDocs将合成的良性与对抗性文本片段进行混合，选取部分良性片段中的关键标记（token）替换为“分散标记”（dispersion tokens），以在嵌入空间中扩散恶意文档的表征，同时采用连贯性过滤策略控制文本可读性的下降。实验结果表明，CamoDocs在七种RAG防御机制、三种开源大模型及三个基准测试上均实现了较高的平均攻击成功率（ASR），且有效规避了基于查询重叠的简单检测手段；此外，该方法对专有模型（如GPT-5.4-mini和Claude-Haiku-4.5）仍保持显著攻击效果，平均ASR分别达到61.80%和55.09%。研究还发现，依赖大量擦除操作的聚类型防御（如TrustRAG）虽能降低攻击成功率，但会带来检索依赖型任务（如NeoQA）上的显著性能退化，揭示了安全与实用性之间的权衡。
链接: https://arxiv.org/abs/2608.28389
作者: Jaewon Jung,Haizhong Zheng,Hongsun Jang,Jaeyong Song,Beidi Chen,Jinho Lee
机构: Seoul National University (首尔国立大学); Carnegie Mellon University (卡内基梅隆大学)
类目: Cryptography and Security (cs.CR); Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026
Abstract:Retrieval-augmented generation (RAG) augments LLMs with external documents, but public or user-editable sources expose RAG systems to data poisoning: attackers can inject malicious documents to steer outputs toward targeted answers. Existing poisoning attacks often rely on query inclusion, inserting the target query into poisoned documents to improve retrieval; however, this creates lexical and embedding-space artifacts that make them easy to filter. We propose CamoDocs, a poisoning attack that avoids direct query inclusion by camouflaging adversarial documents among benign content. CamoDocs chunks synthesized benign and adversarial drafts, replaces selected tokens in benign chunks with dispersion tokens that spread poisoned-document embeddings, and applies coherence filtering to limit readability degradation. Across seven RAG defenses, three open-weight LLMs, and three benchmarks, CamoDocs achieves strong average ASR while avoiding query-overlap artifacts exploited by simple query detection. It also remains effective against proprietary models, achieving average ASRs of 61.80% on GPT-5.4-mini and 55.09% on Claude-Haiku-4.5. Finally, we show that erasure-heavy clustering defenses such as TrustRAG can reduce ASR, but only with substantial utility drops on retrieval-dependent benchmarks such as NeoQA. Code is available at this https URL.
[NLP-16] Semantic Head Specialization Guides Hybrid ViT Attention for Multimodal LLM s
【速读】： 该论文旨在解决多模态大语言模型（Multimodal LLMs）中视觉变换器（Vision Transformers, ViTs）缺乏有效混合注意力（hybrid attention）设计的问题，尤其针对当前对为何某些注意力模式表现更优尚无共识的现状。其核心解决方案在于揭示并量化注意力头在全注意力（full attention）设置下呈现出的语义角色分化现象——即部分注意力头专精于感知对象（object-specialist），而另一些则专注背景（background-specialist），这一现象被称为语义头专业化（Semantic Head Specialization, SHS）。研究提出SHS-Index作为量化指标，验证其能有效区分全注意力与分块窗口（chunk-window）ViT架构，并与下游任务性能高度相关。进一步分析发现，窗口交互（window interaction）、token序列化（token serialization）和局部Softmax分配（local softmax allocation）是塑造SHS的三个关键结构因素，并以此为基础构建了Ariadne Attention这一混合注意力机制。该机制在保持与全注意力相当性能的前提下，将注意力计算量降低6.5倍，覆盖22项图像与视频任务。研究成果确立了注意力头专业化为可度量的属性，为大规模多模态LLM中的混合注意力设计提供了可解释、可优化的理论框架与实践路径。
链接: https://arxiv.org/abs/2608.28383
作者: Chenhong He,Lei Li,Shicheng Li,Hanglong Lv,Lingpeng Kong,Qi Liu,Tong Yang,Shuhuai Ren
机构: Peking University (北京大学); Xiaomi Corporation (小米公司); The University of Hong Kong (香港大学)
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注:
Abstract:Hybrid attention dominates frontier LLMs, yet Vision Transformers (ViTs) in multimodal LLMs lack a satisfactory hybrid design, with no consensus on why certain attention patterns work better. To fill this gap, we study ViT attention heads and find they differentiate into object- and background-specialist roles, a pattern most pronounced under full attention; we call this Semantic Head Specialization (SHS). We propose SHS-Index to quantify this specialization, show that it distinguishes full-attention from chunk-window ViTs, and find that it strongly tracks downstream benchmark performance. We then identify three structural factors that shape SHS—window interaction, token serialization, and local softmax allocation—and use them as design principles for hybrid attention. Guided by these factors, we design Ariadne Attention, a hybrid that matches full attention on 22 image and video tasks at 6.5x less attention compute. Our findings establish head specialization as a measurable property for diagnosing and designing principled hybrid ViT attention at the multimodal-LLM scale.
[NLP-17] When Linguistic and Internal Confidence Diverge in Large Language Models
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在用户询问时所表达的言语自信（linguistic confidence）是否真实反映其内部置信度的问题。研究表明，言语自信与基于逻辑值（logits-based confidence）或语义熵（semantic-entropy-based uncertainty）等内在指标之间存在显著偏差，三类评估维度——关联性（association）、数值一致性（magnitude agreement）和校准性（calibration）——常呈现不一致。关键发现在于：尽管在较易样本或更强基线模型上言语自信与内部置信度的关联性有所提升，但整体上实例级关联性较弱；指令微调模型虽常报告更高自信且部分提升关联性，却伴随更大的置信差距和更差的校准表现；提示工程主要影响自信分布而非本质一致性，其中态度线索会虚增自信但无助于对齐，而评分示例在避免置信度坍缩时可保留排序信号。回归分析进一步表明，自信分数的分布特性是解释其与内部置信度对齐模式的主要因素，而模型元数据的影响在控制后相对较小。因此，研究支持“有损信道”（lossy-channel）视角：言语自信的分散分布可能携带有用的排序信息，但无法保证校准性。结论强调，在将言语自信用于下游可靠性流程前，必须采用多轴诊断方法进行系统评估。
链接: https://arxiv.org/abs/2608.28382
作者: Hefan Zhang,Bingquan Zhang,Ming Cheng,Saeed Hassanpour,Weicheng Ma,Soroush Vosoughi
机构: Dartmouth College(达特茅斯学院); Oakland University(奥克兰大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Users often ask large language models (LLMs) to report how confident they are, but it is unclear whether such linguistic confidence tracks the model’s internal confidence. We study this question across 8 classification tasks, 2 generation tasks and 30 models from three families. For classification, we compare linguistic confidence with logits-based confidence along three axes: association, magnitude agreement and calibration. For generation, we test whether linguistic confidence tracks semantic-entropy-based uncertainty. The axes frequently diverge. Instance-level association is weak on average, although it improves on easier items and for stronger base models. Instruction-tuned models often report higher confidence and sometimes show higher association, but they also have larger confidence gaps and worse calibration. Prompt design mostly changes the distribution of reported confidence. Attitude cues inflate confidence without improving alignment, while score exemplars can preserve rank-order signal when they avoid collapsed confidence values. Regression analyses show that distributional properties of confidence scores explain much of the observed alignment pattern, with model metadata playing a smaller role after controls. These results support a lossy-channel view of linguistic confidence. A more dispersed verbal confidence distribution can carry useful rank information, but it does not make the scores calibrated. Linguistic confidence should therefore be evaluated with multi-axis diagnostics before being used in downstream reliability pipelines.
[NLP-18] PersonaForge: Realistic Multi-Turn User Simulation for Agent ic Systems
【速读】： 该论文旨在解决当前大语言模型作为智能体工作流执行者时，训练与评估数据集普遍基于信息完整、单轮交互的假设，与真实用户多轮交互行为之间存在显著脱节的问题。实证分析显示，75.9%的真实用户-智能体会话为多轮交互，暴露出现有系统在模拟真实复杂交互场景上的不足。其解决方案的关键在于提出一种名为PersonaForge的用户模拟框架，通过四维人格空间（four-dimensional persona space）、基于真实用户统计校准的SOUL驱动行为控制，以及基于真实种子查询的逆向深度构建（Reverse Deep Construction）机制，生成高度拟真的多轮用户-智能体交互数据。基于该框架构建的6.3K条记录训练数据集及手动标注的138任务基准测试集PersonaForge-Bench，涵盖20余个专业领域并支持四维评分体系。实验表明，采用PersonaForge训练的Qwen3.5-27B模型在综合得分上提升4.1%，尤其在任务完成度（+6.0%）和响应质量（+6.8%）方面表现突出，且交互效率更高（减少对话轮次与工具调用次数），消融实验进一步验证了SOUL组件与自适应模拟机制的有效性。因此，PersonaForge及其基准体系为实现更贴近真实场景的智能体训练与评估提供了坚实基础。
链接: https://arxiv.org/abs/2608.28378
作者: Hanglong Lv,Dawei Zhu,Lei Li,Bowen Ye,Huaqiu Liu,Yifan Song,Bofei Gao,Weimin Xiong,Jinhao Dong,Chenhong He,Lingpeng Kong,Qi Liu,Tong Yang,Fuli Luo
机构: Peking University (北京大学); Xiaomi(小米); The University of Hong Kong (香港大学); Renmin University of China (中国人民大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Large language models are increasingly used as agentic workflow executors, yet existing training data and benchmarks largely assume informationally complete, single-turn queries. Our analysis of 16K real-world sessions shows that 75.9% of interactions are multi-turn, revealing a substantial gap between how users interact with agents and how such systems are trained and evaluated. We introduce \textbfPersonaForge, a user simulation framework for synthesizing realistic multi-turn user–agent interactions. PersonaForge combines a four-dimensional persona space, SOUL-driven behavioral control calibrated to real-user statistics, and Reverse Deep Construction grounded in authentic seed queries. Using PersonaForge, we construct a 6.3K-record training dataset and \textbfPersonaForge-Bench, a manually annotated 138-task benchmark spanning over 20 professional domains with four-dimensional scoring. Experiments on Qwen3.5-27B show that PersonaForge training improves the composite score by +4.1%, with gains across all four dimensions and the largest improvements in Task Completion (+6.0%) and Response Quality (+6.8%). Further analyses show that PersonaForge-trained agents use fewer turns and tool calls, suggesting improved interaction efficiency, while ablations confirm the contribution of SOUL components and adaptive simulation. Together, PersonaForge and PersonaForge-Bench establish a foundation for training and evaluating agents under realistic multi-turn user interaction.
[NLP-19] BanglaMed-QA: A Question Answering System for Healthcare Support in Bangla
【速读】： 该论文旨在解决低资源语言（如孟加拉语）在医疗问答（Medical QA）领域中因数据集匮乏和专用系统缺失而导致的健康信息获取障碍问题。其核心解决方案在于构建一个面向孟加拉语医学领域的结构化知识库——BanglaMed-QA，包含4,493个涵盖506种疾病、分属9个类别的问答对；通过引入领域特定的词根词典与同义词集合以增强语义理解，并结合词性标注实现代词消解；采用支持向量机（SVM）作为最优分类模型进行问题分类，并融合余弦相似度、杰卡德相似度、BM25及编辑距离等多种相似度度量，结合软投票与硬投票机制实现查询匹配。实验结果显示，系统在自动评估中达到95%的F1分数，人工满意度平均达0.9/1.0，验证了其在真实场景中弥合孟加拉语使用者医疗信息鸿沟的有效性。
链接: https://arxiv.org/abs/2608.28329
作者: Rowzatul Zannat,Abdullah Al Shafi,K. M. Azharul Hasan,Atia Shahnaz Ipa
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Accepted and presented at 3rd International Conference on Big Data, IoT and Machine Learning (BIM 2025)
Abstract:Medical question answering (QA) systems have become crucial tools for providing reliable health information. But they remain very unexplored for low-resource languages like Bangla due to limited datasets and systems tailored to these languages. To address this, we introduce BanglaMed-QA, a robust QA system specifically designed for the Bangla medical domain. The process begins with building a structured medical knowledge base that includes 4,493 QA pairs in 9 categories under 506 diseases. To improve semantic comprehension, domain-specific root word dictionaries and synonym sets are proposed, in addition to part-of-speech tagging for anaphora resolution. We adopt supervised machine learning models in which SVM is found to be the best model to categorize questions. Multiple similarity metrics, including cosine, Jaccard, BM25, and Levenshtein, are applied with soft and hard voting methods for query matching. The performance of the QA system has been evaluated in two aspects, with a 95% F1 score in an automated evaluation and an average human satisfaction rating of 0.9 out of 1.0. This validates the real-world application of BanglaMed-QA in closing the healthcare information gap for Bangla speakers.
[NLP-20] Layered LLM Defenses as an Ensemble: Access Tiers Inference Cost and the Measured Failure Correlation Between Defense Layers
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）安全防御中“防御堆叠（defense stacking）”有效性假设的可信性问题。现有实践普遍认为通过叠加多个防御层可实现安全性能的复合提升，但这一假设成立的前提是各防御组件在不同输入上失效——即“非重叠失败”（fail on different inputs），而该条件在现有LLM安全研究中虽被推荐却从未被实证测量。为此，论文提出两个关键工具：一是攻击者访问层级模型（Adversary Access-Tier Model, AATM），用于量化攻击者权限等级（从A0到A4）；二是成本模型，将防御机制按推理时开销分为五类，并与攻击者层级对齐以实现可比性分析。基于此，论文推导出防御堆叠的行为特性，发现覆盖范围在同层级内趋于饱和、成本随类别上升、误拒率呈并集累积、残余攻击成功率仅在独立条件下呈乘法下降。研究进一步通过实测验证了这种独立性假设的缺失：在一个七层防御堆叠上运行自适应攻击，15个可测量的防御对之间均呈现显著正相关性（φ值介于0.30至0.75），联合残余攻击成功率超出乘法预测上限达0.172。尽管通过行为难度分层可部分削弱相关性，表明其主要源于共因依赖（common-cause dependence），但该依赖关系仍顽固存在于置换推理、多数投票评分标签及外部校准阈值等情境下。值得注意的是，该堆叠拒绝了五分之四的良性提示，却在统计上与最强单层防御无异。研究揭示，这种依赖性源于架构本质——所有成员共享同一基础模型结构，而非采样差异，因此扩大成员池无法缓解该问题。由此得出结论：防御多样性虽有助于选择堆叠成员，但无法准确预测最终堆叠系统的实际表现，必须进行端到端的整体评估。
链接: https://arxiv.org/abs/2608.28327
作者: Abrar Alotaibi,Muhammad Shahid Jabbar,Sadam Al-Azani,Moataz Ahmed
机构: King Fahd University of Petroleum and Minerals (沙特法赫德国王石油与矿业大学); Imam Abdulrahman Bin Faisal University (阿卜杜勒拉赫曼·本·法伊斯尔大学); SDAIA-KFUPM Joint Research Center for Artificial Intelligence (SDAIA-KFUPM 人工智能联合研究中心)
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Practitioners defend large language models (LLMs) by stacking defenses, assuming the layers compound. A stack is an ensemble, and ensembles compound only under a condition the LLM security literature recommends but never measures: the members must fail on different inputs. Two instruments make that measurable. The Adversary Access-Tier Model (AATM) grades an adversary by the access it holds, from system-only (A0) to influence over training data (A4). A cost model sorts defenses into five classes of inference-time overhead; because two classes require training weights or reading activations, they tier the defender as AATM tiers the adversary. From these we derive how a stack behaves, and the quantities a defender cares about diverge: coverage saturates within a tier, cost rises by class, false refusals accumulate as a union, and residual attack success falls multiplicatively only under independence. We measure that independence. Running one adaptive adversary against a seven-layer stack, failure correlation is positive in all fifteen measurable pairs ( \phi from 0.30 to 0.75 ), and the joint residual exceeds the multiplicative prediction by up to 0.172 . Stratifying on behavior difficulty dissolves most of the association, so the dependence is predominantly common-cause, but it survives permutation inference, majority-vote grader labels, and externally calibrated thresholds. The same stack refuses four in five benign prompts while remaining statistically indistinguishable from its strongest single layer. The dependence is architectural rather than sampling-based: members correlate through the model they all wrap, so no wider member pool weakens it. Diversity therefore selects stack members but does not predict what an assembled stack delivers, which has to be measured end to end. Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL) Cite as: arXiv:2608.28327 [cs.CR] (or arXiv:2608.28327v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2608.28327 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Abrar Alotaibi [view email] [v1] Fri, 28 Aug 2026 13:36:06 UTC (760 KB)
[NLP-21] AIM: Anchor Identity Features Then Match for Multimodal Large Language Model Unlearning EMNLP2026
【速读】： 该论文旨在解决多模态大语言模型（Multimodal Large Language Models, MLLMs）在用户请求删除个人身份信息时所面临的隐私泄露问题，尤其针对无法获取保留图像（retain images）或真实答案的现实场景。现有遗忘方法通常依赖于访问保留图像或地面实况答案，这在实际应用中难以实现。为应对这一挑战，论文提出一种无需保留图像即可实现身份遗忘的方法——AIM（Anchor-based Identity Masking）。其核心创新在于：通过分析发现，身份相关问题与视觉感知类问题在微调后的隐藏状态中占据不同的空间区域，并呈现出不同的聚类模式——前者按人物聚集，后者按问题类型聚集。基于此观察，AIM采用两阶段策略：首先利用通用视觉提示锚定一个身份遗忘目标，随后在基于Fisher信息约束下，将视觉编码器匹配至该目标，从而有效抑制身份知识而保留通用视觉感知能力。实验表明，该方法在实现高效身份遗忘的同时，能较好地保持非删除对象的身份、先验知识及对同一图像的视觉理解性能。
链接: https://arxiv.org/abs/2608.28312
作者: Wonjun Lee,Jaehyuk Jang,Kangwook Ko,Hee-Seon Kim,Changick Kim
机构: Korea Advanced Institute of Science and Technology (KAIST); Daejeon, Republic of Korea
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注: Accepted to Findings of the Association for Computational Linguistics: EMNLP 2026
Abstract:Multimodal large language models (MLLMs) can memorize identity-specific facts about people in their fine-tuning data, creating privacy risks when a person requests deletion. Existing MLLM unlearning methods often assume access to retain images or ground-truth answers during deletion, which is unrealistic in many practical scenarios. We study identity unlearning when retain images are unavailable at deletion time. Our analysis shows that identity and visual-perception questions occupy distinct regions in fine-tuned hidden states and are organized differently: identity questions cluster by person, whereas perception questions cluster by question type. This suggests that identity knowledge can be suppressed without erasing general visual perception. Building on this observation, we propose AIM, a two-stage method that anchors an identity-forgetting target with a universal visual prompt and then matches the vision encoder to that target under a Fisher-based constraint. Extensive experiments show that AIM achieves competitive identity forgetting while preserving non-deleted identities, prior knowledge, and visual perception on the same images.
[NLP-22] VISTA: Verifier-Informed Student-to-Teacher Adaptation for On-Policy Self-Distillation
【速读】： 该论文旨在解决在基于策略自蒸馏（On-policy self-distillation, OPSD）框架下，由于教师模型（teacher）分布与学生模型（student）有效推理路径不一致时，导致单向监督信号误导学生模型的问题。标准OPSD将教师分布视为固定目标，仅更新学生参数，但未考虑教师在生成过程中可能因特权信息干扰而提供非最优指导，从而引发推理偏差。为应对这一挑战，本文提出验证器引导的学生到教师适应方法（Verifier-Informed Student-to-Teacher Adaptation, VISTA），其核心创新在于在保持原有学生更新机制的基础上，利用结果验证后的轨迹（outcome-verified rollouts）反向调整教师分布，使教师更贴近学生的真实推理路径。具体而言，VISTA仅在每个验证轨迹中对教师与学生间KL散度最大的前k个位置进行适应性更新，以聚焦关键分歧点。该方法无需额外采样或引入独立奖励目标，直接复用标准OPSD的轨迹与损失函数，实现高效集成。在AIME24、AIME25和HMMT25三个基准上，使用不同规模的Qwen3模型（1.7B、4B、8B）进行实验，VISTA在所有尺度均取得最高平均精度（Avg@12），相比OPSD分别提升0.6、0.7和2.1个百分点，充分验证了基于结果验证的师生双向适配机制的有效性，揭示了学生引导教师适应作为提升生成式推理性能的重要方向。
链接: https://arxiv.org/abs/2608.28306
作者: Zewen Ding,Zezhong Wu,Zhou Tao,Shida Wang,Shizhuo Hou,YongXiang Hua,Haoyu Cao,Linli Xu
机构: 未知
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:On-policy self-distillation (OPSD) improves reasoning by training a problem-only student on its own rollouts using dense token-level supervision from a privileged teacher that also sees a reference solution. However, standard OPSD treats the teacher distribution as a fixed target along the student’s rollout and updates only the student %, although – even though privileged conditioning does not guarantee that the teacher always provides the most appropriate target for problem-only reasoning. This one-way supervision can therefore misdirect the student when the teacher distribution is misaligned with valid student reasoning. We therefore introduce Verifier-Informed Student-to-Teacher Adaptation (VISTA), which preserves the standard OPSD student update while using outcome-verified rollouts to adapt the teacher toward the student distribution. Within each verified rollout, VISTA further restricts this adaptation to the top- k positions with the largest teacher–student KL divergence. Notably, VISTA reuses the rollout and loss function from standard OPSD, introducing no additional sampling or separate reward objective. Across AIME24, AIME25, and HMMT25 with Qwen3 models at 1.7B, 4B, and 8B, VISTA achieves the highest Avg@12 at every scale, improving over OPSD by 0.6 , 0.7 , and 2.1 points, respectively. These results demonstrate the value of student supervision from outcome-verified rollouts and highlight student-to-teacher adaptation as a promising direction for OPSD.
[NLP-23] A Probabilistic Interpretation of KV Cache Eviction
【速读】： 该论文旨在解决大语言模型推理过程中键值（Key-Value, KV）缓存管理中的缓存淘汰（KV eviction）问题，其核心挑战在于如何在不显著降低生成质量的前提下，通过合理淘汰部分缓存条目以提升吞吐量。现有方法多依赖启发式策略选择淘汰目标，缺乏理论支撑且未系统化建模。本文的关键贡献在于从概率推理的视角对KV淘汰问题进行形式化定义，并证明该问题在计算上是难解的；进一步提出将淘汰问题转化为期望估计任务，可通过采样近似求解；更重要的是，基于该概率框架，首次揭示了在解码阶段对被淘汰条目进行修正这一长期被忽视的问题具有可行性；同时指出现有方法本质上为零方差有偏估计器，可经简单调整实现解码时修正。实验表明，结合概率化淘汰与解码阶段修正的方法，在相同压缩预算下相比现有方法更具任务鲁棒性且性能相当，为高效、可靠的KV缓存管理提供了新的理论基础与实践路径。
链接: https://arxiv.org/abs/2608.28293
作者: Renato Geh,Alex Chen,Daniel Israel,Aditya Grover,Guy Van den Broeck
机构: University of California, Los Angeles(加州大学洛杉矶分校)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:The premise and promise of KV (cache) eviction is simple: higher throughput can be achieved by evicting some entries from the KV cache, at a negligible cost to quality. This holds empirically for many existing methods, though most rely on creative heuristics for selecting which entries to drop. Despite recent advances, the problem of KV eviction has remained informal in the literature. This paper aims to properly formalize this problem through the lens of probabilistic reasoning and reveal what can be learned from this perspective. Concretely, we (1) formalize the problem of KV eviction and, unfortunately, prove that it is computationally hard, (2) show that by framing it probabilistically, KV eviction reduces to the problem of expectation estimation, which can be approximated through sampling, (3) show that through this probabilistic interpretation, correcting for evicted entries during decoding—a previously ignored problem—becomes feasible, and (4) reveal that existing methods in the literature are zero-variance biased estimators that can be easily adapted in order to enable decode time correction. In practice, we show that this probabilistic version of KV eviction coupled with decode time correction is more robust to different tasks compared to existing eviction methods and achieves competitive performance at the same compression budget.
[NLP-24] Embedding Models for Stance-Aware Argument Retrieval
【速读】： 该论文旨在解决生成式语义检索中支撑性与攻击性论点的准确匹配问题，核心挑战在于如何在检索过程中同时实现对论点与主张之间主题相关性（topic-relevance）和立场正确性（stance correctness）的双重评估。现有密集嵌入模型（dense embedding models）在实际应用中表现出严重的非对称推理偏差，倾向于过度依赖主题重叠而忽略指令立场，导致检索结果偏离真实语义意图。此外，通过对比学习进行纠正后，模型又会陷入新的失效模式——过度修正，即过度关注极性关键词（如“supports”或“refutes”），从而忽视深层语义主题。为诊断这一现象，论文提出基于词项消融（word-ablation）的度量指标，并提出以数据为中心的解决方案：结合平衡论点课程（balanced argument curriculum）与大语言模型（LLM）增强的立场反转论点，强制嵌入模型学习更深层次的方向性逻辑，而非依赖表层词汇捷径。实验表明，对于足够强大的模型，该方法可有效缓解过矫正问题，显著提升面向立场感知的论点检索性能。
链接: https://arxiv.org/abs/2608.28283
作者: Angelo Sparacino,Francesca Toni,Adam Dejl
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: CMNA’26
Abstract:In computational argumentation, obtaining arguments that explicitly support or attack given claims is a critical precursor to downstream reasoning tasks. When these supporting and attacking arguments are to be retrieved using semantic search methods, they need to be assessed for topic-relevance to the claims of interest as well as for correctness of their (positive or negative) stance towards the claims. In this paper we explore how dense embedding models (hereafter, models), powering modern retrieval pipelines, can serve as the basis of semantic search incorporating this dual assessment. We show experimentally that existing models struggle with asymmetric reasoning, exhibiting a strong bias toward topical overlap while ignoring instructional stance. We also show that correcting this bias via contrastive training triggers a new failure mode where models over-correct, over-fixating on polarity keywords (e.g., “supports” or “refutes”) at the expense of the semantic topic. We thus introduce diagnostic word-ablation metrics to quantify this phenomenon and propose a data-centric solution. By implementing a balanced argument curriculum alongside LLM-augmented, stance-inverted arguments, we force the (embedding) models to learn deeper directional logic rather than exploiting superficial lexical shortcuts. Our evaluation demonstrates that, for sufficiently powerful models, this approach can alleviate the observed overcorrection, achieving further improvements in stance-aware argument retrieval.
[NLP-25] Synth-JDoc: Synthesizing a Japanese Document Image Dataset for OCR with Diverse Layouts and Embedded Images ICDAR2026
【速读】： 该论文旨在解决大视觉语言模型（LVLMs）在阅读日文文档图像中垂直书写文本时性能显著低于水平书写文本的问题。这一差距限制了其在文档视觉问答等应用中的实际效用，尤其在日文这种兼具竖排与横排书写形式的语言环境中尤为突出。现有解决方案依赖于高质量的OCR数据集，但人工构建成本高昂且难以扩展；而通过现有文档图像使用OCR模型提取文本的方式又面临识别误差及图像来源受限等挑战。为此，本文提出一种基于文本合成文档图像的创新方法：利用HTML与CSS生成包含多列布局、融合竖排与横排书写风格的合成文档，并引入由文本到图像模型生成的图像以增强视觉真实性；同时，通过施加噪声和退化滤波来提升模型对真实场景中图像质量下降的鲁棒性。实验表明，基于该合成数据集微调的模型在垂直日文文本识别任务上表现最优，验证了该方法的有效性。其核心贡献在于构建了一个兼具高真实感、多样化排版结构和强鲁棒性的合成OCR数据集，为提升LVLM在复杂书写模式下的文本理解能力提供了有效支持。
链接: https://arxiv.org/abs/2608.28248
作者: Keito Sasagawa,Shuhei Kurita,Daisuke Kawahara
机构: Waseda University (早稻田大学); NII LLMC (日本国立情报学研究所语言模型研究中心); NII (日本国立情报学研究所)
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注: Accepted to ICDAR 2026, 17pages, 5 figures
Abstract:The ability of Large Vision Language Models (LVLMs) to read text within document images is crucial, as it enables various applications such as Document Visual Question Answering. To enhance the text-reading capabilities of LVLMs, high-quality OCR datasets are essential. This need is particularly critical for Japanese documents, which often feature vertically written text alongside horizontally written text. Current LVLMs demonstrate considerably lower performance on vertically written Japanese text than on horizontally written text, necessitating specialized OCR datasets to bridge this gap. However, manually constructing OCR datasets is expensive and difficult to scale. Alternatively, constructing datasets by extracting text from existing document images using OCR models introduces challenges, such as text recognition errors and the prerequisite of sourcing document images. To address these issues, we construct an OCR dataset by synthesizing document images directly from text. Leveraging HTML and CSS, we generate multi-column documents that incorporate both vertical and horizontal writing styles. Furthermore, to ensure the visual realism of the documents, we embed images generated by text-to-image models within the layout. Additionally, to foster model robustness, we apply noise and degradation filters to the synthesized document images. In our experiments, we compared the performance of models fine-tuned on our synthetic dataset against baselines fine-tuned on synthetic datasets from prior work and those generated by a high-performance text-to-image model. Evaluation results demonstrate that our synthetic dataset is the most effective approach for improving LVLM performance on reading vertically written Japanese text. Our dataset and code are publicly available (this https URL). Comments: Accepted to ICDAR 2026, 17pages, 5 figures Subjects: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL) Cite as: arXiv:2608.28248 [cs.CV] (or arXiv:2608.28248v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2608.28248 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-26] Stay Within Your Bounds: Distance-Guided Decoding for Guaranteed Context-Free Grammar Compliance EMNLP2026
【速读】： 该论文旨在解决大语言模型在生成结构化输出（如代码、JSON、SQL）时，因语法约束导致的语法不合法问题。尽管已有基于上下文无关文法（context-free grammar, CFG）的语法约束解码方法通过强制前缀局部可扩展性来保证生成结果的语法有效性，但在存在分词器-文法不匹配及有限解码预算的情况下，仍可能出现看似可行的前缀最终无法被目标文法接受的问题。为此，论文提出一种基于下推自动机（pushdown automaton, PDA）的前瞻引导解码框架：离线阶段，构建带有可达性标签和到接受状态上界距离的有界下推摘要；在线阶段，利用这些预计算的估计信息指导具有前瞻感知的剪枝与束搜索策略。该方案的关键在于通过形式化的方法对文法结构进行静态分析，从而在解码过程中动态评估当前前缀向有效完成演进的可能性，实现语法上的严格正确性（syntactically sound），即所有输出均被目标文法接受。实验表明，该方法在JSON、SQL和线性时序逻辑（LTL）任务中不仅实现了稳定的语法正确性，且在完成质量上优于现有基线方法。
链接: https://arxiv.org/abs/2608.28229
作者: Vincenzo Collura,Karim Tit,Eleonora Giunchiglia,Mike Papadakis,Maxime Cordy
机构: University of Luxembourg(卢森堡大学); Imperial College London(帝国理工学院)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Formal Languages and Automata Theory (cs.FL); Machine Learning (cs.LG)
备注: EMNLP 2026 Findings, Long Paper
Abstract:Grammar-constrained decoding helps large language models produce syntactically valid structured outputs, such as code, JSON, and SQL. For context-free grammars, many practical decoders enforce local prefix feasibility: each token must keep the current prefix extendable to some valid completion. Yet, under tokenizer-grammar mismatch and finite token budgets, feasible prefixes may still fail to reach acceptance. We propose a lookahead-guided decoding framework for context-free grammars based on pushdown automata. Offline, we compute bounded pushdown summaries with reachability labels and upper-bound distances to acceptance. Online, these estimates guide horizon-aware pruning and beam search. The resulting decoder is syntactically sound: every output is accepted by the target grammar. Experiments on JSON, SQL, and Linear Temporal Logic (LTL) show both consistent syntactic validity and improved completion quality over existing baselines.
[NLP-27] xt Restoration of Ancient Documents with Language Models
【速读】： 该论文旨在解决古籍手稿因物理缺损（lacunae）导致文本缺失的修复问题，尤其关注在实际考古与文献复原场景中如何利用语言模型实现有效文本恢复。其核心挑战在于：缺损区域的上下文信息不完整、模型对文本结构的感知能力受限，以及模型分词（tokenization）边界与真实缺损边界之间的错位。解决方案的关键在于提出多种贴近现实的模拟场景，并根据文档结构特征（如公式化内容与非公式化内容）选择适配的语言模型架构；同时设计多类解码策略，以缓解分词边界与缺损区域不一致带来的性能偏差。研究发现，尽管完全自动化修复尚不可行，但该方法可作为辅助工具显著提升古文字学家（paleographers）的工作效率，且模型表现高度依赖于待修复文本的结构性质及缺损长度是否已知。本研究首次系统分析了公式化与非公式化内容在恢复中的差异性影响，以及缺损长度信息的可利用性，为构建面向古籍修复的智能辅助系统提供了实证依据与技术指南。
链接: https://arxiv.org/abs/2608.28170
作者: Shibingfeng Zhang,Edoardo Caraffa,Annafelicia Zuffrano,Maddalena Modesti,Giovanni Colavizza
机构: University of Bologna(博洛尼亚大学); Centre for Digital and Computational Humanities, University of Copenhagen(哥本哈根大学数字与计算人文中心)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Purpose - This study investigates the feasibility of restoring missing text caused by physical lacunae in damaged ancient manuscripts using language models. Methodology - The study proposes different scenarios to replicate real-world conditions. Language models of different architectures are applied according to their suitability to each scenario. We also propose several decoding strategies that further enhance performance and address the discrepancy between lacuna boundaries and the models’ tokenization schemes. Findings - The results reveal that text restoration of these documents cannot be fully automated, but it can serve as a useful tool to assist paleographers in their work. Model performance varies greatly depending on which structural part of the document needs to be restored and whether the character length of missing text is available. Originality - This is the first study and to analyze model performance on formulaic and non-formulaic content and the impact of lacuna length awareness in manuscript restoration. Both are recurring challenges in paleographers’ manual restoration work. Through systematic comparison and both qualitative and quantitative analysis of different models’ performance under varying settings, this study offers a guideline for developing assistive tools to support paleographers. Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI) Cite as: arXiv:2608.28170 [cs.CL] (or arXiv:2608.28170v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2608.28170 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Shibingfeng Zhang [view email] [v1] Fri, 28 Aug 2026 10:36:15 UTC (118 KB)
[NLP-28] FinExam-10K: When Retrieval Helps Financial Reasoning ?
【速读】： 该论文旨在解决专业金融考试（涵盖CFA和FRM全体系）中缺乏统一评估基准的问题，即现有模型难以在融合领域知识、复杂计算与判断推理的综合任务上进行有效评估。其解决方案的关键在于构建并发布FinExam-10K——目前已知规模最大的英文金融专业测评基准，包含10,198道由专家重新标注的题目，覆盖CFA一至三级及FRM一、二部分。为区分题库覆盖范围与上下文完备性对推理的影响，研究提出“全覆盖追踪”（Full-Coverage Track）与“上下文完备推理追踪”（Context-Complete Reasoning Track）双轨评估机制。核心创新在于引入一种仅基于公开数据训练的门控机制（gate），根据问题特征和初始回答决定是否调用FunctionGraph-RAG增强推理流程。实验表明，在5,088道预留测试题上，该门控机制仅在7.9%的问题上调用增强模块，但将准确率从70.83%提升至71.23%（p = .0446），实现统计显著的净增益，验证了轻量级动态调用策略在提升金融专业推理能力中的有效性。
链接: https://arxiv.org/abs/2608.28155
作者: Yan Lin,Jingyu Sun,Zhongliang Guo,Qing Li,Zhuohan Xie,Yuxia Wang
机构: Newcastle University (纽卡斯尔大学); University of Manchester (曼彻斯特大学); University of Melbourne (墨尔本大学); University of Aberdeen (阿伯丁大学); University of Groningen (格罗宁根大学); MBZUAI (MBZUAI); INSAIT (INSAIT); Sofia University “St. Kliment Ohridski” (索非亚大学“圣克莱门特·奥赫里德斯基”)
类目: Computation and Language (cs.CL)
备注:
Abstract:Professional financial examinations require models to combine domain knowledge, calculation, and judgment, yet no benchmark covers the full CFA and FRM structure under one protocol. We introduce FinExam-10K, to our knowledge the largest reported English benchmark for this setting, with 10,198 expert-reannotated questions spanning CFA Levels I-III and FRM Parts I-II. We release 5,110 questions and sequester 5,088 for a quarterly maintained leaderboard. To separate coverage from local answerability, we report a 10,198-item Full-Coverage Track and a 7,625-item Context-Complete Reasoning Track, which is the primary basis for claims about reasoning from the supplied record. Across 17 models, the best accuracy is 85.29% overall. On the frozen Hard band, the best score is 34.68% on the Full-Coverage Track and 54.57% on the 372 context-complete items. All 17 models share 47 context-complete failures. Function-RAG and FunctionGraph-RAG rescue hundreds of errors but also overturn many correct answers, producing little or negative net gain. A gate trained only on public data decides from the question and initial response when FunctionGraph-RAG should run. On the 5,088 held-out items, the gate invokes FunctionGraph-RAG for 7.9% of questions and improves accuracy from 70.83% to 71.23% (p = .0446).
[NLP-29] H-Scale: Hessian-Guided Scale Refinement for NVFP4 Sub-Byte LLM Inference
【速读】： 该论文旨在解决生成式 AI（Generative AI）在NVIDIA Blackwell架构上进行大语言模型（LLM）推理时，因采用超细粒度的NVFP4格式所引入的每组缩放因子（per-group scaling factors）选择难题。现有后训练量化（PTQ）方法主要聚焦于优化量化权重值，而忽视了对敏感且庞大的缩放因子空间的有效优化。针对这一问题，本文提出H-Scale，一种轻量级的后处理方法，用于实现NVFP4格式下每组缩放因子的精细化调整。其核心创新在于不直接最小化权重重建误差，而是基于校准激活数据构建一个对角线二阶代理（diagonal second-order proxy），以选择符合硬件约束的缩放因子，从而更直接地控制层输出扰动。H-Scale可作为多种NVFP4量化流程中RTN型缩放选择的即插即用替代方案，仅需少量离线校准，且在推理阶段引入严格零开销，实验表明其在主流LLM上能显著提升多种NVFP4基线性能，并使多个变体逼近BF16参考精度。
链接: https://arxiv.org/abs/2608.28113
作者: Hao Yu,Zheng Li,Dayiheng Liu,Jianwei Zhang
机构: Qwen Team, Alibaba Inc.(通义实验室，阿里巴巴集团)
类目: Computation and Language (cs.CL)
备注:
Abstract:The NVIDIA Blackwell architecture, with native support for the ultra-fine-grained NVFP4 format, opens new opportunities for accelerating large language model (LLM) inference. NVFP4’s micro-block design, such as a group size of 16, offers strong representational flexibility for capturing local weight distributions and isolating outliers, but it also introduces a large and highly sensitive space of per-group scaling factors. Existing post-training quantization (PTQ) methods primarily focus on refining quantized weight values, leaving this scale-selection step underexplored. To address this gap, we propose \textbfH-Scale, a lightweight post-processing method for NVFP4 per-group scale refinement. Instead of minimizing plain weight reconstruction error, H-Scale selects hardware-valid group scales using a diagonal second-order proxy derived from calibration activations, thereby targeting layer output perturbation more directly. It is designed as a drop-in replacement for RTN-style scale selection in diverse NVFP4 pipelines, requires only modest offline calibration, and introduces strictly zero overhead at inference time. Under a fixed evaluation protocol, experiments on mainstream LLMs show that H-Scale generally improves a broad range of NVFP4 baselines and brings several variants closer to the BF16 reference.
[NLP-30] Speculative Probing: LLM Monitoring at Speculative-Decoding Cost
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）推理过程中实时分类任务在准确性与效率之间难以兼顾的核心问题。现有方法要么因缺乏上下文感知能力（如仅基于单个隐藏状态向量进行判断，忽略位置间交互）而影响精度，要么通过引入专用分类器模型（如Llama Guard、Qwen Guard）或对所有标记的隐藏状态进行计算并聚合（如MultiMax），导致显著的计算开销，形成内在的效率-精度权衡。其解决方案的关键在于复用近期大模型中已有的推测解码（speculative-decoding）模块，通过在目标序列末尾附加一个经过训练的软提示（soft prompt），将该模块重构为高效的序列分类器。由于推测解码流程中键值缓存（KV cache）已在GPU内存中，因此该方法在推理阶段添加分类任务几乎不引入额外开销。实验结果表明，所提出的轻量级探测器在四个不同任务上均持续优于零样本的GPT-5.4-mini，并在多语言提示安全检测任务中达到甚至超越专用8B规模安全分类器（如Qwen3Guard-Gen-8B、Llama-Guard-3-8B）的性能，且无需运行完整的大语言模型，实现了高效率与高质量分类的统一。
链接: https://arxiv.org/abs/2608.28099
作者: Collin Zhang,Tingwei Zhang,Vitaly Shmatikov
机构: Cornell Tech(康奈尔科技校区)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Real-time classification during language model inference is valuable for safety filtering, behavioral analysis, and model monitoring, but current approaches force a trade-off between accuracy and efficiency. Hidden-state probes are fast but limited: they are either not context-aware: operating on a single vector and cannot model interactions across positions; or they are very costly: having dedicated classifier models (Llama Guard, Qwen Guard, LLM-as-judge) or performing computation on hidden states for all tokens and then pooling the results (MultiMax). This shows an intrinsic trade-off between efficiency and accuracy. However, we find that the speculative-decoding module in recent LLMs can be repurposed for efficient high-quality classification. By appending a trained soft prompt at the end of the target sequence, we can repurpose the speculative-decoding module into a sequence classifier. At inference time in a speculative-decoding pipeline, the KV cache is already in GPU memory, so classification adds negligible overhead. We evaluate on four classification tasks across four models (Qwen3.5-4B, 9B, 27B, MiniCPM4.1-8B). Our small probes consistently outperform zero-shot GPT-5.4-mini and, on multilingual prompt safety, match or beat specialized 8B safety classifiers (Qwen3Guard-Gen-8B, Llama-Guard-3-8B) without running a full LLM. Subjects: Artificial Intelligence (cs.AI); Computation and Language (cs.CL) Cite as: arXiv:2608.28099 [cs.AI] (or arXiv:2608.28099v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.28099 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-31] CNeo-Bench: Diagnosing Large Language Models on Chinese Neologisms
【速读】： 该论文旨在解决中文新词（Chinese neologisms）在大语言模型（LLM）中理解与生成能力不足的问题，尤其关注其独特的语言机制，如音近替换（如“886”代表“拜拜”）和字形拆解等。其核心挑战在于模型不仅需准确描述新词含义，还需掌握其底层生成机制并还原原始形式。解决方案的关键是提出首个系统性的基准测试CNeo-Bench，包含4,759个中文新词及其定义，并按语言机制划分为五类顶层与九类子类别；同时设计双层评估框架，区分模型对新词的描述能力与对其生成机制的操作能力。实验表明，当前18个主流大模型在定义生成任务中平均表现低于40%，且在多个子类别中存在“识别-操作鸿沟”：模型可正确描述新词，但在源形式还原任务中常以语义等价的改写替代原始形式，无法恢复原生表达。对1,058个高难度样本的少样本分析进一步揭示，尽管上下文示例能缓解部分难题，仍存在显著残余错误，表明该问题远超提示工程（prompting）所能解决的范畴，亟需更深层的语言认知建模。
链接: https://arxiv.org/abs/2608.28053
作者: Kaiyan Zhao,Zhongtao Miao,Zheyong Xie,Shaosheng Cao,Yoshimasa Tsuruoka
机构: The University of Tokyo(东京大学); Xiaohongshu Inc.(小红书公司); Tsinghua University(清华大学)
类目: Computation and Language (cs.CL)
备注: Work in progress
Abstract:Chinese neologisms exploit diverse and unique linguistic mechanisms, such as phonetic substitution (e.g., 886 for ``bye-bye’') and visual character decomposition that are rare in other languages. We introduce CNeo-Bench, a benchmark of 4,759 such neologisms with reference definitions, organized into five top-level categories and nine subcategories by the linguistic mechanism behind each expression. CNeo-Bench is paired with a two-tier evaluation framework that separates whether a model can describe a neologism from whether it can operate on its underlying mechanism. Evaluating 18 LLMs, we find that Chinese neologisms remain an open challenge; most models fall below 40% on definition generation, and on several subcategories a systematic recognition-manipulation gap emerges: models describe neologisms correctly but, in source-form restoration tasks, substitute a semantic equivalent (paraphrase) for the source form rather than producing the source form itself. A few-shot analysis on 1,058 hard items shows that in-context examples can solve many difficult cases, but leave a noticeable portion of errors remaining, indicating challenges beyond prompting alone can address.
[NLP-32] SimpCue: Cue-Based Prompting for Multilingual Text Simplification
【速读】： 该论文旨在解决多语言句子级易读性简化（Easy-to-Read simplification）中如何有效提升生成质量的问题，尤其关注在提示（prompt）中引入显式句法复杂度语言学线索（linguistic cues）是否能改善大语言模型的简化效果。其解决方案的关键在于对比三种不同提示策略：基础提示（baseline prompt）、基于人工标注黄金线索的提示（gold-cue prompt）以及基于自动预测线索的提示（predicted-cue prompt）。研究发现，尽管自动预测线索提示在四项评估指标（SARI、BLEU、chrF、BERTScore）上整体表现最优，但相较于基线提升有限；而黄金线索提示在不同语言间表现不一致，未能稳定超越基线。这表明，基于线索的提示方法虽对多语言易读性简化具有一定的影响，但其增益效应微弱且高度依赖具体评价指标与目标语言。
链接: https://arxiv.org/abs/2608.28042
作者: Mehrzad Tareh,Horacio Saggion,Stefan Bott
机构: TALN Group, Universitat Pompeu Fabra (UPF); Barcelona, Spain
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted at CLEAR-TEXT 2026: Readability and text simplification workshop at the International Conference Computational Linguistics in Bulgaria (CLIB 2026)
Abstract:Text simplification aims to make complex texts easier to understand while preserving their original meaning. Recent large language models can perform simplification through prompting, but it remains unclear whether adding explicit linguistic information about sentence complexity to the prompt improves their outputs. We investigate this question for multilingual sentence-level Easy-to-Read simplification in Catalan, Spanish, and Italian. Using Qwen3-8B, we compare a baseline prompt, a gold-cue prompt enriched with gold linguistic cues, and a predicted-cue prompt enriched with automatically predicted cues. We evaluate the outputs using SARI, BLEU, chrF, and BERTScore, and complement this evaluation with a manual qualitative analysis. Predicted-cue prompting obtains the best overall scores across all four metrics, although the gains over the baseline are small. Gold-cue prompting does not consistently improve over the baseline, and results vary across languages. These findings indicate that cue-based prompting can influence multilingual Easy-to-Read simplification, but its benefits are modest, metric-dependent, and language-dependent.
[NLP-33] A Shaky Voice Is Not Always a Dodge: Benchmarking Textual and Vocal Evasion Detection in Earnings Calls EMNLP2026
【速读】： 该论文旨在解决传统收益电话会议（earnings calls）中逃逸行为检测仅依赖文本内容、忽视语音维度的局限性问题。现有方法将逃逸视为单一维度现象，而本文指出，口头沟通中的逃逸本质上是多维度的：不仅包括高管所言内容，其表达方式（如语调、语气等声学线索）亦蕴含独立且互补的信息。为此，论文提出DualEvasion，一个融合文本与音频的多模态逃逸检测基准，涵盖60场收益电话会议中的505个问答对，每条数据均包含两个独立标注标签：文本层面的逃逸类型（直接 vs. 逃避）以及语音层面的声学特征——以说话人自信度（confident vs. unconfident）作为代理指标。实验表明，当前最先进的多模态模型在识别低自信响应时表现不佳，其根本原因在于模型倾向于孤立地解析声学线索，而非相对于说话人自身基线进行相对评估。尽管引入说话人级参考信息可带来小幅性能提升，但模型与人类表现之间仍存在显著差距，凸显了理解语音上下文语境在逃逸检测中的关键挑战。
链接: https://arxiv.org/abs/2608.28040
作者: Mirae Kim,Seonghun Jeong,Youngjun Kwak
机构: Financial Tech Lab, KakaoBank Corp.(金融科技实验室，KakaoBank公司); Yonsei University(延世大学)
类目: Computation and Language (cs.CL); Sound (cs.SD)
备注: EMNLP 2026 Findings
Abstract:Existing approaches to evasion detection in earnings calls focus on textual transcripts, treating evasion as a single-dimensional phenomenon. We argue that evasion in spoken communication is inherently multidimensional: beyond what executives say, how they say it carries independent and complementary information. To study these dimensions jointly, we introduce DualEvasion, a benchmark for evasion detection across text and audio in earnings call QA. The benchmark contains 505 annotated question-answer pairs from 60 earnings calls, each with two independent labels: textual evasion (direct vs. evasive) and vocal cues operationalized as speaker confidence (confident vs. unconfident). Our experiments show that state-of-the-art multimodal models struggle to detect vocal confidence, particularly on unconfident responses. Our analysis suggests these models interpret acoustic cues in isolation rather than relative to each speaker’s baseline. Providing speaker-level references yields modest improvements, but a substantial gap with human performance remains.
[NLP-34] win Worlds: Equivariance-Based Abstention for Evidence-Grounded Reasoning EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在知识密集型推理中因依赖内部记忆参数而生成看似合理但缺乏证据支持的答案这一可靠性问题。现有方法多基于不确定性估计或证据充分性检查，但未能有效验证模型的推理过程是否真正基于所提供的证据。其关键缺陷在于：上下文中的实体提及会激活模型内部的记忆关联，导致模型生成与证据无关的合理幻觉。为此，论文提出“孪生世界”（Twin Worlds, TW）框架，通过等变性（equivariance）机制实现更可靠的拒答（abstention）。其核心思想是：若模型推理真正依赖于证据，则当输入中的实体被同类型替换但关系结构保持不变时，输出应相应地发生一致变换；反之，若输出未随实体替换而合理变化，则表明推理过程受模型先验参数干扰，存在非等变性。TW通过构建多个保留关系结构但降低参数先验影响的替代输入世界，以检测等变性违反情况，并将其作为拒答信号。实验表明，在四个基准测试和三种模型架构上，TW能有效识别非可靠推理，显著优于基于不确定性和证据充分性的基线方法。
链接: https://arxiv.org/abs/2608.28018
作者: Vy Nguyen,Ziqi Xu,Jeffrey Chan,Estrid He,Feng Xia,Renqiang Luo,Erik Cambria,Xiuzhen Zhang
机构: RMIT University (皇家墨尔本理工大学); Jilin University (吉林大学); Nanyang Technological University (南洋理工大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Accepted to EMNLP 2026 (Main Conference)
Abstract:Knowledge-intensive reasoning requires Large Language Models (LLMs) to ground answers in provided evidence. When evidence is insufficient, it is desirable that models abstain rather than confidently generating unsupported answers. Existing abstention methods rely on uncertainty estimation or evidence sufficiency checks, but neither tests whether the reasoning process for generation, driven by the interaction of provided evidence and the model’s internal memory parameters, is actually grounded in the evidence. A key contributing factor is that entity mentions in context activate memorised associations, causing models to generate plausible responses ungrounded in evidence. We propose Twin Worlds (TW), a framework for improving reliability in knowledge-intensive reasoning through equivariance-based abstention: unlike invariance, which requires outputs to remain unchanged, equivariance requires outputs to transform correspondingly under entity substitutions. A model grounded in the evidence should produce answers that shift consistently when entities are substituted while their relations are preserved. TW constructs multiple worlds via typed substitutions of the original input that preserve relational structure while reducing parametric priors, and uses equivariance violations as an abstention signal. Across four benchmarks and three model backbones, TW identifies when answers are not reliably grounded in the provided evidence and outperforms uncertainty- and sufficiency-based baselines.
[NLP-35] Beyond Global Scalars: Synergizing Token-Level Statistics and Deep Semantics for Adversarial AIGC Text Detection EMNLP2026
【速读】： 该论文旨在解决生成式文本检测在对抗性场景下的鲁棒性问题。现有方法主要分为两类：无训练方法依赖全局统计指标（如困惑度），而基于训练的方法则利用语义隐状态，但二者均存在根本性缺陷——前者因对局部概率突变的压缩导致信息丢失，后者则易过拟合特定指纹且易被伪造。为揭示这些弱点，作者构建了MOSAIC，一个涵盖全粒度攻击类型的综合性对抗基准，包含16000个样本。针对上述挑战，论文提出NeuroStat，一种端到端框架，通过单一因果语言模型主干同时捕获未压缩的标记级概率logits与深层语义隐状态，实现统计与语义之间的桥梁。该框架采用宏观状态残差调制（Macro-State Residual Modulation）融合异构信号，以全局不确定性指标自适应校准局部卷积特征，并结合正交与对比损失确保互补表征的学习。大量实验表明，NeuroStat在MOSAIC基准上表现出卓越的鲁棒性，显著优于现有先进方法，确立了对抗性文本检测的新标准。
链接: https://arxiv.org/abs/2608.28009
作者: Peiming Li,Yifan Wang,Zhiyuan Hu,Shiyu Li,Zheng Wei,Yang Tang
机构: Tencent BAC; School of Electronic and Computer Engineering, Peking University (北京大学电子与计算机工程学院)
类目: Computation and Language (cs.CL)
备注: Accepted by EMNLP 2026 Findings
Abstract:The rapid evolution of large language models necessitates robust machine-generated text detection. Existing paradigms typically follow two isolated tracks. Training-free methods rely on global statistical scalars such as perplexity, while training-based methods utilize semantic hidden states. Both approaches exhibit fundamental vulnerabilities in adversarial scenarios. Global scalars act as lossy compressions that obscure local probabilistic burstiness in interleaved texts, whereas pure semantic models overfit to specific fingerprints and remain susceptible to spoofing. To expose these flaws, we introduce MOSAIC, a comprehensive adversarial benchmark comprising 16000 samples across a full-granularity attack spectrum. To address these challenges, we propose NeuroStat, an end-to-end framework bridging the statistical and semantic gap. NeuroStat captures uncompressed token-level probabilistic logits alongside deep semantic hidden states from a single causal language model backbone. We fuse these heterogeneous signals through Macro-State Residual Modulation, which adaptively calibrates local convolutional features using global uncertainty indicators. Orthogonal and contrastive losses further ensure the learning of complementary representations. Extensive experiments demonstrate that NeuroStat maintains exceptional robustness on MOSAIC compared to the severe degradation of state-of-the-art methods, establishing a new standard for adversarial text detection. Code and the MOSAIC benchmark are available at this https URL.
[NLP-36] Predicting Turn-Taking Outcomes in Multi-Party Conversation: Interpretable Modelling of Speech and Gaze Dynamics with Interpersonal Closeness
【速读】： 该论文旨在解决在嘈杂、自然的多人对话环境中，如何准确预测说话人之间话语权转换（即“换言权”）的问题，尤其是面对多个对话者同时存在时，如何通过非语言与语言线索有效识别何时应进入或退出对话。其核心解决方案在于整合多模态行为信号——包括眼神交互（gaze）、语音特征（如说话音量）以及感知的人际亲密度（Interpersonal Openness Score, IOS），并利用可解释的行为学驱动特征构建逻辑回归模型，以分类判断话语权转移是表现为沉默间隙（gap）还是重叠（overlap）。研究发现，眼神特征（如过渡模式、互视、注视分配熵等）具有显著的预测能力，且与音量特征结合后显著提升了模型性能（ROC AUC = 0.76 ± 0.04），其中音量反映说话人控制力，眼神分散程度与指向对象则分别表征听者准备状态与竞争性介入意图。更重要的是，该模型在不同噪声条件下仍保持鲁棒性，表明眼神线索在复杂对话中提供了互补且抗干扰性强的换言权提示机制。
链接: https://arxiv.org/abs/2608.27988
作者: Mark Dourado,Karim Haddad,Henrik G. Hassager,Stefania Serafin
机构: Research Exploration, GN Store Nord, Ballerup, Denmark; Department of Architecture, Design and Media Technology, Aalborg University, Copenhagen, Denmark; Department of Engineering Technology and Didactics, Technical University of Denmark, Kongens Lyngby, Denmark
类目: Computation and Language (cs.CL); Sound (cs.SD)
备注: Part of an industrial PhD collaboration between GN Group and Aalborg University
Abstract:Smooth speaker transitions are fundamental to effective conversation and rely on an interlocutor’s ability to predict when to enter the conversation. This ability depends on accurately interpreting and expressing the verbal and non-verbal cues that signal when a speaker wishes to take or relinquish the floor. The process becomes even more complex in noisy, natural, multi-party settings, with multiple interlocutors available. This study models how gaze and speech, together with perceived interpersonal closeness, signal conversational floor changes in free four-person dialogue. Using the GaMMA corpus, we trained logistic regression models using interpretable, behaviourally motivated features extracted before each turn-taking event to classify floor-transfer outcomes as gaps or overlaps. Predictors included gaze features such as transition motifs and behavioural contrasts, entropy, gaze-based addressee identity, and mutual gaze, alongside speech features derived from speaker loudness, as well as perceived interpersonal closeness (IOS) between speakers. Results show that gaze features capture predictive structure, and that combining them with loudness improves performance (ROC AUC = 0.76 ± 0.04). Loudness reflected speaker control, while gaze dispersion and addressing indexed listener readiness and competitive entry. Performance remained robust across noise conditions, indicating that gaze provides a complementary, noise-resilient cue to turn-taking dynamics.
[NLP-37] QUORUM: QUality-Optimized Routing Using Multiple annotators
【速读】： 该论文旨在解决自然语言处理中数据标注的瓶颈问题，即如何在有限预算下高效获取高质量的人工标注数据。现有基于大语言模型（LLM）的自动化标注方法虽能降低成本，但其可靠性高度依赖于输入实例的复杂性，尤其在需要精细推理或上下文理解的任务上表现不佳。为应对这一挑战，论文提出一种名为QUORUM（QUality-Optimized Routing Using Multiple annotators）的预算感知路由框架，其核心创新在于通过特征驱动的难度估计信号动态决定每个样本应由人类标注员还是LLM进行标注，同时支持对同一实例进行多次标注，并利用基于一致性的奖励机制融合多源标注结果以提升可靠性。该方案不依赖传统的置信度或不确定性估计，显著提升了标注质量与成本效益，在多种闭合式与开放式标注任务及多语言场景下均实现最高达34.4%的标注质量提升，同时降低8.8%的成本，优于现有方法。
链接: https://arxiv.org/abs/2608.27974
作者: Antonio Purificato,Maria Sofia Bucarelli,Andrea Bacciu,Amin Mantrach,Fabrizio Silvestri
机构: Amazon(亚马逊); Sapienza University of Rome(罗马第一大学); Université Côte d’Azur, CNRS, Inria, I3S(蔚蓝海岸大学，法国国家科学研究中心，法国国家信息与自动化研究所，信息与安全研究所)
类目: Computation and Language (cs.CL)
备注: 4 figures, 18 pages
Abstract:Data annotation remains a central bottleneck in natural language processing, requiring human effort to obtain high-quality labels at scale. While Large Language Models (LLMs) offer a fast and cost-effective alternative, their reliability is highly instance-dependent: they perform well on simple inputs but often fail on examples requiring nuanced reasoning or contextual understanding. In this work, we address this challenge with QUORUM (QUality-Optimized Routing Using Multiple annotators), a budget-aware routing framework that dynamically assigns each instance to human or LLM annotators under a fixed annotation budget. Unlike prior approaches relying on model confidence or uncertainty estimates, QUORUM leverages feature-based signals to estimate instance difficulty and supports multiple annotations per instance, combining them through agreement-based rewards to improve reliability. We evaluate QUORUM across diverse closed- and open-ended annotation tasks in English and multilingual settings, and QUORUM improves annotation quality by up to 34.4% while reducing costs by 8.8% over competing methods. Code can be found at this https URL.
[NLP-38] DisCTI: Who Needs to Know Timely? Automated Sector-Aware Cyber Threat Intelligence Dissemination
【速读】： 该论文旨在解决当前网络威胁情报（Cyber Threat Intelligence, CTI）在跨行业分发过程中缺乏精准、自动化领域映射的问题。随着CTI数据量的快速增长，安全分析师面临海量且异构的情报信息，难以及时识别与自身所属行业相关的威胁事件，而现有平台如恶意软件信息共享平台（MISP）虽提供行业标签功能，但实际应用中高达98%的事件未被正确分类，导致关键信息基础设施（Critical Information Infrastructure, CII）相关组织暴露于高风险之中。针对这一挑战，论文将领域目标化的CTI分发问题建模为多标签分类任务，并结合领域专家对CTI结构及各行业特有威胁模式的深度知识，构建了一个包含872个带领域标签的CTI事件的新型数据集。通过引入基于Transformer架构的BERT模型，实现了对CTI事件的自动化领域映射。利用结构化威胁信息表达（STIX）格式保障跨平台互操作性，实验结果表明，该方法在自定义数据集上达到0.89的宏平均F1分数和0.055的汉明损失，即94.5%的单个领域标签预测准确率。其核心解决方案在于将领域专业知识嵌入机器学习模型设计中，有效填补了威胁情报流程中的关键空白，显著提升了情报分发的时效性与上下文相关性，支持更快速、精准的防御响应。
链接: https://arxiv.org/abs/2608.27967
作者: Fajar Wijitrisnanto(1),Alsharif Abuadbba(2),Yansong Gao(2 and 3),Nan Wu(2) ((1) National Cyber and Crypto Agency, Jakarta, Indonesia, (2) CSIRO, Sydney, Australia, (3) The University of Western Australia, Perth, Australia)
机构: National Cyber and Crypto Agency (国家网络与密码局), Jakarta, Indonesia; CSIRO (澳大利亚联邦科学与工业研究组织), Sydney, Australia; The University of Western Australia (西澳大学), Perth, Australia
类目: Cryptography and Security (cs.CR); Computation and Language (cs.CL)
备注:
Abstract:The timely dissemination of cyber threat intelligence (CTI) is critical for organizations to mount swift and effective incident response. When valid CTI is delivered to the right sector at the right time, identical attacks can often be contained or mitigated. However, today’s rapidly expanding CTI landscape overwhelms analysts, who must sift through massive and heterogeneous feeds. Existing platforms such as the Malware Information Sharing Platform (MISP) provide sector tagging features (e.g., energy, finance, government), but in practice, these remain largely unmapped (98% of events are left uncategorized). This lack of automated and timely sector mapping severely limits the operational value of shared intelligence, leaving organizations that belong especially to the critical information infrastructure sector exposed. To address this gap, we formulate sector-targeted CTI dissemination as a multilabel classification problem. Leveraging deep field knowledge of CTI structures and sector-specific threat patterns, we construct a novel data set of 872 sector-labelled CTI events from a threat intelligence platform (TIP). We then apply BERT, a transformer-based model, to automate the mapping of CTI events to sectors. Using the structured threat information expression (STIX) format for cross-platform interoperability, our approach achieves a macro-averaged F1-score of 0.89 at a Hamming loss of 0.055 on the custom dataset, i.e. 94.5% of individual sector-label assignments are correct. These results not only demonstrate the feasibility of sector-aware, automated CTI dissemination but also highlight how embedding expert field knowledge into machine learning design fills a crucial gap in the threat intelligence pipeline, enabling faster and context-relevant defensive action. Subjects: Cryptography and Security (cs.CR); Computation and Language (cs.CL) Cite as: arXiv:2608.27967 [cs.CR] (or arXiv:2608.27967v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2608.27967 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-39] Lexically conditioned realization ambiguity in Korean predicate morphology
【速读】： 该论文旨在解决韩语表层实现（surface realization）与形态分析之间关系的理论问题，具体探讨在形态成分（如词干和词尾）及语法类别标签形式化确定的情况下，是否能够唯一决定对应的表层形式。研究发现，在一类受限制但具有理论意义的韩语谓词中，答案是否定的：形式上相同或近似相同的词干-词尾配置可能产生不同的表层实现，其差异取决于词汇身份（lexical identity）和实现类归属（realization class membership）。该现象被解释为具有屈折分化特征的同形异义（homonymy with inflectional divergence），通过分析规则与双音节不规则对、规则与비읍（bieup）不规则对、以及르이（reu）不规则与료（reo）不规则对等案例，揭示出词干形态与词尾本身不足以完全决定表层实现。关键在于，词汇意义、子范畴特征（subcategorization）以及语义角色结构共同识别目标谓词，进而由谓词决定其所属的实现类，最终由实现类决定具体的表层形式。因此，该研究指出，仅依赖形式化的形态表示存在局限性，强调了语义与句法因素在韩语实现机制中的核心作用。
链接: https://arxiv.org/abs/2608.27966
作者: Wonjun Oh,KyungTae Lim,Jungyeul Park
机构: Korea Advanced Institute of Science and Technology (韩国科学技术院), South Korea (韩国)
类目: Computation and Language (cs.CL)
备注:
Abstract:This paper examines Korean surface realization as distinct from morphological analysis. It asks whether a sequence of canonical morphemes and grammatical category labels uniquely determines the corresponding surface form. The answer is negative for a restricted but theoretically revealing class of Korean predicates. In these cases, formally identical or near-identical stem-ending configurations yield different outputs depending on lexical identity and realization class membership. We analyze this phenomenon as homonymy with inflectional divergence, focusing on regular versus digeut irregular pairs, regular versus bieup irregular pairs, and reu irregular versus reo irregular pairs. These cases show that stem shape and ending alone do not always determine surface realization. Instead, lexical meaning, subcategorization, and semantic role structure help identify the intended predicate; the predicate determines the realization class; and the realization class determines the surface form. Korean realization thus reveals a limit of bare morphological representation.
[NLP-40] Entity-Memory Graph Retrieval Improves Evidence Coverag e in Long-Conversation Question Answering
【速读】： 该论文旨在解决传统对话检索方法在处理长对话时因语义相似性忽略邻近记忆节点而导致的证据召回率下降问题。现有基于密集向量匹配（dense cosine ranking）的方法容易遗漏与查询相关但语义上不直接匹配的相邻对话轮次，从而影响最终答案的准确性。为此，论文提出一种基于实体-记忆图（Entity-Memory graph）的检索框架，其核心创新在于通过显式建模对话中的实体共享关系和时间顺序边，构建结构化记忆网络：将每一轮对话作为原始记忆节点保留，利用共同提及的实体连接相关记忆，并以有向时间边连接相邻记忆。在查询时，检索器沿“实体门控→语义融合→单跳时间恢复→密集补全”的路径进行多阶段推理，有效捕捉被传统方法忽略的邻近记忆。关键在于引入匹配的密集控制机制（matched dense control），使检索结果与阅读器共享相同的向量表示、上下文预算、回答协议及评估标准，从而将图结构的影响与其他变量解耦，实现对图结构本身带来的覆盖增益的独立评估。实验表明，在1,986个来自十个LoCoMo对话的问题上，该方法在top-k=25时将官方证据召回率从79.7468%提升至84.4842%，且在top-k 5至50范围内均保持优势；然而，尽管召回率显著提升，最终答案的F1分数未见统计学支持的改善，说明当前方案主要提升了检索覆盖率，尚未转化为下游任务性能的全面提升。
链接: https://arxiv.org/abs/2608.27925
作者: Shumao Sun
机构: Tsinghua University(清华大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Entity-Memory graph retrieval keeps dialogue turns as verbatim Memory nodes, links repeated mentions through shared Entities, and connects adjacent Memories with directed chronological edges. At query time the retriever moves from Entity gating through semantic fusion and one-hop chronological recovery to dense backfill. The path can keep a neighboring Memory that dense cosine ranking would otherwise omit. A matched dense control shares the Memory and query vectors, context budget, requested answer protocol, and evaluator, isolating graph structure from changes to the reader. On 1,986 questions from ten LoCoMo conversations, graph retrieval raises official evidence recall at top-k 25 from 79.7468% to 84.4842%. The recall advantage is supported from top-k 5 to 50, while no matched cutoff supports an overall final-answer F1 difference. Four paper-eligible requested configurations support empirical robustness across the tested GPT-3.5 and DeepSeek extractors on both outcomes. Embedding robustness is mixed: F1 has no supported contrast, but recall is sensitive to the embedding artifact. The comparison isolates a retrieval-coverage gain from graph structure. It does not establish a final-answer F1 gain, model or embedding equivalence, or cross-dataset generalization. Subjects: Computation and Language (cs.CL) Cite as: arXiv:2608.27925 [cs.CL] (or arXiv:2608.27925v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2608.27925 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-41] What Makes Agent Memory Useful for Reliable Unanswerable Question Handling?
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）驱动的智能体在面对无法回答的问题（Unanswerable Questions, UAQs）时，如何通过记忆机制实现可靠处理这一关键问题。尽管记忆在智能体系统中被广泛使用，但其在UAQ处理中的具体作用与有效性尚不明确。论文提出在统一的代理式检索增强生成（agentic RAG）框架下，系统性地研究四种代表性记忆方法在三个与UAQ相关的数据集及两种基础模型上的表现。研究发现，记忆虽能在某些场景下提升UAQ处理性能，但这种增益具有选择性而非普适性，且在数据分布迁移（dataset shift）下仍显脆弱。值得注意的是，跨模型的记忆复用通常比跨数据集迁移更可行，表明答案可回答性模式的变化对记忆复用构成更大挑战。进一步分析表明，相较于轨迹塑造（trajectory shaping），通过决策引导（decision guidance）更能有效保留UAQ处理的收益；同时，记忆的有效性高度依赖于表征方式——尤其是过程化（procedural）和规则驱动（rule-based）的记忆形式在支持UAQ处理方面表现出更强的可靠性。当过程化指导与互补的行为信号相结合时，记忆组合策略效果最佳。总体而言，研究揭示：可靠的UAQ记忆并非依赖于存储更多经验，而更在于保持可迁移的行为引导能力。
链接: https://arxiv.org/abs/2608.27924
作者: Chuanyuan Tan,Junjie Yu,Yuxin Wang,Yining Zheng,Xipeng Qiu,Wenliang Chen
机构: Soochow University (苏州大学); Suzhou City University (苏州城市学院); Fudan University (复旦大学); Shanghai Innovation Institute (上海创新研究院)
类目: Computation and Language (cs.CL)
备注:
Abstract:Reliable handling of unanswerable questions (UAQs) is critical for trustworthy LLM-based agents. Although memory is widely used in agent systems, its role in reliable UAQ handling remains unclear. We present a systematic study of agent memory for UAQ handling under a unified agentic RAG framework, evaluating four representative memory methods across three UAQ-related datasets and two base models. We find that memory can improve UAQ performance in some settings, but such gains are selective rather than universal and remain fragile under dataset shift. Interestingly, cross-model memory reuse is often more feasible than cross-dataset transfer, suggesting that shifts in answerability patterns pose a greater challenge to memory reuse than changes in the base model itself. We further find that UAQ gains are more strongly preserved through decision guidance than through trajectory shaping, and that memory effectiveness depends strongly on representation. In particular, procedural and rule-based memories often provide the most reliable support for UAQ handling, while memory composition is most effective when procedural guidance is combined with complementary behavioral signals. Overall, our findings suggest that reliable UAQ memory depends less on storing larger amounts of experience and more on preserving transferable behavioral guidance. Subjects: Computation and Language (cs.CL) Cite as: arXiv:2608.27924 [cs.CL] (or arXiv:2608.27924v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2608.27924 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-42] AI Alignment through a Game-theoretic Lens: A Survey EMNLP-2026
【速读】： 该论文旨在解决在高风险场景中部署大语言模型与日益复杂的AI代理时，如何有效对齐其行为以符合复杂、动态且多主体参与的人类价值观这一核心挑战。现有对齐方法虽在提升助手性、无害性和可控性方面表现良好，却难以捕捉情境依赖、非传递性以及由多方互动演化而成的真实世界偏好。论文提出从博弈论视角重新审视AI对齐问题，其解决方案的关键在于将近期研究进展系统地归纳为博弈论中的核心要素，并围绕三个关键挑战展开综述：偏好多样性、对齐优先级以及时间动态性。该框架不仅揭示了当前对齐方法在博弈论分析中真正受益的领域，也指出了理论应用尚不严谨之处，从而为构建更鲁棒、自适应且可验证的AI系统提供了清晰的研究路径与未解难题。
链接: https://arxiv.org/abs/2608.27910
作者: Yanan Cai,Zhongrui Zhao,Zhigang Lu,Ickjai Lee,Wei Emma Zhang,Minhui Xue,Yihong Zhang,Shuchao Pang,Wei Xiang
机构: James Cook University(詹姆斯库克大学); Western Sydney University(西悉尼大学); Adelaide University(阿德莱德大学); CSIRO(澳大利亚联邦科学与工业研究组织); The University of Osaka(大阪大学); Nanjing University of Science and Technology(南京理工大学); Macquarie University(麦考瑞大学); La Trobe University(拉筹伯大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Computer Science and Game Theory (cs.GT)
备注: This paper has been accepted by EMNLP-2026 as a main conference paper
Abstract:As large language models and increasingly capable AI agents are deployed in high-risk settings, aligning them with complex human values has become a central challenge. Existing alignment methods, while effective in improving helpfulness, harmlessness, and controllability, often struggle to capture real-world preferences that are context-dependent, non-transitive, and shaped by dynamic multi-party interactions. This survey reviews AI alignment through a game-theoretic lens. Specifically, it organizes recent progress around key game-theoretic elements and synthesizes the literature along three challenges: preference diversity, alignment priority, and temporal dynamics. This perspective clarifies where current alignment methods genuinely benefit from game-theoretic analysis, where the framework is looser, and what challenges remain in building robust, adaptive, and verifiable AI systems.
[NLP-43] LandingAgent : A Reference-Annotated Dataset and Agent ic Generation Framework for Landing Pages
【速读】： 该论文旨在解决生成式网页设计中目标导向性不足与说服力缺失的问题，即现有大语言模型在直接根据自然语言提示生成落地页时，常产生通用模板化内容且包含未经验证的营销主张。其核心解决方案在于提出一种以目标为基准、参考为引导的落地页生成方法，关键创新在于构建了LandingBench——一个将真实落地页抽象为模块化段落序列、布局模式、语气描述、视觉重点及行动号召（CTA）结构的参考型数据集，并基于此设计LandingAgent，一个三阶段智能体框架：首先对目标进行画像分析，其次构建参考驱动的线框图，最后通过批判性反馈进行精细化优化。该方法显著提升了生成页面的目标契合度、呈现质量与布局多样性，在忠实性、简洁性、可读性、美学性和结构多样性等维度上均优于直接提示法。
链接: https://arxiv.org/abs/2608.27902
作者: Injun Baek,HyeongSeok Lee,Yearim Kim,Junhoo Lee,Nojun Kwak
机构: Seoul National University (首尔国立大学); Samsung Electronics (三星电子)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 30 pages, 8 figures
Abstract:Landing pages are goal-oriented web interfaces that must communicate a target-specific value proposition while organizing information flow, visual hierarchy, and calls to action (CTA). Although large language models can generate plausible webpage code from natural-language prompts, direct generation often yields generic templates and unsupported persuasive claims. We study target-grounded, reference-guided landing-page generation, where a system must create an executable page for a new target by adapting reusable patterns from real pages without copying them. We introduce LandingBench, a reference-profile dataset that abstracts real landing pages into section sequences, layout patterns, tone descriptors, visual emphasis, and CTA structure. Building on LandingBench, we propose LandingAgent, a three-phase agentic framework that profiles the target, constructs a reference-guided wireframe, and refines the page through critique-guided polishing. We evaluate LandingAgent against direct prompting on faithfulness, conciseness, readability, aesthetics, and structural diversity. Experiments show improved target grounding, presentation quality, and layout diversity. Code is available at this https URL.
[NLP-44] OpenStamp: A Watermark for Open-Source Language Models
【速读】： 该论文旨在解决开放源代码大语言模型（LLM）在生成内容中难以有效嵌入鲁棒水印的问题。现有水印技术通常通过调整令牌采样概率来隐式嵌入信号，但这类方法在白盒访问场景下极易被用户在推理阶段禁用，因而不适用于开源模型。其解决方案的关键在于提出OpenStamp，一种将水印逻辑直接编码至模型权重中的新方法，仅通过修改模型的最后一层投影（或称未嵌入层，unembedding layer）即可实现。该方法使水印与模型参数深度耦合，显著增强了对改写攻击和事后微调擦除的抵抗能力。实验表明，相较于已有开源水印方案，OpenStamp在保持模型性能几乎不受损的前提下，具备更优的检测精度与鲁棒性，并已开源多个主流开源模型的水印版本及完整实现代码，便于开发者部署应用。
链接: https://arxiv.org/abs/2608.27899
作者: Miroojin Bakshi,Saksham Rastogi,Danish Pruthi
机构: Indian Institute of Science (印度科学研究所); Carnegie Mellon University (卡内基梅隆大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Published at COLM 2026
Abstract:With the growing prevalence of large language model (LLM) generated content, watermarking is considered a promising approach for attributing text to LLMs and distinguishing it from human-written content. A prominent class of techniques embeds subtle but detectable signals in generated text by modifying token sampling probabilities. However, such methods are unsuitable for open-source models, where users have white-box access and can easily disable watermarking during inference. In this work, we introduce OpenStamp, a watermarking technique that encodes the watermarking logic directly into the model weights by modifying only the final projection, or unembedding, layer. Through experiments across two models, we show that OpenStamp achieves superior detection performance, with minimal degradation in model capabilities compared to prior methods. The implanted watermark is explicitly designed, and empirically confirmed, to be more robust to paraphrasing attacks and harder to scrub off through post-hoc fine-tuning than prior open-source watermarks. To enable developers to watermark their models, we release our code alongside watermarked versions of 4 popular open-source models.
[NLP-45] AI Writers Have a Consistent Stylometric Footprint but AI Editors Do Not EMNLP
【速读】： 该论文旨在解决生成式人工智能（Generative AI）在文本生成与文本编辑两种不同应用场景下，其产生的语言风格特征是否具有可区分性的核心问题。研究发现，尽管两类操作均涉及AI干预，但它们在语言风格上留下的“风格痕迹”（stylometric footprint）存在本质差异：在文本生成场景中，以熵值和词汇多样性为核心的一小部分特征能够稳定地区分AI生成文本与人类写作，且这一模式在8种大语言模型（LLM）和5个不同领域中保持一致；而在文本编辑场景中，AI修改后的文本仅表现出词汇多样性轻微上升和熵值下降的特征，未呈现生成场景中的双增趋势，反而凸显了词汇密度（lexical density）作为主要判别信号。这表明，生成与编辑所引发的风格变异机制不同，因此不能将“AI文本”视为单一现象进行统一建模。研究的关键在于揭示生成与编辑行为在风格特征上的根本性差异，并强调应针对不同应用场景分别设计检测策略。
链接: https://arxiv.org/abs/2608.27855
作者: Zhengyang Shan,Yukyung Lee,Sophie Hao
机构: Boston University (波士顿大学)
类目: Computation and Language (cs.CL)
备注: EMNLP Main 2026
Abstract:Text generated by large language models (LLMs) has been shown to be stylometrically distinct from human-written text \citepandreDetectingAIAuthorship2023, shahDetectingUnmaskingAIGenerated2023, oparaStyloAIDistinguishingAIGenerated2024, soto2024fewshot, liLinguisticDifferencesAI2025, selviogluFeatureExtractionAnalysis2025. But LLMs are increasingly used not only to generate text but also to edit human writing, and it is unclear whether the two leave the same trace. We show that AI generation leaves a consistent stylometric footprint'': a small subset of features, primarily entropy and lexical diversity, consistently separates AI-generated text from human writing across 8 LLMs and 5 domains, while the remaining features depend heavily on the domain and generator. AI editing, however, does not reproduce the same footprint. Relative to their human-written sources, AI-edited texts show only a small increase in lexical diversity and a decrease in entropy, rather than the joint increase that characterizes AI generation. Lexical density, which contributes little to generation, instead becomes the dominant editing-associated signal. Stylometric features therefore separate AI-edited text from AI-generated text but are substantially less effective at separating it from human-written text. Our results suggest that AI text’’ is not a single phenomenon: generation and editing leave qualitatively different stylometric traces and should be studied separately.
[NLP-46] Is Prosody Lost in Translation? Fine-Grained Cross-Lingual Prosody Similarity Across Languages EMNLP2026
【速读】： 该论文旨在解决跨语言语音韵律（prosody）在语音到语音翻译（S2ST）中如何迁移与适配的问题，尤其关注不同语言间韵律模式的相似性与差异性。现有研究虽在表达性S2ST方面取得进展，但对韵律特征在跨语言间的共性与变异性缺乏系统认知，制约了韵律信息的有效建模与迁移。本文提出基于多语言配音数据（涵盖英德、英西、英法三组语言对）的细粒度跨语言韵律分析方法，系统比较源语言与目标语言在基频（pitch）、能量（energy）和时序（temporal）特征上的对应关系，并探究语言类型及句法对齐等因素对韵律相似性的影响。研究发现，特定语言间存在固有的韵律结构相关性，揭示了韵律可迁移性的潜在基础。其关键贡献在于首次通过实证数据揭示了跨语言韵律模式的规律性，为未来表达性S2ST系统中韵律的跨语言迁移提供了可量化的指导依据。
链接: https://arxiv.org/abs/2608.27848
作者: Haopeng Xie,Ismail Rasim Ulgen,Sofia Son,Berrak Sisman,Philipp Koehn
机构: Johns Hopkins University (约翰霍普金斯大学)
类目: ound (cs.SD); Computation and Language (cs.CL); Audio and Speech Processing (eess.AS)
备注: Accepted to EMNLP 2026 Findings
Abstract:Prosody plays an important role in speech translation, conveying information such as emphasis, emotion, and intent beyond lexical content. However, despite recent progress in expressive speech-to-speech translation (S2ST), little is known about how prosodic patterns are similar/different across languages. Understanding these cross-lingual similarities and differences is crucial for effectively incorporating prosody into expressive S2ST systems. In this work, we present the first fine-grained cross-lingual analysis of prosody using multilingual dubbing data across English-German, English-Spanish, and English-French language pairs. We analyze the similarity of pitch, energy, and temporal feature patterns between source and target speech and investigate the linguistic and alignment-related factors affecting this similarity. Our analysis reveals inherent cross-lingual correlations in prosodic structure between certain languages. The findings provide important insights into the transferability of prosody across languages and offer empirical guidance for future expressive speech-to-speech translation systems.
[NLP-47] EvoHarmBench: Breaking Content Moderation with Iterative Human-Like Evasion EMNLP2026
【速读】： 该论文旨在解决现有有害内容检测评估体系中静态基准测试与真实平台动态对抗生态之间存在的显著脱节问题。传统静态基准无法反映用户在内容审核反馈后持续迭代表达策略的交互式对抗特性，导致离线评估得分与线上实际防护效果之间存在巨大性能差距。为此，论文提出EvoHarmBench——首个面向内容审核系统的动态对抗评估框架，其核心创新在于构建了一个语义聚类层级上的迭代优化循环机制，能够在保持人类可读性的前提下，持续演化逃逸策略以提升攻击成功率。该框架对基于大语言模型（LLM）的防御模型进行了系统性评估，覆盖5个违规类别下的229个语义子聚类，基于从真实内容平台收集的5,002个对抗样本进行测试。实验结果揭示了即使是最先进的商用系统也存在严重脆弱性：在12轮优化迭代后，受限于可读性的攻击成功率仍高达80.3%。该研究的关键贡献在于推动内容安全研究从静态基准向动态对抗评估范式转变，并公开完整数据集、评估框架与代码，以促进更贴近真实场景的鲁棒性评测。
链接: https://arxiv.org/abs/2608.27844
作者: Ruijie Jian,Benlei Cui,Ting Ma,Haidong Ding,Kangwei Liu,Ziwen Xu,Longtao Huang,Hui Xue,Ziqiang Zhu,Junjie Li,Haiwen Hong
机构: Alibaba Group(阿里巴巴集团); University of Chinese Academy of Sciences(中国科学院大学)
类目: Computation and Language (cs.CL)
备注: Accepted to the Findings of EMNLP 2026
Abstract:Existing evaluations of harmful content detection rely predominantly on static benchmarks, which struggle to reflect the interactive adversarial ecosystem of real-world content platforms where users continuously revise their expressions in response to moderation feedback. This mismatch creates a significant performance gap between offline benchmark scores and online deployment effectiveness. To the best of our knowledge, we present EvoHarmBench, the first dynamic adversarial evaluation framework for content moderation systems. The framework employs an iterative optimization loop that evolves evasion strategies at the semantic-cluster level, while simultaneously optimizing for evasion success and human readability. We systematically evaluate LLM-based defense models which are widely used in real world moderation systems. The evaluation covers 229 semantic sub-clusters across five violation categories, derived from 5,002 real-world adversarial samples collected from content platforms. Our experiments reveal substantial vulnerabilities even in leading commercial systems: after twelve optimization iterations, the attack success rate under readability constraints reaches 80.3% within SOTA LLM moderators. We will release the full benchmark data, evaluation framework, and code to encourage a shift from static benchmarking toward dynamic adversarial evaluation in content safety research.
[NLP-48] Auditing Generative Audio Calls for Known-Task Audio-LLM Evaluation
【速读】： 该论文旨在解决语音与音频大模型（Speech and audio LLMs）评估中存在的一种混淆问题：当前常用评估方法将波形提示（waveform prompt）与自动语音识别（ASR）转录文本进行对比，但这一比较同时混杂了声学信息的可访问性与调用生成式音频模型（generative audio model）的需求，导致难以准确衡量生成式模型的真实贡献。为厘清这一混淆，论文提出将评估问题形式化为一个受控的“决策调用”（call-decision）任务，明确区分声学编码器提供的证据与生成式模型调用的价值。其解决方案的关键在于设计一个策略（policy），在给定样本中从三种路径中选择最优动作：保留转录文本标签、使用来自对比语言-音频预训练（CLAP）、音频谱图变换器（AST）或WavLM等编码器的声学证据，或调用Qwen2-Audio、Qwen2.5-Omni、MOSS-Audio等生成式音频模型。通过关键消融实验——移除所有生成式调用动作而保持选择器和开发协议不变——发现，在已知闭集任务下，仅使用监督学习的CLAP和WavLM编码器即可达到0.850和0.854的准确率，显著优于仅依赖转录文本的0.296；而引入生成式调用的策略虽能提升至0.925（仅12.5%调用比例），但与不调用生成模型的对照组（0.921）差异极小（配对差值0.004，95%置信区间[-0.025, 0.033]），表明生成式调用在已有转录文本与编码器证据基础上的边际增益有限。此外，尽管集成策略中的聚合特征（agreement and stacking features）可增强较弱选择器的表现，仍无法超越最强的无生成调用控制组。因此，研究结论强调：对于已知任务的终点判断（endpoint claims），真正关键的是在已利用转录文本与编码器证据后，生成式调用所能带来的边际价值，而实证结果表明该价值微乎其微。
链接: https://arxiv.org/abs/2608.27817
作者: Mengzhe Geng
机构: 未知
类目: ound (cs.SD); Computation and Language (cs.CL); Audio and Speech Processing (eess.AS)
备注:
Abstract:Speech and audio LLMs are often evaluated by asking whether a waveform prompt beats an automatic speech recognition (ASR) transcript. For known closed-set tasks, that comparison conflates two factors: access to acoustic evidence and the need to call a generative audio model. We evaluate this distinction as a controlled call-decision problem. For each example, a policy chooses among keeping a transcript label, using encoder evidence from Contrastive Language-Audio Pretraining (CLAP), Audio Spectrogram Transformer (AST), or WavLM, and calling Qwen2-Audio, Qwen2.5-Omni, or MOSS-Audio; the decisive ablation removes all generative actions while keeping the selector and development protocol fixed. On VocalSound, transcripts reach 0.296 accuracy, so waveform information is needed. Yet supervised CLAP and WavLM controls reach 0.850 and 0.854 with no generative audio calls. A selector with generative actions reaches 0.925 accuracy using 12.5% calls, compared with 0.921 for the matched no-call selector (paired difference 0.004; 95% CI [-0.025,0.033]). Agreement and stacking features improve weaker selectors but do not beat the strongest no-call control. For known-task endpoint claims, the relevant quantity is the marginal value of the generative call after transcript and encoder evidence have already been used.
[NLP-49] PersonaEdit: Representative Sample Selection for Personalized Model Editing EMNLP2026
【速读】： 该论文旨在解决大语言模型（LLM）个性化应用中现有基于检索的方法在长期交互场景下性能下降的问题，其核心挑战在于检索质量依赖性强且难以适应持续更新的用户偏好。为应对这一问题，论文提出一种名为PersonaEdit的解决方案，其关键在于通过隐藏表示聚类与比例分层抽样相结合的样本选择策略，从大量用户数据中筛选出具有代表性的编辑样本，从而在显著降低计算开销的同时避免多条编辑之间的干扰。实验表明，该方法在保持接近全量编辑性能的前提下大幅减少了所需编辑样本数量；此外，研究还发现将模型编辑与基于检索的提示增强相结合可进一步提升个性化效果，因编辑后的知识与检索到的上下文信息具有互补性。结果证明，模型编辑是一种高效且可扩展的个性化技术路径。
链接: https://arxiv.org/abs/2608.27816
作者: You-Mei Huang,Chung-Chi Chen,An-Zi Yen
机构: National Yang Ming Chiao Tung University (国立阳明交通大学); National Institute of Informatics (日本信息研究所)
类目: Computation and Language (cs.CL)
备注: Accepted by EMNLP 2026 Findings
Abstract:Personalization has attracted growing interest in LLM applications, yet existing retrieval-based approaches depend heavily on retrieval quality and degrade in long-term interactions. Model editing, which directly modifies internal model parameters to incorporate new knowledge, has demonstrated effective knowledge modification capabilities in factual knowledge editing tasks and may provide a potential solution for personalization. However, scaling model editing to personalization is non-trivial. Editing large amounts of user data increases computational cost and causes interference among edits, motivating the need for effective sample selection. To address this issue, we propose, PersonaEdit, a hidden representation clustering strategy that selects representative editing samples through proportional stratified sampling. Experiments show that model editing is effective for personalization, and that our selection strategy preserves most of the performance while substantially reducing the number of required editing samples. Beyond standalone editing, we find that combining model editing with retrieval-based prompt augmentation further improves personalization, as edited knowledge and retrieved context provide complementary information. These results demonstrate the potential of model editing as an efficient and scalable approach for LLM personalization.
[NLP-50] Representation of syntax in LLM s through the lens of linear distance and similarity-aware entropy
【速读】： 该论文旨在解决语言模型中句法结构表征的抽象程度及其对嵌入空间几何特性依赖性的问题。传统结构探针（structural probes）通过计算在标注语料上正确重构的句法树边比例来评估模型性能，通常以无向无标签依存得分（UAS）作为衡量指标。然而，该研究指出，这一综合指标掩盖了不同句法关系之间显著的重建差异。为此，论文提出将评估粒度细化至每种句法关系的无向带标签依存得分（UASL），从而揭示句法关系间的细微差异，并识别出影响其重建准确率的关键因素：一是相关词之间的线性距离（以对数尺度计算）的均值与离散程度，二是句法关系中核心词（head）的多样性（即考虑相似性的熵）。这两个因素共同解释了大部分句法关系在UASL上的变异。研究结果在多种模型规模与架构下具有普适性，表明语言模型中的句法表征不仅具有一定的抽象性，且其质量高度依赖于嵌入空间中的几何结构特征。
链接: https://arxiv.org/abs/2608.27813
作者: Juan Pablo Vigneaux,Mary Kennedy,Khalil Iskarous,Robert Frank,Matilde Marcolli
机构: Northwestern University(西北大学); University of Southern California(南加州大学); Yale University(耶鲁大学); California Institute of Technology(加州理工学院)
类目: Computation and Language (cs.CL)
备注: 29 pages (9 main text, 18 appendix), 20 figures, 7 tables. Code and data: this https URL
Abstract:Structural probes were introduced by Hewitt and Manning to reconstruct syntactic trees from a neural language model’s latent representations. They are evaluated by calculating the proportion of syntactic tree edges correctly reconstructed over an annotated corpus (as measured by undirected unlabeled attachment score). Here, we disaggregate this measure, considering undirected attachment score by label (UASL), which assesses the reconstruction accuracy of each syntactic relation separately, establishing important differences among relations that overlap linguistic distinctions. Moreover, we identify two factors that predict most of UASL’s variability across relations: (i) the mean and dispersion of the linear distance (on a log scale) between the related words, and (ii) the diversity (similarity-aware entropy) of the syntactic relation’s head. These results, which hold across a range of model sizes and architectures, shed light on the degree of abstraction of the representation of syntax in language models and the dependence of such representation on geometric properties of the embedding space.
[NLP-51] CEDAR: Automata as Verifiable Interfaces for Language-Guided Embodied Action
【速读】： 该论文旨在解决具身智能体在自然语言任务指令下，如何有效维持随环境动态变化而持续存在的约束条件这一关键问题。传统基于代码生成的大语言模型（LLM）代理虽能生成看似合理的行为程序，但其自由形式的代码缺乏稳定结构，难以进行验证、组合新约束或从失败轨迹中修复。为此，论文提出CEDAR框架，其核心创新在于将自然语言指令语义映射为环境事件轨迹上的正则语言（regular languages），并利用语言模型进行语义判断、执行轨迹实现纠错，最终将技能与规范统一表示为确定性有限自动机（deterministic finite automata, DFA）。这一方法使约束条件转化为可执行的有限状态对象，通过集合交运算可直接合成满足多约束的控制器，实现约束的构造性保障而非依赖反复提示。在Minecraft环境中，相较于基准方案，CEDAR不仅有效保持了时空约束，且通过复用已学技能显著降低了累积的LLM调用次数，表明正则语言为自然语言指令与具身智能体策略之间提供了一种兼具可验证性与实用性的中间层。
链接: https://arxiv.org/abs/2608.27797
作者: Lekai Chen,Alvaro Velasquez,Ashutosh Trivedi
机构: University of Colorado Boulder (科罗拉多大学博尔德分校)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Formal Languages and Automata Theory (cs.FL)
备注:
Abstract:Natural-language tasking of embodied agents is rarely just goal specification: users also impose constraints that must persist while the world changes. Code-generating LLM agents can produce plausible behaviors for such instructions, but their free-form programs provide no stable object to verify, compose with new constraints, or repair from a failing trace. We present CEDAR, a counterexample-guided framework that grounds instructions as regular languages over environment event traces. CEDAR uses a language model for semantic judgments and execution traces for correction, then represents both skills and specifications as deterministic finite automata. This turns constraints into executable finite-state objects: a learned skill can be intersected with a learned sleep at night or stay in this biome specification, yielding a controller that enforces the learned constraint by construction rather than by repeated prompting. In Minecraft, with the same simulator/API observations available to a program-generating baseline, CEDAR maintains temporal and spatial constraints that the baseline fails to preserve and amortizes reuse of learned skills, reducing cumulative LLM queries. These results suggest that regular languages offer a practical verification layer between natural-language instructions and embodied-agent policies.
[NLP-52] Compositional Failure in Audio-Visual LLM s: Late-Layer Prior Dominance Under Cross-modal Conflict ICML2026
【速读】： 该论文旨在解决多模态大模型（AV-LLMs）在面对视听冲突场景下的组合泛化能力问题，即当音频与视频在语义上不一致但时间同步时，模型能否正确判断二者是否匹配。其核心挑战在于模型在输入存在矛盾信息时仍表现出对内部先验模式的过度依赖，导致决策偏差。解决方案的关键在于揭示并验证“先验主导”（prior dominance）这一失败模式：尽管模型输出分布发生显著偏移，但在深层网络中仍集中于约25.5 ± 1层进行早期且不可逆的决策，该决策主要基于内部偏好而非对冲突输入的合理整合。研究通过机制可解释性分析发现，增强时间对齐虽能改变答案倾向，却无法提升模型在组合性冲突情境下的推理能力，从而指出当前模型在跨模态融合中的根本局限。
链接: https://arxiv.org/abs/2608.27785
作者: Adarsh Sudheer,David Li,Omar Elbanna,Ishaan Kodarapu,Arjun Bahuguna,Vasu Sharma
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted to the 2nd Workshop on Compositional Learning at ICML 2026. 7 pages, 4 figures
Abstract:We study audio-visual conflict as a compositional generalization test for AV-LLMs: the model must combine synchronized but semantically incompatible audio and video evidence and decide whether the pair matches. On VideoLLaMA 2-7B-AV, three alignment configurations remain nearchance on the scored exact-string Yes/No subset of AVHBench, even though their output priors shift substantially. Similarly, off-the-shelf InternVideo2 experienced a 32.3% accuracy decrease specifically under cross-modal conflict, accompanied by a 17.3% instruction-following failure. We call this failure mode prior dominance: late-layer commitment to an internally preferred answer pattern that is weakly grounded in the conflicting inputs. To explain this behavior, we conduct a mechanistic interpretability analysis and find that commitment remains concentrated at 25.5 \pm 1 layers. We show that stronger temporal alignment changes answer bias, but do not improve compositional conflict resolution. Code and data to reproduce our mechanistic audit and behavioral evaluations are available at this https URL.
[NLP-53] Memorization Is Not Extraction: Tight Differential-Privacy Bounds and Audit Blind Spots
【速读】： 该论文旨在解决大语言模型中记忆现象的度量与隐私保护之间的核心矛盾问题，具体聚焦于两种具有实际意义的记忆形式——反事实记忆（counterfactual memorization）与自适应提取（adaptive extraction）之间的关系及其在差分隐私（Differential Privacy, DP）框架下的边界控制。其解决方案的关键在于精确刻画了在f-DP（一种广义差分隐私形式）下，这两类记忆行为的可控制性边界：研究发现，自适应提取的成功概率受控于一个依赖于先验猜测能力κ的函数1−f(κ)，且该界在密集基线集上是紧的，表明差分隐私对提取的控制严格依赖于秘密被先验猜测的能力阈值；同时，最小熵（min-entropy）提供了分布无关的保障，在纯ϵ-DP下，只要满足H∞≥ϵlog2e+log2(m/τ)，即可将提取风险控制在τ≤1/2以下，且在均匀先验下达到紧界。在记忆层面，f-DP对任意有界得分的记忆优势进行上界约束，其上限为优势函数η(f)，在纯DP下等于tanh(ϵ/2)；当存在k≥2个重复副本时，传统线性缩放ϵ↦kϵ所对应的tanh(kϵ/2)无法实现，真实最优常数为通过几何噪声计数构造的闭式阶梯函数。这一理论边界在实践中常用的局部得分类别中可被达成，从而揭示出两类记忆机制的本质分离：某些机制虽被模型“记住”但无法被自适应提取，另一些则完全可被提取却对基于损失的评分指标“不可见”。这种双向盲区导致基于损失的审计与遗忘验证方法在千亿参数模型中仍存在严重失效风险——即尽管某敏感触发词可通过单一提示完整恢复，但当前主流审计工具却会误判其为无记忆状态。
链接: https://arxiv.org/abs/2608.27782
作者: Xujun Che,Depeng Xu,Shuhan Yuan
机构: University of North Carolina at Charlotte (北卡罗来纳大学夏洛特分校); Utah State University (犹他州立大学)
类目: Cryptography and Security (cs.CR); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Memorization in large language models is measured through a zoo of definitions whose formal relations are unknown, and differential privacy (DP) is treated as a proxy against all of them at once. We pin down the exact DP constant for the two that carry the practical weight, counterfactual memorization and adaptive extraction, and show that they do not control each other. Under f -DP, every adaptive extraction protocol with list budget m succeeds with probability at most 1-f(\kappa) for the oblivious baseline \kappa , and the bound is tight on a dense set of baselines: DP uniformly controls extraction exactly up to a threshold in how well the secret can be guessed a priori. Min-entropy certifies that baseline distribution-free, since H_\infty\ge\epsilon\log_2 e+\log_2(m/\tau) holds extraction below a risk level \tau\le1/2 under pure \epsilon -DP for every prior, and is exact on uniform priors. On the memorization side, f -DP caps the counterfactual memorization of any bounded score at an advantage functional \eta(f) , equal to \tanh(\epsilon/2) under pure DP; for k\ge2 duplicated copies the naive \epsilon\mapsto k\epsilon bound \tanh(k\epsilon/2) is unattainable, the exact constant being a closed-form staircase attained by geometric noisy counting. That cap is attained inside the local score class used in practice, and it is there that the two measures separate: one mechanism is memorized yet unextractable, another fully extractable yet exactly invisible to every loss-based score. The two-sided blind spot this opens for loss-based auditing and unlearning verification survives on billion-parameter models: a reserved-trigger release is recovered verbatim from one prompt while the audits practitioners deploy certify it clean.
[NLP-54] Why Didnt It Check? Unsupported Final Claims and Their Repair in Two Tool-Equipped Language Models
【速读】： 该论文旨在解决生成式 AI（Generative AI）在具备工具调用能力时，仍可能做出缺乏证据支持的最终断言（unsupported claim）的问题，即使存在单一工具调用即可消除不确定性，且模型指令明确禁止假设与猜测。其核心解决方案的关键在于将这一失败行为分解为两个可量化、精确定义的指标：发生率（occurrence），即模型在无外部干预下自主产生无证据支持断言的频率，仅基于可见证据与最终断言进行评估；以及条件修复率（conditional repair），即当缺失证据被提供后，这些原本不支持的断言中有多少能够被纠正。实验表明，在特定设置下，尽管模型在33个案例中做出了无证据支持的断言，但一旦提供正确证据，所有断言均被成功修复，而无关信息则无法修复任何错误；同时，引入自动校验规则后，通过增加21次工具调用，实现了对全部10个错误断言的修正，并保持了11个偶然正确的断言不变，未引入新错误。此外，在另一模型配置中，模型始终调用工具且未产生无证据支持的断言，表明工具调用机制本身可有效预防此类问题。研究揭示了两种固定模型配置下的局部现象，但尚未证明该问题在真实场景中的普遍性或其是否为跨模型共享的通用机制。
链接: https://arxiv.org/abs/2608.27768
作者: Justin Bronder
机构: Independent Researcher(独立研究员)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 18 page, 1 figure, 5 tables
Abstract:A language model with access to tools can commit to a final claim unsupported by the evidence it has seen, even when a single available tool call would resolve the uncertainty and its instructions explicitly forbid assumptions and guesses. We separate this failure into two precisely defined quantities: occurrence, how often the model makes an unsupported claim on its own, measured from the visible evidence and final claim without using the hidden correct answer; and conditional repair, how often those same naturally occurring unsupported claims are repaired when the missing evidence is supplied. On one fixed Qwen3-32B setup, 33 of 512 first responses to 256 new prompt templates ended with an unsupported established claim. We replayed each case from an exact copy of the state in which the claim occurred; within each matched replay, the alternative tool responses had the same structure and length and differed only in a one-character response code. Resolving evidence repaired 33 of 33 claims; a matched response carrying no useful information repaired 0 of 33. When the evidence supported the original answer, the model preserved 33 of 33, with no observed harm. In a separate experiment, on 64 cases where evidence was needed, an automatic checking rule added 21 evidence calls, corrected all 10 wrong unsupported claims, preserved the 11 that were correct by accident, and never changed a correct answer into a wrong one. On a fixed Gemma 4 setup using the same sampling settings, the model called the tool in all 512 first responses and never made an unsupported final claim, so conditional repair could not be measured for that setup. These results describe two local fixed model setups on two synthetic task families. They do not show how common this failure is in real-world deployments, nor that it reflects a general mechanism shared across models.
[NLP-55] Fast Weight Attention for Continual Learning FAST
【速读】： 该论文旨在解决序列建模中如何在保持固定维度递归状态的前提下，有效压缩不断扩展的上下文信息，并实现在线学习的挑战。其核心问题是：在自回归生成框架下，如何设计一种可微、稳定且具备良好泛化能力的状态更新机制，以支持长程依赖建模与长度外推（length extrapolation）。解决方案的关键在于提出一种基于快速权重记忆（fast-weight memories）与选择性状态空间模型（selective state-space models）的统一框架，通过局部快速记忆机制在每个时间步引入前缀对齐的输入-输出对 (xt,yt)=(ϕ(kt−1),vt)，从而实现因果性约束下的动态状态更新。作者推导了针对平方误差回归与负内积目标的归一化一阶更新规则，发展出Falcon系列算法（如Falcon-1/2/3及其对应内积变体Falcon-1A/2A/3A），并提供递归式、掩码并行和分块并行等多种计算形式，结合数值稳定的正衰减重归一化策略，确保训练稳定性。该框架明确分离了时间对齐、可塑性、遗忘机制与有限回放等关键组件，在语言建模任务中保持竞争力，并显著提升变量位数加法任务中的长度外推性能。
链接: https://arxiv.org/abs/2608.27763
作者: Yifan Zhang,Steve Ta,Jasper Zhang,Jichen Feng,Shuzhen Li,Yongxin Zhang,Yifeng Liu,Huizhuo Yuan,Mengdi Wang,Quanquan Gu,Andrew Chi-Chih Yao
机构: 未知
类目: Machine Learning (cs.LG); Computation and Language (cs.CL); Machine Learning (stat.ML)
备注: Project Page: this https URL
Abstract:Recurrent fast-weight memories and selective state-space models compress an expanding context into a fixed-size recurrent state, making the state transition an online learning rule. We study this rule under read-after-write autoregressive semantics. For the prefix-prediction objective considered here, the local fast-memory example revealed at step t is the prefix-aligned pair (\mathbfx_t,\mathbfy_t)=(\phi(\mathbfk_t-1),\mathbfv_t) . The common same-step association (\phi(\mathbfk_t),\mathbfv_t) remains causal, but optimizes a different internal objective. We derive normalized first-order updates for squared-error regression and negative inner-product objectives. The regression family comprises Falcon-1 (a scalar NLMS update), Falcon-2 (its per-column extension), and Falcon-3 (a sliding-window mini-batch update); Falcon-1A/Falcon-2A/Falcon-3A are the corresponding inner-product variants. We provide recurrent, masked-parallel, and chunk-parallel forms, together with numerically stable positive-decay renormalization. Representative variants remain competitive in language modeling and improve length extrapolation on variable-digit addition. This framework separates temporal alignment, plasticity, forgetting, and bounded rehearsal in recurrent sequence models.
[NLP-56] Informational Antilocality and the Locality Bias in LLM s
【速读】： 该论文旨在探究基于Transformer的大型语言模型（LLM）在学习一类称为k-反局部语言（k-antilocal languages）的语言时的能力，这类语言在任意连续k个符号的跨度上均无互信息。研究通过构建随k增大而逐渐增强反局部性的语言，发现尽管模型在不同反局部性水平下的交叉熵损失表现相近，即学习成功率相当，但其收敛速度在更高度反局部的语言中显著减慢。这表明非局部依赖关系的学习难度更高，但支持这一认知的证据主要来源于学习速度的差异，而非最终学习成功与否。因此，该研究的关键在于揭示了模型对非局部结构的隐式学习偏倚，其核心机制体现为学习效率随依赖距离增加而下降，而非无法完成学习任务。
链接: https://arxiv.org/abs/2608.27760
作者: Andrew McInnerney,Shane Storks,Steven Abney,Richard L. Lewis
机构: University of Michigan (密歇根大学); Eastern Michigan University (东密歇根大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:We consider the ability of transformer-based language models (LLMs) to learn what we call k-antilocal languages, i.e., languages that have no mutual information across any span of k contiguous symbols. We construct such languages with increasing k , finding that LLMs trained on them achieve comparable cross-entropy loss regardless of antilocality, but converge more slowly on more antilocal languages. Our findings support the idea that non-local dependencies are more difficult to learn, but the evidence for this bias comes from learning speed rather than learning success.
[NLP-57] Load-Bearing Context: The Question Damage Score for Evaluating Context Reliance in Linguistic Reasoning
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在回答问题时依赖上下文信息还是自身先验知识这一核心挑战。其关键解决方案在于构建一个基于自包含语言奥林匹克谜题的诊断框架，通过系统性地移除单个上下文示例来评估模型对特定信息的依赖程度。研究采用53道英国语言学奥林匹克竞赛题目，生成两种修改版本：一种为随机删除（uniform random deletion），另一种为针对性删除（targeted deletion），后者受纠错码思想启发，专门移除承载关键结构信息的“负载核心”示例。通过引入“问题损伤评分”（Question Damage Score）量化删除操作的影响，可将谜题分类为脆弱型或鲁棒型。实验结果表明，在指令要求信息不足时应放弃回答的前提下，三款前沿大语言模型仍极少选择放弃，且在移除负载核心上下文后仍能持续输出正确答案，揭示其高度依赖记忆而非真正基于上下文推理。该框架不仅推动了对上下文推理、先验知识与语言推断机制的深入探究，还支持细粒度分析，包括因果干预、停止集分析、定向污染实验及机制可解释性研究，为理解模型内部决策过程提供了有效工具。
链接: https://arxiv.org/abs/2608.27756
作者: Neh Majmudar,Elena Filatova
机构: CUNY(纽约市立大学)
类目: Computation and Language (cs.CL)
备注: 21 pages
Abstract:Determining whether large language models derive answers from context or prior knowledge remains a fundamental challenge. Self-contained linguistic olympiad puzzles provide a controlled setting where all answers derive solely from expert-designed context examples without external knowledge. Removing individual context examples can eliminate information needed for specific questions while leaving the rest of the puzzle unchanged. We leverage this to introduce a diagnostic framework for analyzing individual context examples. Using 53 UK Linguistics Olympiad puzzles, we generate two modified variants by deleting a single context example: (1) uniform random deletion, and (2) targeted deletion (inspired by error-correcting codes) to remove a structurally load-bearing example uniquely carrying necessary information. We formalize this impact using a Question Damage Score to classify puzzles as fragile or robust. Evaluating three frontier LLMs under instructions to abstain when information is insufficient, we find they rarely abstain, often continuing to produce correct answers after load-bearing context is removed. These findings motivate further investigation into context-based reasoning, prior knowledge, memorization, and linguistic inference. Beyond abstention, the framework enables fine-grained analyses of context reliance, including causal interventions, stopping-set analysis, targeted contamination studies, and mechanistic interpretability.
[NLP-58] he Calls are Coming from Inside the Model: Investigating Probe-based Detection of Tool-Calling Errors in LLM s
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在与外部世界交互时，难以通过输入输出分析有效检测工具调用错误的问题，尤其关注那些因参数值错误但类型正确而未被标准日志系统记录的隐蔽性错误。其核心解决方案是利用线性探测（linear probes）技术，从模型的隐藏状态中提取潜在的行为与知识信息，以识别多种类型的工具调用异常。研究表明，该方法在伯克利函数调用排行榜（Berkeley Function Calling Leaderboard）上评估的18个工具调用模型中均表现出良好效果，关键成功因素包括模型规模、探测所处的网络层位置以及模型后训练类型；此外，探测器具备对新型错误的泛化能力，这对于实际部署中的安全性与鲁棒性至关重要。
链接: https://arxiv.org/abs/2608.27750
作者: Eric Yeats,Brendan Kennedy,Loc Truong,John Buckheit,Jung Lee,Jesse Friedbaum,John Emanuello,Henry Kvinge
机构: Pacific Northwest National Laboratory(太平洋西北国家实验室); National Security Agency(国家安全局)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注: 4 main pages, 8 pages of references and appendices
Abstract:The hidden states of large language models (LLMs) are known to capture rich information relating to model knowledge and behavior that can be hard to extract from examination of input and output alone. As LLM-based systems increasingly interface with the external world, one area of concern is detecting incorrect or improper use of tools. Motivated by this, we study the effectiveness of using linear probes to detect incorrect tool-calls, measuring probe efficacy across 18 tool-calling LLMs evaluated on the Berkeley Function Calling Leaderboard. Overall, we find that probing is an effective means to catch a range of different tool-calling errors, including errors arising from using an argument that has the wrong value but the correct type, which might not be recorded by standard logging frameworks. Important factors in success include model size, probing layer, and model post-training type. We also show that probes are capable of generalizing to novel types of errors, which is critical in real world deployments.
[NLP-59] Below the Noise Floor: Bimodal Seed Collapse and Distinct Failure Modes in Small-Model Knowledge Distillation
【速读】： 该论文旨在解决小模型知识蒸馏（Knowledge Distillation, KD）在函数路由（Function Routing）任务中因单次种子评估导致的性能误判问题。具体而言，现有研究多基于单一随机种子评估KD方法效果，但在实际应用中，不同种子间的训练结果存在显著差异，而这种变异性未被充分揭示，从而掩盖了真实性能表现。其解决方案的关键在于采用多种子（three to six seeds）的系统性评估框架，以揭示和量化小模型知识蒸馏中的核心失败模式。研究发现：（i）单个种子的标准差可达2.8至48.7个百分点，远超5个百分点的宣称增益，使得多数KD优势无法成立；（ii）七种KD变体中有三种出现双峰坍塌（bimodal collapse），部分种子准确率低于55%而其余正常，另有第四种表现出更高方差；（iii）坍塌具有特定模式——ce_kd与ce_paraphrase表现为错误函数选择，而reasoning_kd则呈现此前未被记录的输出截断模式，模型生成推理过程但未输出函数名（准确率仅0.9%）；（iv）仅有progressive_kd与rank_kd在所有测试种子下避免坍塌，且标准差为3.9个百分点；（v）原本通过输入增强获得+3.78个百分点的增益，在控制种子数量的交叉验证中转为-2.70个百分点。因此，该研究强调，仅依赖单种子评估会严重低估小模型知识蒸馏中的系统性失败风险，必须采用多种子、受控实验设计才能准确评估方法可靠性。
链接: https://arxiv.org/abs/2608.27729
作者: Dipto Sumit,Sakib Ul Haque,Farig Sadeque
机构: BRAC University (布拉大学), Bangladesh; Department of Computer Science and Engineering
类目: Computation and Language (cs.CL)
备注:
Abstract:Function routing – selecting the correct API call from a fixed catalog given a natural-language request – is a deployment problem where small students are attractive but knowledge distillation gains are typically reported single-seed, at scales where seed variance is unknown. On a 740-instance healthcare API routing task with a 1.5B Qwen student and a 20B teacher, we compare eight KD variants against supervised cross-entropy, using three to six seeds for key configurations. We find: (i) per-seed standard deviation ranges from 2.8 to 48.7 percentage points, swallowing every claimed KD gain below five points; (ii) three of seven KD variants exhibit bimodal collapse, with at least one in three to five seeds falling below 55% accuracy while the others train normally, and a fourth showing elevated variance; (iii) collapse has distinct modes – wrong-function selection for ce_kd and ce_paraphrase, and a previously undocumented output-truncation mode for reasoning_kd, where the model emits reasoning but terminates before producing a function name (0.9% accuracy); (iv) only progressive_kd and rank_kd avoid collapse across observed seeds, with sigma = 3.9 pp; (v) a naive cross-split +3.78 pp gain from input enrichment reverses to -2.70 pp under controlled within-split multi-seed re-testing. Single-seed evaluation is therefore unable to detect central failure modes in small-model KD.
[NLP-60] First Make It Playable Then Make It Good: Staged Interaction Learning for Small Dialogue-Game Agents EMNLP2026
【速读】： 该论文旨在解决小规模语言模型在对话式游戏交互中表现不佳的问题，特别是如何在有限参数量下提升模型的游戏可玩性（playability）与决策能力。其核心挑战在于：如何在不依赖复杂增强策略的前提下，有效提升模型在真实游戏场景中的连贯性、逻辑性与得分表现。解决方案的关键在于采用三阶段微调策略：首先基于Playpen数据集中成功的游戏轨迹进行标准监督微调（SFT），其次引入加权的逐轮级SFT以强化关键决策节点的表现，最后通过教师指导式SFT（teacher-guided SFT）优化输出格式并评估样本质量。值得注意的是，教师模型仅用于格式修正与评分反馈，未生成新的黄金动作（gold actions），从而避免引入偏差。实验结果表明，该方法显著提升了模型性能——在公开验证集上获得57.12的ClemScore与42.68的StatsScore，且在官方挑战赛中相较基线模型实现约+36的ClemScore增量，位列第二。研究进一步揭示，模仿完整游戏轨迹有助于提升可玩性，而逐轮与教师引导训练则更有效地优化决策质量；相比之下，诸如回放修复（replay-repair）和硬例挖掘（hard-example mining）等程序化增强手段并未带来收益，说明对于小模型而言，精心设计的数据筛选与训练策略比激进的工程改造更为有效。
链接: https://arxiv.org/abs/2608.27672
作者: Syed Mahbubul Huq,Pranava Madhyastha
机构: City, University of London(伦敦城市大学); The Alan Turing Institute(艾伦·图灵研究所)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted at the LMP Challenge (EMNLP 2026 Workshop)
Abstract:We present Qwen-GuidePlay-2B, a 2B-parameter language model for dialogue-game interaction. We fine-tune Qwen3.5-2B using three steps: a) SFT on only successful game trajectories from Playpen, b) weighted turn-level SFT, and c) teacher-guided SFT. The teacher model (which is a larger model) is only used to fix formatting and evaluate examples, but does not create new gold actions. Our final model scores 57.12 clemscore and 42.68 statscore on the public Playpen validation. In the officially released challenge results, our model obtains the second-highest Playpen clemscore delta among submitted systems (which is approximately +36 over its base model). Our findings suggest that imitating full trajectories helps with playability, while turn-level and teacher-guided training usually improve decision-making and increase the overall score. Alternative procedurally heavy approaches like replay-repair and hard-example mining did not help, which suggests that small models are performant simply by using careful curation strategies rather than aggressive changes. We make available both the model and the code for reproducibility.
[NLP-61] Semantic Watermarking with Order-Robust Detection over Sub-sentence Units
【速读】： 该论文旨在解决生成式语义水印（semantic watermarking）在面对内容保持型篡改攻击时的鲁棒性不足问题，特别是针对重述（rewording）、重排（reordering）和重新分段（resegmentation）等导致嵌入位移（embedding displacement）的攻击手段。其核心挑战在于：水印检测器仅能观测攻击者提供的文本，而攻击者可通过上述操作在不改变语义的前提下使文本嵌入空间发生偏移，从而规避检测。为此，论文提出自适应嵌入位移攻击（Adaptive Embedding Displacement Attack, EDA），通过统一优化目标最大化嵌入位移，利用公开的改写工具与代理编码器，在无需访问水印生成器或密钥的情况下，成功在4种水印方案中实现32.6%至47.9%的标记移除率，显著优于传统被动改写攻击。为应对这一威胁，作者设计了k-SwordStamp：一种基于子句单元的语义水印机制，具备对文本结构扰动的鲁棒性检测能力，显著降低对攻击者可选结构的敏感度，仅以轻微质量损失为代价。实验表明，即使面对适配k-SwordStamp的增强版EDA攻击，其成功率也仅为10.8%，远低于对k-SemStamp的65.5%攻击成功率，验证了新方案的有效性。
链接: https://arxiv.org/abs/2608.27666
作者: Abdulrahman Diaa,Jonathan Petit,Florian Kerschbaum
机构: University of Waterloo(滑铁卢大学); Qualcomm(高通公司)
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 20 pages, 10 figures, 5 tables
Abstract:Semantic watermarks tie the mark to sentence meaning rather than token choices, promising robustness to content-preserving edits. However, the detector only observes attacker-supplied text, which can be reworded, reordered, or resegmented to evade detection without content loss. Rewording, reordering, and resegmentation all cause embedding displacement: detection tests embeddings different from those selected during watermarking and can therefore lose the mark. Our adaptive embedding displacement attack (EDA) admits all three edits under a single objective that maximizes this displacement. It uses a public paraphraser and surrogate encoder without access to the provider’s generator or secret key. At a 5% false-positive rate (FPR) and content-preservation threshold \barq=90% , EDA successfully removes the mark on between 32.6% and 47.9% of documents across four schemes, the highest among the tested attacks. Therefore, EDA evaluates the schemes’ robustness more thoroughly than passive paraphrasing. To address these vulnerabilities, we design (k)-SwordStamp: semantic watermarks with order-robust detection over sub-sentence units, reducing sensitivity to attacker-chosen structure at a small quality cost. Against k-SwordStamp, the strongest no-box attack we test is an EDA variant adapted to its design, with a 10.8% attack-success rate. A stronger EDA with access to the provider’s detector and secret key reaches a 39.7% attack-success rate, compared with 65.5% on k-SemStamp. Our code is available at this https URL. Comments: 20 pages, 10 figures, 5 tables Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL) Cite as: arXiv:2608.27666 [cs.CR] (or arXiv:2608.27666v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2608.27666 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-62] Knowing Before Answering: Decoding Language Models for Reliable RAG
【速读】： 该论文旨在解决检索增强生成（Retrieval-Augmented Generation, RAG）系统中因检索结果不足或存在冲突信息而导致回答不可靠的问题。现有系统缺乏对输入文档是否足以支持回答的判断能力，无法识别信息不足或矛盾的情况。为此，作者提出将此问题建模为三分类任务——判断提供的信息为“充分”、“不足”或“冲突”，并构建了一个受控的基准数据集，模拟真实RAG场景并标注相应标签。其解决方案的关键在于利用语言模型内部的隐藏激活（hidden activations）和注意力衍生特征作为输入，训练一个轻量级线性分类器（即特征路由模块），以实现对三类状态的准确区分。在16种不同架构与规模的语言模型上验证表明，该基于特征的路由方法显著优于基于提示（prompting-based）的基线方法及专用RAG模型。进一步分析揭示，最具判别性的信号集中于模型中间层，其中隐藏激活状态的表现普遍优于注意力值或MLP特征输出。研究结果表明，语言模型内部确实编码了关于检索证据是否充分的可解码信号，该信号可被可靠提取用于RAG系统的初步诊断与决策。
链接: https://arxiv.org/abs/2608.27661
作者: Syed Mahbubul Huq,Christopher Child,Tillman Weyde,Pranava Madhyastha
机构: City St George’s, University of London(伦敦城市圣乔治大学); The Alan Turing Institute(艾伦图灵研究所)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted at the Third Conference on Language Modeling (COLM 2026)
Abstract:In Retrieval-Augmented Generation (RAG), retrieval may provide insufficient or conflicting information needed to answer a question. The system should not only know when to answer but also be able to identify cases in which the documents provided in RAG are insufficient or contain conflicting information. This can be framed as a three-way classification problem, where we use the model’s internal signals to determine whether the provided information in the input can be classified as sufficient, insufficient, or conflicting. We create a controlled benchmark dataset that replicates a RAG setup with fictitious information and labels each instance as answerable, insufficient, or conflicting. We use hidden activations and attention-derived features as inputs to train a lightweight linear model to distinguish among the three classes. Across 16 language models spanning different architectures and a range of model sizes, our feature-based router consistently outperforms prompting-based baselines and the performance of specialised RAG-models. We further conduct analyses into the information dynamics of the models. We show that the most informative signals for the classification are available in the middle layers, with hidden activation states being more effective than attention values or the MLP-feature outputs in most of the tested models. Overall, our results suggest that language models internally encode whether retrieved evidence is sufficient to support answering, and that this signal can be decoded reliably for RAG triage.
[NLP-63] When Tokenizers Fail: Byte-Level Chunking for Zero-Shot Transfer to Low-Resource Languages EMNLP2026
【速读】： 该论文旨在解决低资源语言处理中因子词分词（subword tokenization）导致的分布偏差问题，即主流语言的频率模式被强制映射到共享书写系统的语言变体上，进而影响模型性能。同时，针对字节级模型在非拉丁语系中因粒度不匹配而引发的词级任务表现不佳问题，提出一种改进方案。其核心解决方案是构建一种自适应的层次化字节网络框架，通过直接从冻结的子词基础模型中初始化字节嵌入，利用块对齐损失（chunk alignment loss）将动态分组的字节块投影至预计算的子词目标，辅以轻量级词性标注（POS）监督引导边界检测，从而在无需大规模训练的情况下弥合模态差异。实验结果表明，该无分词器方法在六种语言上的词级形态任务中显著提升性能，词性标注任务最高提升达13.3%。
链接: https://arxiv.org/abs/2608.27658
作者: Sanjeev Kumar,Atsuki Yamaguchi,Nikolaos Aletras
机构: Indian Institute of Technology Bombay (印度理工学院孟买分校); University of Sheffield (谢菲尔德大学)
类目: Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026
Abstract:Subword tokenization hinders low-resource language processing by imposing frequency patterns from dominant languages onto script-sharing variants. Byte-level models bypass this issue by processing raw UTF-8 characters, yet they create a granularity mismatch for word-level tasks in non-Latin scripts. Hierarchical byte-level architectures address this mismatch by grouping bytes into word-aligned chunks. However, these architectures require massive training data and suffer from representational misalignment when paired with frozen subword-based language models. In this paper, we propose an adapted hierarchical network framework that bridges this modality gap without extensive training. Our method initializes byte embeddings directly from the subword representations of a frozen base model. We apply a chunk alignment loss to project dynamically grouped byte chunks toward precomputed subword targets, and interleave lightweight part-of-speech (POS) supervision to guide boundary detection. Experiments across six languages demonstrate that our tokenizer-free approach improves performance for word-level morphological tasks, yielding up to a 13.3% improvement on POS tagging.
[NLP-64] rajectory-Level Speculative Decoding for Diffusion Language Models
【速读】： 该论文旨在解决基于扩散的语言模型（dLLMs）在低置信度情况下解码策略退化为单标记生成的问题，从而严重限制了推理吞吐量。其核心挑战在于，与自回归模型中按固定左右顺序进行标记推测不同，dLLMs需对具有显式位置和去掩码顺序的多标记更新轨迹进行推测。为此，论文提出一种轨迹级推测框架，通过分层置信度树探索构建草稿去噪轨迹，并利用双向注意力掩码实现块级并行验证；进一步引入块间推测机制，借助扩散模型的双向结构实现跨块前瞻。研究形式化刻画了该方法的精确性条件，并指出轨迹漂移是提升并行性带来的根本代价。基于Fast-dLLM的双缓存架构，该方法将去噪迭代次数减少30%-40%，每步生成令牌数从2.6提升至4.3，在推理和代码任务基准上实现7-14倍于原始dLLMs的加速，较Fast-dLLM提升1.3倍，且准确率损失低于1%。
链接: https://arxiv.org/abs/2608.27514
作者: Tianxiang Pan,Baitao Gong,Mo Guang,Hongwei Yong,Tianpeng Jiang,Yaqian Li,Zheng Cao,Kaiwen Long
机构: Li Auto Inc.(理想汽车); Shanghai, China
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Diffusion-based language models (dLLMs) enable parallel token generation through iterative denoising, but existing decoding strategies collapse to single-token generation under low confidence, severely limiting throughput. Unlike autoregressive models where speculative decoding operates on token sequences in a fixed left-to-right order, dLLMs require speculating over denoising trajectories-sequences of multi-token updates with explicit positions and unmasking orders. We develop a trajectory-level speculative framework that constructs draft denoising trajectories via confidence-stratified tree exploration and verifies them through blockwise parallel evaluation with bidirectional attention masking. Our method further introduces inter-block speculation, exploiting diffusion models’ bidirectional structure to perform cross-block lookahead. We formally characterize when this approach is exact and identify trajectory drift as the fundamental cost of increased parallelism. Building on Fast-dLLM’s dual-cache infrastructure, our framework reduces denoising iterations by 30-40% and increases tokens-per-step from 2.6 to 4.3, achieving 7-14x speedup over vanilla dLLMs and 1.3x over Fast-dLLM with less than 1% accuracy change across reasoning and code benchmarks.
[NLP-65] Quantization-Triggered Backdoors in Language Models: Cross-Quantizer Transferability and the Validation–Deployment Gap
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在边缘部署过程中，因后训练量化（Post-training Quantization）引入的语义不一致性问题。传统做法通常假设量化是语义中立的优化手段，仅在高精度（如FP16）源模型上进行验证后即直接应用量化，忽略了量化作为参数空间上的多对一映射可能导致行为等价性丧失这一关键风险。其核心问题是：源模型在全精度下的行为认证无法保证量化后部署配置的行为一致性，从而形成“结构化验证-部署差距”（structural validation–deployment gap）。解决方案的关键在于提出量化行为等价类（Quantization Behavioral Equivalence Classes, QBECs），并证明QBEC成员身份并不蕴含行为等价性，为量化触发的后门攻击提供了理论基础。研究进一步构建了一个三阶段对抗微调框架，将隐蔽恶意载荷嵌入满足源精度验证条件的模型中，使其在INT8或4比特压缩后激活特定对抗性行为。实验在战术机器翻译与政治内容分析两类实际场景中验证了该威胁，结果显示，经量化的翻译模型在敌我识别任务中的误判率从0%升至最高85.02%，立场分类器的意识形态偏移量ΔBias达0.33。跨量化器迁移性分析表明，攻击持久性受量化方案与模型架构共同影响，而非仅由名义位宽决定。研究结论强调，仅依赖源精度审计不足以排除量化引发的行为异常，可信边缘人工智能必须将最终部署配置纳入行为认证范围。
链接: https://arxiv.org/abs/2608.27512
作者: Jacopo Dardini,Claudio Stanzione,Giordano Colò,Giuseppe Fenza
机构: University of Bologna (博洛尼亚大学); Luiss Guido Carli University (卢伊士·吉多·卡里大学); Live Tech (Live Tech); Department of Artificial Intelligence (人工智能系); University of Salerno (萨莱诺大学)
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Cryptography and Security (cs.CR)
备注: Accepted at the 21st International Conference on Availability, Reliability and Security (ARES 2026)
Abstract:Post-training quantization is often treated as a semantically neutral optimization for edge deployment of Large Language Models. When a full-precision source checkpoint is evaluated and quantization is applied downstream without equivalent re-evaluation, this workflow creates a structural validation–deployment gap: because quantization is a many-to-one mapping over parameter space, source-precision certification does not guarantee behavioral equivalence in the deployed configuration. We formalize this gap through Quantization Behavioral Equivalence Classes (QBECs) and prove that QBEC membership does not imply behavioral equivalence, providing a theoretical basis for quantization-triggered backdoor attacks. Building on a three-stage adversarial fine-tuning framework, we embed latent malicious payloads into models that satisfy the source-precision checks used in our evaluation, yet activate targeted adversarial behavior upon INT8 or 4-bit compression. We evaluate this threat in two operationally motivated scenarios, tactical machine translation and political content analysis, extending prior work from decoder-only causal LMs to multilingual encoder-decoder sequence-to-sequence models. Results show that backdoored translation models move from zero measured friend–foe corruption at repaired FP16 to up to 85.02% inversion after quantization, and that a paired stance classifier measures an ideological shift of up to \Delta\mathrmBias=0.33 upon compression. A cross-quantizer transferability analysis further shows that attack persistence varies across quantization schemes and model architectures, rather than being determined by nominal bit-width alone. These findings demonstrate that source-precision auditing alone does not rule out quantization-triggered behavior and that the final deployed configuration must be included in behavioral certification for trustworthy edge AI.
[NLP-66] How Do Linear Probes Emerge? A Circuit-Tracing Framework with Concept-Targeted Attribution EMNLP2026
【速读】： 该论文旨在解决生成式模型中内部概念表征的可解释性问题，即如何明确揭示模型在生成文本过程中为何产生特定概念表征，而不仅限于解释其对下一个词元（next token）高概率输出的原因。传统注意力图（attribution graphs）通常以生成目标为导向，难以分离出与内部概念相关的具体计算机制。为此，论文提出概念目标导向归因（Concept-Targeted Attribution, CTA），通过将归因图训练于线性探测器（linear probe）的方向上，构建针对特定概念探测任务的可解释电路。其核心创新在于：CTA能够生成与探测任务直接相关的神经路径结构，从而独立于最终生成的词元，解释为何某一内部概念在输入提示中被激活。研究采用跨层编码器（Cross-Layer Transcoders）验证了该方法的有效性，发现图级别的特征能高度预测四种典型概念类别上的探测准确率（ρ = 0.91，R² = 0.84），局部特征则精准定位驱动每条提示分类的关键稀疏组件。进一步因果消融实验表明，探测相关特征与词元相关特征所对应的是功能上截然不同的机制——移除探测相关特征会显著降低内部概念得分但基本不改变生成结果，而移除词元相关特征则在92%至100%的案例中改变生成词元且对探测分数影响微乎其微。因此，CTA实现了从行为层面的探测准确率评估向机制层面的可解释性分析的跃迁，为内部概念表征（包括安全关键领域）提供了更精细的审计能力。
链接: https://arxiv.org/abs/2608.27510
作者: Vedant Palit,Florent Draye,Terry Jingchen Zhang,Bernhard Schölkopf,Zhijing Jin
机构: Jinesis Lab, University of Toronto, Vector Institute; EuroSafeAI; Max Planck Institute for Intelligent Systems, Tübingen, Germany; Hector Foundation; ELLIS Institute Tübingen
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: Accepted to EMNLP 2026 (Main Conference). 29 pages, 18 figures, 9 tables
Abstract:Transcoder attribution graphs are usually trained to explain why a model assigns high probability to a particular next token. We introduce Concept-Targeted Attribution (CTA), which instead trains attribution graphs with respect to a linear probe direction. CTA therefore yields probe-specific circuits that explain why an internal concept representation arises in a prompt, independently of whether it is expressed in the generated token. Using Cross-Layer Transcoders, we show that these probe-targeted graphs contain predictive structure: graph-level features predict probe accuracy across four widely studied concept categories ( \rho = 0.91 , R^2 = 0.84 ), while local features identify the sparse components driving per-prompt classification. This connects probe performance to interpretable circuit structure, allowing us to ask not only whether a probe works, but which internal computations make it work. Causal ablations further show that probe-targeted and logit-targeted graphs capture functionally distinct mechanisms. Removing probe-relevant features reduces internal concept scores while largely preserving generated tokens, whereas removing logit-relevant features changes the generated token in 92% to 100% of cases with near-zero effect on probe scores. CTA provides a framework for moving from behavioral probe accuracy to mechanistic explanations of probe performance, enabling more detailed audits of internal concept representations, including safety-critical ones. Our code is available at this https URL
[NLP-67] A Survey on Rubric-Guided Reinforcement Learning for Language Models EMNLP2026
【速读】： 该论文旨在解决传统基于人类反馈的强化学习（Reinforcement Learning from Human Feedback, RLHF）中依赖标量奖励信号所导致的可解释性差及难以全面捕捉生成内容质量多维度特征的问题。其核心解决方案在于引入结构化、可解释的评估标准——即“评分量表”（rubrics），作为奖励设计、反馈生成与策略优化的基础，从而实现更精细、可控的对齐过程。该研究的关键创新在于提出一个贝叶斯框架，将“宪法”（constitution）定义为评价标准的先验分布 $ P® $，而具体实例的评分量表则被视为在给定输入 $ x $ 条件下的后验实例化 $ R_x \sim P(R|x) $，由此构建了一个统一的理论视角。在此基础上，论文系统地提出了沿先验-后验轴划分的评分量表引导强化学习（rubric-guided RL）分类体系，涵盖宪法式AI、实例特定评分量表、过程级监督、自演化评分量表及其代理式与多模态扩展。此外，鉴于评分量表本身是自然语言构造物，研究进一步通过语言学分析揭示了粒度权衡、语义漂移及语言奖励劫持等现象对对齐可靠性的影响，识别出若干亟待深入探索的关键开放问题。
链接: https://arxiv.org/abs/2608.27505
作者: Zifei Shan,Fangning Shao
机构: WeChat(微信); Tencent(腾讯); Independent Researcher(独立研究者)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted to Findings of EMNLP 2026
Abstract:Reinforcement learning from human feedback (RLHF) has become the dominant paradigm for aligning large language models (LLMs) with human preferences. However, traditional RLHF relies on scalar reward signals that lack interpretability and fail to capture the multifaceted nature of response quality. Rubric-guided reinforcement learning addresses these limitations by introducing structured, interpretable evaluation criteria, or rubrics, as the backbone of reward design, feedback generation, and policy optimization. In this survey, we introduce a Bayesian framework that defines constitutions as prior distributions P® over evaluation criteria and rubrics as conditional instantiations R_x \sim P(R|x) . Under this unified view, we present a taxonomy of rubric-guided RL along the prior-posterior axis, covering constitutional AI, instance-specific rubrics, process-level supervision, self-evolving rubrics, and their agentic and multimodal extensions. Furthermore, as rubrics are natural-language artifacts, we present a linguistic analysis of how granularity trade-offs, semantic drift, and linguistic reward hacking impact alignment reliability, identifying key open problems for future research.
[NLP-68] INSPIRE: An Internalize-Then-Improve Approach for Example-Driven Mathematical Reasoning EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在数学推理任务中过度依赖最终答案正确性优化，导致模型可能仅通过记忆解题模式而非真正内化数学概念的问题。其核心挑战在于：现有方法难以有效构建用于偏好优化的高质量示例对，且模型需经历从策略习得至正确应用的渐进式能力发展过程。为此，论文提出INSPIRE框架，其关键创新在于“先内化、再提升”（Internalize-Then-Improve）的双阶段策略：首先采用参考引导的学生内化（Reference-Guided Student Internalization, RGSI），在策略模型自身分布下生成高质量的偏好候选样本；随后通过分阶段的评分标准偏好训练（stage-wise rubric preference training），将学习过程分解为方法导向与正确性导向两个阶段，逐步实现对基于示例推理（example-based reasoning）等深层数学理解能力的建模。实验表明，该方法在多规模、多架构模型上均取得一致性能提升，甚至超越更大规模开源模型，且在分布外基准测试中未出现泛化能力退化，验证了其有效性与鲁棒性。
链接: https://arxiv.org/abs/2608.27501
作者: Shuai Wang,Jiayi Kuang,Yinghui Li,Haojing Huang,Xinnian Liang,Ying Shen,Liang Lin
机构: Sun Yat-sen University(中山大学); Tsinghua University(清华大学); ByteDance Inc.(字节跳动公司); Peng Cheng Laboratory(鹏城实验室)
类目: Computation and Language (cs.CL)
备注: EMNLP 2026
Abstract:Mathematical reasoning has seen rapid progress in large language models (LLMs), yet existing methods optimize predominantly for final-answer correctness, raising the question whether models truly internalize mathematical concepts or merely memorize solution patterns. In human mathematics education, example-based reasoning such as constructing counterexamples to test theorem boundaries reflects deep conceptual understanding, but remains underdeveloped in current LLMs. Enhancing this capability through preference optimization presents two key challenges: (1) the model’s limited example-based reasoning ability makes constructing effective preference pairs inherently difficult; and (2) capability acquisition is progressive, as the model must first learn to adopt this strategy before learning to apply it correctly. Therefore we propose INSPIRE, an Internalize-Then-Improve approach combining Reference-Guided Student Internalization (RGSI), which produces high-quality preference candidates under the policy model’s own distribution, with a stage-wise rubric preference training strategy that decomposes learning into method-oriented and correctness-oriented stages. Experiments across multiple model scales and families demonstrate consistent improvements, even surpassing larger open-source models, while evaluations on out-of-distribution benchmarks confirm no degradation in general mathematical reasoning ability.
[NLP-69] XHotpotQA: A Benchmark for Cross-Lingual Knowledge Composition in Multi-Hop Question Answering
【速读】： 该论文旨在解决多跳知识问答（multi-hop question answering）中跨语言知识组合的挑战，尤其针对现有多语言基准数据集在翻译整个示例时隐藏了语言边界处推理失败的问题。其核心解决方案是提出XHotpotQA，一个受控的跨语言知识组合基准，通过将每个实例建模为带有显式语言标注的证据依赖图（evidence-dependency graph），明确区分问题、桥接证据、答案承载证据及干扰项的语言属性。该基准包含15,661个训练与7,405个验证样本，提供句级支持监督和预设干扰项。实验表明，当问题与答案证据跨越不同语言时，完全不匹配导致的Unicode感知答案F1分数下降达10.25至15.79分，不同书写系统证据导致的性能损失更高达11.98至23.70分，而适配后的选择器仅表现出1.71至1.78分的差异，说明阅读器对语言不匹配的敏感度远高于选择器。因此，该研究的关键在于构建一个能够揭示语言间知识整合缺陷的诊断性、模块化评估框架，为需要跨语言证据融合的知识型系统提供了可审计的测试平台。
链接: https://arxiv.org/abs/2608.27481
作者: Iman Barati,Arash Ghafouri,Behrouz Minaei-Bidgoli
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Knowledge-intensive multi-hop question answering requires systems to select evidence and compose dependent facts, yet multilingual benchmarks usually translate an entire example into one language. This hides failures at language boundaries inside the reasoning chain. We introduce XHotpotQA, a controlled benchmark for cross-lingual knowledge composition over mixed-language evidence. Each instance is modeled as an evidence-dependency graph whose question, bridge evidence, answer-bearing evidence, and distractors have explicit language assignments. The audited resource contains 15,661 training and 7,405 validation instances, with sentence-level support supervision and supplied distractors. In validation, 99.81% of items cross the question-to-gold-evidence language interface and 95.60% use gold paragraphs in different languages. Across three reader artifacts, full question-evidence mismatch is associated with 10.25 to 15.79 lower Unicode-aware answer F1 than partial alignment, and different-script evidence with deficits of 11.98 to 23.70 points; the corresponding adapted-selector contrasts are 1.71 and 1.78 points. Under this supplied-candidate design, the evaluated readers therefore show substantially larger condition-associated deficits than the selector. XHotpotQA provides role-aware diagnostics, modular evaluation, and an audited test bed for knowledge-based systems that must integrate evidence across languages.
[NLP-70] Retrieving Relations Detecting Fallacies: A RAG Approach to Political Debate Analysis
【速读】： 该论文旨在解决在高敏感性语境（如重大政治辩论）中自动识别谬误（fallacy）的难题，其核心挑战在于谬误判断依赖于超出文本表面的上下文知识，包括议题相关的世界知识以及论证之间支持与攻击关系等论辩结构信息。现有方法虽已证明论辩结构有助于提升分类性能，但通常仅将其作为静态特征输入，缺乏灵活性。为此，本文提出一种基于引导式检索增强的谬误检测与分类方法，其关键创新在于利用论证间的支持与攻击关系动态引导外部知识库的文档检索过程，从而实现对相关背景信息的精准提取。实验在ElecDeb60to20基准上进行，覆盖42种检索配置与14种模型，在15GB政治相关文档知识库上验证，结果表明，该方法在谬误检测上的宏平均F1最高达0.864，在分类任务上最高达0.725，显著优于非检索基线，证明了论辩引导式检索机制在引入外部知识方面的有效性与优越性。
链接: https://arxiv.org/abs/2608.27471
作者: Deborah Dore,Greta Damo,Elena Cabrio,Serena Villata
机构: Université Côte d’Azur, CNRS, INRIA, I3S (蔚蓝海岸大学，法国国家科学研究中心，法国信息与自动化研究所，信息与安全研究所)，France (法国)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Fallacies are arguments that employ invalid reasoning, making their automatic detection critical in sensitive contexts such as high-stakes political debates, where public opinion is shaped. Spotting a fallacious argument requires contextual knowledge beyond its pure surface text. This entails world knowledge pertaining to the subject matter under discussion, as well as knowledge of the relationships that exist between arguments within the argumentative discourse. Prior work on fallacy analysis has shown that argumentative discourse structure can beneficially improve classification performance. However, such structure is typically encoded only as static classifier features, limiting its flexibility. Building on this intuition while addressing this limitation, we introduce a guided retrieval-augmented methodology for fallacy detection and classification that leverages argumentative relations of support and attack to dynamically steer the extraction of relevant documents. We evaluate our approach on the ElecDeb60to20 benchmark across 42 retrieval configurations and 14 models, performing retrieval over a 15GB knowledge base of collected political-related documents. Our approach improves macro-F1 up to 0.864 for fallacy detection and up to 0.725 for classification over non-retrieval baselines. These results show that incorporating external knowledge significantly enhances fallacy detection and classification when retrieval is argumentatively guided.
[NLP-71] Select Dont Train: The Benefits of Modular Entity Disambiguation with LLM -Based Selection ISWC2026
【速读】： 该论文旨在解决实体消歧（Entity Disambiguation, ED）任务中传统方法将候选实体检索与正确实体选择合并为单一任务所导致的性能瓶颈问题。现有神经模型通常在共享嵌入空间中联合优化两个子任务——高召回率的候选检索与细粒度的实体选择，这种耦合设计迫使表示学习需在两者间权衡，且依赖于需持续维护的训练过的检索器，增加了系统复杂性。本文提出一种系统性的检索策略比较框架，将基于大语言模型（LLM）的实体选择作为统一阶段，而独立评估多种检索策略（包括稀疏检索BM25、Web知识库搜索以及先进的训练型稠密检索器）与不同开源及闭源LLM组合的效果。研究发现，一旦选择阶段由高性能LLM承担，训练检索器带来的增益有限；仅使用无需训练的BM25检索器与LLM结合即可在ZELDA基准上达到新的最优表现，使inKB micro-F1从82.3提升至86.3（+4），若搭配训练过的稠密检索器则进一步提升至88.5。更重要的是，解耦检索与选择暴露了当前ED系统的根本缺陷：当正确实体未被检索到时，系统被迫预测错误实体。为此，本文框架引入“拒答”机制，在检测到检索失败时允许系统选择不输出，从而避免错误预测。在奖励正确拒答的评估设置下，该训练免费的BM25 + LLM流水线实现了90.7的F1，显著优于传统方法。其核心解决方案在于通过解耦检索与选择，并利用强大LLM实现智能选择与拒答，既降低了维护成本，又提升了鲁棒性与准确性。
链接: https://arxiv.org/abs/2608.27470
作者: Fina Polat,Daniel Daza,Pengyu Zhang,Klim Zaporojets,Paul Groth
机构: University of Amsterdam (阿姆斯特丹大学); Vrije Universiteit Amsterdam (阿姆斯特丹自由大学); Aarhus University (奥胡斯大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Databases (cs.DB)
备注: Accepted at the 25th International Semantic Web Conference (ISWC 2026), Research Track
Abstract:Entity Disambiguation (ED) is a key task for constructing and using knowledge graphs. State-of-the-art neural approaches commonly model ED as a single task, although it consists of two distinct subproblems: retrieving candidate entities and selecting the correct one given context. Dual-encoder models optimize for both within a shared embedding space, forcing representations to balance high-recall retrieval with fine-grained selection, and they require trained retrievers, which are costly to maintain as knowledge graphs change. While recent work has begun to combine retrievers with LLM-based selectors, the interplay between the two stages has not been studied systematically. In this paper, we present a systematic comparison of retrieval strategies for candidate generation under a shared LLM-based selection stage, combining sparse retrieval (BM25), Web KB search, and a state-of-the-art trained dense retriever with several open- and closed-source LLMs. We show that, once selection is delegated to a capable LLM, training the retriever provides only modest additional value: a fully training-free BM25 retriever paired with an LLM selector reaches a new state of the art on the ZELDA benchmark, raising inKB micro-F1 from 82.3 to 86.3 (+4); pairing the same LLM with a trained dense retriever reaches 88.5. Decoupling retrieval from selection also exposes a limitation of current ED systems: when the correct entity is missing from retrieved candidates, they are forced to predict an incorrect entity. In contrast, our framework allows for abstention when retrieval failure is detected. In an evaluation setting that rewards correct abstentions, the training-free BM25 + LLM pipeline reaches 90.7 F1.
[NLP-72] PACE: Publisher-Adaptive Content Extraction via Agent ic Automation
【速读】： 该论文旨在解决大规模网页内容抽取中准确率、可扩展性与适应性难以兼顾的问题。现有方法在面对特定出版商的复杂布局及包含元数据、图片、表格等多模态信息的抽取目标时，普遍存在局限：通用提取器泛化能力虽强但对特定布局鲁棒性差；基于大语言模型（LLM）的直接抽取方法灵活性高但存在成本与延迟过高的问题；而人工设计的专用解析器虽精度高，却需大量人力投入维护。为此，论文提出PACE（Publisher-specific Agentic Configuration Extractor），一种基于智能体（agentic）框架的自动化学习方法，通过分析代表性页面和用户需求，自动生成针对特定出版商的内容抽取配置。其核心在于：在训练阶段利用LLM分析页面结构并归纳可复用的抽取模式；在推理阶段则将学习到的配置固化为确定性的抽取模板，实现无需额外调用大模型的高效、可扩展抽取。实验表明，PACE在文章正文、元数据及多模态内容（如图像、表格）抽取上均显著优于非人工的可扩展基线方法，且接近人工定制解析器的性能，验证了智能体配置学习在构建面向大模型（LLM-ready）的高质量网页表示方面的有效性。
链接: https://arxiv.org/abs/2608.27466
作者: Zhanlin Liu,Munirathnam Srikanth
机构: ProRata.ai
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Web content extraction is essential for reliable LLM data pipelines, yet existing methods often struggle to jointly satisfy accuracy, scalability, and adaptability. General-purpose extractors can be applied broadly, but they are often brittle on publisher-specific layouts and richer extraction targets such as metadata, images, and tables. Direct LLM-based extraction offers greater flexibility, but incurs substantial cost and latency at scale, while manually engineered publisher-specific parsers can achieve high accuracy but require substantial human effort to build and maintain. We introduce PACE, an agentic framework for learning publisher-specific extraction configurations from representative pages and user requirements. During training, PACE uses LLMs to analyze page structure and aggregate reusable extraction patterns. At inference time, the learned configurations instantiate a fixed deterministic extractor template, enabling scalable extraction without additional LLM calls. Experiments spanning article-body, metadata, and multimodal extraction show that PACE outperforms scalable non-manual baselines while approaching the quality of manually engineered publisher-specific parsers. PACE achieves stronger extraction of article text, metadata, images, and tables, demonstrating that agentic configuration learning can automate publisher-specific extraction for LLM-ready page representations beyond article text. Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI) Cite as: arXiv:2608.27466 [cs.CL] (or arXiv:2608.27466v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2608.27466 Focus to learn more arXiv-issued DOI via DataCite
[NLP-73] Sledgehammer or Scalpel? A Fine-grained Adaptive Framework for Implicit Hate Speech
【速读】： 该论文旨在解决隐性仇恨言论（implicit hate speech）在在线内容审核中难以检测的问题，尤其针对那些通过隐喻和语境暗示隐藏恶意、表面合规的表达形式。现有基于预训练语言模型（PLM）或大语言模型（LLM）的方法通常对所有样本采用统一的推理流程，忽视了语言层面的细粒度差异，导致对简单样本产生不必要的计算开销。为应对这一挑战，论文提出了一种细粒度自适应隐性仇恨言论检测框架（Fine-grained Adaptive Implicit Hate Speech Detection, FAID），其核心在于根据仇恨言论的复杂程度将其划分为三类：浅层型（Shallow）、目标指向型（Targeted）和上下文依赖型（Context-Dependent）。针对不同类别，FAID采用差异化处理策略：对浅层样本采用轻量级提示调优（prompt-tuning）实现快速分类；对目标指向型样本引入知识增强机制，通过迭代优化揭示隐藏目标；对上下文依赖型样本则构建代理式（agentic）框架，自动生成提示以动态扩展上下文、推断缺失背景并识别模糊的恶意意图。该自适应架构实现了计算资源在复杂样本上的精准投放，避免了对简单样本的冗余推理，实验结果表明其在四个基准数据集上显著优于现有最先进方法。
链接: https://arxiv.org/abs/2608.27462
作者: Han Wang,Yuhu Cheng,Xuesong Wang,Yi Zhu
机构: China University of Mining and Technology (中国矿业大学); Yangzhou University (扬州大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Unlike explicit attacks with obvious profanity, implicit hate speech hides malice within seemingly compliant expressions through metaphors and contextual hints, making its detection in online content review challenging. While existing PLM- or LLM-based methods perform well, they typically apply a single reasoning process to all samples. This overlooks fine-grained linguistic nuances and causes unnecessary computation for simpler cases. We observe that online hate speech is not monolithic but manifests in varied forms. We therefore define three fine-grained categories: Shallow, Targeted, and Context-Dependent. Accordingly, we propose Fine-grained Adaptive Implicit Hate speech Detection (FAID), a novel framework that first performs fine-grained classification and then adapts to specific categories. Specifically, for Shallow samples with surface-identifiable intents, the framework adopts lightweight prompt-tuning for rapid classification; for Targeted comments that bind malicious intent to concealed targets, we design knowledge augmentation to iteratively refine the model and reveal hidden targets; for Context-Dependent comments lacking background information, we utilize an agentic framework that automatically generates prompts to evolve context, infer missing background information and identify ambiguous malicious intents. This adaptive architecture focuses computational resources on complex implicit samples while avoiding redundant reasoning for shallow samples. Experiments on four benchmark datasets demonstrate that FAID significantly outperforms SOTA baselines.
[NLP-74] SciReC: Diagnostic Evaluation of Multimodal Multi-Turn Relational Reasoning with Adaptive Interaction
【速读】： 该论文旨在解决多模态大语言模型（Multimodal Large Language Models, MLLM）在复杂关系推理任务中的性能评估与缺陷诊断问题，尤其关注其在类比、结构及因果等高级认知关系理解上的表现。为实现这一目标，研究构建了SciReC——一个面向学术对话的模型自适应多模态基准测试集，以系统评估模型在多种关系推理场景下的能力。解决方案的关键在于提出一种基于缺陷分析的诊断框架DMRA（Deficit-based Diagnostic Framework），通过量化视觉理解、知识呈现与记忆召回等多维度组件对推理失败的贡献度，精准识别导致模型表现不佳的核心原因。实验结果表明，尽管Claude 4.6在整体关系推理得分上表现最优（73%），但所有模型均普遍存在关系推理能力不足的问题，其次为记忆限制，且不同模型在空间、层级与序列关系上的短板呈现显著差异，揭示出当前MLLM在高层次语义整合与跨模态关联建模方面仍存在系统性瓶颈。
链接: https://arxiv.org/abs/2608.27461
作者: Nilay Yilmaz,Naga Sai Abhiram Kusumba,Stella Wenxing Liu,Yezhou Yang
机构: Arizona State University (亚利桑那州立大学); Capital One (资本一)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Relational reasoning requires the process of perceptual understanding, comparing, and integrating the underlying relationships between concepts. This ability consists of multiple categories, such as analogical, structural, and cause-effect, each capturing a different aspect of higher-order understanding. To examine the performance of multimodal large language models (MLLM) on these relational inference tasks, we developed SciReC, a model-adaptive multimodal academic dialog benchmark. As the relational reasoning process involves multiple representations and various factors (visual understanding, exhibiting knowledge, and memory recall), we propose DMRA, a deficit-based diagnostic framework that quantifies the contribution of these components to identify the primary cause of unsuccessful cases. Claude 4.6 achieved the best performance on the overall relational score with 73%, followed by GPT 5.4 with 68%. Performance trends indicate that open-source models achieve their lowest scores on spatial relations, while proprietary models struggle more with hierarchical and sequential relations. Across domains, model performance is lowest on Astronomy and highest on Psychology. The results of DMRA reveal that relational reasoning is the primary source of error across all models, followed by memory limitations.
[NLP-75] Accelerating LLM Inference via Vector Index Based Output Embeddings ICML2026
【速读】： 该论文旨在解决自回归解码过程中因大规模输出嵌入矩阵导致的内存带宽瓶颈问题，尤其是在具有大容量多语言词汇表的紧凑型大语言模型（LLM）中尤为显著。其核心解决方案是将传统的输出投影与top-k token选择过程重新建模为在词元嵌入上的最大内积搜索（Maximum Inner Product Search, MIPS），并用基于HNSW（Hierarchical Navigable Small World）的向量索引替代原有的稠密词汇投影层。该方法通过高效检索高分候选词元集合，大幅减少计算和内存开销，并可通过将检索到的对数（logits）散射至稀疏全词汇张量的方式无缝集成至现有解码流水线。在CPU推理环境下，针对Gemma 3、Llama 3.2和Qwen 3等模型的实验表明，该方法可使输出投影阶段的性能显著提升，端到端单批次解码吞吐量最高提升达82%（以Gemma 3 270M为例），同时在AlpacaEval评测下保持生成质量不变。结果表明，在延迟敏感的小批量解码场景中，近似检索是一种可行且高效的稠密输出投影替代方案。
链接: https://arxiv.org/abs/2608.27460
作者: Martin Loretz,Sepp Hochreiter
机构: Martin Loretz; Sepp Hochreiter
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: ICML 2026 - AdaptFM Workshop
Abstract:Large output embedding matrices create a significant memory bandwidth bottleneck during autoregressive decoding, especially for compact LLMs with large multilingual vocabularies. We reformulate the output projection followed by top-k token selection as a maximum inner product search over token embeddings and replace the dense vocabulary projection with an HNSW-based vector index. The resulting output head retrieves only a small candidate set of high-scoring tokens and can be integrated into existing decoding pipelines by scattering retrieved logits into a sparse full-vocabulary tensor. On CPU inference with Gemma 3, Llama 3.2, and Qwen 3 models, our method substantially accelerates the output projection and improves end-to-end batch-size-one decoding throughput by up to 82% for Gemma 3 270M, while preserving generation quality under AlpacaEval evaluation. These results suggest approximate retrieval is a practical alternative to dense output projections in latency-sensitive small-batch decoding.
[NLP-76] Benchmarking large language model agent societies against human behavioural distributions
【速读】： 该论文旨在解决大规模语言模型代理（Large Language Model Agents）在作为实验社会进行研究时所面临的三大核心疑虑：一是模型代理的行为是否真实反映人类行为，二是研究结果是否对实验装置的微小变化具有鲁棒性（即规则不变但外部设置改变），三是观察到的社会动态是否真正源于交互行为，还是仅是模型对训练数据中已有实验模式的复现。针对这些问题，论文提出并引入了SILICA——一个开放可复现的评估工具，用于系统性检验上述三类问题。其解决方案的关键在于设计了五个包含已发表人类基准数据（human anchors）的环境，每个环境均配备两种变体：一种通过扰动重置相同规则下的情境（如动作顺序调整），另一种则改变收益结构以偏离模型可能记忆的预期结果。在单块消费级显卡上运行12个开源权重模型后发现，仅有8个模型在初始轮次的公共品贡献上与人类数据处于等效区间，而所有模型均无法匹配人类合作的终态或合作走廊。进一步实验表明，仅调整两个动作的呈现顺序就导致某一模型的合作水平下降58点；当给予固定报价序列时，仅一个经过推理训练的模型将接受阈值置于激励要求的位置，其余模型或部分调整、或反向调整、或完全未形成有效阈值。此外，惯例的形成依赖于对行动名称的共享先验而非协商过程，但一旦该先验被破坏，协商行为又会重新出现。基于文中提出的认证阶梯（certification ladder），当前硅基社会仅能支持探索性假设，尚不足以支撑确定性结论。
链接: https://arxiv.org/abs/2608.28182
作者: Raad Bin Tareaf
机构: XU Exponential University of Applied Sciences( XU 未来应用科学大学); Potsdam, Germany(德国波茨坦)
类目: Physics and Society (physics.soc-ph); Computation and Language (cs.CL)
备注:
Abstract:Populations of large language model agents are increasingly used as experimental societies. Three doubts shadow every such result: whether the agents behave like the humans they stand in for, whether a finding survives changes to the apparatus that leave the rules untouched, and whether apparent social dynamics are interaction at all rather than the reproduction of experiments the models have read. This article introduces SILICA, an open instrument that tests all three. Five environments carry published human anchors, each paired with perturbations that re-render the same rules and with variants whose payoffs point away from the memorised result. Twelve open-weight models were run through it on a single consumer graphics card. Agreement with human data is confined to starting points: first-round public-goods contributions fall inside the equivalence margin for eight of eleven models, while no model matches end-state contributions or the human corridor of cooperation. Merely swapping the order in which two actions are listed costs one model 58 points of cooperation. Presenting responders with a fixed schedule of offers shows that only one model, the sole reasoning-trained one, places its acceptance threshold where the incentive requires; two move theirs part of the way, two move them the wrong way, and three never acquire one. Conventions form through a shared prior over the names rather than through negotiation, though negotiation reappears once that prior is disrupted. On the certification ladder defined here, current silicon societies support exploratory claims and no more.
[NLP-77] SURE-Challenge: Evaluating Speech Evidence Before Speech-LLM Generation
【速读】： 该论文旨在解决语音大模型（Speech LLMs）在推理过程中对非支持性输入（如静音、噪声、合成音等）缺乏有效过滤的问题，即在模型生成回答前，系统应具备判断输入是否应被拒绝的前置筛选能力。现有评估方法仅基于模型输出进行评分，忽略了输入质量对结果可靠性的影响，导致潜在的“生成错误模式”未被识别。为此，论文提出语音不支持拒绝评估挑战赛（Speech-Unsupported Rejection Evaluation Challenge, SURE-Challenge），构建了一个包含多种非支持性输入类型（如静音、有色噪声、合成音、源模糊的人声背景）的基准测试集，并采用分离的训练/测试数据划分以确保评估严谨性。其解决方案的关键在于设计并验证一种基于能量特征与Whisper得分组合的固定规则（energy-plus-Whisper-score rule），该规则在前处理阶段实现对无效输入的高效拒止。实验表明，该规则在474条经泄漏筛查的测试样本中可将204个非支持性输入中的196个正确拒绝，而原始模型（Qwen2-Audio）仅能识别其中15个，且不影响支持性输入的生成准确率。外部验证进一步确认了该规则的鲁棒性：随着Whisper得分阈值收紧，Common Voice数据集保留率下降，且不同随机种子下对无速度人声背景的拒绝数量稳定在18至24之间。结果揭示了仅依赖答案评分会遗漏此类前生成阶段的错误模式，凸显了引入输入级拒绝机制的重要性。
链接: https://arxiv.org/abs/2608.27783
作者: Mengzhe Geng
机构: 未知
类目: Audio and Speech Processing (eess.AS); Computation and Language (cs.CL); Sound (cs.SD)
备注:
Abstract:Speech LLMs are usually graded after they answer, although an operating system first has to decide whether a waveform should be sent to the model. We define the Speech-Unsupported Rejection Evaluation Challenge (SURE-Challenge) for this admission step. The benchmark pairs LibriSpeech-derived transcription and first-word question answering with unsupported silence, colored noise, synthetic tones, and source-ambiguous babble under disjoint source splits. Front-end ablations use Qwen2-Audio; the selected energy-plus-Whisper-score rule is then replayed before six speech/audio LLMs. On the 474-row leakage-screened SURE-Extended test set, raw Qwen2-Audio rejects 15/204 unsupported inputs, whereas the fixed rule rejects 196/204 and leaves supported accuracy unchanged. External checks delimit this number: Common Voice retention drops as the Whisper-score threshold is tightened, and no-speed babble gives 18 to 24 rejected clips out of 54 across regenerated seeds. The result identifies a pre-generation error mode missed by answer-only scoring.
信息检索
[IR-0] PULSAR: Pooled Unified Late-Interaction Search and Retrieval for Enterprise Visual Document RAG EMNLP2026
链接: https://arxiv.org/abs/2608.28572
作者: Benjamin Constable,Anup Roy,Vishal Sharma,Rishabh Upadhyay,Robin Mills,Aidan Millar
类目: Information Retrieval (cs.IR)
备注: Accepted at EMNLP 2026 (Industry Track)
Abstract:Institutional investors search visually dense pitch decks, board packs, and diligence materials that change hourly near deal closing. OCR followed by figure verbalisation is costly to refresh at this scale and can lose chart detail. We present PULSAR, a production vision-first retrieval system deployed at Mubadala Investment Company. PULSAR indexes page images with a frozen ColPali-style backbone and uses a pooled two-stage late-interaction index: compact page summaries support initial retrieval, followed by exact MaxSim rescoring over a finer pooled representation. On ViDoRe V3, this design reduces median vector-search latency by 15.1 times against an unpooled configuration with less than 0.01 absolute NDCG@10 and Recall@10 loss; production median vector-search latency is 156 ms. Under concurrent load, the pooled index sustains approximately 88 times higher QPS than an unpooled index. The event-driven ingestion path is estimated to be approximately 20 times cheaper per page than the OCR+verbalisation baseline it replaced. Since March 2026, PULSAR has served 78 thousand documents and approximately 2.4 million pages across more than 3,000 deals. At the production top K, it more than doubles answer-fact recall over the OCR+verbalisation baseline.
[IR-1] QUEST: A Query and Extraction System for Topics in Asylum Law Application Decisions
链接: https://arxiv.org/abs/2608.28555
作者: Maria Vlachou,Anna Murphy Høgenhaug,Mohammad N. S. Jahromi,Galadrielle Humblot-Renaux,Thomas Gammeltoft-Hansen,Thomas B. Moeslund,Desmond Elliott
类目: Information Retrieval (cs.IR)
备注:
Abstract:Legal decisions on asylum applications consist of long, complex, and heterogeneous documents, covering narrative applicant interviews, original decisions, and additional supporting materials. If an application is rejected, a critical question in processing an appeal is whether the credibility of the information in the original application was a factor that determined the original decision. In this paper, we present the QUEST system (Query and Extraction System for Topics) to extract and identify factors relating to credibility assessments in two datasets of Danish asylum application appeals. QUEST frames this problem as an information retrieval task, combining synthetic query generation, topic extraction, and relevance assessment to identify information related to credibility indicators in appeals board application materials. In addition to standard retrieval evaluation metrics, we propose a new type of domain-specific assessments distinct from the traditional relevance to evaluate the performance of the tested systems with respect to credibility factors. In this way, we obtain insights about how well automatic methods can return answers for different types of indicators appearing in asylum appeals. Our results indicate that there is an increased challenge when estimating performance using credibility-based relevance assessments, thus pointing to the difficulty of the task.
[IR-2] SG-UMP: Sequence-Guided Universal Multimodal Prioritization Calculation Framework
链接: https://arxiv.org/abs/2608.28503
作者: Xinyi Zhang,Yutong Li,Peijie Sun
类目: Information Retrieval (cs.IR)
备注: Accepted as a Full Paper at MM 2026
Abstract:Multimodal sequential recommendation (MSR) improves recommendation by incorporating heterogeneous information such as text, images, and user interactions. However, existing MSR methods often fail to capture user-level preference heterogeneity and dataset-level modality bias, limiting their adaptability across users and datasets. To address this issue, we propose \textbfSequence-\textbfGuided \textbfUniversal \textbfMultimodal \textbfPrioritization Calculation Framework (\textbfSG-UMP), a plug-and-play plugin for enhancing multimodal information processing in MSR. SG-UMP includes a Module Combiner for flexible multimodal processing and a Module Router for dynamic module ordering, enabling adaptation to both user preferences and dataset characteristics. Experiments on four real-world datasets show that SG-UMP consistently improves recommendation performance across different backbones and multimodal settings. The code is available at this https URL .
[IR-3] Every Article Deserves a Video: Contextual Video Matching for Digital Publishers
链接: https://arxiv.org/abs/2608.28359
作者: Arnaud Corone,Brice Pierre de la Briere,Gladys Roch,Samuel Leonardo Gracio,Yassine Bouher,Parvati Chauchaix
类目: Information Retrieval (cs.IR)
备注:
Abstract:As digital publishers face the challenge of managing massive content catalogs, the ability to effectively embed relevant video within text-based articles has become critical for both monetization and user retention. However, manual selection is impractical for large scale publishers, especially when navigating their own extensive video libraries or the entire global Dailymotion catalog. In this paper, we present the “Contextual Video Matching” system, a solution that automatically matches relevant videos with text-heavy web pages and articles. By leveraging Large Language Models (LLMs) and textual embeddings, we provide a scalable solution for publishers to efficiently combine video content with their articles. We discuss in detail the motivations, architecture, evaluations, and deployment of this system within Dailymotion’s production environment. Since its launch, the system has been adopted by hundreds of publishers, significantly increasing user engagement and enriching user experiences with highly relevant video content.
[IR-4] Nested Byte-Level Vocabularies Are Cheap to Deploy and Expensive to Share: A Pre-Registered Negative Result
链接: https://arxiv.org/abs/2608.28151
作者: Christos Koutsiaris
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
备注: 5 pages, 2 figures, 4 tables. Pre-registered study. Code and reproducibility materials: this https URL
Abstract:A byte-level BPE tokenizer is an ordered list of merge rules, so applying only a prefix yields a vocabulary whose token identifiers are the first rows of the full vocabulary. This prefix nesting allows one language model to operate at several vocabulary sizes, use a control token to indicate the active size, and be deployed at any trained size by slicing its embedding and output head. We pre-registered five claims, including margins, seeds, contrasts, and a stop rule, and trained 30 models with 3.1M- and 10.6M-parameter bodies on 200M tokens each. Slicing is numerically exact: across 76 checks, a sliced model reproduces the restricted full model’s logits bit for bit and removes 66% of deployed weights without changing latency. However, the shared model trails a fixed-cap specialist by 3.64% bits per byte at 32k against a 1% margin, and by 2.96% at 8k against a 2% margin. A 2x2 ablation separating the control token from output restriction finds that the token changes performance by +0.07% to +0.13%, with all intervals crossing zero, while output restriction costs +0.47% to +1.19%; the factors are substitutes rather than complements. Multi-cap training nevertheless improves robustness: under typographical noise, the same checkpoint degrades 12.5–15.4 points less in its fine mode and outperforms each fixed-cap specialist at that specialist’s vocabulary size. A control with neither cap token nor output restriction is equally robust, attributing this benefit to multi-granularity training rather than conditioning. The per-cap penalty tracks each cap’s share of training rows, yielding a falsifiable prediction for future work.
[IR-5] HubMixer: Progressive Latent Hub Mixing for Parameter-Efficient Feature Interaction in Recommendation
链接: https://arxiv.org/abs/2608.27991
作者: Jie Zhou,Zixian Gong,Wenhao Li,Chang Liu,Enzhao Shen,Bo Liu,Xu Guo,Fei Pan,Peng Jiang
类目: Information Retrieval (cs.IR)
备注:
Abstract:Learning effective feature interactions is central to industrial recommendation and advertising ranking systems. Recent token-mixing architectures simplify self-attention with lightweight mixing operators, improving hardware efficiency and enabling large-scale deployment. However, recommendation tokens are fundamentally heterogeneous: user profiles, item attributes, behavioral sequences, context features, statistical signals, and business-side features live in different semantic spaces and interact in sparse, sample-specific patterns. Directly mixing all tokens in the raw heterogeneous token space may therefore be parameter-inefficient, as the model must implicitly discover which feature groups should interact and how such interactions should be routed. In the paper, we propose HubMixer, a parameter-efficient latent hub mixing architecture for feature interaction in recommendation. Instead of directly mixing raw feature tokens, HubMixer introduces a small set of learnable latent hubs to organize feature interactions through an induction--interaction--readout paradigm. First, hub induction summarizes heterogeneous tokens into compact latent hubs, where latent hubs query input tokens through cross-attention. Second, hub interaction performs high-order interaction in the cleaner latent hub space. Third, token-conditioned readout lets each original token selectively read from the interacted hubs, injecting global interaction semantics while preserving token-level field identity. Extensive offline experiments on industrial recommendation tasks show that HubMixer outperforms the SOTA models. Online A/B testing in the Kuaishou short-video recruitment business further shows a statistically significant 5.48% improvement in resume submission conversion rate, and HubMixer has been fully deployed in production.
[IR-6] Information-Guided Selective Modality-Interest Alignment for Multimodal Recommendation CIKM2026
链接: https://arxiv.org/abs/2608.27950
作者: Wenze Ma,Chenyu Sun,Yanmin Zhu,Qiwen Gu,Xuhao Zhao
类目: Information Retrieval (cs.IR)
备注: Accepted at the 35th ACM International Conference on Information and Knowledge Management (CIKM 2026)
Abstract:Multimodal recommendation (MMRec) aims to enhance recommendation performance by leveraging rich item content from multiple modalities. However, directly incorporating all modality information does not necessarily lead to better preference modeling, since user interests are often more related to a subset of modality signals, while other signals may be weakly aligned with user preferences or even introduce noise. Although recent MMRec methods improve modality utilization through invariant learning, attention mechanisms, graph refinement, or contrastive learning, their alignment processes are often implicit or heuristic and lack a clear objective for selecting modality signals that better match user interests. In this paper, we propose AMUR, an information-guided selective modality-interest alignment framework for multimodal recommendation. Inspired by an information-theoretic view, AMUR aims to enhance modality information that is more related to user interests while reducing the influence of less aligned signals. Specifically, AMUR first refines modality graph structures towards user behavior, and then selectively aligns shared interest-related semantics across modalities. This enables AMUR to improve modality-interest alignment while preserving useful modality-specific complementary information. Extensive experiments on three real-world datasets demonstrate the effectiveness of AMUR over competitive baselines. The code is available at this https URL. Comments: Accepted at the 35th ACM International Conference on Information and Knowledge Management (CIKM 2026) Subjects: Information Retrieval (cs.IR) Cite as: arXiv:2608.27950 [cs.IR] (or arXiv:2608.27950v1 [cs.IR] for this version) https://doi.org/10.48550/arXiv.2608.27950 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[IR-7] ITER: Interaction-Aware Retrieval for Agent ic Search
链接: https://arxiv.org/abs/2608.27912
作者: Haodong Chen,Shuai Wang,Yu Yin,Shengyao Zhuang,Guido Zuccon,Teerapong Leelanupab
类目: Information Retrieval (cs.IR)
备注:
Abstract:Deep-research agents answer complex user questions through an iterative sequence of search steps, where the agent autonomously formulates sub-queries to retrieve the evidence needed at each stage. However, existing retriever training typically relies only on the sub-query and its corresponding search results at the current step as training signals, leaving the information accumulated from previous interactions largely underutilized. We introduce iter, an agent interaction-aware dense retriever trained using agent trajectory learning signals. iter represents each query by incorporating not only the current sub-query, but also the main question and preceding sub-queries, and is trained using trajectory-relative learning signals derived from the agent’s interactions. Across six agent backbones from three model families, iter consistently outperforms the existing agent-trajectory-trained dense retriever, LRAT, achieving an average improvement of 7.5% on InfoSeek-Eval and 13.5% on BrowseComp-Plus. iter also demonstrates stronger cross-agent robustness than AgentIR, a deep-research retriever that relies on external LLM-judge signals and the agent’s pre-search reasoning. Ablations further show that the main question and previous sub-queries provide the most robust query representation, while previously visited and useful documents, used as redundancy negatives in subsequent searches, provide the strongest trajectory-relative supervision. Code is available at this https URL. Subjects: Information Retrieval (cs.IR) Cite as: arXiv:2608.27912 [cs.IR] (or arXiv:2608.27912v1 [cs.IR] for this version) https://doi.org/10.48550/arXiv.2608.27912 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[IR-8] An Empirical Evaluation of Cross-City POI Recommendation on a Large-Scale Benchmark
链接: https://arxiv.org/abs/2608.27840
作者: Peibo Li,Yang Song,Hao Xue,Maarten de Rijke,Flora D. Salim
类目: Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
备注:
Abstract:Cross-city point-of-interest (POI) recommendation is crucial for navigating unfamiliar urban environments, yet its progress has historically been constrained by data limitations. Using the recently proposed large-scale benchmark Trip World, we empirically re-examine whether conclusions drawn on small prior benchmarks still hold under worldwide coverage, low home-destination region overlap, and large, semantically rich POI inventories. Our evaluation surfaces three bottlenecks of representative state-of-the-art methods: (1) hometown-aware models appear to rely more on destination-region priors than on user-specific preference transfer; (2) their accuracy-efficiency trade-off degrades at this scale, where the simplest model is among the strongest; and (3) existing mechanisms for integrating semantic metadata yield little benefit. We further include a diagnostic pilot on agentic methods adapted from next-POI recommendation, finding that naive adaptation trails a simple popularity prior even though the relevant semantic signal is present in the data. These results highlight the need for task-specific designs that support cross-city preference transfer, semantic grounding, and scalable reasoning over unseen destination inventories.
[IR-9] Personalized and Multi-View Representation for Federated Cold-Start Recommendation
链接: https://arxiv.org/abs/2608.27826
作者: Jaehyung Lim,Wonbin Kweon,Woojoo Kim,Junyoung Kim,Dongha Kim,Hwanjo Yu
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注:
Abstract:Federated recommendation (FedRec) enables personalized modeling without centralizing users’ interaction histories, but most existing methods assume a fixed item pool and thus overlook the practical cold-item setting where new items continuously arrive. Under the dual-sided constraint, where the server cannot access clients’ interactions while clients cannot access the server’s proprietary item attribute features, prior federated cold-start recommendation approaches suffer from three structural limitations: a lack of personalization, compositionality failure caused by forcing heterogeneous semantics into a single embedding space, and training- and communication-inefficiency arising from explicit alignment between separate collaborative and attribute representations. To address these challenges, we propose Personalized and Multi-view Representation for Federated Cold-Start Recommendation (PMFRec). PMFRec learns a personalized representation generator to produce user-specific item representations from attribute features, and introduces a global multi-view encoder with item-adaptive gating and an orthogonality objective to capture complementary semantic views while reducing cross-view redundancy. In addition, PMFRec fuses collaborative and attribute knowledge into a single exchanged item representation, eliminating the need for an explicit client-side regularizer and reducing communication overhead. Extensive experiments on real-world datasets show that PMFRec consistently outperforms strong baselines in cold-item recommendation and further improves user-level fairness, warm-scenario adaptability, and robustness under Local Differential Privacy (LDP).
[IR-10] LINE Conversation History Retrieval for Personal Memory RAG : Evaluating Search Representations and Hybrid Retrieval
链接: https://arxiv.org/abs/2608.27809
作者: Akito Hattori
类目: Information Retrieval (cs.IR)
备注: 16 pages, 6 figures, 10 tables. Exploratory single-user case study
Abstract:As an initial step toward personal memory retrieval-augmented generation (RAG) for large language models (LLMs), this study presents a retrieval-only case study over one user’s LINE conversation history. We segmented 358,896 messages into 22,329 temporally coherent chunks and constructed three search representations: raw_text, a generated summary, and embedding_text, which combines a summary with a raw-text excerpt and other fixed text. We compared BM25, dense vector retrieval, and linear hybrid retrieval on 100 evaluation questions verified by a single annotator. Among individual retrievers, embedding_text_bm25 achieved the highest point estimate, with Recall@5 of 0.584. We then explored six retriever pairings and 21 weights, for 126 configurations on the same evaluation set. The selected combination of embedding_text_bm25 and embedding_text_vector at beta = 0.45 achieved Recall@5 = 0.697, MRR@5 = 0.595, and nDCG@5 = 0.575. Its Recall@5 exceeded that of embedding_text_bm25 by 0.113, with a question-level paired percentile-bootstrap 95% confidence interval of [0.048, 0.184]. This interval is conditional on fixing the configuration selected on the same 100 questions and does not account for uncertainty from configuration selection or weight search. The difference from a summary-based hybrid at beta = 0.50 was 0.050, with a 95% confidence interval of [-0.013, 0.115], so no clear difference could be established. The 17 aggregate questions also yielded lower point estimates than the other question types, suggesting that flat chunk-level retrieval struggles when evidence is distributed across multiple times and conversations. This evaluation is an exploratory single-user, single-annotator study conducted on the same question set used for configuration search; it does not evaluate final answer generation or generalization to unseen questions.
[IR-11] NormasTCU — A Brazilian Portuguese IR Dataset and an Evaluation of LLM -as-a-Judge for Relevance Assessment
链接: https://arxiv.org/abs/2608.27746
作者: Leandro Carísio Fernandes,Marcus Vinícius Borela de Castro,Leandro dos Santos Ribeiro,Leonardo Augusto da Silva Pacheco,Edans Flávius de Oliveira Sandes
类目: Information Retrieval (cs.IR)
备注:
Abstract:Portuguese Information Retrieval (IR) lacks public datasets, and relevance assessment for specialized collections remains costly. While Large Language Models (LLMs) increasingly support relevance assessment, their reliability in non-English specialized domains remains unclear. We introduce NormasTCU (this https URL), a Brazilian Portuguese IR dataset with 14,469 legal documents, 46 queries, and 3,048 human judgments over 812 query-document pairs. Using NormasTCU, we evaluated LLM-as-a-judge for relevance assessment by prompting three models with two prompt techniques to grade these pairs. We then compared the rankings of 15 IR systems derived from LLM-generated and human reference qrels. LLMs consistently showed a positive scoring bias (mean absolute error: 0.46–0.66 on a 0-2 scale). Furthermore, pair-level agreement with human judgments achieved only fair to moderate levels, with Cohen’s kappa ranging from 0.32 to 0.53. Despite this bias, LLM-generated judgments often yielded highly similar system rankings for nDCG@10 and MRR (observed Kendall’s tau greater than or equal 0.90, although the bootstrap confidence intervals did not always remain above this threshold), but were less reliable for P@10 and R@10. Notably, LLM-based rankings were sometimes more strongly correlated with the reference ranking than individual human annotations were. As a practical implication, our results suggest that LLMs could effectively support scalable relevance assessment in specialized Portuguese corpora when evaluated using nDCG or MRR (rank-aware metrics), but they should be avoided when relying on precision or recall.
[IR-12] A Versioned Unified Graph Index for Dynamic Timestamp-Aware Nearest Neighbor Search
链接: https://arxiv.org/abs/2608.27663
作者: Jun Woo Chung,Weijie Zhao
类目: Information Retrieval (cs.IR)
备注: 15 pages, 8 figures, 3 algorithms
Abstract:We present TiGER (Time-Integrated Graph for Efficient Retrieval), a novel approach for performing fast time-aware approximate nearest neighbor searches on dynamic vector datasets with flexibility over any possible time range. Our proposed algorithm builds and maintains a unified graph for all vectors by leveraging an index structure based on integrated versioned connectivity, allowing arbitrary time intervals to be queried directly on the unified graph without having to traverse invalid vectors. This forgoes the need for post-search filtering or merging, or separate graphs for each possible composite range. Empirical evaluations show that our method attains up to a 5x improvement in queries per second (QPS) without compromising accuracy over baselines based on filtering or per-time-segment sub-graphs. We believe that this method will enable efficient temporal analysis across evolving datasets in real-time recommendation systems, log analysis, and any scenario requiring fast similarity search over dynamic, time-segmented data.
[IR-13] Beyond the Vacuum: Combinatorial Strategy Selection for Competitor-Aware Generative Engine Optimization
链接: https://arxiv.org/abs/2608.27631
作者: Vaibhav Sourirajan,Yao Zhang,Himanshu Kumar,Sahil Wadhwa,Mann Patel,Amirfarrokh Iranitalab
类目: Information Retrieval (cs.IR)
备注: 20 pages, 2 figures
Abstract:Generative Engine Optimization (GEO) has emerged as a novel paradigm for transforming content to increase visibility in Large Language Model (LLM) responses. Traditional GEO methods, however, select rewriting strategies in isolation, ignoring a critical externality: as adoption of content optimization grows, optimal strategies for rewriting content change. We formalize GEO as a competitor-aware strategy selection problem and propose a two-phase pipeline to solve it: (1) We use Bayesian Optimization of Combinatorial Structures (BOCS) to efficiently search the space of rewriting strategies, (2) We generate preference pairs and grounded reasoning traces from the BOCS black-box observations to fine-tune a language model to analyze a document corpus and propose optimal rewriting strategy combinations. We achieve state-of-the-art performance across several impression metrics over existing agentic and single-heuristic methods on both geo-bench and our synthetically augmented competitive dataset geo-bench_comp. Our method also transfers to multiple out-of-distribution datasets, proving effective across domains, queries, and document types.
[IR-14] LitCurate: A Configuration-Driven AI-Assisted Framework for Scientific Database Construction with an Application to Lower-Mantle Equation-of-State Data
链接: https://arxiv.org/abs/2608.27629
作者: Abin Shakya,Wilson Samuels,Dominica Wilson,Gioia A. Marchi,Israa Draz,Chenxing Luo,Renata M. Wentzcovitch
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Geophysics (physics.geo-ph)
备注: 26 pages, 7 figures
Abstract:The growing scientific literature contains decades of experimental and computational results that could support data-driven and physics-based modeling, yet much of this infor- mation remains locked in publications and is not readily usable for large-scale analysis or sci- entific software. Building structured databases from the literature is particularly challenging whenrelevantstudiesmustfirstbediscoveredamonglargecollectionsofpapersandreported quantities must be extracted with enough scientific context to remain usable. We present LitCurate, an open-source framework for building scientific databases from the literature using large language models within an auditable, stage-wise curation workflow. LitCurate integratesliteraturediscovery, relevancescreening, full-textprocessing, andstructuredinfor- mation extraction while retaining intermediate results and provenance, allowing researchers to inspect and revise individual stages rather than treating automated curation as a black- box process. We apply LitCurate to construct an equation-of-state database of lower-mantle and lower-mantle-relevant high-pressure mineral phases from experimental and theoretical studies, comprising 1,334 entries from 205 papers. The resulting dataset links reported equation-of-state parameters to mineral phases, compositions, equation formulations, meth- ods, and parameter constraints, and labels values as source-reported or citation-reported when provenance can be determined. The records are available through a searchable web application. By connecting scientific literature to traceable, machine-readable data, LitCu- rate provides a reusable approach for transforming accumulated literature into resources for scientific analysis and computational modeling.
人机交互
[HC-0] Between Algorithm (AI) and Intuition (Human): Preserving Designer Agency in AI-Assisted Sensemaking of Qualitative UX Data
链接: https://arxiv.org/abs/2608.28420
作者: Md Haseen Akhtar
类目: Human-Computer Interaction (cs.HC); Emerging Technologies (cs.ET)
备注: Accepted as a workshop paper in Proceedings of (CHI 26 Workshop on Sensemaking and AI 2026: Uses, Behaviors, Design, and Recommendations, April 13th, 2026, Barcelona, Spain) ACM, New York, NY, USA, 5 pages. Workshop website: this https URL
Abstract:The integration of AI into qualitative design research presents a fundamental tension: how do we leverage AI while preserving the subjective, intuitive judgments that define design expertise? This paper examines this question through a case study of analyzing 20 user responses about video conferencing platforms for educational contexts. We argue that AI sensemaking tools risk flattening the rich data patterns, amplifying contradictory textures of user feedback into sterile categories thereby transforming design research from an interpretive craft into a mechanical sorting exercise (rigid and formal). Through comparative analysis of AI-assisted sensemaking versus human-centered approaches to the same dataset, we identify when algorithmic efficiency enhances understanding and when it diminishes the designer’s interpretive agency (uncovering hidden needs, critical enquiry, what if enquiries, making decisions, having trade-offs). We present a framework for augmented sensemaking that positions AI as an instrument for amplifying human judgment rather than replacing it. Our findings suggest that the most valuable role for AI in design research is not to eliminate subjectivity, but to make it more intentional, reflective, and accountable.
[HC-1] AI as Teammate: Rethinking Task Distribution in Medical Training
链接: https://arxiv.org/abs/2608.28373
作者: Fendi Tsim,Alina Gutoreva,Anthony Weiss,Nicole Dubosh
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注:
Abstract:Integrating Artificial Intelligence (AI), particularly generative AI, into medical training has prompted concerns about learner over-reliance, misuse, and erosion of foundational clinical competencies. We propose a conceptual reframing at the decision level: the problem is not misuse but misclassification - a mechanistic failure of real-time metacognitive evaluation in selecting a subzone-inappropriate AI interaction mode. Drawing on “SCAN” (Substitute, Complement, Aid, Non-Negotiable), a human-centric decision-making framework for generative AI task allocation grounded in Vygotsky’s Zone of Proximal Development and metacognition, we advance the emerging social-constructivist conversation around AI in medical education by offering a testable account of AI’s role in clinical reasoning development. This framework yields testable predictions for how misclassification can be detected, mitigated, and, more importantly, prevented in the clinical learning environment. Regarding clinical reasoning development, we show how trajectories of skill acquisition (upskilling) and failure (the triad of skill failure: de-skilling, never-skilling, and mis-skilling) operate at the individual task level in ways that fixed-phase, cohort-wide treatments fail to capture. We further identify passive engagement within correctly classified AI-scaffolded tasks as a particularly insidious, detection-resistant pathway to mis-skilling - one requiring subzone re-identification from AI assistance to expert assistance, with human experts serving as epistemic auditors. The paper operationalizes SCAN for clinical curriculum design, supervision, and assessment, and opens an empirical research agenda grounded in cognitive science. This paradigm shift from misuse to misclassification is not semantic: it offers educators a clear perspective on what to look for, what to assess, and what to intervene on.
[HC-2] It Takes Three to Converse: Empirical Observations on How the Developer the Convener and the Participant Shaped 119 Polis Conversations
链接: https://arxiv.org/abs/2608.28368
作者: Lodewijk Gelauff
类目: Computers and Society (cs.CY); Human-Computer Interaction (cs.HC)
备注:
Abstract:Polis is a popular democratic innovation tool that allows asynchronous citizen engagement through atomic statements: short statements that together describe a complex question, inviting the citizen to vote Agree or Disagree on each. This paper uses 119 conversations with 100 or more participants and an extensive data export, drawn from a wider set of 271 collected processes. The paper asks what determines the output of such a process. Three parties shape the result. The developer of the platform has made important design choices that restrict the outcome: the number of groups the platform is able to report (restricted to 2–5) and which statements are prioritized. The convener defines the assignment: the initial statements that set the tone, the policy that accepts or rejects new statements and who can be invited. Finally, the participant works within these boundaries. With access to less than half of the generated statements, they end up responding to more statements when their conversation seems to have an achievable number of statements to complete, than when they are presented with more statements. Due to choices such as warm path clustering, the exported resulting clustering cannot be reproduced based on the voting data. Conveners may want to re-analyse their own conversations once the process is closed, to consider the data in its entirety, and make their own analysis priorities explicit. Subjects: Computers and Society (cs.CY); Human-Computer Interaction (cs.HC) Cite as: arXiv:2608.28368 [cs.CY] (or arXiv:2608.28368v1 [cs.CY] for this version) https://doi.org/10.48550/arXiv.2608.28368 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Lodewijk Gelauff [view email] [v1] Fri, 28 Aug 2026 14:24:34 UTC (232 KB)
[HC-3] Generative AI Alignment with Hinduisms Theological Plurality and Sacred Representation
链接: https://arxiv.org/abs/2608.28228
作者: Dipto Das,Arpita Kundu,Nusrat Jahan Mim,Shion Guha,Syed Ishtiaque Ahmed
类目: Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Human-Computer Interaction (cs.HC)
备注:
Abstract:Generative AI systems are increasingly used to answer personal questions and mediate everyday practices, including religion. However, existing discussions around AI alignment and ethics have largely centered secular, Western, and Abrahamic assumptions about religion, offering limited attention to other faith-based traditions. In this paper, we examine how Hindu users engage with generative AI systems in relation to their religious knowledge, belief, and practice. Drawing on 15 semi-structured interviews with Bangladeshi Hindu participants, we analyze how users interpret AI-generated religious representations, scriptural explanations, devotional interactions, and synthetic religious media. We found that AI can be both accessible and ethically troubling. While AI supported scriptural inquiry, devotional visualization, and religious storytelling, our study also identified concerns about theological flattening, cultural misrepresentation, devotional manipulation, and the simulation of sacred presence and authority. We conclude by arguing that religious alignment in generative AI requires interpretive alignment: systems that disclose their limits, preserve plurality, and avoid simulating sacred authority and sycophantic personalization.
[HC-4] CrabOS: An Operating System for Human-AI Co-inhabitation
链接: https://arxiv.org/abs/2608.28165
作者: Qi Yang,Yun Ma
类目: Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC); Operating Systems (cs.OS)
备注:
Abstract:AI agents are evolving into long-running computational entities that can invoke tools, maintain memory, and complete complex tasks across applications. In real-world settings, completing a task often requires humans and AI to take turns leading its execution. Such alternation depends on the seamless handoff of the work state of the task between humans and AI. Existing agent systems, however, provide humans and AI with separate work environments. AI agents must therefore rely on additional bridges to continue work: either developers build task-specific interfaces to access the work state, or users manually transfer relevant parts of it through screenshots or textual descriptions. Both approaches make handoffs costly and scale poorly. We propose Human-AI Co-inhabitation, a type of work environment that enables humans and AI to seamlessly take turns continuing work on the same task, and design and implement CrabOS to realize this concept. CrabOS represents the work state as natural-language-readable text objects shared by humans and AI, allowing both to access and manipulate it directly through the same auditable interface without bridges. Case studies show that CrabOS elevates support for complex tasks with alternating human and AI leadership from bridge-dependent application-level solutions to native operating-system capabilities, which provide a new foundation for developing and running AI agents. Subjects: Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC); Operating Systems (cs.OS) Cite as: arXiv:2608.28165 [cs.AI] (or arXiv:2608.28165v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.28165 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[HC-5] Gen-TAS: A Generative AI-Aided Hardware-Software Task Allocation Framework for FPGA-GPP Heterogeneous Systems
链接: https://arxiv.org/abs/2608.28160
作者: Mary Kong,Yuqin Zhao,Semih Vazgecen,Cristian Sestito,Themis Prodromakis
类目: Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注:
Abstract:FPGA-GPP heterogeneous systems combine software flexibility with the performance and energy efficiency of reconfigurable hardware. However, determining which application tasks should execute on the GPP or FPGA requires extensive expertise and design-space exploration, particularly when user objectives vary across latency, communication, resource utilisation, and power. This paper proposes Gen-TAS, a knowledge-grounded LLM framework for user-specific FPGA-GPP task allocation. By combining task-graph analysis with RAG, Gen-TAS grounds LLM reasoning in historical implementation knowledge and generates multiple explainable strategies tailored to the specified objectives. Human-in-the-loop selection and a deterministic backend connect LLM-generated decisions to reproducible FPGA SoC implementations. Experiments on CNN and SDR workloads across multiple LLMs demonstrate stable, requirement-driven allocation. Under latency-oriented objectives, implementations following the selected strategies achieve speedups of up to 2.45 \times and 92.53 \times , respectively, relative to the corresponding all-GPP baselines while other objectives select strategies that trade some acceleration performance for FPGA-GPP communication, resource utilisation, or FPGA power.
[HC-6] User Preferences for UI Anchoring in MR: Effects of Task Mobility and Interface Properties
链接: https://arxiv.org/abs/2608.28064
作者: João Belo,Sina Elahimanesh,Anna Maria Feit
类目: Human-Computer Interaction (cs.HC)
备注: Accepted to IEEE International Symposium on Mixed and Augmented Reality 2026
Abstract:Anchoring - the choice of frame of reference for mixed reality (MR) interface elements - is a critical design decision involving trade-offs between accessibility, interaction comfort, and visual interference. Despite its importance, user preferences for anchoring across different mobility contexts and interface properties remain poorly understood, as prior work has largely focused on specific tasks or fixed interface configurations. We address this through a mixed-methods user study in which participants configure anchoring strategies across different mobility conditions and interface types. Combining behavioral analysis with structured qualitative inquiry, we analyze how participants select and reason about anchoring modes. Our results show a clear transition from world-anchored interfaces in stationary contexts to body-anchored interfaces during locomotion. However, no single body anchor consistently dominates, highlighting the personal nature of anchoring strategies. Our qualitative analysis reveals the factors users consider in their anchoring decision, including interface accessibility, stability during interaction, visual clutter, and individual mental models. These findings inform the design of adaptive and controllable MR interfaces and highlight the importance of supporting user customization.
[HC-7] oo Much of the Same: From Algorithmic to Human Bias in Learning to Defer
链接: https://arxiv.org/abs/2608.28050
作者: Dario Pesenti,Alessandro Bogani,Stefano Teso,Andrea Pugnana
类目: Human-Computer Interaction (cs.HC)
备注: 10 pages, 19 pages with Supplementary Material, 7 images
Abstract:Learning to Defer (LtD) extends supervised learning by allowing a Machine Learning (ML) model to defer harder or less confident decisions to a human expert. Despite being geared for human-AI collaboration, LtD strategies neglect the potential negative interference of human cognitive biases. Our contribution is twofold. First, we demonstrate that standard LtD strategies show class-dependent sampling bias in classification tasks in practice, and thus may disproportionately defer the minority classes when applied to imbalanced datasets. Second, we show that such asymmetries in task delegation may trigger human biases, ultimately leading to poorer downstream decision making. Specifically, we conduct a user study ( N=226 ) where participants complete a classification task on a set of deferred items, with conditions presenting different levels of class imbalance. Our results show that participants exposed to a highly imbalanced rejection set achieved lower classification accuracy in the majority class compared to those exposed to a more balanced set, regardless of which class constituted the majority. Exploratory analyses suggest that this may be an instance of the Test-taker’s effect, which stems from a mismatch between the actual distribution of classes and the participants’ expectations about that distribution. Finally, we discuss the implications of these findings for the deployment of LtD algorithms.
[HC-8] Graphionale: How Graph Visualizations of LLM Rationales Affect Human Decision Making
链接: https://arxiv.org/abs/2608.27932
作者: Xinru Wang,Zhexuan Ma,Ming Yin,Shuai Ma,Thomas W Malone
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:Large Language Models (LLMs) are increasingly equipped with augmented reasoning capabilities to generate rationales that support human decision-making. Yet these text-dense rationales often impose substantial cognitive burdens. Building on a formative co-design study that identified user preferences for non-linear reasoning representations, we developed Graphionale as a testbed for empirically studying argument-map-style rationale visualization. This system transforms linear LLM rationales into interactive, multi-level graphs. It explicitly structures logical relationships (e.g., conclusions, premises, support, and objections), while further extracting entities and relations within each statement to construct condensed node-link representations. We conduct a large-scale online user study (N = 204) to examine when graphical rationales are more effective than textual ones, across varying task modality (verbal vs. visual reasoning), rationale format (textual vs. graphical), and question difficulty (easy vs. hard). Our results show that graphical rationales do not help uniformly: they improve trust calibration for verbal reasoning yet feel more cognitively demanding and less satisfying; for visual reasoning, they impair calibration yet feel more engaging and helpful. In each modality, the format that better supports calibrated decisions is not the one users prefer, highlighting that matching rationale format to task modality is key to effective AI explanation design. Our findings contribute empirical design knowledge about when and how graphical rationales support human decision making, and inform the next-generation reasoning-aware AI interfaces.
[HC-9] Guidelines Are Not Rules: Characterizing Terminologies around Visualization Design Guidelines IEEE-VIS2026
链接: https://arxiv.org/abs/2608.27842
作者: Anna L. Chinni,Md Dilshadur Rahman,Bon Adriel Aseniero,Petra Isenberg,Kushin Mukherjee,Ghulam Jilani Quadri,Paul Rosen,Karen B. Schloss,Daniel Weiskopf
类目: Human-Computer Interaction (cs.HC)
备注: Accepted at BELIV @ IEEE VIS 2026
Abstract:A common expectation in visualization research is that outcomes recommend how researchers and practitioners take action or make design decisions. We often express these as “guidelines.” Yet, the term “guideline” is both ambiguous and loosely defined, and what one researcher considers a guideline may be too broad, too loose, or too strict for another. We take a closer look at a broader set of terms that can express desirable results around visualization research, and untangle how these words are understood in the community in relation to other similar terms. We base our work on an exploratory study with experts, followed by a crowdsourcing study with a separate mapping phase (n=30) and rating phase (n=42) targeting input from the broader visualization community, and an analysis of the use of terminology in 3,877 IEEE VIS papers published from 1990 to 2024. Based on our findings, we call for more nuanced, precise discussions of research outcomes and their communication to the broader community, including practitioners and students.
[HC-10] How Much Can AI Understand? Toward AI-Assisted Sensemaking of Collaborative Discussion in Groups with Shared History
链接: https://arxiv.org/abs/2608.27799
作者: Soobin Cho,Mark Zachry,David W. McDonald
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注:
Abstract:AI tools that support collaborative discussion typically treat the discussion as a standalone task, focusing only on its content and setting aside the social context of the group having it. But it is groups with a shared history, with their own norms, hierarchies, and relationships, where the most tangled and complex discussions tend to arise. These discussions cannot be understood apart from that context, and AI that overlooks it risks failing to convey what a discussion means, or even misrepresenting it. Drawing on two studies of how experienced Wikipedia editors read and make sense of discussions, we propose an AI-Assisted Sensemaking Model for Collaborative Discussions, which captures not only a discussion’s arguments but also the norms and participants behind it, along with the context that gives each meaning. In this model, the system supports the early stages of the sensemaking process, and the degree to which it performs interpretive work can range from low to high. We argue that higher interpretive work reduces the burden on users but increases their reliance on the system’s judgment. We then discuss the risks of an insufficiently intelligible system, what it would take to make one more intelligible, and the safeguards it still requires.
[HC-11] Deceptive Patterns as a Sociotechnical Phenomenon: Review Catalog and Discussion
链接: https://arxiv.org/abs/2608.27684
作者: Luiz Adolpho Baroni,Alisson Puska,Roberto Pereira
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:Background: Deceptive patterns are interface design strategies aimed at misleading users or favoring specific interests, compromising user experiences and ethical privacy principles. These patterns involve exploit different issues of the interaction between humans and technology, revealing gaps in the scientific literature regarding their understanding and working mechanisms. Purpose: This research characterizes deceptive patterns as a sociotechnical phenomenon, integrating human and technical dimensions while offering a self-explainable interactive catalog to raise awareness among users and designers. Methods: To characterize deceptive patterns as a sociotechnical phenomenon, a systematic literature review in Computer Science was conducted, and the Semiotic Framework was applied to analyze and organize the sociotechnical aspects of identified patterns in an integrated manner. An exploratory evaluation of the interactive catalog was conducted, combining heuristic evaluation and focus group. Results: The research identified multiple deceptive patterns in scientific literature. Analysis revealed that existing studies frequently addressed specific aspects of patterns, such as typologies or effects, and do not focus on sociotechnical issues. The exploratory evaluation suggested that the self-explainable catalog is useful and easy to use, with potential to inform users and promote awareness regarding the deceptive patterns existence and functioning. Conclusion: The study contributes to an informed understanding of deceptive patterns, highlighting the need for an approach that integrates human and technical dimensions. The self-explainable catalog is a promising tool to inform and promote awareness about the topic. By exposing the risks and mechanisms of these patterns, the research seeks to promote awareness among users and designers towards ethical interface design practices.
[HC-12] Generative AI Expands the Intellectual Reach of Course Based Undergraduate Research Experiences (CUREs)
链接: https://arxiv.org/abs/2608.27638
作者: Aditi Babar,Kristin J. Davin,Alex Dornburg
类目: Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注:
Abstract:Course-based undergraduate research experiences (CUREs) broaden access to authentic scientific inquiry through responsive instructor support as research problems become increasingly complex. Generative artificial intelligence (GenAI) may extend this support by providing individualized assistance that can adapt as student needs change. However, how embedding GenAI within a CURE to provide support across the research process impacts student inquiry, collaboration, and scientific reasoning remains unresolved. Here we use longitudinal qualitative data collected across three semesters of a bioinformatics and genomics CURE to show that GenAI expanded the intellectual reach of the research experience in three distinct ways. First, personalized, on-demand scaffolding allowed students to move beyond the boundaries of instructor expertise and transform their own interests into researchable inquiry, with all teams developing distinct self-directed projects rather than selecting instructor-provided topics. Second, GenAI became part of the distributed cognitive system of research teams, helping novice researchers communicate and coordinate across differentiated expertise without eliminating specialization. Third, expanded capability did not replace the need for disciplinary judgment. Students increasingly validated, revised, or rejected AI-generated contributions, such that research independence emerged through retained intellectual responsibility. Together, these findings suggest that GenAI can extend the reach of CUREs by expanding what novice researchers can investigate, how they can collaborate, and the level of responsibility they can assume while preserving human judgment central to authentic scientific inquiry.
[HC-13] UIC-AIHealth4All at ArchEHR-QA 2026: Answer-First Evidence Grounding for Clinical Question Answering LREC2026 ALT
链接: https://arxiv.org/abs/2608.27467
作者: Mohammad Arvan,Hossein Haeri,Natalie Parde,Rebecca T. Feinstein
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注: 10 pages, 2 figures, 6 tables. System description paper for the UIC-AIHealth4All submission to the ArchEHR-QA 2026 shared task, presented at the CL4Health workshop, LREC 2026
Abstract:We describe the UIC-AIHealth4All system for ArchEHR-QA 2026, a shared task on grounded question answering from electronic health records. We participated in Subtasks 2 (evidence identification), 3 (answer generation), and 4 (answer-evidence alignment). For Subtasks 2 and 3, we propose an answer-first pipeline in which the model generates candidate answers citing specific note sentences before classifying the full evidence set, exploiting the asymmetry between judging relevance in the abstract versus relative to a generated answer. For Subtask 4, we apply self-consistency voting over five independent model calls, retaining links by vote threshold. Our pipeline ranked third on evidence identification (Strict Micro F1 62.90), ninth on answer generation (Overall 31.90), and fifth on answer-evidence alignment (F1 79.81). A post-hoc linguistic analysis of 45 stylistic features reveals that model outputs remain 3.2 Flesch-Kincaid grade levels harder to read than clinician-authored references despite matching their word and sentence counts, suggesting readability warrants explicit optimization in clinical NLP systems. Code and prompts are available at this https URL.
[HC-14] he Effect of Emotional Context on Large Language Models Endorsement of Premature Decisions: Comparing Emotional Vulnerability Across Six Commercial Models
链接: https://arxiv.org/abs/2608.27465
作者: Cheolho Shin,Yoojin Han,Donghun Shin,Kunho Lee
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Human-Computer Interaction (cs.HC)
备注:
Abstract:As large language models (LLMs) are increasingly used for everyday decision-making advice, whether a model shifts the direction of its advice according to the user’s emotional state has become an important safety problem. We test whether emotional expression increases a model’s endorsement (encouragement to proceed) when a user, holding the same objective information, is overconfident about a premature decision (e.g., quitting a stable job on weak evidence). As a key control, we include a no-emotion multi-turn (neutral) condition that holds factual content and the number of conversational turns constant, isolating the effect of emotion from that of conversation length. We exposed six commercial models (top-tier and mid-tier models from OpenAI, Anthropic, and Google) to three scenarios (career change, business expansion, emigration) across three conditions (cold/neutral/distress) with six repetitions each, yielding 324 conversations, and measured endorsement strength (0-100) via an eight-item rubric-based automated scoring. Emotional expression significantly increased endorsement (neutral 18.6 to distress 31.5, +12.9 points; mixed-effects \beta = +12.9 , p .001 ; Cohen’s d = 0.51), and this was not explained by conversation length (cold-neutral difference non-significant, p = .083 ). Critically, the vulnerability varied by individual model rather than by price tier: five of six models showed a significant emotion effect, including the top-tier flagships Gemini 3.1 Pro and GPT-5.5, while only Claude Opus showed no significant change. Results were reproduced with an independent non-Google judge model ( \rho = .89 ) and agreed in rank with two human coders ( \rho = .70 ). Through a controlled design that separates emotion from conversational context, we show that emotional context increases LLM sycophancy even in top-tier flagship models.
[HC-15] Not All Explanations Are Sought: Information-Seeking Psychology for Human-Centered XAI
链接: https://arxiv.org/abs/2608.27464
作者: Andrea Beretta,Salvatore Rinzivillo
类目: Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注: Accepted at the 6th Workshop on Human-Centered Explainable AI (HCXAI), CHI 2026
Abstract:This position paper argues that human-centered explainable AI (HCXAI) should incorporate insights from the psychology of information seeking. Drawing on Sharot and Sunstein’s framework of information-seeking motives, we propose that people evaluate whether to engage with explanations based on three types of expected utility: instrumental (will it help me act better?), hedonic (will it make me feel better?), and cognitive (will it improve my understanding?). Each utility is estimated through a lens shaped by well-documented cognitive biases, including illusion of control, automation bias, unrealistic optimism, impact bias, overconfidence, and confirmation bias. These biases can lead to two failure modes: excessive information-seeking that fragments attention without improving decisions, and insufficient information-seeking that leaves critical risks and misunderstandings unexamined. This challenge is particularly acute for agentic AI systems, where explanations must support not just understanding a single output but anticipating cascading actions, assessing risks, and deciding when to intervene. By integrating information-seeking psychology into HCXAI, we advocate for a shift from making explanations available to making them sought: designing systems that account for when and why users actually want to know.
计算机视觉
[CV-0] SignRR: Retrieve and Refine Real Motion for Sign Language Production BMVC2026
链接: https://arxiv.org/abs/2608.28568
作者: Fidel Omar Tito Cruz,Angie Sanchez Marquina,Summy Farfan,Gissella Bejarano
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at the 37th British Machine Vision Conference (BMVC 2026)
Abstract:Sign language production (SLP) aims to generate continuous signing motion from spoken language, often through gloss-to-pose generation. Prior work mainly follows two paradigms. Generative models synthesize motion from a learned prior or from noise, without reference to an observed signing instance, making rare hand configurations and signer-specific articulation difficult to preserve. Retrieval-based methods reuse real, well-articulated motion segments, but concatenating segments from different signers and co-articulation contexts can introduce rhythm and style inconsistencies across the full sequence, not only at segment boundaries. These limitations suggest a complementary solution: use retrieval to provide realistic articulation, and use learned refinement to impose the global coherence that retrieval alone lacks. We therefore propose retrieve-and-refine, a paradigm that starts from real retrieved motion and refines it into a globally coherent signing sequence rather than generating motion from scratch. Our framework, SignRR, initializes motion from a dictionary of real sign segments and refines the full sequence with a part-aware Residual VQ-VAE, where residual quantization preserves fine hand articulation and temporal length differences are handled in the latent space. Experiments on PHOENIX14T and CSL-Daily show that SignRR achieves state-of-the-art back-translation performance while maintaining competitive pose quality.
[CV-1] GeBDA: Building Damage Assessment as Text-Based Sequence Prediction
链接: https://arxiv.org/abs/2608.28567
作者: Olivier Dietrich,Krishna Sapkota,Konrad Schindler,Genady Beryozkin
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Conventionally, Building Damage Assessment (BDA) is tackled either with dedicated network architectures or by fine-tuning geospatial image foundation models. In this work, we ask whether a general-purpose Vision-Language Model (VLM) can localize buildings and grade their damage through autoregressive sequence generation alone. We cast BDA as predicting a variable-length set of bounding boxes, each specified by its coordinates and a damage label. Our preliminary implementation, based on the open Gemma model, achieves promising damage mapping results from only bi-temporal satellite images and a suitable text prompt.
[CV-2] Video Generative Models as Geometry Learner
链接: https://arxiv.org/abs/2608.28549
作者: Haosen Yang,Jifei Song,Zhensong Zhang,Xiatian Zhu,Jiankang Deng
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 19 pages, 4 figures, 5 tables. Project page: this https URL
Abstract:Recent generative approaches to geometry estimation adapt pretrained image diffusion models and treat the task as image-conditioned generation. Leveraging off-the-shelf image diffusion models, they either (i) train task-specific geometry models (for depth and surface normal estimation) independently, losing the opportunity of exploring the intrinsic correlation of these geometric targets, or (ii) jointly fine-tune modified image diffusion backbones (e.g., altered self-attention), which typically demands substantial labeled data. To overcome these limitations in a principled fashion, we repurpose pretrained video generative models as a unified and data-efficient framework for geometry estimation, formulated innovatively as a next-frames prediction task. Our method, GeoNeXt, inherits naturally structured knowledge and richer priors from the video model, while further adapting them for joint modeling of images and geometry targets (image - geometry), enabling more data efficient and effective learning of geometry. Extensive experiments validate our method for zero-shot monocular depth and surface normal estimation across diverse datasets, outperforming both previous task-specific and unified generative competitors while using substantially less training data. Notably, our method rivals discriminative state-of-the-art approaches trained on over 100x more data and even standouts on several benchmarks.
[CV-3] xture Image Classification Using DWT AlexNet Feature Fusion and Deep Neural Networks
链接: https://arxiv.org/abs/2608.28524
作者: Arun D. Kulkarni
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 15 pages, 15 figures
Abstract:Texture image classification plays a significant role in computer vision applications, including industrial inspection, medical image analysis, remote sensing, and object recognition. Handcrafted features can capture local texture characteristics but may have limited capability to represent complex visual patterns. In contrast, deep learning models automatically learn discriminative representations but may not fully exploit the multiscale spatial-frequency information inherent in texture images. This paper proposes a hybrid feature fusion framework, termed DWT_AlexNet_DNN, which combines Discrete Wavelet Transform (DWT) features with deep features extracted using AlexNet for texture image classification.
[CV-4] Learning the Target Priors Before Image Translation: A Decoupled Training Paradigm for Cross-Modal Image Translation in Remote Sensing
链接: https://arxiv.org/abs/2608.28517
作者: Keyan Hu,Mingtao Wang,Ziyu Zhou,Tiandong Shi,Haifeng Li,Ji Qi,Chao Tao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 26 pages, including supplementary material
Abstract:Cross-modal image translation in remote sensing must preserve source-observed content while matching the target-domain distribution. Existing methods jointly learn the target prior and cross-modal dependence from scarce paired data, overlooking a key asymmetry: only the latter intrinsically requires cross-modal correspondence. We formalize this distinction through conditional-score and denoising-risk analyses and propose Learning the Target Priors Before Image Translation (LTP-BIT), a prior-first paradigm that decouples the two learning tasks. LTP-BIT first learns a target-domain generative prior from large-scale unpaired imagery, then retains the pretrained backbone weights and learns source-conditioned control through P-DART, a parameter-efficient dual-stream architecture. Controlled experiments show that prior matching and scaling primarily improve target-domain realism, whereas instance fidelity relies more strongly on conditional adaptation. LTP-BIT achieves state-of-the-art performance across SAR-to-RGB and NIR-to-RGB benchmarks using only 9.81% task-specific parameters. On QXS-SAROPT, it retains near-full-data instance fidelity with only 25% of the paired samples.
[CV-5] Anatomy-Aware Promptable Segmentation with Online Interactive Training for AUTOPET V
链接: https://arxiv.org/abs/2608.28461
作者: Pablo Lozano-Jimenez,Sergio Romero-Tapiador,Ruben Tolosana
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:We present an anatomy-aware, promptable model for whole-body lesion segmentation in FDG and PSMA PET/CT, developed for the AUTOPET V challenge. The proposed method is built as family of nnU-Net-based models and trained in two stages: i) a pre-training stage that produces a strong initial segmentation, and ii) an online interactive stage that learns to exploit scribble prompts, refining the prediction over successive interactions. Anatomical context is incorporated through organ supervision using a single shared head that predicts lesions and organs from the same features, which reduces false positives arising from physiological uptake. Also as the tracer (i.e., FDG/PSMA) is not provided at inference, we add a tracer classifier based on image processing and a random forest over coronal MIP features, routing each study to a combined FDG+PSMA model or to a PSMA-specific model. Across four-fold cross-validation the organ-supervised model achieves the best and most stable performance, the interactive stage improves the Dice score monotonically with each prompt, and PSMA-specific training yields the strongest tracer-wise results.
[CV-6] LayerRecall: A State-Conditioned Memory Router for Long-Horizon Consistency in Video Generation
链接: https://arxiv.org/abs/2608.28460
作者: Yixuan Ding,Jiahao Kong,Wei Huang,Ruijie Quan,Yi Yang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 15 pages, 11 figures
Abstract:Autoregressive video diffusion enables scalable long-video generation by producing chunks from a bounded recent context. While recency-based caching preserves local continuity, it evicts historical cues needed when subjects, objects, scenes, or attributes reappear. Existing memory mechanisms expose models to nonlocal history, but access alone does not ensure effective use. Our analysis reveals that video DiT layers exhibit distinct preferences for current, recent, and distant context, suggesting that long-range memory requires deciding both what to retrieve and where to use it. We introduce LayerRecall, a current-conditioned, layer-selective memory router that retrieves relevant historical K/V states and injects them only into backbone-specific memory-sensitive layers while preserving local attention elsewhere. To reduce reliance on scarce high-quality long-horizon videos and explicit memory-allocation labels, we further propose Cross-Horizon Prediction Matching (CHPM), which uses a privileged long-context reference to supervise the bounded-memory router in prediction space. Across 100 multi-shot evaluation prompts, LayerRecall achieves the best overall results on MemoBench and MovieBench while matching its backbone on VBench-Long, demonstrating stronger long-range recovery without sacrificing local continuity. Qualitative analyses further reveal memory-guided self-correction, whereby initially mismatched local attributes return to their historical appearance without resetting ongoing motion or scene structure. Additional analyses show cross-backbone portability and negligible inference overhead.
[CV-7] ARC-CT: Anatomy-Routed Contrastive Vision-Language Learning for 3D Chest CT MICCAI2026
链接: https://arxiv.org/abs/2608.28455
作者: Huseyin Umut Isik,Mehmet Alp Ozaydin,Sila Kurugol,Şeyda Ertekin
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted to the Thoracic Image Analysis (TIA) Workshop at MICCAI 2026
Abstract:Contrastive vision-language learning uses paired chest CT volumes and radiology reports to learn abnormality classifiers without manually annotated labels. However, two characteristics of chest CT challenge conventional global contrastive learning. First, many critical abnormalities are small or anatomically localized, and pooling an en- tire volume into a single embedding may dilute their visual evidence. Second, the standard contrastive objective treats every other scan in a batch as a negative. Because many chest CTs share abnormalities, this objective incorrectly pushes co-positive pairs apart. We propose Anatomy-Routed Contrastive Learning for 3D Chest CT (ARC-CT), a region-aware framework that addresses these limitations using only la- bels extracted from reports by an LLM, with no manual annotations or bounding boxes. ARC-CT combines three components: (1) an Anato- myQFormer localizing evidence via queries constrained by automatically generated organ masks; (2) a label-Jaccard soft InfoNCE objective in- tegrating the standard one-hot target with the label-set overlap of each pair, which reduces false-negative penalties between studies that share clinical findings; and (3) an organ-level alignment loss connecting mask- pooled visual features to organ-specific report text extracted offline with a large language model. ARC-CT achieves a 0.86 mask-free macro AUC across 18 abnormalities using a compact 3D ResNet-18 backbone. Over- all, ARC-CT outperforms both comparable efficient baselines and sev- eral larger transformer models. Our code and weights are available at this https URL.
[CV-8] Prompt-Guided Interactive Segmentation of Interstitial Lung Disease in Thoracic CT
链接: https://arxiv.org/abs/2608.28453
作者: Vasilis Dedousis,Lubnaa Abdur Rahman,Lorenzo Brigatο,Ethan Dack,Andreas Christe,Christoph Frank,Manuela Funke-Chambour,Justus Roos,Adrian Huber,Lukas Ebner,Stavroula Mougiakakou
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Accurate segmentation of interstitial lung disease (ILD) patterns is essential for quantitative disease assessment and longitudinal monitoring. However, existing approaches remain limited by relying on dense annotations and producing static predictions that cannot be refined, motivating interactive approaches. While promptable models show promise in interactive segmentation, their adaptation to ILDs remains largely unexplored. To address this gap, we investigate prompt-guided foundation models for ILD refinement and present, to the best of our knowledge, the first adaptation of MedSAM2 for interactive 3D ILD segmentation on thoracic CT. We investigate three fine-tuning strategies and multiple clinically motivated prompts: bounding-boxes (BBox), point, lasso, and scribble. On a dataset spanning seven ILD patterns and healthy lung tissue, full model fine-tuning performed best, improving the average Dice score by 4.7 percentage points over this http URL BBox prompts achieve the strongest performance, non-native MedSAM2 interactions such as lasso and scribble prompts also prove effective. Finally, we present and evaluate a proof-of-concept end-to-end workflow in which MedSAM2 is initialized from an automatic segmentation prior and subsequently refined using radiologist prompts. Model weights and plug-ins made available at: this https URL.
[CV-9] Lossy Event Compression: From Event Stream Distortion to Task Performance
链接: https://arxiv.org/abs/2608.28429
作者: Zahra Rezaee,Catarina Brites,João Ascenso
类目: Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
备注:
Abstract:Event cameras generate asynchronous, sparse data streams with microsecond temporal resolution, but in moderate-to-high motion scenes they can produce as many as hundreds of millions of events per second, creating significant bandwidth and storage challenges. Lossy compression is therefore essential for practical deployment, yet existing event stream distortion metrics fail to reliably predict compression-induced degradation at the task level, forcing codec optimization to rely on expensive task-specific evaluations. To address this gap, this paper introduces two fundamentally different event compression pipelines: i) an aggregation-based pipeline that converts the event stream into polarity-based histogram frames for compression with the conventional image codec JPEG 2000, and ii) a frame-free point cloud-based pipeline that codes events natively as 3D points using the octree-based codec G-PCC. Both pipelines are then assessed within a unified task-driven evaluation framework that relates event stream distortion to downstream application performance across four representative tasks: i) video reconstruction, ii) object detection, iii) optical flow estimation, and a delay-sensitive task iv) asynchronous feature tracking under a reference-relative protocol. Building on this framework, five classification-based distortion metrics are applied to event compression for the first time, to the best of the authors’ knowledge, and benchmarked against existing event stream metrics. Experimental results demonstrate that the proposed metrics reliably predict compression-induced task degradation across different coding frameworks. This demonstrates that event stream distortion assessment can be an efficient alternative to repeated task-specific evaluation, providing direct guidance for the development and optimization of future event data coding solutions.
[CV-10] Post-Training VLMs for Video Mistake Detection BMVC2026
链接: https://arxiv.org/abs/2608.28406
作者: Federico Spurio,Olga Zatsarynna,Lars Doorenbos,Emad Bahrami,Gianpiero Francesca,Juergen Gall
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: Accepted at BMVC 2026
Abstract:Human mistakes are inevitable when following instructions, yet they can lead to severe consequences. As such, there has been an increased interest in developing methods for detecting mistakes in videos, with current methods mostly focusing on closed-set protocols. While successful in controlled settings, the closed-set assumption limits their wider applicability, as any changes to the task require collecting new data and re-training models. Instead, we argue that mistake detection methods should learn the general concept of a mistake, rather than overfitting to step-specific details. To reflect this, we introduce the Mistake Detection Video Question Answering (MD-VQA) protocol and accompanying benchmark. MD-VQA tests whether methods can discern if a step was executed correctly with respect to its description, for both seen and unseen actions. To address this important challenge, we propose the first video-language-model post-training technique for mistake detection. Our method uses a tailored reward function to encourage the model to identify discrepancies between an instruction and the corresponding video. Extensive evaluations demonstrate that this approach outperforms zero-shot, supervised fine-tuning, and post-training baselines. Notably, our method generalizes especially well to unseen procedures, for instance, with an improvement of up to 11.6% over the best-performing baseline on EP-VQA, paving the way toward general mistake detection. We release our code and benchmark at this https URL.
[CV-11] How Far Can 5500 Hours of Driving Take You? A Scaling Law Analysis of Video Diffusion Models
链接: https://arxiv.org/abs/2608.28404
作者: Victor Besnier,Anh-Quan Cao,Elias Ramzi,Spyros Gidaris,Tuan-Hung Vu,Andrei Bursuc,Eloi Zablocki,Matthieu Cord
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Video generation for autonomous driving cannot follow the web-scale route: driving data is expensive to collect, bound by privacy requirements, and cannot be scraped at will, so models must make the most of a fixed corpus. We present a systematic scaling-law study of video diffusion models trained from scratch on driving data: a family of models from 1M to 9B parameters, trained at different exposures on up to 5,500 hours of driving. Validation loss follows consistent power laws in both model size and training exposure, answering the questions that shape a training budget: whether compute is better spent on longer training or on a larger model, and whether more data is needed. Loss improves much faster with training exposure than with model size, making longer training the most effective way to improve a fixed model under limited compute. However, larger models continue to achieve lower asymptotic loss, so compute-optimal scaling still favors increasing model size when sufficient compute and data are available. Guided by these laws, we train a 9B-parameter model, to our knowledge the largest video diffusion model trained from scratch on driving data: it sets a new open-source state of the art for driving video generation, as measured on nuScenes. Our code and pretrained models are available at this https URL. NATIX is separately releasing the underlying driving data in stages.
[CV-12] GraspHOI: Full-Body 3D Human-Object Reconstruction with Finger-Level Grasps from a Single In-the-Wild Image
链接: https://arxiv.org/abs/2608.28386
作者: Semin Kim,Haechan Shin,Jongyoo Kim
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Existing monocular full-body 3D human-object interaction (HOI) methods do not combine explicit finger-level grasp optimization with category-agnostic object reconstruction. Despite plausible body-object configurations, their fingers may float from or penetrate objects instead of forming a grasp. We present GraspHOI, the first framework that reconstructs a full-body 3D HOI from a single image while explicitly optimizing finger articulation against the reconstructed object. GraspHOI recovers object geometry directly, without predefined meshes or a fixed category vocabulary. It reconstructs the body, hands, and object separately, aligning them in metric camera space via depth-based registration and image-space alignment. Occlusion-aware palmar correspondences seat the object against the grasping hand, and contact-aware optimization refines arm and finger articulation to form surface contact without excessive penetration. Across four benchmarks and six baselines, GraspHOI improves relative human-object placement, hand accuracy, and contact plausibility. Full pipeline code will be released.
[CV-13] Real-Time Musculoskeletal Surrogates for Pediatric Cerebral Palsy: a Credibility Pilot
链接: https://arxiv.org/abs/2608.28371
作者: Mohammad Arif Ul Alam
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Real-time musculoskeletal (MSK) surrogates could support personalized rehabilitation for children with cerebral palsy (CP), but their credibility depends on subject-wise evaluation, low inference latency, and calibrated uncertainty. We develop a subject-conditioned causal neural surrogate using OpenSim-derived static parameters, temporal joint kinematics, true muscle capacities, and training-only perturbations. On a real pediatric CP gait dataset comprising nine children, we use leave-one-subject-out validation on six development subjects and evaluate a frozen configuration once on three locked test subjects. The surrogate accurately reproduces musculotendon lengths (R-square = 0.92 in development validation and approximately 0.95 on locked subjects; nRMSE 8%) while requiring only sub-millisecond to few-millisecond neural inference, well below a 100 ms interactive-rehabilitation target. In contrast, direct muscle-force estimation remains unstable at this small, heterogeneous scale: pooled metrics can overstate within-subject, per-muscle accuracy. A Monte Carlo credibility pilot further shows that propagating only +/-5% anthropometry and muscle-capacity variation produces severely overconfident nominal 90% intervals (approximately 4% force coverage and below 1% MT-length coverage). These results establish a leakage-free evaluation and credibility framework for pediatric MSK surrogates, while identifying force modeling and epistemic uncertainty as the central next challenges for clinically credible digital twins.
[CV-14] Denoising-Aware Temporal Point Cloud Completion for 3D Crop Architecture Recovery and Phenotypic Trait Extraction
链接: https://arxiv.org/abs/2608.28343
作者: Mrudul Mittal,Soumyashree Kar
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 25 pages of pdf for manuscript , total 12 files including bbl and tex
Abstract:High-throughput phenotyping depends on accurate 3D reconstruction of plants across growth stages, yet the development and evaluation of temporal completion methods are limited by the lack of datasets with complete geometric ground truth. To address this challenge, we introduce SynthCrop4D, a procedurally generated synthetic dataset of temporally evolving plant point clouds that provides controllable noise, occlusion, and complete plant geometry for benchmarking reconstruction methods. Using this dataset, we evaluate a two-stage pipeline that combines spatial denoising and temporal point cloud completion. First, a denoising module removes structural artifacts from raw laser-scanned point clouds. The resulting data are then processed by an Adaptive Temporal PoinTr model that reconstructs the current growth stage (t) using information from the previous stage (t-1), enabling recovery of regions missing due to self-occlusion. We evaluate the proposed framework on both SynthCrop4D and the real-world Pheno4D dataset (tomato and maize) under settings with and without denoising. Results show that denoising substantially improves reconstruction quality, with the best configuration achieving a Chamfer Distance of 0.0061 on SynthCrop4D (Temporal PoinTr + Mamba-DG) and an F-Score of 0.2080 on Pheno4D (Vanilla PoinTr + Mamba-DG). We further demonstrate the use of completed point clouds for phenotypic trait extraction, including plant height, canopy width, and convex hull volume, obtaining hull-volume MAEs of 0.021 on synthetic data and 0.343 on real data. Together, SynthCrop4D and the proposed pipeline provide a benchmark and methodology for temporal plant reconstruction and high-throughput crop phenotyping.
[CV-15] Cross-Spectral Dense Correspondence for Multimodal Spectral Medical Imaging ECCV2026
链接: https://arxiv.org/abs/2608.28341
作者: Eric L. Wisotzky,Jost Triller,Simon W. Härtl,Oliver T. Bruns,Peter Eisert,Anna Hilsmann
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted at 2nd Data Curation Augmentation in Medical Imaging Workshop at ECCV 2026
Abstract:Precise dense correspondence is a fundamental prerequisite for multimodal spectral imaging systems that fuse disparate wavelength ranges for subsequent analysis in medical and scientific imaging. Corresponding image points are often observed with non-overlapping spectral sensitivities, leading to wavelength-dependent contrast changes, intensity inversions, and appearance shifts for which dense ground truth is difficult to obtain and conventional RGB-based training data provides only limited supervision. We address this data gap by introducing a sensor-agnostic cross-spectral modulation protocol on established correspondence benchmarks with intensity input projection, and by proposing a synthetic cross-spectral correspondence benchmark simulating physically plausible radiometric differences. Evaluation on several modern dense correspondence backbones trained with our unified cross-spectral protocol showed substantial improvements under severe spectral mismatch while maintaining performance on standard RGB benchmarks. Ablation experiments show that view-dependent channel selection and nonlinear radiometric transformations provide complementary robustness, indicating that the primary limitation of existing models is not their structural matching capacity but the mismatch between training distribution and spectral characteristics of the target image pair. Qualitative evaluations on heterogeneous medical spectral acquisition systems demonstrate the practical relevance of the proposed training data augmentation protocol as an enabler for spatially coherent spectral fusion in HSI workflows.
[CV-16] Abstract4D: A Large-Scale Dataset and Framework for Understanding the Visual Language of Abstract Art
链接: https://arxiv.org/abs/2608.28339
作者: Haowei Zhang,Yuanpei Zhao,Ji-Zhe Zhou,Mao Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Artificial intelligence can classify artistic styles and synthesize images, but it still lacks a model of the visual language that gives art meaning. Abstract painting minimizes object semantics and foregrounds structural cues, making it an ideal testbed for computational perception. We introduce \textbfAbstract4D, the largest dataset of abstract paintings to date: more than 120,000 images paired with rich metadata and multi-dimensional prompts that capture each work’s perceptual attributes—\textitform, color, texture, and composition. Annotations are produced by a hybrid human–VLM pipeline for quality and consistency. Using Abstract4D, we (i) analyze the semantic structure of abstract art through large-scale embedding visualization, uncovering how perceptual relationships organize artistic meaning, and (ii) establish benchmark tasks for classification, cross-modal retrieval, and text-to-image generation to evaluate how AI models perceive and reproduce abstract visual language. Together, these analyses demonstrate how Abstract4D enables both exploration and quantitative assessment of AI’s ability to represent and interpret abstract art.
[CV-17] Conditional Visual Evidence Utility: State-Dependent Rank Reversals in Frozen Vision-Language Encoders
链接: https://arxiv.org/abs/2608.28316
作者: Yunxuan Fang,Xinhe Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Static importance scores compress visual evidence into a single ranking, but the value of remaining evidence can change after one cue has been observed. We study this possibility in controlled compositional visual search, where color, shape, and texture evidence can be independently exposed and their conditional marginal utility measured across acquisition states. In a held-out confirmation on 800 scenes, frozen OpenCLIP and SigLIP exhibit robust state-dependent rank reversals that concentrate in candidate-overlap regimes designed to induce ordering changes. The structure persists across two evidence-accumulation constructions and ten equivalent query wordings, but disappears under query-scene derangement. We also ask whether these reversals matter for decisions. In a post-confirmation exploratory matched-first-action analysis, reranking only after the first acquisition yields positive step-2 utility when decisions are selected under one evidence mode, wording, or backbone and evaluated under another. Together, these results show that evidence importance is state-dependent in this controlled setup and that updating an evidence ordering can retain decision-relevant value across evaluator changes. They motivate evaluating vision-language evidence use conditionally rather than through a single static ranking, while providing a measurable target for future adaptive evidence-selection methods.
[CV-18] FUSED: Forensic-Semantic Mixture-of-Experts for AI Inpainting Detection and Localization
链接: https://arxiv.org/abs/2608.28302
作者: Anton Nuzhdin,Marcel Worring,Ivona Najdenkoska
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Diffusion-based inpainting models modify only a localized part of an image, while many AI-image detectors rely on global artifacts and do not localize. These artifacts vary across generators, limiting detector transfer under distribution shifts. Recent work shows that restoring the authentic pixels outside the inpainted region removes these cues and can degrade pretrained detectors. To address this, we present FUSED, a unified framework for the joint detection and localization of AI-generated inpainting. FUSED combines low-level forensic cues with high-level semantic features using a sparsely-gated Mixture-of-Experts architecture, enabling the model to adaptively prioritize the most relevant signal for each token. For each input, FUSED predicts both an image-level manipulation score and a pixel-level mask of the inpainted area. On the OpenSDID cross-generator benchmark, FUSED achieves the best average detection and localization, with the largest gains on unseen generators. The same model transfers directly to the held-out AutoSplice and CocoGlide benchmarks, more than doubling localization performance. Evaluating each held-out benchmark with and without the global generator artifact further shows that all evaluated methods, ours included, partly read the artifact as evidence of manipulation, and FUSED remains the strongest under both conditions. Code and pretrained models are available at this https URL.
[CV-19] GeoFF3D: Coordinate-Anchored Feed-Forward Reconstruction for Large-Scale UAV Mapping
链接: https://arxiv.org/abs/2608.28288
作者: Xiang Yang,Yongli Wang,Yunsheng Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Existing feed-forward 3D reconstruction methods typically process a bounded number of images and recover cameras and geometry in local or internally normalized frames. Extending them to large-scale UAV mapping requires scalable multi-chunk processing and reliable aggregation, while full Sim(3) alignment can become unstable for near collinear trajectories. We present GeoFF3D, which combines a coordinate-anchored model with a spatial large-scale reconstruction framework (SLRF). The model uses georeferenced camera translations and optional geometric priors to predict camera poses and dense point maps directly in a gravity-aligned Z-up metric frame. SLRF partitions images into spatially overlapping chunks, propagates shared-view priors, and aggregates local reconstructions hierarchically, while remaining applicable to different bounded-view models. Across nine aerial mapping blocks, GeoFF3D achieves the best average reconstruction quality, improving F@5 from 0.829 for Pi3X + SLRF to 0.877. On long UAVScenes sequences, it reaches 0.848, compared with 0.687 for Pi3X + SLRF and 0.451 for the strongest evaluated SLAM/streaming baseline. GeoFF3D reconstructs 2,000 images in approximately five minutes, demonstrating scalable and robust large-scale UAV this http URL code is available at this https URL.
[CV-20] Non-Uniform Quantisation for 3DGS Compression
链接: https://arxiv.org/abs/2608.28272
作者: Bert Van hauwermeiren,Patrice Rondao Alface,Adrian Munteanu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:3D Gaussian Splatting (3DGS) has emerged as a powerful technique for novel view synthesis, yet its high bitrate requirements pose significant challenges for storage and transmission. To enable practical applications and ensure interoperability within the 3DGS ecosystem, standardised compression formats are essential. In this paper, we propose a novel non-uniform quantisation scheme specifically tailored for 3DGS models. Our approach adapts to the underlying data distribution by applying importance-weighted quantisation and eliminating post-voxelisation redundancy through importance weighted merging. Extensive evaluations on benchmark datasets demonstrate that our method achieves state-of-the-art compression performance. Furthermore, the proposed scheme is compatible with any point-cloud-based representation and is intended as a formal contribution to the upcoming MPEG 3DGS compression standardisation activities.
[CV-21] Physics-Guided Flow Matching for CT Image Reconstruction
链接: https://arxiv.org/abs/2608.28256
作者: Davide Evangelista
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注: 16 pages, 6 figures, 2 tables
Abstract:Deep generative models have recently emerged as powerful priors for solving ill-posed inverse problems in CT, with diffusion-based approaches achieving state-of-the-art reconstruction performance. However, diffusion models typically rely on stochastic sampling procedures, long inference trajectories, and carefully tuned noise schedules, which can limit computational efficiency and numerical stability, especially at high spatial resolutions. In this work, we investigate Flow Matching as an alternative generative prior for CT reconstruction. We train a high-resolution Rectified Flow Matching model on 256x256 chest images from the Mayo Clinic Low-Dose CT dataset. To mitigate overfitting and limited anatomical variability, we employ a two-stage training strategy consisting of an initial phase with strong, anatomically informed data augmentation, followed by a fine-tuning phase with reduced or no augmentation to refine structural fidelity. The resulting model is capable of generating high-quality and anatomically coherent CT-like images, serving as a strong learned prior. We then evaluate multiple reconstruction methods specifically designed for Flow Matching models, including Plug-and-Play Flow, FlowDPS, Flower, and Flow-Priors (ICTM), and compare them against state-of-the-art diffusion-based reconstruction algorithms such as DDRM, DPS, and DiffPIR. Experimental results across several CT inverse problem settings show that Flow Matching-based approaches consistently outperform diffusion-based methods in terms of PSNR, SSIM, and perceptual quality, while requiring fewer sampling steps. Finally, we publicly release the trained Flow Matching model and accompanying code to facilitate reproducibility and future research. Overall, this work demonstrates that Flow Matching provides a stable, efficient, and effective alternative to diffusion models for high-resolution CT image reconstruction.
[CV-22] A comprehensive and trustworthy benchmark of AI methods for change detection in Earth observation
链接: https://arxiv.org/abs/2608.28247
作者: Tadej Tomanič,Alice Baudhuin,Jan Sotošek,Jure Brence,Panče Panov,Nikola Simidjievski,Dragi Kocev
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Change detection in Earth observation (EO) is critical for monitoring land surface transformations, yet recent research in the field is constrained by inconsistent evaluation protocols and a narrow focus on predictive accuracy without regard for computational efficiency. To address this, we present a standardized, open-source benchmark for evaluating state-of-the-art (SOTA) deep learning methods for Earth observation change detection. We conduct a comprehensive analysis of ten representative model architectures, ranging from convolutional networks (CNNs) to vision transformers (ViTs), across ten heterogeneous change detection datasets. We rigorously evaluate these models with identical experimental protocols, comparing models trained from scratch against those utilizing pre-trained weights. Furthermore, we evaluate predictive performance alongside computational efficiency, including parameter counts and inference latency. Our findings reveal that well-optimized classical architectures, such as Siamese U-Nets, frequently outperform more complex contemporary models when computational efficiency is factored in, and that pre-training consistently provides a significant performance boost with no additional inference cost. To ensure complete transparency and reproducibility, all experimental resources, including standardized data splits, training scripts, training logs, and model checkpoints are publicly available and adhere to FAIR principles (Findable, Accessible, Interoperable, and Reusable).
[CV-23] WilLaGS: Latent-Conditional 3D Appearance Fields for Robust Gaussian Splatting In-the-Wild ECCV2026
链接: https://arxiv.org/abs/2608.28240
作者: Yuhao Bai,Qianqiu Tan,Lilong Chen,Huanhuan Lv,Lijun Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ECCV2026
Abstract:3D Gaussian Splatting (3DGS) delivers real-time and high-fidelity rendering but remains challenged by unconstrained in-the-wild scenes, where drastic appearance variations and transient objects violate multi-view consistency. Existing methods are fundamentally limited by independent and discrete embeddings that struggle to capture continuous environmental changes or model spatially-varying local illumination. To address these limitations, we propose \textbfWilLaGS, a unified framework for robust 3D scene reconstruction and generative appearance synthesis under unconstrained settings. Specifically, we introduce a generative appearance model where a \beta -VAE learns a structured and continuous manifold of global appearance. Conditioned on the latent code, we construct a 3D neural appearance field that generates dynamic Tri-Plane features to encode spatially-varying local illumination effects. Furthermore, to suppress transient artifacts, we present a self-supervised perceptual masking mechanism that leverages a Teacher-Student (EMA) architecture to derive a stable scene consensus, robustly identifying inconsistent regions via perceptual discrepancies. Extensive experiments on multiple datasets demonstrate that \textbfWilLaGS achieves state-of-the-art performance in reconstruction quality and novel view appearance synthesis, while maintaining real-time rendering efficiency.
[CV-24] RASA: Disentangled Spatial-Motional Priors for Cross-Identity Character Animation ECCV2026
链接: https://arxiv.org/abs/2608.28219
作者: Zhen Xiao,Zhen Shen,Zhaofan Qiu,Ting Yao,Xueliang Liu,Tao Mei
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ECCV 2026. 15 pages, 7 figures
Abstract:Cross-identity character animation aims to drive a target identity from a reference image to follow the motion of a source character from a driving video. The core challenge lies in the inherent entanglement of two capabilities: cross-identity spatial mapping (aligning position, scale, and skeletal proportions) and motion control (refining joint articulation, volumetric consistency, and view coherence). We introduce Reference-Aware Structural Alignment (RASA), a framework that disentangles spatial mapping from motion control by injecting structured priors into a Diffusion Transformer (DiT). Our approach has two stages. First, a Spatial Prior Calibrator (SPC) fuses reference identity with driving pose to generate a spatially grounded initial noise latent, ensuring correct positioning, scaling, and alignment with the driving skeleton. Second, an Inherent Motional Guider (IMG) encodes shape-agnostic SMPL articulation parameters into a semantic motion vector beyond appearance-biased 2D keypoints. Injected into intermediate DiT layers, this vector complements the base pose condition for anatomically consistent articulation and view-aware volumetric refinement. We curate CIM-Bench, a high-quality benchmark with rigorous curation, for evaluation. Extensive experiments show RASA significantly outperforms state-of-the-art methods in motion fidelity and visual quality. Our work establishes a new paradigm showing disentangled spatial and motional priors are key to robust character animation. Project page: this https URL
[CV-25] Focus Where It Counts: A Salience-Driven Vision-Language Model for Low Vision Assistance
链接: https://arxiv.org/abs/2608.28218
作者: Jiazhao Liang,Hao Huang,Shuaihang Yuan,Congcong Wen,Geeta Chandra Raju Bethala,Giles Hamilton-Fletcher,Yu Hao,John-Ross Rizzo,Mengyu Wang,Anthony Tzes,Yi Fang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Vision-language models (VLMs) are rapidly progressing and offer promising capabilities for assistive technologies supporting persons with blindness or low vision. However, existing VLMs are primarily designed for general-purpose captioning and do not explicitly model human perceptual priorities, thereby limiting their ability to emphasize the most relevant information in a scene. To address this gap, we propose a salience-driven captioning framework that prioritizes scene elements according to their importance for human-centered assistance. We curate three salience-aware datasets, namely, Salience COCO, Salience Flickr, and Salience VizWiz, with object-level salience annotations designed to reflect the visual information most relevant to low vision users across different environments. Building on these datasets, we introduce Salience-LLaVA, a salience-aware VLM that incorporates salience cues to generate captions in which important elements are mentioned in the order of importance. Our work makes four main contributions. We build salience-aware datasets verified by low vision participants, propose Salience-LLaVA to describe objects in the order of importance, introduce SCMI to evaluate ordering accuracy, and deploy the system on assistive glasses to demonstrate real-world practicality. Code and datasets are available at: this https URL
[CV-26] WALDO: One-Shot Exemplar-Conditioned Object Detection in Cluttered Scenes
链接: https://arxiv.org/abs/2608.28216
作者: Kishor Datta Gupta,Ahmed Rafi Hasan,Md. Mahfuzur Rahman,Md. Sadman Haque,Mohd Ariful Haque
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Locating a specific object instance in a cluttered scene using a single reference image and a short description, and reporting when that instance is absent, large vision-language models usually address this task. We ask whether the same capability is available far more cheaply, from representations already learned by a world-model pretraining objective. We present WALDO, a one-shot exemplar- and language-conditioned detection head with 3.4M trainable parameters that reads frozen V-JEPA 2.1 features to jointly predict object localization and target presence, with no gradient on the backbone. Because exemplar-conditioned supervision is scarce, we synthesize training episodes from instance annotations, mining exemplars from ground-truth boxes and constructing absence cases that exclude the referenced instance while leaving same-category distractors in view. This is easy to get wrong: in the obvious implementation, crop size alone predicts the label, and a head trained on it reaches 0.9998 absence AUROC without ever consulting the exemplar, and we report the negative controls that close the shortcut. On 35 held-out cluttered scenes, WALDO achieves a 0.461 catalogue AP@50, compared to 0.306 for a prompted Grounding DINO baseline under an identical scorer. Substituting DINOv3 for V-JEPA under a matched 576-token grid drops within-category absence AUROC from 0.880 to 0.726 and instance AP@50 from 0.201 to 0.141, isolating the pretraining objective rather than input resolution as the source of the gain. Instance-level Success@1, however, reaches only 0.190 against a 0.190 category-chance floor: world-model features transfer to localization precision and absence detection but not to instance identity.
[CV-27] Explainable Diabetic Retinopathy Classification Using Vision Foundation Models
链接: https://arxiv.org/abs/2608.28207
作者: Abhishek Verma,Anila Krishna,Abhishek Gajanan Bankar,Juan Miguel Lopez Alcaraz
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 11 pages, 4 figures, source code under this https URL
Abstract:Diabetic retinopathy (DR) is a major cause of preventable blindness, creating a need for accurate and trustworthy automated screening. This study investigates an explainable DR classification framework using vision foundation models and multiple transfer learning strategies. Three backbones, DINOv2, CLIP, and Vision Transformer (ViT), were evaluated using full fine-tuning, linear probing, and Low-Rank Adaptation (LoRA). Models were trained and internally evaluated on the ODIR dataset and externally evaluated on APTOS to assess generalization. DINOv2-LoRA achieved the highest internal AUROC of 0.758, while DINOv2 full fine-tuning and ViT full fine-tuning achieved the highest external AUROC of 0.920. Calibration was further assessed using reliability analysis after isotonic regression. For explainability, Grad-CAM and HiResCAM were evaluated against expert-annotated lesion masks from the IDRiD dataset using Dice, Intersection over Union (IoU), and Pointing Game metrics. The results demonstrate that foundation models, particularly DINOv2, can provide strong predictive performance, while LoRA offers a parameter-efficient alternative to full fine-tuning. Quantitative evaluation of explanation maps further supports the assessment of whether model attention corresponds to clinically relevant retinal lesions.
[CV-28] NumBench: Diagnosing Counting Failures in Text-to-Image Models
链接: https://arxiv.org/abs/2608.28206
作者: Sandeep Wadhwa,Mayank Vatsa,Richa Singh,Parrva Chirag Shah,Prakhar Galriya
类目: Computer Vision and Pattern Recognition (cs.CV); Databases (cs.DB)
备注: The compiled main paper has seven technical-content pages; references start on page 8. The compiled supplement has three pages
Abstract:Text-to-image (T2I) models often generate the wrong number of objects, yet existing benchmarks are too small or weakly controlled to explain why. We introduce \textbfNumBench, a benchmark of 640,000 prompts spanning 1,600 categories and counts from 1 to 100. Its factorial design varies object composition, spatial guidance, and appearance conditions while balancing counts and category exposure. We also develop a process model in which requested instances compete for a finite set of resolvable image regions. The model predicts a near-quadratic collision deficit at low occupancy and shows how coordinated placement reduces it. For scalable evaluation, we propose the Confidence-Weighted Numeric Precision Score (\cwnps), which aggregates three calibrated detectors and discounts uncertain proposals. Across five commercial systems, two open models, and two specialized counting methods, performance declines sharply with requested count; all evaluated methods are weak above 50 objects. Count range has the largest measured effect, followed by layout and composition. Grid guidance is strongest among guided layouts, consistent with the coordination prediction, although the analysis does not establish collision as the sole cause. A 14,400-image human study supports automated evaluation through count 50, while results on 243 natural-language prompts show transfer beyond NumBench templates.
[CV-29] Cut-ViT: Task-Specific Model Pruning via Gram Anchoring Subspace Consistency ECCV2026
链接: https://arxiv.org/abs/2608.28205
作者: Jianjian Yin,Liulei Li,Tao Chen,Yi Chen,Yazhou Yao,Wenguan Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ECCV2026
Abstract:Pruning visual foundation models has attracted considerable attention. However, existing methods focus on rigid point-to-point token alignment on a single dataset for pruning, suffering from two limitations: i) robustness degradation, and ii) task-specificity deficiency. To address these limitations, we propose a task-specific pruning pipeline, named Cut-ViT. Specifically, we first construct gram anchoring matrices from both spatial and semantic perspectives, and perform the subspace decomposition to extract the corresponding subspace bases. Basis-agnostic and residual constraints are then adopted to align the gram subspaces between the native and pruned DINOv3 models along spatial and channel dimensions, enabling subnetworks to inherit robust feature representations of native DINOv3. Furthermore, we design spectral entropy adaptation, which quantifies the information density of feature manifolds along spatial and channel dimensions, thereby adapting the pruning objective to specific downstream tasks. Experiments show that Cut-ViT requires approximately one minute on a single A100 GPU to obtain subnetworks at various sparsity levels, using only 20.9% of the time and 45.5% of the GPU memory compared with previous methods, while achieving SOTA performance on six tasks across nine datasets.
[CV-30] UniLipi: A Unified Multi-Script OCR for Historical Indic Manuscripts ICDAR2026
链接: https://arxiv.org/abs/2608.28195
作者: Tathagata Ghosh,Sai Madhusudan Gunda,Simran Singh Sandral,Ravi Kiran Sarvadevabhatla
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: To be published in ICDAR 2026
Abstract:Optical character recognition (OCR) for handwritten Indic manuscripts is essential for large-scale digitization and computational access to manuscript heritage. However, existing approaches are typically developed for one script at a time and require substantial script-specific customization. This limits scalability and practical deployment across diverse collections. We present UniLipi, a unified multi-script OCR model for handwritten Indic manuscripts trained jointly across 13 Indic scripts within a single framework. UniLipi directly handles realistic manuscript conditions, including extreme variation in line geometry, large variation in line length, and partial interruptions caused by non-textual manuscript entities such as holes, stains, or pictorial illustrations. To operate effectively under ultra low-resource conditions, the model leverages script-aware synthetic manuscript data generation, substantially reducing reliance on large volumes of real annotated data. Beyond historical manuscripts, we show that UniLipi serves as an effective foundational pretrained model. Specifically, its learned representations enable good OCR performance for contemporary Indic handwriting and extend to several non-Indic scripts, including Tibetan, Italian, Latin, and Chinese scripts. In addition to transcription, UniLipi predicts script identity and per-line native character counts, supporting practical manuscript cataloging workflows.
[CV-31] Locate Anything in Videos: Rethinking Efficient Generative Spatio-Temporal Video Grounding
链接: https://arxiv.org/abs/2608.28192
作者: Hanoona Rasheed,Haania Siddiqui,Ming-Hsuan Yang,Fahad Shahbaz Khan,Salman Khan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Spatio-temporal video grounding (STVG) requires models to identify when a referred event occurs and localize the target entity throughout that interval. Existing multimodal large language models typically serialize dense localization trajectories autoregressively, causing decoding latency to grow with tube length and allowing localization errors to propagate across time. We introduce Parallel Tube Decoding (PTD), a generative formulation that decomposes grounding into a temporal block followed by time-conditioned spatial blocks decoded simultaneously. This removes both token-level and trajectory-level dependencies, reducing the sequential decoding depth to a fixed 1 + 1 rounds, independent of tube length. To enable parallel spatial generation, we introduce Decoupled Block Attention, which preserves access to shared video-query context while eliminating cross-box dependencies, together with localization-aware policy optimization for temporal boundaries and spatial geometry. On VidSTG, PTD reduces Tube Completion Latency by 79x and increases spatial decoding throughput by 92x over standard autoregressive decoding, while also improving grounding accuracy. With a compact 4B backbone, our model performs favorably well on VidSTG and HC-STVG, and generalizes zero-shot to temporal grounding, grounded VideoQA, and referring video object tracking. Our results show parallel tube generation is an efficient and effective alternative to autoregressive localization in videos.
[CV-32] EXPOSE: Explainable and Domain-Robust Embeddings from Pathology Vision Foundation Models using Sparse Autoencoders
链接: https://arxiv.org/abs/2608.28191
作者: Anja Witte,Maximilian Lennartz,Jan Baumbach,Guido Sauter,Stefan Bonn,Patrick Fuhlert,Marina Zimmermann
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Vision Foundation Models (VFMs) are widely used in computational pathology but remain sensitive to domain shifts arising from variations in staining, tissue preparation, and scanner hardware. A key limitation is that VFM embeddings entangle biological with domain-specific information, hindering cross-domain generalization. We propose Explainable Probing of Cross-Domain Sparse Embeddings (EXPOSE), a framework that uses Sparse Autoencoders (SAEs) as an explainable bottleneck to identify and suppress domain-specific components in VFM embeddings. We train a sparse representation of VFM features, use a linear classifier to identify domain-specific latent dimensions, and mask these features prior to downstream relapse prediction without retraining the backbone model. Experiments on a large prostate cancer dataset with multiple acquisition domains show that SAE features capture both domain- and task-specific information, which are partially disentangled in the latent space. Removing domain-specific features improves cross-domain performance and increases embedding robustness as measured by the Domain Robustness Index (DoRI). Code is available at this https URL .
[CV-33] Memory-efficient GPU pipelines for real-time non-line-of-sight reconstruction
链接: https://arxiv.org/abs/2608.28183
作者: Alfonso López-Ruiz,Diego Royo
类目: Distributed, Parallel, and Cluster Computing (cs.DC); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Non-line-of-sight (NLOS) imaging reconstructs scenes hidden around a corner from indirect light recorded by a single-photon avalanche diode (SPAD). A single reconstruction is a large inverse problem: billions of photon timestamps must be binned, moved through memory, transformed and inverted. As SPAD arrays raise acquisition throughput, reconstruction becomes the limiting stage. We rebuild the GPU execution of two established wave-based algorithms, f-k migration and phasor-fields, for both streaming and offline processing. On the phasor-fields side we assemble the ring-and-radius kernels of previous work once and offline, using the analytic Fourier transform of a ring, so the propagation kernel never exists in dense form at runtime, reducing the memory and bandwidth. We reorganize the pipeline of both algorithms with fused kernels, warp-level photon binning, batched transforms, CUDA graph replay, and FP16 storage applied only where it reduces the actual bottleneck. Our implementations are up to 42x faster than the reference streaming pipeline and up to 14x faster than the fastest published GPU baseline, all while using a fraction of the memory (down to 2.5%), enabling vastly larger and finer reconstructions on the same hardware, or comparable ones within a much lower memory budget. We report an ablation of each implementation choice and propose three denoising strategies enabled by the resulting frame budget for next-generation NLOS video processing.
[CV-34] Manifold4D: Denoising on Point Cloud Rendered Manifolds for Video Re-shooting
链接: https://arxiv.org/abs/2608.28174
作者: Yongqi Mao,Zijia Dai,Zhishuo Liu,Wei Xu,Kaiwei Wang,Guotao Meng
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 20 pages, 9 figures, 10 tables. Project page: this https URL
Abstract:Video re-shooting re-renders a monocular video of a dynamic scene along a user-specified camera trajectory, and the dominant recipe supplies the target geometry explicitly: per-frame depth lifts the source video into a 4D point cloud, which is rasterized along the trajectory into a point cloud render. Because the render and the source video are both handed to the network as visual conditions, they compete at every denoising step, leaving the model with a trust dilemma — how much of the render to believe — which can degrade trajectory control or visual quality on data outside the training distribution. We argue that a render already pixel-aligned with the target view does not need to be supplied as an explicit conditioning stream at all. We propose MANIFOLD4D, which injects the render directly into the initial noise of flow matching, so that generation no longer departs from standard Gaussian noise but from a new noise manifold carrying geometric information, leaving the source video as the only visual condition. The render is thus used exactly once, and the network is never asked to learn how to read it; in subsequent denoising steps the model can focus on the source video. On our DAVIS-Traj benchmark and on the Vista4D evaluation set, MANIFOLD4D attains the best camera-control accuracy on every metric, lowering rotation error by 25% and 27% and translation error by up to 32% over the strongest baseline, while matching it in video fidelity and leading on real-world novel-view photometric quality. In a user study, our method achieves clear advantages in trajectory following and dynamic consistency. The gap widens as the yaw amplitude grows past the training range, and the model still recovers correct dynamic motion from the source video when the render is deliberately corrupted, confirming that the geometric prior guides generation without overriding it.
[CV-35] Empowering Local Agriculture: A Deep Learning-Powered Web System for Identifying Bangladeshi Mango Varieties
链接: https://arxiv.org/abs/2608.28161
作者: Monowar Islam,Safaruzzaman Shovo
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: Accept in Journal of Bangladesh Academy of Sciences, Volume 50, Supplement 1, April 2026. 2 authors
Abstract:Mango variety identification in Bangladesh is challenging because closely related cultivars can have similar visual characteristics and images are often captured under varying real-world conditions. This work presents a deep learning-based web system for automatic identification of Bangladeshi mango varieties. We collected 2,013 high-quality mango images (3024x4032 pixels) from local markets and farms and organized them into nine classes, combining Bari-4 and Bari-7 as a single Bari class. The dataset was divided into training (70%), validation (15%), and test (15%) sets, with image augmentation applied to improve model generalization. Three pretrained CNN architectures, ResNet18, ResNet50, and EfficientNetB0, were fine-tuned under consistent training settings. EfficientNetB0 achieved the best performance, obtaining 98.01% validation accuracy and 97.36% test accuracy, compared with 86.47% and 78.55% test accuracy for ResNet18 and ResNet50, respectively. Class-wise F1-scores for EfficientNetB0 ranged from 0.93 to 0.99, while the Bari class achieved an F1-score of 0.97. The selected EfficientNetB0 model has approximately 4 million parameters, making it suitable for lightweight deployment. We integrated the model into a Streamlit web application that enables users to upload a mango image and receive a predicted variety with class probabilities. The system provides an accessible, practical tool for mango identification and demonstrates the potential of deep learning for supporting agricultural applications in Bangladesh.
[CV-36] Dual-Stream Semantic Guidance with Prototype Anchor Calibration for Source-Fully-Free Adaptation of Vision-Language Models
链接: https://arxiv.org/abs/2608.28145
作者: Weiwei Xiang,Shun Peng,Guangyi Xiao,Hao Chen,Lei Yang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Source-Fully-Free Domain Adaptation (SFF-DA) has emerged as a strategic paradigm to adapt Vision-Language Models (VLMs) without any access to source data or task-specific source models. However, we identify a critical Dual Semantic Drift that hinders this process: static drift arising from the rigidity of fixed class embeddings, and dynamic drift stemming from the divergence of generated captions, causing severe semantic misalignment that intensifies the stability-plasticity dilemma. To address this, we propose DSSG (Dual-Stream Semantic Guidance), an end-to-end framework that reconciles fine-grained plasticity with global stability. Our core contribution is the Dual Semantic Guidance (DSG) module, which integrates a caption stream for domain-specific knowledge with a class-anchor stream to anchor global categorical consistency. Furthermore, a Dynamic Cross-Modal Knowledge Distillation (CMKD) module is introduced to leverage the evolving teacher distribution for calibrating teacher-student consistency. Building upon DSSG, we further introduce Prototype Anchor Calibration (PAC), yielding DSSG-PAC, which periodically calibrates prototype anchors and caches them until the next calibration. This design reduces redundant text-side computation while preserving the adaptability of class guidance to the evolving text space. We further establish SFF-DA risk bounds that relate student risk to semantic-teacher quality and teacher–student discrepancy. Extensive experiments demonstrate that DSSG consistently outperforms current state-of-the-art methods across multiple benchmarks, while DSSG-PAC largely preserves its adaptation performance with 18.9% lower total adaptation time. The code is available at this https URL.
[CV-37] oken-Budget Distillation: Transferring Full-Token Semantics to Compressed Video Vision-Language Models ACM-MM2026
链接: https://arxiv.org/abs/2608.28138
作者: Xiaoyang Guo,Guoping Luo,Jusheng Zhang,Keze Wang,Wenhao Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: ACM MM 2026
Abstract:Adapting video vision-language models (VLMs) is computationally expensive because video inputs produce a large number of visual tokens, making both fine-tuning and inference costly. Although visual token compression can reduce this overhead, direct adaptation on compressed inputs often causes semantic drift and noticeable performance degradation. We present Token-Budget Distillation (TBD), a parameter-efficient fine-tuning framework for adapting video VLMs under a fixed token budget. TBD freezes the pretrained backbone, updates only LoRA adapters, and integrates FlashVID-based visual token compression into the video pathway. To preserve full-token semantics under compression, TBD employs a dual-path teacher-student design, where a full-token teacher provides stable supervision and a compressed student is optimized with task loss, answer-region KL distillation, GT-anchored margin distillation, and reliability-aware KD control. This design enables the student to recover the semantic behavior of the full-token model while remaining efficient under aggressive token reduction. We evaluate TBD on three video VLM backbones, including LLaVA-Video, LLaVA-OneVision, and Qwen3-VL-8B-Instruct, across four video understanding benchmarks. TBD consistently outperforms compression-only baselines under both moderate and aggressive compression. On LLaVA-Video at retention ratio R = 10 percent, TBD preserves 97.0 percent of the Vanilla model’s average accuracy; on LLaVA-OneVision at R = 10 percent, it achieves an average score of 58.4 and matches 100.0 percent relative accuracy.
[CV-38] DeicticVLA: Unifying Instruction Modes Based on Language and Deictic Gestures in a Single VLA
链接: https://arxiv.org/abs/2608.28108
作者: Kango Yanagida,Tatsuya Aoki,Yuichiro Yoshikawa,Takato Horii
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Vision-Language-Action models (VLAs) allow users to specify manipulation tasks in natural language, but distinguishing a target or placement goal among objects of the same category or similar appearance requires detailed expressions that VLAs may not use reliably. We propose DeicticVLA, which canonicalizes Language Instruction (LI), Vision-Language Instruction (VLI), and Visual Instruction (VI) into a text prompt and deictic masks through text-prompt completion and deictic gesture grounding, enabling a single pretrained VLA to handle all three instruction modes. With a shared backbone, demonstrations, and matched training steps, we compare two RGB visual prompting methods, two separate-channel mask prompting methods, and three training strategies in simulation. Under two-stage training, the four prompting methods achieve high in-distribution success but differ in their ability to use deictic masks in unseen layouts. Across methods, training-strategy ablations show that two-stage training improves such use, while retaining second-stage LI data mitigates forgetting without reducing VLI and VI performance. In three real-world tasks, one policy supports all modes. VLI and VI outperform LI under unseen expressions, appearance changes, and novel objects. For unseen categories, both achieve 100% success, compared with 16.7% for jointly trained LI. These results demonstrate the unified three-mode interface and guide DeicticVLA design.
[CV-39] What Will This Copper Look Like Later? Forecasting Surface Appearance and Rendering It as a PBR Material
链接: https://arxiv.org/abs/2608.28102
作者: Teejuta Sriwaranon,Borworntat Dendumrongkul,Tanapat Chamted,Pizzanu Kanongchaiyos
类目: Graphics (cs.GR); Computer Vision and Pattern Recognition (cs.CV)
备注: 10 pages, 5 figures, 3 tables. (Repository this https URL )
Abstract:Digital design requires predicting how a metal surface will look later in its oxidation; this paper presents such a pipeline for copper. Given a fixed-camera observation, the system forecasts appearance 10 accelerated units ahead and converts it into the albedo, normal, roughness and metallic maps a renderer consumes. Forecasting is evaluated as an authoring tool would use it, on a copper specimen the system has not observed: an entire recording is held out, so training and checkpoint selection use one specimen and the test set is the whole of a second, recorded on a different day and condition. Under this protocol a learned spatio-temporal model with a monotone oxidation state, the most accurate forecaster within a single recording, is less accurate than copying the last observed frame on an unseen specimen, in both directions, as are three further trained architectures. The only forecaster that transfers is a closed-form global color extrapolation with no trained parameters, improving on copy-last-frame by 13.4% and 50.6%, with a margin that increases with horizon to +16.7% and +55.5% at t+10. Two controls qualify this: correcting every frame for the photometric drift measured on a non-oxidizing reference region leaves both margins intact, ruling out uncontrolled exposure as their source, and a moving-block bootstrap over the 6 independent windows each recording contains separates the larger margin from zero but leaves the smaller one not individually significant. The mechanism is measured: a learned susceptibility map encodes where corrosion begins on the training specimen and misleads on a new one, whereas the global color trajectory is what specimens share. The pipeline therefore deploys the closed-form forecaster for unseen specimens and the learned model only for continuing one already observed. Code, splits, protocol and leakage audit are released.
[CV-40] Ex-Sim(3)-Reg: 2D-3D Correspondence Pruning via Extended Sim(3) Registration ECCV2026
链接: https://arxiv.org/abs/2608.28096
作者: Pei An,Muyao Peng,Junfeng Ding,Jiaqi Yang,Liangliang Nan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ECCV 2026
Abstract:Learning-based image-to-point-cloud (I2P) registration has garnered increasing attention in recent years. Nevertheless, existing methods still struggle with severe outliers under challenging scenarios with unseen, low-inlier, or distorted cases. A fast and robust 2D-3D correspondence pruning method is therefore highly desirable. Recently, a promising scheme lifts 2D-3D correspondences to 3D-3D correspondences using depth priors, casting correspondence pruning as a Sim(3) registration problem. However, depth priors estimated from monocular images are inherently noisy, which undermines the reliability of this scheme. In this paper, to explicitly model non-negligible depth noise, we reformulate correspondence pruning as an extended Sim(3) registration problem and propose a simple yet effective pruning algorithm termed Ex-Sim(3)-Reg. We further provide a theoretical analysis to justify the effectiveness of our method. Extensive experiments on the 7-Scenes, RGBD-V2, ScanNet, and TUM datasets demonstrate that Ex-Sim(3)-Reg achieves up to \textbf24.7% improvement in registration recall over state-of-the-art baseline methods. Code is released at this http URL
[CV-41] Ada-TokenCom: Rate-Adaptive Token Communications via Large-Model-Driven Token Compression and Generation
链接: https://arxiv.org/abs/2608.28086
作者: Zijun Zhang,Li Qiao,Mahdi Boloursaz Mashhadi,Zhen Gao,Mehdi Bennis,Kaibin Huang
类目: Information Theory (cs.IT); Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
备注:
Abstract:Token Communications (TokenCom) has recently emerged as a new paradigm in which tokens serve as unified units for communication and computation, enabling efficient multimodal semantic and goal-oriented transmission. In this paper, we develop Ada-TokenCom, a rate-adaptive TokenCom framework based on large autoregressive models, which integrates next-token prediction with arithmetic coding to achieve ultra-low bitrate semantic communication at the token level. We propose a mixed reconstruction/generation scheme, where the transmitter encodes and transmits the highly informative tokens at the beginning of the token sequence leveraging a pre-trained autoregressive large model, while the receiver uses an identical model to predict the rest. Moreover, we design a Lyapunov-based algorithm to dynamically optimize both the source compression rate and the modulation and coding scheme, adapting to time-varying network conditions. Simulation results demonstrate that our proposed Ada-TokenCom framework outperforms both digital and deep joint source-channel coding-based semantic communication baselines.
[CV-42] Attribute Token Arithmetic: Disentangled and Continuous Semantic Control for Visual Autoregressive Models
链接: https://arxiv.org/abs/2608.28082
作者: Xindi Yang,Yicheng Wu,Cheng Zhang,Jianfei Cai,Tien-Tsin Wong
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Autoregressive text-to-image generation has recently achieved remarkable progress, offering high-fidelity synthesis via a unified generative framework. However, fine-grained semantic control remains challenging due to the attribute entanglement and the misalignment between textual and fine-grained visual representations. In this paper, we introduce Attribute Token Arithmetic (ATA), a method that enables disentangled and continuous attribute control in visual autoregressive modelling. Inspired by the vector arithmetic property observed in word embeddings, ATA identifies semantic directions corresponding to visual attributes (e.g., aging, fatness, emotion) directly within the pretrained autoregressive latent space. These directions are learned from a single reference image, without model retraining or large-scale supervision. During generation, attributes can be continuously adjusted and compositionally combined through simple arithmetic operations with other attribute tokens. Extensive experiments demonstrate that ATA achieves identity-preserving, fine-grained, and multi-attribute adjustment, outperforming existing autoregressive editing baselines in controllability, generality, and computational efficiency. Our code will be available at this https URL.
[CV-43] Cyc3D: Evaluating Cyclic Structural Stability and Asset Usability in Image-to-3D Generation
链接: https://arxiv.org/abs/2608.28080
作者: Liwen Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Image-conditioned 3D generation has advanced rapidly, yet existing evaluation protocols largely judge rendered-view plausibility and semantic alignment, overlooking whether a generator forms a stable 3D interpretation and produces assets usable in graphics pipelines. We introduce Cyc3D, a multidimensional benchmark that evaluates image-to-3D generation along two complementary axes: Cross-View Object Consistency and Representation Quality. At the asset level, Cyc3D measures whether object identity remains semantically coherent across rendered viewpoints. At the model level, we propose View-Cycle Structural Consistency, a closed-loop render-regenerate-align protocol that repeatedly re-observes a generated asset from novel views and quantifies geometric, perceptual, and semantic drift across generations. To assess native asset usability beyond rendered appearance, Cyc3D further evaluates geometric structure, reference-image fidelity, mesh discretization and efficiency, and UV parameterization quality. Together, these diagnostics expose failures obscured by a single perceptual score and provide interpretable evidence of both model instability and representation defects. Experiments on five representative image-to-3D systems show that closed-source feed-forward models consistently outperform open-source optimization-based baselines in geometric fidelity, mesh quality, and cycle stability. Nevertheless, even the strongest methods achieve cycle-stability scores below 48, revealing a persistent gap between visually plausible generation and robust 3D object understanding.
[CV-44] ask-State Adaptation with Prototype Memory for Multi-Task Dense Prediction
链接: https://arxiv.org/abs/2608.28078
作者: Yangyang Xu,Haobo Yuan,Yuzhu Wang,Duo Su,Xi Ye,Yibo Yang,Jun Zhu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Preprint
Abstract:Vision foundation backbones provide strong representations for dense prediction, yet a single shared feature still needs to support tasks with different, image-dependent adaptation requirements. We propose MemMTL, a multi-task dense prediction framework that estimates a compact task state from global visual context and refines it through a learnable task-state prototype memory. The refined state is converted into task-conditioned expert logits and combined with token-level logits before sparse top- k selection over a local expert bank shared by all tasks. A separate task-agnostic residual bank provides a common adaptation path, and both paths are added once to the backbone feature before task-specific prediction. We specify a matched evaluation protocol on NYUD-v2 and PASCAL-Context with SAM 3 and ViT-L backbones to measure predictive quality, computational cost, and the contributions of task-state conditioning, prototype retrieval, and sparse routing. The numerical record in the present working draft predates this canonical implementation and must be regenerated before it can support empirical claims.
[CV-45] CF-YOLO: Context-Aware Feature Refinement for Camouflaged Industrial Micro-Defect Detection
链接: https://arxiv.org/abs/2608.28070
作者: Xinda Yu,Kunxin Zheng,Chunan Yu,Qingbo Song,Hao Xiao,Ying Zang,Jie Liu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Automated detection of surface micro-defects on industrial components, such as copper tubes, is critically important for quality assurance but remains challenging due to the minute scale of anomalies and their visual camouflage against complex backgrounds. These factors lead to weak feature representations and high rates of false positives and missed detections. To address these issues, we propose a novel real-time detection framework designed for efficient context perception and feature refinement. Our method integrates a Context-Perception Aggregation Module (CPAM), which synergises large-kernel perception for macro-texture context and small-kernel aggregation for sharp boundary delineation, effectively breaking the background camouflage. Furthermore, a Feature Additive Refinement Module (FARM) employs a linear-complexity additive token mixer to globally verify and refine the representation of fine-grained anomalies, suppressing noise-induced errors. To support research in this domain, we introduce the Copper Tube Defect Dataset (CTDD), a manually annotated benchmark containing 1,847 images and 4,898 boundingbox defect instances from copper-tube inspection scenarios. Extensive experiments demonstrate that our detector achieves strong and consistent performance on CTDD, outperforming representative baseline detectors, including YOLOv11, by 2.2% in mAP@50 and 3.9% in Precision while maintaining real-time inference speed. This work provides a robust and efficient solution for high-precision industrial inspection, bridging the gap between contextual understanding and detailed feature analysis. Our code and model are available at: this https URL
[CV-46] VersaGauss: A Versatile Framework for Generating Multiphase Dynamics with 3D Gaussians
链接: https://arxiv.org/abs/2608.28069
作者: Ruijie Su,Lingxiao Yang,Xiaohua Xie,Jianhuang Lai
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Recent progress has been made in 3D Gaussian representation for reconstruction, generation, and physical simulation. However, current approaches mainly concentrate on physics-based dynamic generation of solid objects and only handle single-phase collision interactions. We introduce VersaGauss, a unified framework for generation, simulation, and rendering that supports versatile physics-based dynamic generation, particularly for multiphase interactions. Our system takes a few images as input and produces a realistic, physics-driven 3D dynamic scene with multiple objects. To optimize the Gaussian kernel distribution, we develop a particle pruning algorithm. We also propose the Coupled Multiphase Point Method (CMPM) to effectively model and generate multiphase interactions. Additionally, harmonic interpolation within CMPM and a Gaussian evolution strategy are introduced to achieve realistic fluid rendering. Extensive experiments demonstrate that our framework can simulate interactions among various materials such as fluid, rubber, sand, snow, and others. Code is available at this https URL.
[CV-47] A Controlled Audit of Architectural Complexity in Uncertainty-Aware Multi-Organ Ultrasound Classification
链接: https://arxiv.org/abs/2608.28063
作者: Yang Song,Pengbo Sun,Shichang Feng,Ye Zhu,Xin Xu,Ziran Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Supplementary material is available as an ancillary file on this arXiv page
Abstract:Multi-organ ultrasound classifiers increasingly combine attention, mixture-of-experts routing, uncertainty gating, and evidential deep learning (EDL) objectives to address heterogeneous anatomy and acquisition. Yet a plausible design rationale does not by itself establish that an added component improves the trained system. We contribute a controlled complexity-audit framework, applied to the deployment decision between the maximal evidential candidate Full-EDL and simpler alternatives. Six candidates were evaluated on the primary dataset and three in an internal replication, using ten matched seeds, frozen image-level partitions, capacity- and optimisation-aware comparisons, symmetric temperature scaling, paired decision rules, and a separate out-of-distribution (OOD) veto. Retaining Full-EDL did not establish a reliable macro-F1 gain on either dataset, while the simplified alternatives remained inconclusive under the non-inferiority margin. Simple cross-entropy with temperature scaling (Simple-CE+TS) met the calibrated negative log-likelihood criterion on both datasets and showed favourable selective-risk ordering. The raw calibration advantage of evidential training disappeared after temperature scaling and did not recur on the second dataset. The gate had negligible observable influence at the audited checkpoints, and deleting the Full-only chain revealed no stable task or calibrated-loss benefit. Simple-CE nevertheless triggered the OOD veto against the fetal probe but not the lung probe, precluding an unconditional OOD-safety claim. We therefore selected Simple-CE+TS for the evaluated in-distribution objective while retaining Full-EDL as the maximal reference. Components should earn retention through functional and retraining-based evidence, and calibration and distribution-shift reliability should be evaluated separately.
[CV-48] Dynamic Alignment Compensation for Hallucination Mitigation in Large Vision-Language Models EMNLP2026
链接: https://arxiv.org/abs/2608.28058
作者: Kairong Yu,Zixin Zhu,Le Yu,Hongwei Wang
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted by EMNLP2026 Findings
Abstract:Large Vision-Language Models (LVLMs) remain prone to hallucinations, producing responses that are irrelevant or inconsistent with the multimodal input. Existing mitigation methods mainly rely on external supervision, output calibration, or attention regulation, leaving the internal representation dynamics of autoregressive generation underexplored. We identify an inference-time failure mode in which cross-modal representations degrade across decoder layers and drift across generation steps, destabilizing token prediction and increasing hallucination risk. We propose \emphDynamic Alignment Compensation (DAC), a training-free inference-time method that detects representation divergence and selectively applies lightweight residual compensation. DAC combines Layer-wise Semantic Compensation to mitigate inter-layer degradation with Sequential Semantic Correction to constrain temporal drift. Experiments on nine hallucination-focused and general-purpose multimodal benchmarks across multiple LVLM backbones show that DAC consistently reduces hallucinations while maintaining strong overall performance.
[CV-49] ZipMVS: Multi-View Stereo with Compressed Cost Volumes
链接: https://arxiv.org/abs/2608.28033
作者: Guanglin Jin,Hongshan Yu,Javier Civera,Zhaoxin Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 14 pages, 8 figures
Abstract:Multi-view stereo (MVS) methods typically deliver highly accurate 3D reconstructions from multiple registered RGB images, thanks to the highly informative, geometric constraints between them. However, their substantial memory requirements remain a major obstacle for deployment in domains such as aerospace and autonomous systems, where resource efficiency is critical. In this work, we introduce ZipMVS, an MVS method specifically designed for efficient high-quality reconstruction. We propose a novel depth-hypothesis strategy that enables substantial compression of the cost volume, hence greatly reducing GPU memory consumption while preserving reconstruction accuracy. Experiments on the DTU and Tanks and Temples datasets show that ZipMVS achieves competitive reconstruction quality compared with other efficiency-oriented MVS methods, while achieving a competitive balance between reconstruction quality and GPU memory usage. The code is available at this https URL
[CV-50] 3D-USE: From Image-Level to Scene-Level Underwater Enhancement
链接: https://arxiv.org/abs/2608.28020
作者: Jieyu Yuan,Yuanlin Zhang,Jihong Li,Chunle Guo,Huimin Lu,Chongyi Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Project page: this https URL
Abstract:Underwater 3D reconstruction faithfully reproduces the color shifts and visibility loss of captured views, while physical inversion may leave estimation errors in the recovered scene appearance. We formulate Underwater Scene-level Enhancement (USE) as learning a persistent, visibility-enhanced 3D scene representation from degraded multi-view underwater observations, enabling consistent enhanced rendering. Realizing USE requires both a reliable scene representation for enhancement and a consistent enhancement target without paired enhanced 3D data. Therefore, we present 3D-USE, a two-stage framework. First, the Medium Radial Basis Anchor Representation (MediumRBF) establishes a medium-aware Gaussian scene by representing water effects with shared radial-basis anchors and explicitly decomposing object and medium contributions. Based on this fixed scene representation, Appearance Transition Consensus (ATC) transfers paired 2D underwater image enhancement (UIE) knowledge into scene-global and Gaussian-local targets, avoiding direct supervision from inconsistent enhanced views. An Underwater Bilateral Appearance Field (U-BAF) then realizes these targets in Gaussian radiance and medium appearance. The scene directly renders enhanced novel views without a 2D UIE model at inference. Experiments on real underwater scenes show improved visibility and cross-view consistency while preserving reconstruction quality.
[CV-51] Visual Token Coding for Video Multimodal Large Language Models
链接: https://arxiv.org/abs/2608.28008
作者: Chenxin Fang,Tao Chen,JunChao You,Jun Peng,Yiyi Zhou,Rongrong Ji
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 9 pages, 4 figures
Abstract:In this paper, we propose a new token compression paradigm for video Multimodal Large Language Models (MLLMs), termed Visual Token Coding (VTC). Inspired by classical video coding principles, e.g., HEVC, VTC performs structured compression by predicting the I/P frames of a video and measuring their frame-wise residuals to estimate token redundancy. Based on this baseline framework, we also enhance VTC with a set of novel dynamic designs, such as Dynamic Resolution Input (DyRSO), Dynamic Token Allocation (DyTA), and Spatial Coverage Top-K (SC-TopK), and term this new approach VTC_Dy . To validate VTC, we apply it to three MLLMs and conduct experiments on multiple video understanding benchmarks. The experimental results show that VTC _\mathrmDy achieves an average performance retention of 100.1% with a 50% token budget for Qwen3-VL, while still retaining 97.8% of the average performance when the token budget is reduced to 25%. Moreover, as a plug-and-play design, VTC requires no additional tuning of MLLMs for token coding. Our code is available at this https URL.
[CV-52] A-PAIR: A Benchmark and Identity-Consistent Grounding Framework for Air-Ground Cross-View Referring Person Detection
链接: https://arxiv.org/abs/2608.27997
作者: Zhoupeng Guo,Xinjie Yao,Yunqi Zhu,Zhihe Fan,Siqi Zhao,Jianjun Chen,Yichen Dong,Yan Fan,Pengfei Zhu
类目: Computer Vision and Pattern Recognition (cs.CV); Multimedia (cs.MM)
备注:
Abstract:Air-ground cross-view referring person detection is a necessary component in the language-to-perception-to-control chain of collective embodied intelligence, grounding a language command into the same physical target before ground and aerial agents can coordinate downstream actions. Existing referring expression comprehension and open-vocabulary grounding methods do not jointly account for cross-view identity consistency, making them insufficient for Air-Ground Cross-View Referring Person Detection (AGCV-RPD), which involves similar pedestrian distractors, weak aerial appearance cues, and cross-view identity consistency. To study this problem, we introduce Air-Ground Paired Identity-Aware Referring (A-PAIR), the first comprehensive AGCV-RPD benchmark, containing 22,137 cross-view referring samples. To construct A-PAIR efficiently, we propose Factorized Annotation and Referential Alignment (FARA), a semi-automatic annotation framework that generates factorized referring descriptions and identity-consistency supervision at reduced cost. We propose Identity-Consistent Referring Grounding (ICRG), a framework that combines factorized referential grounding, candidate-completeness supervision, and cross-view consistency calibration for joint air-ground pair selection. ICRG improves ground, aerial, and pair-level detection over strong baselines, increasing pair F1 from 16.65% to 22.28%. These results show that AGCV-RPD requires paired detection and identity-consistent reasoning.
[CV-53] GAN-Based Semantic Communication for Image Transmission in IoV
链接: https://arxiv.org/abs/2608.27989
作者: Ruixing Ren,Shan Chen,Junhui Zhao,Xiaoke Sun
类目: Computer Vision and Pattern Recognition (cs.CV); Systems and Control (eess.SY)
备注: 8 pages, 7 figures
Abstract:For cooperative perception in the internet of vehicles, this paper proposes a generative adversarial network-based semantic communication framework to address the efficiency and fidelity bottlenecks of traditional communication systems in visual data transmission under limited bandwidth and dynamic channel conditions. At the transmitter, the framework adopts a pyramid attention network to extract semantic label maps and introduces a semantic priority preservation mechanism. It assigns differentiated weights to distinct semantic categories based on driving safety, guiding bit allocation and loss function design. At the receiver, an image reconstruction module integrating a coarse to-fine multi-resolution generator and multi-scale discriminator is designed. Combined with the temporal consistency branch, spatial pyramid pooling and class-aware convolutional layers, it achieves high-fidelity reconstruction of high-quality images from corrupted semantic labels. The model is trained with combined adversarial, feature matching and perceptual losses, effectively improving semantic consistency and visual realism of generated images. Experimental results on the Cityscapes dataset show that the proposed method outperforms existing counterparts in both semantic segmentation accuracy and reconstructed image quality, and maintains stable reconstruction performance under AWGN and Rayleigh channels.
[CV-54] GAAT: Geometry-Aware Alignment Transformer for Multimodal UAV Perception
链接: https://arxiv.org/abs/2608.27971
作者: Jingpu Yang,Debin Tang,Yilin Sun,Fengxian Ji,Jiahua Zhu,Wenrui Ding,Yufeng Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Unmanned aerial vehicle (UAV) multimodal perception integrates visible (RGB), infrared (IR), synthetic aperture radar (SAR), and depth sensors for scene understanding under diverse conditions. However, differences in optics, resolution, and mounting often limit practical systems to global or image-center alignment. After tokenization, parallax, platform motion, and lens distortion can shift corresponding patch centers across modalities, weakening the spatial correspondence assumed by dense contrastive learning and cross-modal fusion. We propose GAAT (Geometry-Aware Alignment Transformer), an alignment-first pretrained model that estimates local correspondence reliability before cross-modal interaction. GAAT introduces syncPATC, which learns patch-center consistency under synchronized view transformations without correspondence annotations. It emits geometric priors, including token and query confidence, query centers, and sub-token offsets, that identify reliable local anchors across residual misalignment. Guided by these priors, MG-Sparse-MMA performs query-mediated sparse fusion over top-K_s reliable regions, replacing dense all-patch interaction with geometry-calibrated local updates. RA-QCGCL aligns pretraining supervision with this sparse query bottleneck through reliable patch-to-patch, patch-to-query, and query-to-query contrastive branches. We introduce UAVMeta and StateBench, which provide four acquisition-state scores derived from platform telemetry and image statistics: camera reliability, observation scale, viewpoint stability, and flight maneuver complexity. Extensive experiments across six downstream tasks demonstrate consistently superior transfer performance, establishing GAAT as a state-of-the-art multimodal foundation model for UAV perception. StateBench further enables a systematic diagnosis of real-world acquisition conditions.
[CV-55] raining-Free Temporal Abstraction for General Video Understanding NEURIPS2026
链接: https://arxiv.org/abs/2608.27929
作者: Etienne Casanova,Sevan Brodjian,Pietro Perona
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Submitted to NeurIPS 2026
Abstract:Videos are expensive to analyze frame by frame, yet many video understanding tasks depend on knowing where relevant moments occur. A system may need to find when an action changes, locate the segment described by a sentence, or choose a few frames for a vision-language model. Existing methods often solve these problems separately, using task-specific training data or specialized architectures. We study whether a pretrained video-text model can provide enough temporal structure to support several of these tasks at once. We present STITCH, a training-free method that divides a video into semantically meaningful temporal chunks. STITCH embeds short video windows with a frozen video-text backbone and detects changes in the resulting embedding sequence. These chunks are computed once per video and reused across tasks. We evaluate STITCH on generic event boundary detection, language-based moment retrieval, and frame selection for long-video VLM reasoning. Across all three settings, STITCH remains competitive with more specialized methods while requiring no task-specific training, with especially clear gains when only a small number of frames or tokens can be processed. These results suggest that reusable temporal abstraction is a promising direction for general video understanding, allowing dense video streams to be converted once into semantic units that can be localized, retrieved, sampled, or reasoned over by downstream systems.
[CV-56] PCBnet: A Dataset and Automatic Construction of SPICE Netlists from Schematic Images
链接: https://arxiv.org/abs/2608.27923
作者: Zhen Huang,Yuhao Gao,Yuzhi Liu,Daian Cheng,Chengyuan Shao,Yucheng Chen,Yongjian Jia,Futing Zhang,Yichen Shi,Wenhao Wang,Zuyan He,Yangbo Wei,Zhanfei Chen,Jinlong Yan,Yu Zhang,Haoying Wu,Ting-Jung Lin,Lei He
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted at the 2026 IEEE International Conference on LLM-Aided Design (ICLAD 2026)
Abstract:Printed circuit boards (PCBs) are fundamental to modern electronic systems, yet AI-driven PCB design automation remains constrained by the lack of large-scale paired schematic-netlist datasets. PCB schematics are particularly challenging due to diverse component types, complex wiring topologies, and noisy textual annotations. To address this gap, we present PCBnet, a large-scale PCB schematic dataset comprising over 300 real-world designs with annotated pins and paired SPICE netlists. It contains more than 50,000 component instances, 150,000 wires, 100,000 text regions, and 400,000 characters. We further develop an automated schematic-to-netlist pipeline that combines visual recognition, topology construction, and domain-knowledge-guided multi-agent correction. The proposed method achieves 94.54% component detection mAP, 98.57% text recognition accuracy, and 84.47% end-to-end connectivity accuracy. PCBnet provides a benchmark and data foundation for future AI-driven PCB design automation.
[CV-57] DensityKV: Density-Guided KV Cache Compression for Long Video Generation
链接: https://arxiv.org/abs/2608.27922
作者: Wenqu Zhao,Xuemin Chi,Xin Zhang,Guoqing Ma,Baorun Li,Jianjie Fang,Peizhi Tang,Chen Gao,Wei Wu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 17 pages, 9 figures, 2 tables. Code: this https URL
Abstract:Autoregressive video diffusion models enable streaming generation through sliding-window attention, but each generated block is conditioned on previously generated content, causing appearance and motion errors to propagate recursively over time. Historical key-value (KV) memory preserves earlier subject and scene states and helps maintain long-horizon consistency. However, retaining every generated state creates a historical archive that grows continuously with the rollout, while recurrent states repeatedly add redundant coverage. To address this problem, we propose DensityKV, a training-free historical KV bank management strategy. DensityKV maintains a separate token-level KV bank for each attention head and measures local redundancy among the post-RoPE keys that directly parameterize attention routing using Soft-Riesz density. By constraining neighborhood-density growth after states enter the bank, DensityKV limits repeated historical accumulation while preserving coherent states from each completed generation block. Experiments across three autoregressive video generation backbones and multiple generation lengths show that, at the same upper bound on historical KV capacity, DensityKV improves long-horizon consistency and generation stability while keeping persistent historical storage bounded independently of rollout length.
[CV-58] CommerceVibe: Learning to Design E-Commerce Creatives as Executable Visual Code via Dual-Feedback Reinforcement Learning
链接: https://arxiv.org/abs/2608.27893
作者: Yajiao Xu,Jin Zhang,Jiangbo Ai,Tao Jiang,Mo Xu,Lina Huang,Chengfu Huo
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 20 pages, 13 figures, 12 tables; includes supplementary material
Abstract:High-quality e-commerce creatives are essential for presenting products and conveying marketing messages. Recent diffusion models enable scalable creative generation and produce visually compelling images, but their flattened raster outputs often contain distorted text and inconsistent product details, requiring refinement before deployment. Moreover, without explicit structure, the resulting creatives are difficult to edit and reuse, while complex design requirements remain challenging to encode as verifiable training signals. To address these challenges, we present CommerceVibe, which represents creatives as executable visual code and formulates generation as conditional HTML/CSS program synthesis. Given product images, design requirements, and product information, it produces renderable, editable, and reusable creatives. We further introduce dual-feedback reinforcement learning, in which rule-based feedback evaluates rendered programs for text readability, product visibility, and layout validity, while visual feedback from a vision-language model (VLM) assesses rendered creatives against input specifications across six perceptual and commercial dimensions. Together, these complementary feedback signals improve both constraint satisfaction and perception-dependent quality. We perform supervised fine-tuning (SFT) of Qwen3.5-9B on over 28,000 e-commerce examples, followed by dual-feedback reinforcement learning. On a 1,300-case benchmark, the optimized CommerceVibe model achieves a weighted score of 94.0/100, compared with 87.3 for the SFT-only variant, and outperforms strong external models. Blind evaluations by five e-commerce design experts further validate these improvements. CommerceVibe supports controllable, editable, and scalable e-commerce creative production.
[CV-59] hread-Efficient Decoding for Neural Texture Compression
链接: https://arxiv.org/abs/2608.27888
作者: Janarbek Matai,Sho Ikeda,Lukasz Lipski,Takahiro Harada
类目: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR)
备注: 14 pages, 7 figures,
Abstract:Neural texture compression (NTC) achieves higher compression ratios than BCn formats but suffers from GPU thread divergence, which significantly reduces runtime performance. In this work, we propose a shared decoder MLP architecture – trained with a gradual decoder freezing schedule – combined with texture clustering to reduce thread divergence by 25%-52% while preserving rendering quality. We evaluate our method on over 500 textures and multiple real rendering scenes, demonstrating up to 8.48x speedup on the Radeon RX 9070 XT GPU compared to non-shared baselines. Our key contributions include: (1) a unified shared decoder architecture that reduces divergence by grouping textures; (2) a training recipe with gradual decoder freezing that improves stability and reconstruction accuracy; (3) a semantic clustering strategy using CLIP embeddings that groups similar textures for effective decoder sharing; and (4) comprehensive performance and ablation studies validating our approach.
[CV-60] StreamEMS: Streaming Video Understanding with Self-Evolving Memory Scheme for Vision-Language Models
链接: https://arxiv.org/abs/2608.27881
作者: Yuxin Liu,Peiqin Zhuang,Yali Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Recently, many streaming video understanding methods have been proposed by constructing an external memory to store historical data for computational reduction. Most methods focus on optimizing the injection procedure of current data (write) and retrieving informative historical data (read) from memory, while overlooking the opportunity to further enhancing the representational capability of memory itself. In this work, we present StreamEMS, a general mechanism for improving streaming video understanding by re-structuring the historical data stored in memory through self-evolving memory scheme, enabling more informative and robust memory representations. Specifically, we first introduce a Semantic Evolution Module to evolve the memory into more information-dense representations by exploiting informative memory entities discovered via progressively shrinking semantic scales from coarse to fine. In addition, we further introduce a Prior-informed Evolution Module to evolve memory into more robust representations by leveraging prior memory distributions to refine the current memory state. We validate the effectiveness of our proposed designs on widely-used streaming video understanding datasets, i.e., OVO-Bench and StreamingBench, and the results showcase that our method performs better than other methods. Moreover, the advantage of our method becomes consistently evident even under high token usage drop rate settings, indicating the effectiveness and robustness of our method in unleashing the potential of the memory itself.
[CV-61] What Do Interaction Representations Actually Measure? Pre-Event Separability in Weakly-Supervised Violence Detection
链接: https://arxiv.org/abs/2608.27879
作者: Parishruthi Ganesh
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Articulated human pose provides detailed body-configuration information beyond coarse spatial relationships, but whether this detail yields greater discriminative information when the downstream pipeline is held fixed remains unclear. We examine this through early violence detection. Holding the tracker, temporal head, supervision, folds, and evaluation fixed, we compare five interaction representations spanning coarse bounding-box geometry, a matched handcrafted pose analogue, enriched pose descriptors, and a matched-capacity encoder learned from raw joints, under video-level evaluation with cluster-bootstrap intervals. No pose-based representation outperforms coarse geometry, though with fifteen anomalous videos this subset cannot rule out small effects. Extending the pipeline to frozen visual encoders, and repeating the comparison on XD-Violence (137 anomalous videos, nine times our UCF-Crime sample), person-crop appearance and whole-frame context both exceed geometry by a wide margin, yet context matches appearance on UCF-Crime and exceeds it on the larger split: cropping to the interacting people yields no advantage over encoding the whole frame. This prompts a direct test of what the benchmark measures. Scoring anomalous videos using only frames preceding the annotated onset, under a control removing sequence length as a cue, retains 39-91% of above-chance separation on both benchmarks, including for seven hand-designed geometric channels. Inspection of the tightest pre-onset windows identifies concrete provenance artifacts: editorial title cards and platform watermarks absent from the surveillance footage supplying the normal class. Video-level AUC here is thus a composite of event evidence and pre-event source cues, a shared source of discrimination that can obscure differences between representations. The diagnostic requires only annotations these benchmarks already ship.
[CV-62] Relational Knowledge Distillation Brings DNN Representations Close Enough to Humans to Be Aligned Without Supervision
链接: https://arxiv.org/abs/2608.27877
作者: Yuria Shimizu,Soh Takahashi,Takato Horii,Masafumi Oizumi
类目: Computer Vision and Pattern Recognition (cs.CV); Neurons and Cognition (q-bio.NC)
备注: 53 pages, 6 figures, 5 tables, including Supplementary Information (15 pages, 1 figure, 4 tables)
Abstract:Linking the internal representations of deep neural networks (DNNs) to human mental representations is important for using DNNs as computational models of human vision. Existing DNN representations remain insufficiently similar to human mental representations, which are not directly observable and are therefore commonly measured through large-scale similarity judgments of object images. A natural approach to narrowing this gap is to directly transfer the relational structure of human representations into DNNs, and previous studies have reported improved human-DNN representational similarity. However, whether this improvement holds under stricter evaluation remains untested in two respects: fine-grained alignment at the individual-object level, and generalization to a human embedding derived from a dataset independent of the training data. Here, we employ an unsupervised comparison method, Gromov-Wasserstein optimal transport (GWOT), which estimates human-DNN correspondences from the internal distance structure alone and thereby tests fine-grained alignment. We further assess generalization on a curated test set of concepts non-overlapping with the training data. We show that fine-tuning pre-trained DNNs with Relational Knowledge Distillation (RKD), an established relational transfer method, brings DNNs close enough to humans to be aligned at the individual-object level on this test set. We also show that this improvement is driven by a more human-like global structure, as reflected in the ordering of distances among coarse categories, while the local human-DNN nearest-neighbor overlap rate remains largely unchanged. These findings indicate that relational transfer from humans brings the global structure of pre-trained DNNs close enough to the human structure to enable fine-grained human-DNN alignment without supervision.
[CV-63] mporal Tree of Thought: Reasoning -Guided Visual Cue Search for Long-Video Understanding EMNLP2026
链接: https://arxiv.org/abs/2608.27871
作者: Ziling Huang,Shin’ichi Satoh
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accept by EMNLP2026
Abstract:Long-video understanding remains challenging for Multimodal Large Language Models (MLLMs) due to limited context length. Uniform sampling may miss crucial moments, while agent-based frame video understanding methods often evaluate frames independently, overlooking the temporal organization of videos. Ideally, evidence selection should mimic how humans answer questions about long videos: first locating the relevant segment from the global context, then zooming into local objects and details. We propose Temporal Tree of Thought T^3, a training-free framework for adaptive coarse-to-fine long-video understanding. T^3 constructs a question-agnostic hierarchical temporal tree via recursive temporally constrained clustering, where each node represents a contiguous segment with an informative key frame. During inference, T^3 performs an answer-retrieve-explore loop: it reasons over coarse representative frames, generates a search statement when evidence is insufficient, and expands relevant branches for finer-grained evidence. This process adaptively shifts the search target from temporal regions to specific objects and visual details to help video understanding. Experiments on VideoMME, LongVideoBench, and LVBench show that T^3 improves Qwen2.5-VL-7B by 0.5%, 4.6%, and 4.4%, respectively, under the same frame budget, demonstrating the effectiveness of structured temporal reasoning.
[CV-64] Iron: Intent-Aligned and Retrospective Dual Learning Framework for Enhancing Generalist Virtual Agents
链接: https://arxiv.org/abs/2608.27866
作者: Jiahe Ying,Wendong Bu,Kaihang Pan,Bingchen Miao,Siyu Chen,Wen Wang,Xueming Jiang,Juncheng Li,Siliang Tang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 11 pages, 5 figures, and 4 tables
Abstract:Achieving virtual agents capable of automating tasks across diverse digital environments remains a pivotal challenge in Embodied AI. While Multimodal Large Language Models (MLLMs) offer enhanced visual perception and reasoning, their agentic deployment faces three challenges: costly data annotation, imprecise action-intent alignment, and inefficient exploration from discarded failed trajectories. To address these, we introduce Iron, an intent-aligned, self-improved, and annotation-efficient framework for training GUI agents. Iron employs a novel dual learning strategy that utilizes a stepwise cycle-consistent (SCC) reward to achieve fine-grained alignment between low-level actions and high-level intents, thereby improving instruction grounding and intent understanding. Concurrently, Iron introduces a hindsight reproduction mechanism to repurpose failed trajectories for training, improving both learning efficiency and task diversity. Extensive experiments demonstrate that Iron-trained generalist agents consistently improve performance on cross-environment and cross-device tasks, outperforming models trained with three times more data. Iron also achieves a substantial 25.06% relative improvement on unseen web tasks, with further gains observed on inherently complex tasks, demonstrating the feasibility of building more capable virtual agents.
[CV-65] From Perspective to Fisheye Depth Estimation and Open-Vocabulary Segmentation
链接: https://arxiv.org/abs/2608.27860
作者: Rit Gangopadhyay,Alex Wong
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Vision foundation models are capable of generalizing across 3-dimensional (3D) scenes with high-fidelity estimates; their empirical success can be attributed to training on large-scale datasets of perspective images. However, when transferred to wide field-of-view (FoV) images, such as those captured by fisheye cameras, they return erroneous outputs due to a covariate shift stemming from the radial distortion on the image pixels. We propose a method to generalize vision foundation models to fisheye cameras. The crux of our method lies in a set of learnable parameters, termed Distortion Extenders (DEX), that model the fisheye distortion coefficients and the distributional shift between fisheye and perspective images encoded in the latent space. By minimizing a self-supervised alignment loss, DEX transforms the latent embeddings of fisheye images to resemble those of perspective images to recover high-fidelity estimates. DEX is architecture- and task-agnostic: We demonstrate DEX on monocular depth estimation and open-vocabulary segmentation for convolution- and Transformer-based architectures, where we consistently improve over baselines across indoor and outdoor fisheye datasets. As a byproduct, the activations of DEX can also be decoded to distortion coefficients to support camera calibration. Code available at: this https URL.
[CV-66] CURA: Certified Runtime Alarms for Computer-Use Agents
链接: https://arxiv.org/abs/2608.27808
作者: Divake Kumar,Sina Tayebati,Devashri Naik,Amanda Sofie Rios,Nilesh Ahuja,Omesh Tickoo,Ranganath Krishnan,Amit Ranjan Trivedi
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Self-report is the cheapest oversight channel a deployer has, and on capable computer-use agents (CUAs) it fails precisely where oversight matters. On 361 OSWorld tasks our pipeline, a read-only feasibility gate, a planner, and a GUI executor, reaches a mean task score of 82.9, above the 72.4 human reference, yet 64 of its 71 failures (90%) end with a success claim, 61 acknowledging no blocker, and the explicit failure affordance is never used in roughly 9,100 calls. We introduce CURA (Certified Runtime Alarms for Computer-Use Agents), an external monitor that reads only harness-visible telemetry, with no model internals, extra LLM calls, or prompt changes, and turns the running trajectory into a sequential test with certified false-alarm control. At alpha = 0.10 its CUSUM alarm detects 42.3% of failures a median of 31 steps before termination at a realized false-alarm rate of 0.066, and risk is partly resolvable before the first action (gate probe, 0.69 AUROC). Retrospectively the composite reaches 0.828 AUROC (fold-internal floor 0.802), but its margin over a total-token baseline is not significant (Delta = +0.026, p = 0.101); the separation is online, where CURA recalls more at matched certified budgets: 0.41 versus 0.34 at alpha = 0.10, 0.56 versus 0.38 at alpha = 0.20. Alarm-gated mid-execution oversight recovers 23 of 70 failures while spending a frontier overseer on 38, giving a deployable cascade at mean score 86.8 and 84.5% full-solve (305 of 361). The certificate bounds false alarms only. We also report where behavioral monitoring is uninformative.
[CV-67] uScenes: A Multimodal RGB and 3D Sonar Dataset for Underwater Robot Perception
链接: https://arxiv.org/abs/2608.27795
作者: Trung Tien Dong,Zhenqi Wu,Aditya Penumarti,Zi-Hao Zhang,Micaiah Bartlett,Jane Shin,Xiaomin Lin
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Robust perception is essential for the deployment of autonomous underwater robots. However, optical cameras become unreliable under poor illumination and backscatter. Forward looking (2D) acoustic sensors remain effective under these conditions, but they measure range and bearing while leaving elevation unresolved, creating an ambiguity that prevents individual sonar returns from being localized in three dimensional (3D) space. This complicates the sensor use for 3D scene understanding and precise object detection. We introduce \textbfuScenes, a multimodal underwater dataset containing synchronized 3D multibeam sonar point clouds and RGB imagery. The dataset contains 110 scenes and 95,834 synchronized observation, representing 277.6 minutes of data collected across multiple field sessions. uScenes establishes a foundation for underwater sensor fusion, cross modal representation learning and 3D scene understanding. Code and datasets are given at this https URL.
[CV-68] What Can Low Resource Languages Learn From Each Other? DATE ICDAR2026
链接: https://arxiv.org/abs/2608.27753
作者: Achyuth P,Kahaan Shah,Chetan Arora
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 16 pages, 4 Tables, 6 Figures. To appear at ICDAR 2026. This version predates reviewer revisions. A revised version will be posted
Abstract:Despite the rapid advancement of Vision-Language Models (VLMs), their linguistic reach remains largely confined to high-resource languages, leaving the majority of the world’s 7,000+ living languages on the wrong side of a growing digital divide. This disparity is especially pronounced in Optical Character Recognition (OCR), where low-resource scripts lack the massive datasets required for traditional scaling laws. We investigate OCR adaptation in extreme data-scarce regimes (10K real and 250K synthetic images), demonstrating that conventional fine-tuning strategies often reach a performance ceiling. Our key finding reveals a structural inefficiency in language-specific adaptation: while higher layers of specialized models diverge to capture unique script nuances, the lower layers learn redundant, highly similar features. Motivated by this observation, we propose PSMC (Pre-train, Specialize, Merge, and Co-train), a data-efficient framework that capitalizes on a cross-script “transfer effect”. Our approach first derives language-specific experts from a high-resource base model, then employs task arithmetic to fuse these experts into a unified, high-performance multilingual back- bone. Extensive evaluation across 10 Indian scripts (supporting 20+ languages) shows that PSMC achieves a ~2% average improvement in Word Recognition Rate (WRR) over individual specialist models without increasing parameter count. Our results indicate that joint training in the merged latent space facilitates a constructive knowledge transfer that benefits all constituent scripts, providing a scalable pathway for inclusive VLM development. Source code and datasets will be released post publication.
[CV-69] ABCD: Alpha-Composited Block Coordinate Descent: Constant-VRAM Training for Large Radiance Fields SIGGRAPH2026
链接: https://arxiv.org/abs/2608.27735
作者: Ka Heng Shiu,Kartic Subr
类目: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR)
备注: Presented at ACM SIGGRAPH 2026 Posters
Abstract:We present ABCD (Alpha-Composited Block Coordinate Descent), an out-of-core training framework for alpha-composited radiance fields, instantiated here for 3D Gaussian Splatting. Our method reformulates training as block coordinate descent over spatial partitions: only one block of parameters is active at a time, while all others are frozen. By exploiting the associativity of alpha blending, these inactive regions can be pre-rendered and collapsed into foreground and background RGBA images. As a result, for fixed partition size and image resolution, peak VRAM becomes O(1) with respect to total scene extent, rather than growing with full scene size. This enables GPUs with limited memory to train scenes that would otherwise not fit in core. In experiments, our method closely preserves the reconstruction quality of 3DGS, with less than 5% PSNR degradation, while ABCD with compositing ablated suffers roughly 40% degradation. Our code can be found at this https URL Comments: Presented at ACM SIGGRAPH 2026 Posters Subjects: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR) ACMclasses: I.3.7 Cite as: arXiv:2608.27735 [cs.CV] (or arXiv:2608.27735v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2608.27735 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Journalreference: ACM SIGGRAPH 2026 Posters (SIGGRAPH Posters '26), Article 62, 3 pages, 2026 Related DOI: https://doi.org/10.1145/3799825.3818779 Focus to learn more DOI(s) linking to related resources
[CV-70] Report Supervision
链接: https://arxiv.org/abs/2608.27668
作者: Pedro R. A. S. Bassia,Wenxuan Li,Jakob Wasserthal,Jieneng Chen,Xinze Zhou,Zheren Zhu,Chuntung Zhuanga,Sergio Decherchi,Andrea Cavalli,Kang Wang,Yang Yang,Alan Yuille,Zongwei Zhou
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Published in Medical Image Analysis, 2026
Abstract:Segmentation models can surpass radiologists, classification models, and vision-language models in tumor detection. Importantly, segmentation models outline tumors, allowing radiologists to better verify and trust the AI output. Their main limitation is the scarcity of tumor masks: creating one 3D tumor mask takes up to 30 minutes, so most public CT datasets contain only a few hundred masks, and even the largest private datasets contain only a couple of thousand. Tumor masks are not produced in clinical routine, but radiology reports are. Public datasets contain tens of thousands of CT-Report pairs, and hospitals contain hundreds of thousands. These reports describe tumors in detail, providing large-scale, informative training data. Here, we introduce Report Supervision (R-Super), a training framework that uses reports to directly supervise and improve tumor segmentation. R-Super introduces new loss functions that teach segmentation models to segment tumors that match report descriptions of tumor count, sizes, and locations. Reports are only used for training. We evaluated R-Super on kidney and pancreatic tumor segmentation, exploring diverse training data sizes, up to 41,418 CT-Report plus 3,488 pancreatic tumor CT-Mask pairs. On external validation, R-Super increased tumor detection F1-Score and segmentation DSC by up to +15% with respect to mask-only training. It also surpassed alternative methods such as CLIP and multi-task learning. Leveraging numerous readily available reports to supplement scarce masks, R-Super strongly improves AI performance when very few training masks are available (e.g., 50), and when many masks are available (e.g., 3,488), unlocking scale in tumor segmentation.
[CV-71] Curvature-Aware Radius Shrinkage for Adaptive Nearest Neighbor Classification
链接: https://arxiv.org/abs/2608.27634
作者: Alexandre L. M. Levada
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (stat.ML)
备注: 28 pages, 2 figures and 3 tables
Abstract:Nearest neighbor classification relies fundamentally on how locality is defined, yet conventional k -NN imposes the same neighborhood cardinality throughout the feature space. This assumption can be inadequate for data whose local geometry varies substantially across the underlying manifold. We introduce Curvature-Aware Radius Shrinkage for Adaptive Nearest Neighbor Classification (CARSANN), a geometry-driven framework that adapts the spatial support of each neighborhood according to local geometric complexity. CARSANN first estimates intrinsic dimensionality using TwoNN and constructs an intrinsic representation through principal component analysis. Local mean curvature is then estimated using a shape-operator-based formulation and controls neighborhood scale: highly curved regions receive stronger radius shrinkage, whereas approximately flat regions retain broader spatial support. Unlike methods that modify only the number of neighbors or the local metric, CARSANN explicitly adapts the spatial extent of local evidence. Experiments on more than 70 real-world OpenML datasets show that CARSANN consistently improves upon standard k -NN and is competitive with adaptive nearest-neighbor methods. In a controlled comparison using the same base neighborhood size, CARSANN achieves higher balanced accuracy on 40 of 45 datasets, increasing mean balanced accuracy from 0.6506 to 0.7528. The advantage also persists against k -NN with fixed k=5 . Friedman and Nemenyi tests confirm that the improvements are statistically significant. These results indicate that local manifold curvature can serve as an effective geometric control variable for adapting neighborhood support, providing a complementary paradigm to cardinality-based nearest-neighbor adaptation.
[CV-72] Depth-Aware Pothole Detection Using YOLO and RT-DETR at the Edge
链接: https://arxiv.org/abs/2608.27633
作者: Md Monjurul Ahsan Prodhan,Md Nour Hossain
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 6 pages, 7 figures, IEEE 9th International Conference on Multimedia Information Processing and Retrieval, Presented on Monday, 10 August 2026
Abstract:Pothole detection and its severity measurement is still an important challenges in urban infrastructure management, where late maintenance directly contributes to vehicle damage, road accidents, and escalating repair costs. Existing automated approaches depend on 2D RGB images and cannot measure physical depth of potholes. In this paper, we present a depthaware pothole detection framework and then compare five architectures: YOLOv8n, YOLOv8nSeg, YOLOv9t, RTDETRL, and RTDETRX for RGB-D sensor fusion-based detection and automated depth measurement. A custom offline augmentation pipeline is used here to simulate adverse road monitoring conditions. All models are trained on the PothRGBD dataset with an 80% training and 20% validation split and evaluated using Precision, Recall, mAP@50, and mAP@50_95. Before measuring the depth data, all depth maps are corrected for camera tilt using RANSAC ground-plane orthorectification and all zero-valued sensor pixels are cast to NaN before any statistic is computed. YOLOv8nSeg achieves the highest mAP@50 of 0.9556 and mAP@50_95 of 0.6758 with the most accurate depth estimate of 2.96 cm with the pixel-precise Dseg algorithm. YOLOv8n achieves the fastest inference at 3.6ms. RTDETRX achieves the highest detection confidence at 92.70%. An important finding is that even after full RANSAC orthorectification, bounding box models overestimate pothole depth by 0.16 to 0.21 cm compared to pixel precise segmentation masks. This confirms that the pavement inclusion bias is structural rather than a calibration artifact.
[CV-73] ShiftSplit-AD: Separating Domain Shift from Defects in Foundation-Feature Visual Anomaly Detection
链接: https://arxiv.org/abs/2608.27610
作者: Muhamathu Ameer Ali Aacaas Muhamath
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 7 pages, 5 figures. Independent research manuscript
Abstract:Visual anomaly detectors based on frozen foundation-model features commonly score distances from test patches to a memory of normal features. Benign acquisition changes can also enlarge these distances, confounding domain variation with defects. We investigate whether structured decomposition of nearest-normal DINOv2 residuals can suppress shift-induced evidence while retaining unseen defects. ShiftSplit-AD decomposes the patch residual matrix into low-rank and row-sparse components and scores the sparse component, with an optional low-rank/sparse fusion. The experiments expose a central trade-off rather than a universal separation: genuine defects can contain correlated, low-dimensional structure, so filtering broad residual activity may also remove defect information. On AeBAD-S, using settings fixed after Bottle development, sparse-only scoring improves image AUROC from 0.6780 to 0.7294 and AUPRC from 0.8052 to 0.8465. Paired bootstrap 95% intervals for the improvements are [0.0238, 0.0808] and [0.0170, 0.0650], respectively. However, sparse-only scoring reduces mean clean AUROC from 0.9890 to 0.9133 on four held-out MVTec categories and degrades Bottle localization. These findings show that residual decomposition can help when domain shift strongly contaminates anomaly evidence, but preserving defect structure remains the limiting problem.
[CV-74] Quanta Perception as Probabilistic Events
链接: https://arxiv.org/abs/2608.27584
作者: Varun Sundar,Pavan Thodima,Sacha Jungerman,Mohit Gupta
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: For project webpage, see this https URL
Abstract:Autonomous systems rely on extracting information from light, yet remain brittle in extreme environments, from nighttime navigation to high-speed robotics. Conventional sensors aggregate photons over fixed exposures, imposing trade-offs between sensitivity, dynamic range, and temporal resolution that degrade perception when photons are scarce or dynamics are rapid. Quanta sensors detect individual photons, but their streams exceed real-time compute and latency budgets by orders of magnitude. Here we introduce \textitprobabilistic events , a computational primitive for real-time quanta perception from individual photon detections. By computing the posterior over the time since the last intensity change, we represent photon streams as recursive belief states. Rather than fixed-threshold event-camera triggers, this recursive Bayesian formulation yields three low-latency signals: motion-adaptive scene flux, high-fidelity activity maps, and entropy-based perceptual uncertainty. This representation enables perception in extreme conditions, including pose estimation of a running person at \sim 0.05 lux—without retraining vision models. Our approach processes input streams exceeding 50,000 quanta frames per second on commodity GPU hardware—yielding kilohertz-scale outputs up to four orders of magnitude faster than state-of-the-art quanta reconstruction baselines, even for megapixel arrays. By replacing frame reconstruction with direct probabilistic inference over photon streams, this work bridges photon-counting quanta sensing with robotic vision. Comments: For project webpage, see this https URL Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) Cite as: arXiv:2608.27584 [cs.CV] (or arXiv:2608.27584v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2608.27584 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-75] VidParse: Online Parsing of Egocentric Procedures Like a Pro ECCV2026
链接: https://arxiv.org/abs/2608.27562
作者: Anubhav Gupta,Archit Kambhamettu,Vatsal Agarwal,Pulkit Kumar,Abhinav Shrivastava
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at ECCV 2026
Abstract:Translating continuous, noisy egocentric video streams into discrete, temporally ordered action steps is fraught with visual challenges. Heavy ego-motion, transient occlusions, and the high intra-class variability of unscripted human-object interactions cause standard frame-level online temporal models to struggle, often resulting in severe over-segmentation and structural collapse. To bridge the gap between unstable low-level perception and high-level procedural logic, we present VidParse, an online, training-free framework that treats activity understanding as a graph-constrained inference problem. Rather than relying on learned temporal filters, we dynamically identify semantic transitions using a temporal similarity matrix over manipulation-anchored features, which are extracted from frozen foundation models to prioritize foreground hand-object interactions. A beam search decoder then leverages an induced procedural task graph to explicitly enforce valid action transitions and prune impossible trajectories. By anchoring robust visual segments to hard procedural constraints, our approach preserves long-range state transitions and achieves up to a 10x improvement in complex multi-step parsing accuracy over strong online baselines, all without requiring a single gradient update.
[CV-76] Beyond Data Scaling: Representation-Centric Continued Pre-training for Vision-Language-Action Models
链接: https://arxiv.org/abs/2608.27550
作者: Senqiao Yang,Chengyao Wang,Yuxin Chen,Zixuan Wang,Longxiang Tang,Haokun Gui,Jinhui Ye,Changsheng Lu,Xiaoyang Wu,Mingkang Zhu,Pengguang Chen,Shu Liu,Zhuotao Tian,Hengshuang Zhao,Bei Yu,Jiaya Jia
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注: All models and training pipelines are publicly available at this https URL
Abstract:Scaling robot data is crucial for building generalist Vision-Language-Action (VLA) models, yet robot trajectories are harder to scale than web-scale image-text data because embodied collection is costly and sparsely covers the physical world. This makes representation quality a central bottleneck: under a fixed robot-data budget, continued pre-training must turn limited trajectories into transferable visual-action knowledge rather than merely fit actions. We propose VLAct, a VLA-oriented VLM backbone trained on broad, heterogeneous, multi-embodiment robot data before task-specific fine-tuning. VLAct preserves the broad VLM prior and encourages shared action semantics across embodiments through VLM-prior preservation, multi-head continuous action co-supervision, and a partially unified cross-embodiment action layout, while allowing task-specific action heads during fine-tuning. Across simulation, real-world, and unseen-embodiment transfer, VLAct consistently improves downstream performance under fixed fine-tuning protocols. On LIBERO-Plus and RoboTwin 2.0, VLAct surpasses industrial VLA systems including ABot-M0 and LingBot-VLA, achieving success rates of 82.6% and 92.5%. On RoboDojo, VLAct ranks sixth among all policies by success rate and outperforms all explicitly designated world-action model (WAM) entries on both metrics. Most notably, on RoboCasa-GR1, an unseen humanoid embodiment, VLAct using only 20% of downstream trajectories outperforms the full-data GR00T-N1.6 baseline. These results are obtained using fully open-source data and only a 16-GPU training setup, showing that representation-centric continued pre-training can deliver highly competitive performance under a modest compute budget and is an important independent axis of VLA progress beyond data scaling.
[CV-77] Code as Worlds: Agent ic Discovery of Executable World Representations for Physical Reasoning
链接: https://arxiv.org/abs/2608.27549
作者: Hanyang Wang,Yimo Cai,Weiliang Chen,Jiawei Chi,Haowen Sun,Qiyu Dai,Yi-Hsin Hung,Xingzhuo Guo,Jinshan Ren,Runmao Yao,Ziwei Liu,Mingsheng Long,Yueqi Duan,Jun Gao,Jiangran Lyu,Fangfu Liu,Jialong Wu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Project Page: this https URL
Abstract:Physical understanding and reasoning depend on forming compact and generalizable representations of the world. While modern vision-language models can recognize and explain diverse physical events, they often lack explicit representations of the underlying mechanisms-such as object states, physical parameters, and governing dynamics-needed for reliably reasoning how the world evolves and responds to interventions. In this work, we introduce Code-as-World, a paradigm that represents physical worlds through executable world representations. By expressing physical composition, dynamic evolution, and visual appearance as executable code, Code-as-World provides a compact, quantitatively grounded, and controllable abstraction of the physical world. To construct such representations from multimodal observations, such as natural-language descriptions or real-world videos, we develop an agentic discovery loop inspired by abductive reasoning, where an agent proposes, executes, renders, verifies, and iteratively refines executable world hypotheses. As a concrete application, we use verified executable worlds to provide scalable physical supervision for training vision-language models on quantitative physical reasoning. Experiments show that Code-as-World-VL achieves state-of-the-art performance on QuantiPhy and surpasses leading proprietary models, highlighting the potential of executable world representations as a scalable foundation for physical intelligence.
[CV-78] Fully Unleashing the Multimodal Attacker: Meta-Adaptive Jailbreaking of Vision-Language Models EMNLP2026
链接: https://arxiv.org/abs/2608.27531
作者: Benlei Cui,Shen Pang,Yuke Wang,Xuemei Dong,Yuwen Zhai,Jingqun Tang,Haiyang Yu,Hui Xue,Longtao Huang,Haiwen Hong
类目: Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by EMNLP 2026 Main Conference
Abstract:The safety of large vision-language models is increasingly stress-tested by multimodal jailbreaks, yet existing attacks remain largely static at the meta level: template-based attacks freeze the image–text layout, while iterative attacks adapt only the image–text content with fixed attack strategies and frozen attacker parameters. We propose Meta-Adaptive Multimodal Jailbreaking (MAMJ), which instead optimizes the attacker itself along two axes: an attack strategy prompt (ASP) \theta governing attack iteration and attacker weights \phi determining attack effectiveness. Across groups of multimodal attack trajectories, an LLM-based critique first refines \theta , after which group-aggregated attack-success-rate (ASR) rewards update \phi . On MM-SafetyBench, MAMJ achieves 81.0% , 78.9% , and 82.3% ASR against GPT-4o, Gemini-3-Pro-Preview, and Seed 2.0, respectively, outperforming the strongest sample-level baseline by up to 24.1 percentage points. The learned attacker (\theta^\star,\phi^\star) also transfers without retraining to unseen victims and remains effective under representative defenses. These results reveal a systemic vulnerability of frontier VLMs to meta-adaptive jailbreaks and motivate defenses against meta-level adversaries. Code is available at this https URL.
[CV-79] Revisiting Local Context for Long-Horizon Streaming 3D Reconstruction
链接: https://arxiv.org/abs/2608.27529
作者: Jiarong Han,Jincheng Xiong,Yuzhou Liu,Linzhe Shi,Changjie Wu,Ning Guo,Mu Xu,Hang Zhang,Ming Qian
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Project Page: this https URL , Code: this https URL
Abstract:Streaming 3D reconstruction from extremely long videos requires estimating camera motion and scene geometry online under bounded memory and computation. Early streaming models achieve causal, bounded-cost inference using finite context buffers or compact recurrent states, yet their estimates often deteriorate as sequences grow. Recent methods improve long-horizon stability by coupling short-range context with persistent or multi-level long-range memory. We pursue a different route: we keep the learned temporal state strictly local and formulate predictions whose targets remain independent of sequence length. We present ABot-Recon, a simple streaming model that caches KV features from only the preceding 11 frames. It predicts a point map in the current camera coordinate system together with an adjacent-frame relative pose. These predictions remain equivariant under changes of reference frame, and global poses and geometry are recovered through sequential composition. To reduce accumulated drift, a lightweight temporal refiner improves relative rotations using recent visual and motion context, while a composition-aware pose loss supervises multi-step pose composition. Extensive evaluations on challenging long-sequence benchmarks demonstrate the superior long-horizon performance of our local-context approach. On Oxford Spires, ABot-Recon achieves an ATE of 4.35 m and an RPE-R of 0.12^\circ , reducing both errors by approximately 40% relative to the best prior results.
[CV-80] FVeinSyn: Synthetic Finger Vein Image Generator
链接: https://arxiv.org/abs/2608.27527
作者: Yifan Wang,Jie Gui,Adams Wai Kin Kong,Baosheng Yu,Changsheng Chen,Qi Li,Zhenan Sun,James Tin-Yau Kwok,Alex Kot
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:A major challenge in finger vein recognition is the lack of large-scale public datasets. Existing datasets contain few identities and limited samples per finger, restricting the advancement of deep learning-based methods. To address this, we propose FVeinSyn, a large-scale controllable synthetic data generation framework for finger vein. It explicitly decouples synthesis of vascular topology and imaging appearance to mitigate the limitations caused by insufficient training samples, such as inadequate identity diversity and restricted realism. Specifically: first, a finger vein identity generator models vascular topology under physiological and geometric constraints using stochastic L-systems, producing anatomically valid and identity-distinctive vascular patterns. Then, a cascaded region-aware GAN renders the topological maps into realistic near-infrared images. Finally, an intra-class diversity generator introduces geometric and optical perturbations to simulate realistic intra-class variations. Using FVeinSyn, we generated 500,000 images (10,000 vein identities, 50 samples per identity) and conducted extensive evaluations. Results show that FVeinSyn holds significant advantages in realism, identity diversity, vascular pattern consistency, and intra-class diversity. Models trained with FVeinSyn outperform real-data-only baselines a cross eight public datasets, achieving an average accuracy improvement of 27.43%. The code is available at: this https URL.
[CV-81] A Deeper Analysis of Block-Sparse Featurizers
链接: https://arxiv.org/abs/2608.27515
作者: Alexandru-Iulius Jerpelea,Amith Ananthram
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注: 9 pages, 12 figures, 2 tables
Abstract:The recently introduced block-sparse featurizer (BSF; Fel et al., 2026) is similar to a sparse autoencoder (SAE), but its atomic unit is a small subspace (a block of directions) rather than a single direction. It is designed for features that live on low-dimensional manifolds, which are especially frequent in vision. This work studies the BSF’s strengths and weaknesses, finding how it still somewhat suffers from classic SAE failure modes, like feature splitting and composition. We propose several architectural changes to the BSF, including a Tournament Top-K selection rule that significantly reduces feature splitting, and we also extend the block paradigm to the crosscoder.
[CV-82] Image Augmentation as Test Generation for Deep Learning-Based Image Retrieval Systems
链接: https://arxiv.org/abs/2608.27502
作者: Yehan De Silva,Anirudh Sridhar,Armin Lotfy,Nafiseh Kahani,Yvan Labiche,Ziyu Wang,Frank Ouyang,Clare Carty,Azalia Shamsaei
类目: oftware Engineering (cs.SE); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Ensuring the reliability of deep learning-based image retrieval systems is a software engineering challenge. This paper presents a dual contribution: (1) a literature review of augmentation and generation techniques which resulted in the identification of 50 techniques which we organized into a ten-category taxonomy, and (2) a large-scale empirical study that evaluates these techniques as test generators for embedding-based image retrieval systems. Augmented images are embedded using Amazon Titan and OpenCLIP, and evaluated across four analytical dimensions: (1) embedding-space similarity, (2) embedding uncertainty measured via four estimators, (3) semantic realism scored by LLaVA, and (4) retrieval failure rate. Experiments are performed on three datasets: CIFAR-10, ImageNet-1K, and a dataset from an industrial partner (March Networks). Across all evaluated datasets and embedding models, and under the single severity level tested for each technique, weather simulation and SaSPA are the image augmentation/generation techniques that produce the highest embedding uncertainty and failure rates while maintaining a favorable balance between performance stability, visual realism, and augmentation effectiveness. The results we discuss are configuration-specific and may shift under milder or stronger perturbation settings. In contrast, GAN-based augmentation techniques are among the lowest in realism, indicating the presence of synthetic artifacts and perceptual inconsistencies that reduce their suitability to produce realistic test inputs. Overall, our findings provide practical guidelines for selecting augmentation techniques that maximize test diversity while preserving realistic image characteristics, thereby enabling the construction of comprehensive and effective test suites for image retrieval systems while reducing the cost of manual data labeling through the use of metamorphic testing.
[CV-83] Can Tainted Pixels Expose Deepfake Videos?
链接: https://arxiv.org/abs/2608.27492
作者: Juan Hu,Shaojing Fan,Sanjay Saha,Marc Herrera,Terence Sim
类目: Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Publicly-acceesible face-manipulation tools have made deepfake creation accessible to non-expert users. Against these, existing defenses are mostly post-hoc, detecting only after forgery has occurred, and operating on still images rather than videos. Research is lacking in i) the proactive protection of published facial videos against black-box manipulation tools, and in (ii) understanding its perceptual effect on human viewers. We introduce TaintedPixels, a proactive video-protection method built around an asymmetric visibility trade-off: the embedded watermark should remain inconspicuous in the published video but become obvious once a downstream tool manipulates the video. TaintedPixels injects structured periodic perturbations into the blue channel of facial regions and refines them under stripe-visibility, color-cast, and video-level LPIPS budgets, with lightweight motion-adaptive deployment. We believe TaintedPixels is the first proactive defense designed specifically against black-box manipulation tools rather than image-level pipelines or specific surrogate generators. Across three publicly available off-the-shelf video manipulation tools and two off-the-shelf detectors, TaintedPixels attains the highest forgery fake rate while keeping perturbations small (LPIPS = 0.0042). Our non-expert human study, conducted on a diverse set of 300 video stimuli spanning different lighting conditions, backgrounds, and skin tones, shows that protected source videos draw a 3.26% suspicion rate, while forgeries from protected sources are identified as fake much more often than forgeries from unprotected sources (90.72% vs. 56.71%). This validates the effectiveness of TaintedPixels.
[CV-84] Benchmarking General Mobile Assistants in Challenging Real-World Scenarios
链接: https://arxiv.org/abs/2608.27477
作者: Yiqi Zhu,Feiyu Gao,Jiaxing Fan,Jiahui Zeng,Minggang Wu,Chenliang Li,Haiyang Xu,Peng Li,Ming Yan,Yang Liu
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Graphical user interfaces have emerged as an important environment for evaluating autonomous AI agents on multimodal interactive tasks. Existing benchmarks such as AndroidWorld and MobileWorld provide strong foundations for mobile agent evaluation, but their application coverage and task design do not yet fully capture the diversity and complexity of realistic mobile use. We present GMA, a benchmark for evaluating general mobile assistants in challenging real-world scenarios. GMA introduces seven applications based on open-source projects, spanning domains such as lifestyle sharing and travel planning, and 300 tasks across four difficulty tiers, from atomic actions to complex multi-step workflows. We evaluate eight frontier models and find that performance declines substantially as task complexity increases, with current agents remaining far from reliably handling realistic user requirements. We further conduct controlled ablation studies of agent harness choices, including context retention and explicit state tracking, under a shared environment, model setting, and task taxonomy. Results show that appropriate harness design can meaningfully improve performance, particularly on demanding workflows, while the effectiveness of specific designs can vary across foundation models. Overall, GMA complements existing benchmarks by expanding application coverage and task complexity, providing a challenging testbed for evaluating mobile agents and studying how harness design supports reliable execution in complex mobile workflows.
[CV-85] CheXtriev: Anatomy-Centered Representation for Case-Based Retrieval of Chest Radiographs MICCAI2024
链接: https://arxiv.org/abs/2608.28137
作者: Naren Akash,Arihanth Tadanki,Jayanthi Sivaswamy
类目: Image and Video Processing (eess.IV); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: Accepted at the 27th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI 2024)
Abstract:We present CheXtriev, a graph-based, anatomy-aware framework for chest radiograph retrieval. Unlike prior methods focussed on global features, our method leverages graph transformers to extract informative features from specific anatomical regions. Furthermore, it captures spatial context and the interplay between anatomical location and findings. This contextualization, grounded in evidence-based anatomy, results in a richer anatomy-aware representation and leads to more accurate, effective and efficient retrieval, particularly for less prevalent findings. CheXtriv outperforms state-of-the-art global and local approaches by 18% to 26% in retrieval accuracy and 11% to 23% in ranking quality. The code is available at this https URL.
[CV-86] Do Medical Vision Models Reason About Anatomy? Probing the Spatial Inductive Biases of Learned Visual Representations
链接: https://arxiv.org/abs/2608.28092
作者: Naren Akash,Neeraja Ramanan
类目: Image and Video Processing (eess.IV); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Interpreting a CT scan means comparing structures on either side, judging how far apart organs sit, and knowing where each one belongs. Medical vision encoders are evaluated on diagnostic accuracy, or through assembled multimodal systems where a failure is hard to attribute, so it remains unclear whether their representations support any of this. We construct SPAR-Bench, eight probes over multi-organ abdominal CT that separate coordinate localization, relational reasoning, and spatial queries, and apply them to five architectural configurations and three medical foundation models, frozen and finetuned. Probes that ask for a comparison within the slice stay at chance, and neither pretraining scale, finetuning, nor architecture closes the gap. Probes that appear solved in domain fall to chance under zero-shot transfer, indicating that their accuracy reflects recall of canonical anatomy rather than computation over the image. Reading the same frozen features with a pooled head rather than the full set of tokens moves relational recovery from 0.7% to 67.8%, so pooled probing understates what a representation holds. Questions the encoders answer well are answered at chance by four open-weight MLLMs. Our results suggest these encoders carry a map of where organs usually lie, and little of the machinery for comparing structures within a particular patient. Code and data will be available at this https URL.
[CV-87] CARDINAL Predicts Cardiovascular Risk From Non-contrast Cardiac CT
链接: https://arxiv.org/abs/2608.27690
作者: Roy Gabriel,Nattakorn Kittisut,Jamshid Hassanpour,Michael Galarnyk,Abanoub Abdelmalak,Marly van Assen,Carlo N. De Cecco,Arshed Quyyumi,Ali Adibi
类目: Image and Video Processing (eess.IV); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Robotics (cs.RO)
备注:
Abstract:Cardiovascular risk prediction remains limited by incomplete clinical data and imaging biomarkers that reduce computed tomography (CT) to a small number of handcrafted features. We developed CARDINAL (Cardiovascular Assessment via Representation learning from Deep Imaging with Nested Anatomical Latent embeddings), a clinically grounded framework that learns compact representations from routine non-contrast cardiac CT for major adverse cardiovascular event (MACE) prediction. In 17,659 patients, CARDINAL was evaluated for 1-, 3-, 5-, and 10-year MACE prediction against American Heart Association (AHA) pooled cohort equations (PCE), AHA predicting risk of cardiovascular disease events (PREVENT), coronary artery calcium (CAC), segmentation-derived CT biomarkers, and 70-feature structural radiomics. Gains were largest at longer horizons. At 10 years, CARDINAL (joint) achieved an area under the receiver operating characteristic curve (AUROC) of 0.866 \pm 0.020 and an area under the precision-recall curve (AUPRC) of 0.890 \pm 0.015, compared with an AUROC of 0.826 \pm 0.023 and an AUPRC of 0.826 \pm 0.022 for structural radiomics, the strongest baseline. CARDINAL also achieved the highest survival concordance index (C-index), 0.753 \pm 0.015, and high-versus-low risk-tertile hazard ratio, 10.78 \pm 3.16, with favorable reclassification and exploratory calibration. These findings suggest that non-contrast cardiac CT contains prognostic information beyond conventional risk equations, CAC scoring, and engineered imaging biomarkers.
[CV-88] Destroy Me: Automatic Artifact Generation for Histopathology Images
链接: https://arxiv.org/abs/2608.27516
作者: Zuzanna Krawczyk-Borysiak,Adam Krawczyk,Mateusz Miller,Gabriela Kaczmarek,Sławomir Pakuło,Małgorzata Sokół,Żaneta Swiderska-Chadaj
类目: Image and Video Processing (eess.IV); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Deep learning’s diagnostic utility in pathology is constrained by model vulnerability to real-world data imperfections. While current strategies favor “perfect data” by filtering low-quality regions, which can lead to the loss of valuable diagnostic context, we propose a paradigm shift: engineering models to thrive in imperfect environments using “Destroy Me”, a hybrid framework for realistic artifact synthesis and robust data augmentation. Our approach combines Stable Diffusion, fine-tuned to preserve morphological continuity by realistically integrating artifacts with the underlying tissue architecture, with physics-based procedural modeling to synthesize six common artifact types: tissue folds, precipitates, blur, stitching errors, dust, and pen markers. Artifact fidelity is assessed using Kernel Inception Distance (KID) and color Wasserstein distance metrics. Validating this strategy on lung adenocarcinoma pattern classification with an nnU-Net, we confirm that models trained on “destroyed” patches consistently outperform baselines on independent real-world datasets. Specifically, we observed a 10.5% relative improvement in macro F1-score and a 15% relative increase in the Cohen’s Kappa ( \kappa ) coefficient. Crucially, our results demonstrate that selective, impact-weighted augmentation is vital for balancing practical robustness with the preservation of subtle diagnostic features.
[CV-89] Climate Physics Dynamic Matching
链接: https://arxiv.org/abs/2608.26907
作者: Gurjeet Sangra Singh,Frantzeska Lavda,Alexandros Kalousis
类目: Applications (stat.AP); Computer Vision and Pattern Recognition (cs.CV); Logic in Computer Science (cs.LO)
备注:
Abstract:Deep generative models such as flow matching and diffusion models have shown potential for learning complex dynamical systems, but typically act as black boxes that neglect underlying physical structure, while physics-based models governed by partial differential equations are often incomplete due to missing source terms, or uncertain parametrisations. We present Climate Physics Dynamic Matching (ClimPhyDM), a variational simulation-free dynamics informed framework for weather forecasting that combines an advection-type physics prior with data-driven components in a variational framework. % to capture the stochasticity and multi-modality of unresolved atmospheric dynamics. On the ERA5 benchmark at hourly (42-hour) and monthly (5-month) resolutions, ClimPhyDM outperforms ClimODE, and GB-DM, keeping the lower error at extended horizon, indicating improved temporal stability and resistance to error accumulation, while its simulation-free paradigm also enables training on a single modest 12 GB consumer GPU.
人工智能
[AI-0] Aero Hand Open: A Simulation-Ready Tendon-Driven Hand for Dexterous Manipulation Learning
链接: https://arxiv.org/abs/2608.28578
作者: Nan Wang,Mohit Yadav,Jonathan Wulff,Aidan Rosenbaum,Kezhou Chen,Yuvan Sharma,Xu Dong,Yiwei Tao
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 20 pages, 9 figures. Project page: this https URL
Abstract:Tendon-driven hands are anthropomorphic, and moving the actuators off the joints is what makes a hand of this capability affordable to build. Two effects produce that saving. Routing force through a cable removes the requirement that a motor fit inside the joint it drives, so smaller and cheaper motors suffice, and one motor can drive several joints through a single cable, so fewer motors are needed. They are also harder to learn on than a direct-drive hand. The underactuated transmission that produces the saving is itself difficult to represent in a simulator, and the joints one cable drives are not independently commandable. We present Aero Hand Open, a tendon-driven anthropomorphic hand that is released simulation-ready. Three things ship with it. A simulation model reproduces the cable transmission itself. An identified actuation map connects that model to the motor commands in both directions, including the three-way coupling of the thumb. A reinforcement learning package trains policies for the hand. Together they let a policy be trained entirely in simulation and run on the hand with no fine-tuning and no state estimation. We release the mechanical design, the simulation model, the identified mapping, the training environment and the deployment stack.
[AI-1] Blog: Survey of Optimizers
链接: https://arxiv.org/abs/2608.28557
作者: Ruoran Xu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Neural-network optimization in 2025-2026 is no longer well described as a succession of new Adam variants. The design space has expanded from coordinates to matrices and layers, from fixed training horizons to policies over time, and from mathematical update rules to state representations that must survive sharding and low-precision computation. This survey organizes recent optimizers and training optimization methods along four largely independent axes: temporal estimation, update geometry, horizon management, and representation and systems. It connects the spectral normalization of Muon, the historical matrix statistics of Shampoo and SOAP, adaptive and hybrid matrix methods, memory-efficient optimizers, schedule-free training, small-batch corrections, and quantized optimizer states. The central empirical conclusion is deliberately non-triumphal: matrix-aware methods represent a genuine advance, but there is no context-independent replacement for AdamW. Rankings change with model scale, data-to-parameter ratio, batch size, schedule, parameter partition, tuning budget, and whether the target metric is tokens, FLOPs, wall-clock time, or memory. The practical consequence is a compositional view of optimizer design and a stricter protocol for evaluating optimizer claims.
[AI-2] 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)
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-3] InstructMesh: Selective Refinement of Generative 3D Models for Fabrication
链接: https://arxiv.org/abs/2608.28534
作者: Faraz Faruqi,Ahmed Katary,Demircan Tas,Theresa Hradilak,Ning Zhang,Jiaji Li,Fabian Manhardt,Martin Nisser,Vrushank Phadnis,Ruofei Du,Federico Tombari,Megan Hofmann,Stefanie Mueller
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Recent advances in generative AI allow users to create 3D models from text or images. However, these models prioritize visual plausibility over geometric accuracy, often generating results with flaws that compromise their intended use post-fabrication. We present InstructMesh, an interactive post-generation refinement tool that enables selective repair of generative 3D models through region selection and targeted operations, such as opening or sealing voids, or adjusting local thickness. Users can invoke edit operations via natural language prompts or slider controls. By operating directly on the intermediate latent representation, InstructMesh allows users to apply robust geometric corrections without requiring expert modeling skills. To inform our design, we first analyze common fabrication-related failure modes in outputs from state-of-the-art generative tools. We then conduct two user studies, demonstrating that novices can identify and perform fabrication-relevant repairs on generative outputs using InstructMesh, and revealing user preference for hybrid interfaces that combine slider controls with natural language input.
[AI-4] Conformal Uncertainty Quantification Guarantees for Neural Operators
链接: https://arxiv.org/abs/2608.28515
作者: Tom Stent,Nicolas Boullé
类目: Numerical Analysis (math.NA); Artificial Intelligence (cs.AI); Probability (math.PR)
备注: 19 pages, 6 figures
Abstract:Neural operators provide fast surrogate models for approximating operators between function spaces, but their predictions often lack uncertainty quantification. We develop a split conformal framework to guarantee that a calibrated pointwise band around the neural operator output contains the true solution on at least a 1-\gamma fraction of the evaluation domain, with probability at least 1-\alpha over test and calibration inputs, where \alpha,\gamma\in(0,1) . Our method reduces a normalized residual field to its spatial (1-\gamma) -quantile and computes a scaling factor using a held-out calibration dataset. We prove marginal coverage guarantees for measurable residual fields defined on arbitrary probability spaces, covering both continuum domains and fixed discretizations. Under mild assumptions on the data distribution, we show that the coverage conditional on the calibration set follows a Beta distribution, which we verify with numerical experiments on Darcy flow and Navier–Stokes equations, where our calibration yields bands consistently tighter than existing corrections while retaining the target coverage.
[AI-5] raining Communication-Efficient Mixture-of-Experts Language Models with Layer Re-Configuration
链接: https://arxiv.org/abs/2608.28511
作者: Simeng Sun,Roger Waleffe
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:When training Mixture-of-Experts (MoE) language models with expert parallelism, all-to-all token dispatch and combine collectives can consume a substantial fraction of end-to-end training time. In this work, we study communication-efficient MoE models (CE-MoE), in which we adopt a heterogeneous layer pattern that decouples token-mixing and channel-mixing depth. Compared to conventional models which interleave MoE layers after each token-mixing layer (e.g., attention, Mamba-2), CE-MoE models concentrate expert capacity in a select few routed MoE layers, while maintaining depth by adding additional token-mixing and dense-FFN layers. Across a scaling ladder from 2B to 31.5B total parameters, under matched total and activated parameters, CE-MoE models consistently reduce training cost while matching validation loss and downstream benchmarks with full-MoE baselines. At the 31.5B scale, CE-MoE uses 33.3% fewer GPU-hours while improving average downstream score and inference throughput.
[AI-6] On the Maintenance and Co-evolution of Agent Plugins: An Empirical Study of Claude Code Plugin Marketplaces
链接: https://arxiv.org/abs/2608.28497
作者: Ahmed Hereiz,Yingzhe Lyu,Hao Li,Bram Adams,Ahmed E. Hassan
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: Under review
Abstract:AI coding agents, software tools that automate development tasks through reasoning and tool use, are increasingly extended through plugin marketplaces, yet the structure, maintenance, and co-evolution dynamics of these emerging repositories remain empirically unexplored. Unlike traditional software packages that deliver functionality through source code, agent plugins deliver functionality through a combination of natural-language instruction files, scripts, and configuration files, raising the question of whether these plugins are maintained artifacts that co-evolve across components, or one-off artifacts that developers write once and do not need to revisit. To study the maintenance and co-evolution of agent plugins, we conduct an empirical study of 1,926 repositories hosting Claude Code plugin marketplaces, analyzing 8,351 plugins and 77,773 commits across 2,018 marketplaces. We find that the marketplace is expanding rapidly, plugin-touching commit activity growing 8.8x over six months after the October 2025 launch, and plugins targeting Software Engineering tasks accounting for 61.3% of all plugins. Plugin development is predominantly feature-driven, with feature commits occurring at more than twice the rate of conventional open-source software (OSS) (39.6% vs. 17.2%). Claude co-authors 34.9% of all commits, and four commit types (docs, perf, style, and refactor) carry substantially different meanings in plugin repositories than in traditional software. Most component types evolve independently, but within skills directories, natural-language instruction files and implementation scripts co-evolve at above-chance rates, with 78% of co-changes being functionally coupled, representing a new class of maintenance dependency not observed in traditional software engineering.
[AI-7] AcrossVAM1.0: Particle World Modeling for Text-Assisted Robot Video Prediction
链接: https://arxiv.org/abs/2608.28491
作者: Yafei Zhang,Nan Wu
类目: Artificial Intelligence (cs.AI); Robotics (cs.RO)
备注:
Abstract:Predicting robot videos requires both precise motion reasoning and preservation of high-frequency appearance, yet monolithic pixel models entangle these objectives and often conceal their progress behind a strong last-frame baseline. We present AcrossVAM1.0, a lightweight, text-assisted video action model that factorizes future prediction into object-centric motion and dense appearance. A frozen SAM3-DLP codec decomposes four context frames into semantic particles for the robot, arm, and gripper, together with a background latent. A 0.28M-parameter spatio-temporal Transformer aligns particle identities, rolls their states forward, and is modulated by a frozen OpenCLIP instruction embedding through FiLM. A causal dual-stream decoder combines particle-rendered motion with appearance encoded exclusively from the last observed frame; a residual refiner and learned delivery mask produce five future frames without access to future appearance. On our VRS benchmark constructed from diverse real-robot trajectories, particle dynamics reduce trajectory error by 21.0% over persistence. Across three delivery-mask seeds, AcrossVAM1.0 improves future-frame PSNR/SSIM from 19.97/0.796 to 20.573/0.8004, while raw particle generation improves motion-region PSNR from 11.89 to 13.23. The delivered model does not yet beat persistence in LPIPS, and correct-versus- shuffled language changes trajectory error by only 2.8–3.1%. We report these limitations alongside oracle, negative-control, multi-seed, and per-robot analyses. The results show that explicit particle dynamics are a promising low-dimensional interface for robot video prediction, while robust language grounding and appearance delivery remain the principal open challenges.
[AI-8] LLM -Based Agents for Software and Systems Security: Approaches Applications and Assessment
链接: https://arxiv.org/abs/2608.28490
作者: Jingjing Nie,Jiawei Guo,Krishna Meda,Haipeng Cai
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:Software and systems security workflows are typically procedural: analysts inspect heterogeneous artifacts, form hypotheses, invoke tools, interpret outputs, and revise plans. Large language model (LLM)-based agents, which can plan, use tools, retain state, and revise actions across multi-step workflows, are being rapidly adopted to automate this work. Given the consequences of delegating security decisions to autonomous systems, understanding how such agents are built, used, and assessed is crucial. Yet to this date, there remains a lack of systematic understanding of what has been done and how far we are in this field: the term “agent” is applied inconsistently, applications differ sharply in risk, and assessment protocols are often incomparable. To gain a comprehensive and coherent view of this area hence inform relevant future research, this paper provides a systematic literature review of the (1) technical approaches, including agent architecture, perception, memory, reasoning and planning, action space, orchestration, and self-improvement, (2) applications, with respect to the security tasks served, and (3) assessment, including the datasets, outcome and trajectory metrics, safety measures, and baselines considered, over the peer-reviewed literature spanning the emergence of this area (2023–2026). Our synthesis reveals a field that has built agents able to act but not yet agents whose authority is bounded or whose behavior is auditable. In addition to knowledge systematization, we also extend our insights into the limitations of and challenges faced by current approach, application, and assessment designs, which shed light on potentially promising future research directions.
[AI-9] How Proper Scoring Rules Shape LLM Forecasting
链接: https://arxiv.org/abs/2608.28482
作者: Benjamin Turtel,Paul Wilczewski,Kris Skotheim,Ville A. Satopää,Philip E. Tetlock
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:This paper evaluates how reward function choice shapes the performance and behavior of LLM forecasters. We compare five proper scoring rules as training objectives for binary forecasts of resolved real-world events. Although the rules share the same theoretical incentive for truthful probability reporting, the resulting models differ in calibration, probability use, and estimated profiles of bias, information, and noise, with smaller differences in aggregate accuracy and discrimination. The Brier-trained model has the lowest observed Brier score and highest AUC-ROC, while the log-trained model has the highest observed log score and lowest calibration error. Models with similar aggregate performance also reach that performance through different combinations of bias, information, and noise. Proper scoring rules therefore need not behave interchangeably as training objectives. Reward choice may shape not only how well an LLM forecasts, but how its forecasting errors are structured. Each condition uses a single seed, so some differences may reflect training stochasticity.
[AI-10] COVER: Identifiable Evaluation of Coalition Routing
链接: https://arxiv.org/abs/2608.28475
作者: Raghul Sugumar,Amrit Gopinath
类目: Artificial Intelligence (cs.AI)
备注: 17 pages, 2 figures, 13 tables
Abstract:When a multi-agent system changes its team, it also changes the messages and final answer it produces, so an end-to-end accuracy gap does not by itself identify a routing effect. We introduce method, an evaluation contract that fixes a public information boundary, downstream stack G, and finite legal team family before outcomes are generated. Complete coverage identifies exact finite-benchmark oracle regret conditional on that stack. For any finite collection of frozen policies, executing the union of their distinct selected teams is the minimal assumption-free support for every pairwise policy contrast, though not for absolute oracle regret. Two controlled tables with source-ID-disjoint splits test the instrument. On MuSiQue-12, a pre-specified privileged positive control improves regret from 0.532 to 0.402; a later public-interface control reaches 0.424 versus 0.554 but is retrospective. On HotpotQA-4, a pre-specified public direct scorer improves regret from 0.313 to 0.110. In fixed-stack Llama execution, verified route regret improves by 0.190, while the raw-answer gain is 0.010 with an interval crossing zero. A five-family ToolSandbox variant-shift validation exhaustively evaluates 16 declared teams on 14 untouched task variants (224/224 valid rows): the declared-family oracle reaches 0.768 safe-evidence completion, while the prospectively frozen router gets 0.637 (regret 0.131), failing the predeclared 0.10 criterion. A later retrospective comparator reaches 0.655, matching all-workers with 4.57 versus 5.00 workers on average. Thus COVER exposes selection headroom without manufacturing a routing win. A crossed-stack diagnostic shows absolute scores depend on G but finds no detectable router-by-finalizer interaction. COVER is an auditable measurement methodology, not a claim of stack-invariant or universal agent-routing superiority.
[AI-11] Learning to Use Tools: Reinforcement Learning for Tool-Integrated Mathematical Reasoning
链接: https://arxiv.org/abs/2608.28447
作者: Minghui Xu,Zi Wang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Current large language models (LLMs) increasingly benefit from external tool integration, especially for tasks requiring reliable computation and verification. Motivated by this, we study calculator tool calling for improving mathematical reasoning on the Countdown task. We first analyze reasoning failures and find that calculation errors account for a substantial portion of incorrect responses. We then construct supervised fine-tuning datasets to teach the model useful tool-use patterns and how to interpret returned outputs. Building on this tool-formatted policy, we apply several on-policy reinforcement learning methods, including RLOO, RLOO++, GRPO, and DAPO, using automatically verifiable final-answer rewards. To enable a more reliable evaluation, we construct a fresh 1,024-problem held-out Countdown benchmark with no exact overlap with the training data. Our results show that calculator tool integration consistently improves both SFT and RL baselines, yielding roughly 10 percentage-point gains across pass@k. Among the RL methods, Tool-DAPO achieves the strongest performance, improving pass@1 from 35.8% for Tool-SFT to 66.0%. Further analysis shows that RL encourages more effective tool use even when only final-answer rewards are provided. These findings suggest that tool integration reduces arithmetic and verification errors, while RL increases the probability of correct reasoning traces.
[AI-12] Program Learning with Verifiable Rewards: Symbolic Backpropagation for Post-Training LLM s
链接: https://arxiv.org/abs/2608.28421
作者: Vishvesh Bhat
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Post training a language model to reason means updating its weights. Supervised finetuning and reinforcement learning both place the acquired capability inside the model where it cannot be inspected cannot be checked step by step and cannot be moved to another model. We argue that for tasks whose intermediate steps admit verification, reasoning is better placed outside the base models weights as an explicit program composed from deterministic and neural primitives. We introduce PLVR (Program Learning with Verifiable Rewards): a post training method that learns such programs directly from input-output examples. Its mechanism is symbolic backpropagation: each program layer carries a typed ontology a loss is computed at the output against ground truth and required input ontologies are propagated backward by type inference over primitive signatures: an analogue of the chain rule in which credit assignment is a derivation rather than an estimate. Where RLVR verifies a terminal outcome, PLVRs reward is a per step contract verdict dense over program structure. On LiveCodeBench v6 and Tau2Bench, 30B base models with PLVR outperform RL at matched budget by 27.8 points on average and frontier models an order of magnitude larger by 13.6 points. A single primitive library serves two benchmarks, so the marginal cost of a new task is 100 examples of program search and no new finetuning data. Replacing the loss guided search with uniform sampling over the same type admissible space at equal budget collapses the median program from 65.6 to 17.5, identifying the backward pass rather than the type system as the source of the advantage. We release the symbolic backpropagation library and a conformance checker so the method can be applied to primitive libraries other than our own.
[AI-13] LongPIBench: A Long-Context Benchmark for Prompt Injection EMNLP’26
链接: https://arxiv.org/abs/2608.28411
作者: Yupei Liu,Yuqi Jia,Neil Zhenqiang Gong,Jinyuan Jia
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注: To appear in Findings of EMNLP’26
Abstract:Prompt injection attacks pose a serious security risk to large language models in real-world applications. However, existing prompt injection benchmarks primarily focus on short-context inputs, leaving the attacks and defenses in long-context settings largely unexplored. This gap leads to a substantial overestimation of the effectiveness of current defenses. In this paper, we bridge the gap by introducing LongPIBench, a long-context benchmark for prompt injection covering 4 realistic application scenarios: paper peer review, resume screening, code review, and email summary. For each scenario, we construct a synthetic dataset and a real-world dataset, with context lengths ranging from thousands to tens of thousands of tokens. The evaluation results on LongPIBench reveal significant vulnerabilities of prompt injection defenses under long-context settings: even simple heuristic prompt injection attacks achieve high success rates and frequently bypass state-of-the-art defenses. We hope LongPIBench can serve as a practical benchmark for systematically evaluating prompt injection defenses in realistic long-context scenarios.
[AI-14] VERA-8B: Evidence-Grounded Audit Risk Reasoning from SEC Filings
链接: https://arxiv.org/abs/2608.28402
作者: Menghan Liu,Elynn Chen
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Across audit applications, judgments must be supported by reasonable evidence. However, standard financial language models prioritize fluency over evidence. They are built for general financial reasoning and may produce plausible but ambiguous answers, creating a grounding gap that makes them unsuitable for audit work. We address this gap with VERA-8B, a new end-to-end audit reasoning system that identifies audit risks before enforcement actions occur. Constructing such a model raises several challenges, as no prior machine learning work targets pre-enforcement audit prediction. To our knowledge, we are the first to unify SFT and GRPO for evidence-grounded audit reasoning under one evidence standard, achieving performance that surpasses all evaluated baselines. Because auditing cannot tolerate unsupported claims, we introduce abstention and uncertainty qualification to defer uncertain or evidence-incomplete cases. Finally, we design an AuditBridge to ground model reasoning for practical audit work. It transforms raw filings into verified records and then into reviewer-ready reports, bridging finance and computation with broad generality. Together, these components produce auditable, review-ready outputs suitable for practical audit work.
[AI-15] RetailAgent : Structured Adverse Timing in Self-Conditioned Multimodal LLM Trading Agents
链接: https://arxiv.org/abs/2608.28399
作者: Yupeng Zhang,Liuyuan Jiang,Hongyi Huang,Bingheng Li,Lisha Chen
类目: Artificial Intelligence (cs.AI); Trading and Market Microstructure (q-fin.TR)
备注:
Abstract:In financial markets, a sequential policy that reacts systematically to price movements may become predictable to other market participants. This paper studies whether large language model (LLM) agents exhibit such directional structure through RetailAgent, an experimental framework in which an LLM observes anonymized intraday equity price histories and permitted state, then repeatedly chooses long (hold the stock) or flat (stay out) before the subsequent interval return is revealed. We compare returns during long and flat intervals along the same stock’s intraday path after removing the overall fraction of long decisions. This exposure-matched measure reveals persistent negative timing across modality, horizon, state, and model family. Shuffling saved action sequences substantially attenuates the effect, showing that alignment between actions and subsequent returns drives the negative score. Feeding self-authored memories into decisions further increases policy persistence, while timing becomes more negative among stock-days on which the agent uses both actions. These results reveal stable, recoverable directional structure in sequential LLM financial decisions and a behavioral signal for studying how another participant could respond to a predictable policy.
[AI-16] ming-Aware Repurchase Prediction for Web-Scale E-Commerce: Survival Models for Multi-Surface Grocery Recommendation
链接: https://arxiv.org/abs/2608.28393
作者: Akshay Kekuda,Shreeranjani Srirangamsridharan,Ishan Bhatt,Yanan Cao,Sinduja Subramaniam,Evren Korpeoglu,Kaushiki Nag,Kannan Achan
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: ReSys 2026
Abstract:Repurchase recommenders in e-commerce are commonly framed as a binary question asking “will this customer buy this item within W days”, a formulation that requires a separately trained model for every horizon of interest. We replace this stack with survival models that predict time-to-repurchase directly, and evaluate them on millions of customers from a major grocery e-commerce platform across more than thirty ablation configurations. Our study makes three contributions. First, an empirical hazard analysis reveals a slightly decreasing marginal hazard (k ~ 0.9), differing from the common intuition that grocery items become more likely to be repurchased the longer since the last purchase (increasing hazard, k 1). Log-Normal achieves the best marginal fit (R^2 = 0.998) and the best ranking, despite Weibull providing the best conditional residual fit, revealing an apparent discrepancy we analyze in detail. Second, a single Accelerated Failure Time (AFT) model replaces three per-horizon binary classifiers, matching or exceeding each at its own horizon while using roughly 3x fewer total trees. Feature importance reshuffles under the survival objective: channel-cadence and recency signals rise while aggregate frequency counts fall. Third, a 4-parameter parametric calibration maps raw survival CDFs to per-horizon probabilities with zero cross-horizon monotonicity violations. Calibration quality varies by an order of magnitude across the AFT family: Exponential AFT (Weibull k=1) achieves expected calibration error (ECE) ~1e-4, roughly 10x lower than Log-Normal, while ranking metrics agree within 0.3% relative. We adopt Exponential AFT for probability-consuming surfaces and Log-Normal for pure ranking, exposing a principled calibration-ranking trade-off within a single AFT family.
[AI-17] MAP: A Benchmark on Multimodal Accessibility Planning for Real World Places
链接: https://arxiv.org/abs/2608.28384
作者: Jason Armitage,Ioannis Tsochantaridis,Linda Mazzone,Chuqiao Yan,Srini Narayanan,Sarah Ebling
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:We introduce MAP, the first benchmark to evaluate multimodal AI systems as assistants for users with accessibility requirements when planning visits to places in the real world. In our evaluation, systems are presented with requests to verify or recommend a point of interest meeting an accessibility requirement. MAP contains two novel assessments: Claim verification for accessibility planning assesses if information on places and stated accessibility features is supported and identifies places that satisfy requested accessibility features. Visual evidence retrieval for accessibility planning checks if a multimodal AI system can select visual evidence for the requested place and accessibility feature. Our methodology supports comparison of AI systems in a setting where place information and accessibility information can change over time by evaluating systems and refreshing ground truth data at scheduled times. The benchmark is based on automatic rating and human rating for a proportion of responses.
[AI-18] EvoUndo: Recoverability-Constrained Self-Evolution for LLM Agent Harnesses
链接: https://arxiv.org/abs/2608.28363
作者: Tanmay Sah,Dolly Sah,Harshul Jain,Tanya Sah
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:LLM agents increasingly modify their own prompts, tools, middleware, resources, and execution harnesses at runtime. Such self-evolution can improve capability, but a successful mutation may leave persistent effects that cannot be safely reversed in states different from the one in which it was created. We introduce EvoUndo, a framework for representing, synthesizing, diagnosing, and independently verifying recoverability of model-generated self-modifications across counterfactual states. Across 600 unseen one-shot self-evolution tasks, we identify 197 capability-improving mutations that fail recoverability verification. Under the original recovery representation, conventional repair strategies recover 0/197 of these natural failures. Deterministic oracle analysis recovers 48/197 under the original recovery language L0, while the extended recovery calculus increases empirical oracle recovery to 191/197. A protocol-locked 2x2 grounding-by-expressivity intervention then separates two bottlenecks: exact state-address grounding increases successful recovery from 0/48 to 38/48 (79.2%) when the original language is sufficient, while extending the recovery language enables recovery on 142/143 (99.3%) failures in the oracle-defined S1 stratum. On the primary gpt-oss-120b backbone, adding exact-address diagnostics to the richer language reduces recovery to 133/143 (93.0%); a Qwen3.8-27B replication preserves the grounding and expressivity effects but not this negative interaction, indicating that the latter is model-dependent. These results indicate that reliable agent self-evolution requires co-designing verification, state grounding, witness semantics, and recovery-language expressivity rather than relying on iterative prompting alone.
[AI-19] Optimal Adversarial Testing: Extracting Honest Test Results from Dishonest Test Takers
链接: https://arxiv.org/abs/2608.28362
作者: Owen Cox,April Xu,Weiyu Xu
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computer Science and Game Theory (cs.GT); Signal Processing (eess.SP); Applications (stat.AP); Methodology (stat.ME)
备注: 9 pages
Abstract:In applications, it is often required to test objects or people to determine their qualities in terms of certain metrics. However, besides being naturally noisy, the test results can be corrupted by adversarial behaviors of objects or people being tested (test takers). For example, dishonest test takers can cheat in the exams to distort the test results. With the development of AI technologies, such distortions driven by cheating using AI technologies are becoming more commonplace and severe. In this paper, we propose optimal testing strategies which can still recover needed test results even if there are cheaters polluting the results. The proposed testing strategies will optimally re-test selected group of test takers using different testing security measures. We determine the optimal testing strategies using a dynamic programming method.
[AI-20] GRACE:Gradient-guided Coreset Selection for LLM Unlearning EMNLP
链接: https://arxiv.org/abs/2608.28361
作者: Praveen Bushipaka,Andrea D’Angelo,Lucia Passaro,Tommaso Cucinotta
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 20 pages, 16 tables, 5 figures, accepted to EMNLP Findings
Abstract:Machine Unlearning methods for Large Language Models typically assume pre-specified forget and retain sets. In realistic settings, however, requests may provide only a few examples of undesired behavior, requiring forget and retain sets to be inferred from heterogeneous corpora. We study this data-selection problem and propose GRACE , a gradient-guided coreset selection method that constructs both forget and retain sets for LLM unlearning. GRACE first computes a forget direction from seed examples that elicit the undesired behavior, then selects a compact forget coreset whose gradients approximate this direction using non-negative orthogonal matching pursuit. To preserve model utility, it selects retain examples after projecting out the forget direction and applying clustered orthogonal matching pursuit in the remaining gradient space. Across two target domains, two model families, and four unlearning algorithms, GRACE improves model utility while maintaining comparable forget quality, with particularly consistent gains over prior gradient-based selection methods.
[AI-21] Propagating construction-time knowledge quality into medical question answering: A framework grounded in clinical guidelines
链接: https://arxiv.org/abs/2608.28360
作者: Jie Hu,Junjie Wang,Shan Lu,Yifang Hu,Gong Cheng,Yun Liu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models have facilitated knowledge graph (KG) construction from clinical guidelines, but extracted triples vary in structural validity and evidential support. Meanwhile, graph-augmented question answering (QA) systems typically optimize query relevance during retrieval, with limited reuse of quality information produced during KG construction. This creates a disconnect between construction-time quality control and inference-time evidence use. We investigate whether construction-time triple quality can serve as a persistent signal for downstream evidence selection and presentation. We propose a quality-aware framework that models structural conformance (SchemaConf) and evidential support (EvidScore) as complementary dimensions and fuses them into a per-triple quality signal, Q(t). Rather than using quality solely for filtering, the framework retains Q(t) and derived quality tiers as graph attributes and propagates them into quality-weighted subgraph retrieval and tier-conditioned evidence prompting, while preserving passage-level provenance. Experiments on Chinese diabetes clinical guidelines show that the utility of the quality signal is distribution dependent. Under cross-version and cross-model shift, the fused Q(t) provides stronger triple-quality discrimination than either component alone (AUC 0.748 vs. 0.703 for EvidScore and 0.645 for SchemaConf). In guideline-grounded QA, propagating construction-time quality reduces required-knowledge omission from 16.3% to 5.3% and conflicting outputs from 16.3% to 2.7%, with an evidence-grounded precision of 81.6% and near-zero invalid citations. Blinded clinician ratings favor the full framework over no retrieval (4.68 vs. 4.21 on a five-point scale) and approach the oracle condition (4.80), while cross-generator experiments show consistent trends.
[AI-22] AGENT -O: A Semantic Agent Card Framework for Interoperable and Governed Healthcare AI Agents
链接: https://arxiv.org/abs/2608.28345
作者: Pengze Li,Cui Tao
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:AGENT-O is a modular ontology framework that defines a semantic Agent Card for representing health-oriented AI agent systems and supports assessment of reporting completeness in scientific publications. AGENT-O was developed as an OWL 2/RDF ontology covering runtime, models, workflow, tools, clinical use, evaluation, provenance, governance, and reporting assessment. Evaluation included ontology inventory, OWL-RL reasoning, three SHACL suites, 12 SPARQL competency queries, three cases, and model-assisted reporting-completeness assessment of 279 papers across five dimensions. The ontology contained 1,962 RDF triples and 1,922 Protege axioms, with 252 active classes, 198 active object properties, and 51 datatype properties. All SHACL suites conformed on example graphs, all competency queries returned prespecified evidence, and all 279 papers were scored. Incomplete reporting was highest for runtime/architecture (84.6%), governance/safety (82.8%), and provenance/reproducibility (78.1%), compared with evaluation (25.8%) and benchmark-process alignment (29.8%). AGENT-O supported semantic Agent Card representation and reporting assessment while revealing an evaluation-specification gap: evaluation and benchmark procedures were reported more consistently than runtime architecture, governance, and reproducibility. AGENT-O provides a reusable ontology, semantic Agent Card profile, and reporting-completeness workflow for structured reporting and gap identification, but does not assess agent quality or deployment readiness.
[AI-23] Real-Valued Hyperdimensional Sequence Representations with Hadamard Product Binding and Shift Equivariance
链接: https://arxiv.org/abs/2608.28334
作者: Kenny Schlegel,Dmitri A. Rachkovskij,Denis Kleyko,Amy Loutfi,Stefan Streif,Evgeny Osipov
类目: Artificial Intelligence (cs.AI)
备注: 23 pages, 5 figures
Abstract:Encoding temporal order is a fundamental requirement for sequence representations in Hyperdimensional Computing. Fractional Power Encoding provides similarity-preserving position vectors whose inner products approximate shift-invariant kernels, and it supports shift-equivariant transformations of encoded sequence representations. However, standard formulations of Fractional Power Encoding are primarily designed for binding operations such as circular convolution or complex-valued multiplication, which limits their compatibility with Hadamard product binding of real-valued vectors. This paper develops real-valued position encodings motivated by Random Fourier Features, aiming to retain the desirable properties of Fractional Power Encoding while supporting Hadamard-based operations. We propose three real-valued position-encoding variants: a real-valued baseline based on the inverse Fourier transform, and Sinusoid and Cosine-only representations derived from Random Fourier Features. Among them, the Sinusoid variant provides an explicit algebraic shift operator, allowing temporal shifts to be applied directly to the vector-encoded sequence representation without re-encoding the shifted sequence. Experiments on time-series classification datasets show that the proposed real-valued representations achieve performance comparable to standard Fractional Power Encoding while enabling computationally efficient Hadamard product binding. The Sinusoid variant offers the most favorable trade-off, combining efficient real-valued implementation with exact shift-equivariant transformations.
[AI-24] MAIL: Memory-driven Adaptive Incremental and Literature-grounded Framework for Hypothesis Generation in Chemistry
链接: https://arxiv.org/abs/2608.28315
作者: Mahdi Babaei,Xueshen Li,Yutao Kuang,Jolene P. Reid,Yu Gan
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:The ever-expanding volume of the chemical literature offers unprecedented opportunities to generate novel and impactful hypotheses. However, the bottleneck lies in efficiently navigating this vast knowledge base to formulate high-quality, experimentally meaningful insights. While Large Language Models (LLMs) show promise for this task, existing methods often rely on static inspiration corpora, predefined heuristics, or laborious human-in-the-loop pipelines and decision-support frameworks that limit scalability and novelty. In this work, we propose an automated approach, a Memory-augmented, Adaptive, Incremental, and Literature-grounded (MAIL) framework for hypothesis generation in chemistry. Our MAIL method formulates hypothesis generation as a temporally grounded, memory-driven reasoning process, where hypotheses emerge from an evolving conceptual path that continuously accumulates and reinterprets prior knowledge. We evaluated the MAIL framework on a public TOMATO-Chem dataset and a newly curated and disseminated high-novelty nature/science challenge (HN-NS) dataset. Across both datasets, MAIL generates structurally coherent and mechanistically plausible hypotheses, achieves the highest MIOS and MPOS by more effectively recovering the central ideas and methodological elements of the historical target hypotheses, and obtains the highest overall expert-evaluation scores for scientific quality. These results demonstrate the potential of LLMs to autonomously explore chemical domains and generate hypotheses that are both innovative and chemically plausible. Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2608.28315 [cs.AI] (or arXiv:2608.28315v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.28315 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-25] Deriving Scaling Laws for OpenEuroLLM Models: Learning Rate Batch Size and Loss
链接: https://arxiv.org/abs/2608.28308
作者: Niccolò Ajroldi,Diana Alexandra Onutu,Haider Al-Tahan,Jörg Franke,Sampo Pyysalo,Jenia Jitsev,Aaron Klein
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:We study the scaling behavior of learning rate and batch size in pretraining dense large language models on English-prevalent corpora. Beyond scaling \textitjointly optimal learning rates and batch sizes, we investigate their \textitmarginal evolution with model capacity and data scale and develop a model that captures these relationships. As we employ a Warmup-Stable-Decay learning rate schedule, we further investigate the gains from learning rate annealing over a broad range of hyperparameters settings, models and data budgets, and whether the optimal learning rate and batch size \textittransfer between the stable and decay phases. Finally, we characterize the dependence of loss on model capacity and dataset size, evaluating recently proposed scaling forms that explicitly model their interaction. We find these approaches particularly effective at capturing both undertraining and overtraining regimes across our experiments. This study establishes a first baseline and scaling procedure for the development of future OpenEuroLLM models. We open-source the complete collection of pretraining runs used in this study.
[AI-26] PanelShield: Verifiable Closed-Loop Safe Planning for Robotic Industrial Panel Operation
链接: https://arxiv.org/abs/2608.28305
作者: Guipeng Xin,Jiahe Xu,Chenhui Wan,Jie Liu,Youmin Hu,Zhongxu Hu
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:
Abstract:Industrial panel operation is knowledge-intensive and safety-critical. Beyond control recognition and action generation, execution must satisfy constraints in operation manuals and safety regulations. While foundation-model-based planners show strong semantic capability, they typically lack computable, localizable, and reproducible mechanisms for violation detection and repair. To address this, we propose PanelShield, a verifiable closed-loop safety planning framework for manual-guided industrial panel operation. The framework generates parameterized action primitive sequences from task-relevant manual evidence and applies dual formal verification with LTL and a Safety FSM to enforce cross-step temporal correctness and local transition legality. When violations occur, it outputs a structured counterexample with the earliest violating step and cause, enabling targeted repair and re-verification. We build a multi-level long-horizon planning benchmark covering three representative industrial device panels, and evaluate the framework in simulation and real-world robotic experiments. Results show that PanelShield improves complex safety-constrained task performance over foundation-model-only planning baselines while reducing the violation rate to 2.7%, with 4.1 s total latency. Real-world experiments demonstrate end-toend feasibility. Overall, PanelShield offers a verifiable approach to robotic panel operation that balances flexibility, safety, and auditability.
[AI-27] MaCoPlanner: LLM -Assisted Manual-Compiled Task Planning with Proactive Safety Verification for Robotic Industrial Panel Operation
链接: https://arxiv.org/abs/2608.28300
作者: Guipeng Xin,Jiahe Xua,Mohammad Deghat,Chenhui Wan,Jie Liu,Youmin Hu,Zhongxu Hu
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:
Abstract:Robotic industrial panel operation requires not only accurate control localization but also compliance with operating procedures, safety rules, and device-state constraints distributed across heterogeneous manuals. This study presents MaCoPlanner, a task-planning framework built on knowledge compiled from equipment manuals that converts equipment manuals into a typed intermediate representation, retrieves task- and state-relevant evidence, and uses it to support plan generation. Before actuation, candidate plans are symbolically rolled out and checked against procedural and state-transition constraints; detected violations are localized and returned for targeted repair, while unresolved plans are rejected. A separate execution interface grounds verified symbolic actions to physical controls and updates the device state. Under an independent evaluation oracle, MaCoPlanner achieves a final violation rate of 2.7%, and 26.3% of the runs in the repair analysis are rejected after exhausting the refinement budget. Compared with Raw-Manual, task success increases from 62.8% to 84.4% on Level-2 tasks and from 25.9% to 43.2% on Level-3 tasks. Experiments on a controller-panel simulator without an attached industrial load further demonstrate integrated execution feasibility under representative interaction conditions, without claiming industrial deployment readiness.
[AI-28] Memristive-Friendly Hadamard Reservoir Computing: Structured Multiplier-Free Recurrences at Scale
链接: https://arxiv.org/abs/2608.28295
作者: Andrea Ceni,Gianluca Milano,Carlo Ricciardi,Claudio Gallicchio
类目: Artificial Intelligence (cs.AI)
备注: submitted to Neurocomputing
Abstract:Reservoir Computing (RC) designs Recurrent Neural Networks around a fixed, i.e., untrained, recurrent layer, and is a natural candidate for neuromorphic hardware. Memristive-friendly reservoirs derive the neuron dynamics from memristive-device kinetics, but still rely on dense recurrent matrices, which are expensive to realize physically. In this paper, we replace the dense matrix with a structured orthogonal operator, built from sign diagonals, a permutation, and a fast Walsh-Hadamard transform. The operator is multiplier-free, requires O(N) parameters and O(N\log N) operations per step, and is never materialized as a matrix. We instantiate it in a standard and in a memristive-friendly Echo State Network, with one binary input connection per unit. Our mathematical analysis shows that exact orthogonality yields an echo state condition that is tight in the recurrent scaling, and a noise response that is predictable at design time. Moreover, the operator mixes the whole state in a single application. Experiments on twenty classification and seven regression benchmarks, at reservoir sizes up to N = 8192 , show that the structured models match dense orthogonal reservoirs, and achieve better mean performance than the cycle reservoir by a margin that widens with size. Furthermore, we time the recurrent step on three hardware platforms, where it is up to 50\times faster than a dense product and 10^4\times smaller in memory. Finally, we ablate the operator and measure the response to noise, quantization, device mismatch and discrete faults. Comments: submitted to Neurocomputing Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2608.28295 [cs.AI] (or arXiv:2608.28295v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.28295 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-29] LoopArena: Benchmarking Models as Runtime Controllers for Loop Engineering
链接: https://arxiv.org/abs/2608.28281
作者: Yi Wang,Haopeng Zhang,Chengxiang Huang,Rui Dai,Kaikui Liu,Piotr Koniusz,Xiangxiang Chu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Loop Engineering is emerging as a practice for organizing development work around coding agents. Instead of writing each prompt by hand, practitioners design loops that monitor progress, assign work, run checks, and decide what the agent should do next. Even with a capable coding agent, a loop may trust a stale progress note, skip needed verification, spend its budget in the wrong direction, or stop before the task is safe to submit. Yet the final outcome of one end-to-end run cannot tell whether success or failure reflects the loop’s guidance or the coding agent’s ability to carry out the task. We introduce LoopArena, a benchmark for evaluating how well one model can guide a separate coding agent through a long-running task. The model under evaluation is the \textbfController: after each coding round, it receives a structured summary of the run and instructs a separate, fixed coding agent, the \textbfWorker, on what to do or verify next, or decides whether to stop. LoopArena evaluates this ability in three complementary settings that differ in execution scope and cost. Type I scores next-step Loop Contract selection through execution-validated questions without running the Worker at evaluation time. Type II executes repeated control over a selected slice of a full task, while Type III evaluates the paired full task from its original state. On full tasks, the best observed Strict Success Rate is \textbf24.69%, leaving substantial room for improvement in long-horizon loop control. Across Controllers, the paired reduction in estimated inference cost averages \textbf64.4%, and Type II produces a similar ordering under the main Core criterion (Spearman’s (\rho=\textbf0.9747)). We release the benchmark data and evaluation code at this https URL .
[AI-30] RECAST: Recent Context-Aware Sampling for Test-Time Adaptation in Streaming Biosignals
链接: https://arxiv.org/abs/2608.28271
作者: Yong-Yeon Jo,Junho Song,Joon-myoung Kwon
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Streaming biosignals vary across subjects and drift over time, so population-trained models lose accuracy during long-term monitoring. Test-time adaptation (TTA) enables online personalization by updating the model on incoming samples. But in a stream, a basic question is left open: \emphwhich samples should drive each update? Using all buffered samples blurs the update with irrelevant segments. Using only the latest segment makes the update noisy and unstable. The most useful samples are recent, aligned with the current physiological state, and reliable enough to learn from. We propose \textbfRECAST (REcent \ Context-Aware Sampling for TTA), a lightweight sampling module for buffered TTA frameworks. RECAST builds each adaptation batch from three signals: temporal recency, contextual similarity, and predictive reliability. It changes only which samples are used, leaving the model and the training objective unchanged. On two blood-pressure datasets, RECAST improves estimation accuracy and trend tracking over baselines and ablations. The per-patient gains are statistically significant on both datasets, with broad improvement on the regular benchmark and gains concentrated on the hardest patients in the emergency-department setting. RECAST stays practical, adding only sub-second latency per segment on a single GPU and CPU core.
[AI-31] Spatial-Semantic Reasoning using Large Language Models for Efficient UAV Search Operations
链接: https://arxiv.org/abs/2608.28270
作者: Marin Maletic,Marijana Peti,Tamara Petrovic,Stjepan Bogdan
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: 8 pages, preprint, Published in: 2025 European Conference on Mobile Robots (ECMR), DOI: https://doi.org/10.1109/ECMR65884.2025.11163229
Abstract:We present a real-time semantic navigation framework for Unmanned Aerial Vehicles (UAVs) focused on improving time efficiency in the Object Goal Navigation (ObjectNav) task. Central to our approach is a Large Language Model (LLM) that interprets user-provided natural language instructions and performs semantic reasoning over detected objects and spatial context to prioritize high-probability search regions. The system combines real-time object detection, 3D spatial mapping, and polynomial spline interpolation for smooth and feasible UAV trajectory planning. Unlike prior methods that rely on offline reasoning or simulator-constrained action spaces, our framework can operate in real time, continuously updating semantic relevance based on new observations. Experiments in both simulated and real-world settings demonstrate reductions in mission duration while maintaining high search accuracy, underscoring the effectiveness of LLM-guided reasoning for time- efficient UAV-based ObjectNav.
[AI-32] Finding Where the Buck Stops: An Automated Failure Attribution-Based Reflection Framework for Multi-Agent Collaboration EMNLP2026
链接: https://arxiv.org/abs/2608.28264
作者: Xiaoqing Wang,Keman Huang,Bin Liang,Hongyu Li,Xiaoyong Du,Wuqiong Pan
类目: Artificial Intelligence (cs.AI)
备注: Accepted by EMNLP 2026 main
Abstract:Multi-agent systems (MAS) powered by large language models have shown promise for complex tasks but suffer from high failure rates. Current self-reflection methods for MAS require all agents to reflect upon failure, overlooking a critical reality: failures typically stem from a specific agent leading the task astray, namely the decisive error agent, while others merely fulfill their regular duties. Forcing regular-behaving agents to reflect contaminates their memory with wrong insights. Hence, we propose DoCtOR (Diagnose-then-Correct PPO-enhanced Reflection), a novel reflection framework that enhances multi-agent collaboration. DoCtOR first identifies the decisive error step and decisive error agent through automated failure attribution, then employs counterfactual reasoning to generate a corrected decisive error step, and finally engages only the decisive error agent to produce targeted reflections. Experimental results show DoCtOR achieves 22%, 26%, and 27% improvements over initial success rates on HotPotQA, ChartQAPro, and Mind2Web datasets, outperforming Reflexion, Retroformer, and COPPER. We further establish the generalizability of our diagnose-then-correct paradigm and demonstrate that in low-resource settings, focusing reflection on reasoning steps after the decisive error step achieves comparable quality to reflecting on the complete failure trajectory.
[AI-33] Regime-Aware Portfolio Management via Retrieval-Augmented LLM -Guided Expert Switching
链接: https://arxiv.org/abs/2608.28252
作者: Ahmad Asadi,Reza Safabakhsh
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Financial markets are inherently non-stationary, making the effectiveness of individual portfolio-management strategies highly dependent on changing market conditions. This work proposes a retrieval-augmented expert-switching framework that dynamically selects portfolio management experts based on their historical performance under similar market situations. A dual-stream variational autoencoder represents asset-level and market-wide information, while a retrieval-based knowledge base stores historical situations and expert performance. During inference, an instruction-tuned LLM reasons over the retrieved evidence to identify the most appropriate expert rather than directly generating portfolio actions. We further establish a monotonicity property showing that adding a locally superior expert cannot degrade the switching mechanism’s performance. Experiments across cryptocurrency, stock, and foreign-exchange markets show that the proposed selector achieves the highest cumulative return and Sharpe ratio among the evaluated selection strategies in all three markets. In the stock market, for example, cumulative return increases from 26% for the best fixed expert to 34%, while the Sharpe ratio improves from 0.74 to 0.96. Ablation results confirm the importance of both retrieval and LLM reasoning, while experiments with different expert-pool sizes demonstrate the value of complementary expertise. Overall, the findings support retrieval-grounded expert switching as an effective approach to adaptive portfolio management in non-stationary financial environments.
[AI-34] raining-free Suction Grasp Detection for Deformed Aseptic Cartons Using Vision-Language Models and Geometric Surface Scoring
链接: https://arxiv.org/abs/2608.28246
作者: Marin Maletic,Goran Vasiljevic
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: 6 pages, ICCAS 2026 preprint
Abstract:Robotic sorting of recyclable waste is challenging due to the deformable and geometrically inconsistent nature of target objects. We present a training-free suction grasping system for sorting deformed aseptic beverage cartons, decoupling target identification from grasp-point selection. An open-vocabulary vision-language model detects cartons from a text prompt, SAM2 refines each detection into an instance mask, and a geometric scoring method selects the suction point by combining surface flatness with normal alignment. Three geometric methods are compared: k-nearest-neighbour PCA, Sobel cross-product, and RANSAC plane fitting. Evaluated on a real robot across three deformation levels and 35 cluttered scenes, single-object grasp success reaches 88.2% and end-to-end retrieval in clutter is 72.6%.
[AI-35] Beyond Task-Only Matching: Personalized Skill Routing with Counterfactual Evaluation
链接: https://arxiv.org/abs/2608.28241
作者: Tianle Wang,Yanghe Zou,Xiang Liu,Ziyao Huang,Chenchen Fu,Weiwei Wu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:The rapid expansion of reusable skill repositories makes skill routing a critical capability for large language model (LLM) agents. Existing methods treat routing as task-only semantic matching. However, when users with incompatible constraints issue an identical request, this assumption conflates task relevance with skill suitability: a task-only router can select a semantically plausible skill that is unsuitable for the requesting user. To expose this failure mode, we formulate \textitpersonalized skill routing as profile-conditioned retrieval, in which relevance depends jointly on the task and the user profile. We first introduce a profile-counterfactual benchmark, in which the task is held fixed while changes in the user profile induce changes in the reference skill. We further construct paired counterfactual supervision and propose SkillFeed, a progressive retrieve-and-rerank framework that first establishes task–skill alignment and then learns profile-conditioned discrimination. By retrieving body-level evidence and reranking semantically similar but profile-conflicting candidates, SkillFeed identifies skills that satisfy both task requirements and user constraints. On SkillFeed-Bench, SkillFeed attains 75.1% top-1 retrieval accuracy, a 23.1-point improvement over the corresponding pretrained routing baseline. Adding profile conditioning yields a 35.1-point gain on queries where user profile changes the reference skill. This contrast shows that user profiles are most consequential precisely when they change skill suitability. Our website is publicly available at this http URL .
[AI-36] REINS: Refusal-Enhanced Inhibitory Steering with Sparse Autoencoder Features EMNLP2026
链接: https://arxiv.org/abs/2608.28233
作者: Kai-Xuan Ding,Hao-Xiang Xu,Ji-Hua Peng,Zi-Qi Chen,Jiaqi Wang,Zhen-Hua Ling
类目: Artificial Intelligence (cs.AI)
备注: Accepted at EMNLP 2026 Main Conference
Abstract:Steering with Sparse Autoencoders (SAEs) offers a lightweight inference-time path for adapting the behavior of large language models without retraining. By exposing sparse and interpretable features, SAE steering provides a promising interface for safety control that guides harmful continuations toward refusal. However, we observe that complex wrappers can still undermine existing SAE steering methods on harmful prompts. To evaluate this failure mode systematically, we construct Generalized Undercover Instruction Safety Evaluation (GUISE), a dataset of harmful prompts with complex wrappers. Existing single direction SAE steering methods do not reliably produce refusals on harmful prompts, suggesting that refusal enhancement alone can be too weak when the harmful continuation path remains active. This motivates us to propose Refusal-Enhanced INhibitory Steering (REINS), which suppresses harmful continuation features and enhances safe refusal features in the same SAE feature space. Experiments on GUISE and other datasets show that prior methods either intervene too weakly or achieve only apparent safety through collapse, while REINS substantially reduces harmful responses, markedly improves safe refusals and largely preserves general capabilities.
[AI-37] Performative Privacy: When Differential Privacy Maximizes Utility
链接: https://arxiv.org/abs/2608.28198
作者: Uddalak Mukherjee,Edwige Cyffers,Yann Chevaleyre
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Machine Learning (stat.ML)
备注: Accepted at EuroTDP 2026
Abstract:Privacy-preserving learning is often motivated by the idea that protecting users’ data can preserve trust and thus participation, improving utility in the long term. However, this claim has not been formalized so far. In parallel, performative learning provides a framework for studying learning systems whose deployment affects the data they later observe. In this work, we bring these two perspectives together and introduce \emphperformative privacy, where data leakage reduces future participation. We study a simple model where agents repeatedly contribute data for mean estimation but may leave the system when their data is leaked. Privacy is implemented through differentially private mechanisms, creating a trade-off between estimation noise and future participation. We show, through a theoretical study of the dynamics and numerical experiments, that a finite privacy budget can outperform non-private estimation in the long term when the feedback loop between leakage and participation is sufficiently strong. This provides first evidence that differential privacy can be optimal not only as a protection mechanism, but also from the perspective of long-term utility.
[AI-38] Beyond Flat Netlist: Hierarchical Graph Representation Learning for Scalable Analysis of Sequential Circuits
链接: https://arxiv.org/abs/2608.28188
作者: Jingyi Zhou,Zhengyuan Shi,Jiaying Zhu,Ziyang Zheng,Qiang Xu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Hardware Architecture (cs.AR)
备注:
Abstract:Circuit Representation Learning (CRL) offers a powerful paradigm to guide and optimize core Electronic Design Automation (EDA) tasks, but its practical adoption is hindered by the immense scale of industrial netlists and a failure to explicitly model register-level temporal dynamics. To overcome these barriers, we introduce DeepSeq3, a novel hierarchical framework that abstracts circuits into a two-level representation: fine-grained combinational subgraphs partitioned by flip-flops (FFs), and a high-level Super-Node Graph (SNG) that models the register-transfer structure. A dual Graph Neural Network (GNN) architecture learns representations at both levels, capturing local Boolean logic and global state transitions. Crucially, we introduce a state-centric pre-training scheme that predicts the reachability between FF states, endowing the model with a deep understanding of temporal behavior. Demonstrated on large-scale benchmarks, DeepSeq3’s approach yields superior scalability and richer representations, reducing bounded model checking (BMC) solving time by 18% while guaranteeing correctness.
[AI-39] Expert Knowledge Machine Understanding: Bridging Reactomes Ontology with LLM Semantic Embeddings
链接: https://arxiv.org/abs/2608.28178
作者: Susanna Bravi,Riccardo De Luca,Rosa Sicilia,Christine Nardini,Mario Santoro
类目: Artificial Intelligence (cs.AI)
备注: Accepted at the CIBB 2026 conference ( this https URL )
Abstract:Biological knowledgebases like Reactome provide high-quality pathways that include biological elements’ relationships and textual descriptions (metadata). The quality of such pathways is granted by manual curation, that presents, however, significant scalability challenges. Lately, numerous NLP tools have been proposed to cope with this issue, leveraging textual information to automatically expand biological knowledgebases. However, little exploration has been done so far to assess whether relationships among textual descriptions mirror higher order biological relationships. This study explores whether human-written descriptions in Reactome can be used to infer the experts’ defined global hierarchical structure. To test this, we extracted from Reactome the Homo Sapiens hierarchy of pathways and their reactions (Reactome Hierarchy), and used textual metadata to reconstruct a Semantic Hierarchy, combining a sentence transformer model (SPECTER2) with a modified agglomerative nesting algorithm and a graph reconstruction algorithm. Quantitative (Laplacian Spectral Distance and Bootstrapping) and qualitative (global topological metrics) analyses confirm our hypothesis and indicate that the global hierarchical structure of pathways can be inferred by experts textual metadata.
[AI-40] Under-Mattress Temporal Sensing for Next-Day Agitation Risk Scoring in Dementia Wards
链接: https://arxiv.org/abs/2608.28152
作者: Zhen Liu,Marta Bono,Robbe Decloedt,Ajda Flisar,Maarten Van Den Bossche,Maarten De Vos
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Agitation fluctuates over short time horizons in people living with dementia, yet continuous physiological information for anticipating next-day risk is limited. We assessed whether contactless under-mattress signals from the preceding night inform next-day agitation risk and whether preserving minute-level temporal structure improves performance over conventional nightly summaries. We analyzed 423 patient-nights from 65 subjects in a specialized hospital dementia unit using two under-mattress sensing systems. A unified four-paradigm benchmark compared nightly handcrafted summaries, three-period handcrafted features, full-night sequence modeling, and sliding-window multiple-instance learning. Source-specific preprocessing and five-fold patient-grouped cross-validation were used, with performance estimated from pooled out-of-fold predictions. Evaluation included discrimination, calibration, fixed-threshold metrics, and a comparison of period-signal attribution patterns across two temporal models. Full-night sequence modeling achieved the highest discrimination (AUROC, 0.692; AUPRC, 0.849) and balanced accuracy (0.658). Both minute-level pipelines had higher AUROC than nightly summaries, but differences from three-period handcrafted features were uncertain. Cross-model attribution prioritized activity, heart rate, and respiratory rate during the core overnight period. Calibration remained limited. The preceding night’s signals supported modest next-day risk discrimination, with minute-level temporal modeling outperforming nightly summaries. Prospective calibration and external validation are needed before use in individual care decisions. This patient-grouped benchmark identifies contactless overnight sensing as a promising biomedical engineering direction for agitation-risk research in hospitalized dementia cohorts.
[AI-41] he Approximation Rank of Softmax Attention: Sharp Geometric Laws and Robust Interaction Dimension
链接: https://arxiv.org/abs/2608.28150
作者: Yuhe Sui,Jianing Zhang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 16 pages, 1 figure
Abstract:Which geometry controls the rank complexity of normalized softmax attention? We study maximum-row- \ell_1 approximation rank, exactly the least unrestricted rank preserving every bounded vector-valued output. Two sharp worst-case laws isolate support geometry: for fixed d and error \varepsilon , spherical self-attention has rank \Theta_d,\varepsilon(\min\n,(1+\beta)^(d-1)/2) , while full-ball geometry adds one radial degree and, for \beta\ge\beta_0(d,\varepsilon) and n\ge C_d e^\beta/8 , gives \Theta_d,\varepsilon(\beta^d/2) . For a fixed head, row-softmax quotients out row-scalar logit directions: the remaining visible query–key interaction dimension r yields an r/2 per-instance upper law, and bounded constructions show this exponent is minimax sharp. Approximate interaction subspaces incur an explicit residual output error and yield a tolerance-indexed SVD dimension. On an 84-head BERT-base calibration set, we observe modest effective-dimension reductions across many head–temperature settings, together with positive associations with finite constructive rank upper certificates. Together, these results separate support geometry, which sets worst-case temperature scaling, from softmax-visible interaction geometry, which controls per-head approximation complexity.
[AI-42] Post-Edit Re-Verification in Simulator-Backed Engineering Agents : A Controlled Comparison of Verification-Cadence Guidance
链接: https://arxiv.org/abs/2608.28147
作者: Qingchuan Zhu,Shuyue Tong,Pengju Ren
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: 12 pages, 2 figures
Abstract:Engineering agents that interact with external simulators may need to coordinate design modification with reacquisition of engineering evidence for the modified state. We ask whether first post-edit re-verification changes when explicit verification-cadence guidance is retained versus omitted while verification-relevant state/facts are held constant. Cadence-Guided (CG) retained an instruction to request a new simulation after a substantive modification, whereas Cadence-Omitted (CO) removed that instruction; neither condition used a hard gate. The study therefore measures instruction-conditioned post-edit verification-policy adherence rather than spontaneous recognition that prior evidence has become stale. Using DWSIM as the simulator backend and continuous valve-pressure adjustment, five Alibaba/Qwen models were evaluated on eight synthetic cases; each model-case-condition combination was executed three times via live API calls, yielding 120 evaluation slots per condition. Re-verification was observed in 94/120 CG slots versus 32/120 CO slots; cadence violations occurred in 26/120 versus 87/120; and bounded final success was reached in 95/120 versus 35/120. qwen3.5-35b-a3b showed minimal re-verification (1/24 in CG and 0/24 in CO) and no final success in either condition. Within this bounded protocol, explicit post-edit verification-cadence guidance was associated with more re-verification, fewer cadence violations, and more frequent bounded final success, supporting the treatment of verification cadence as an explicit interaction-protocol component.
[AI-43] he Shape of Power: A Multilingual Framework for Social Power Reasoning in Dialogues
链接: https://arxiv.org/abs/2608.28144
作者: Farah Atif,Sougata Saha,Monojit Choudhury
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Social power plays a fundamental role in shaping human interaction, yet computational studies of power remain limited to narrow linguistic and cultural settings. Existing datasets further lack the demographic and relational depth needed for robust cross-cultural analysis. To address this gap, we introduce a theoretically grounded framework for studying social power in naturalistic multilingual dialogue through movie screenplays. The framework integrates a schema informed by social science theory, a native speaker annotation pipeline refined through pilot studies, and a custom interface for scalable cross-lingual analysis. Using this framework, we constructed an initial corpus containing 15,836 annotated instances from 100 scenes in French and Egyptian Arabic movies. Our analysis reveals strong agreement on observable demographic and contextual attributes, while socially interpretive aspects, such as power asymmetry and intention alignment, remain more contested, highlighting the complexity of social power across cultures. We evaluated 6 Large Language Models (LLMs) and Multimodal LLMs on cross-cultural social power reasoning, finding persistent gaps between human and model agreement in relational and theory-of-mind reasoning. Our work introduces the first extensible multilingual framework for studying social power in dialogues and provides an initial evaluation setting for studying cross-cultural social reasoning.
[AI-44] VICT: Verifier-Instrumented Credit Tracing for Long-Horizon LLM Agent Reinforcement Learning EMNLP2026
链接: https://arxiv.org/abs/2608.28128
作者: Pengcheng Li,Zhengyang Zhang,Dongxu Zhang,Sui Huang,Shaohua Ma
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: accepted by EMNLP2026
Abstract:Fine-grained credit assignment is a central challenge in reinforcement learning for long horizon LLM agents. Standard objectives often train from programmatically verifiable terminal rewards by broadcasting each sparse outcome to every action in a trajectory. Existing methods typically seek finer credit from the rollout side, constructing auxiliary trajectory signals or additional comparisons to estimate action importance. Although useful, these approaches still treat the verifier that judged success as a scalar reward, discarding its internal task structure. Our key insight is that many verifiable tasks already encode the relevant checks inside their terminal verifier. We propose VICT (VerifierInstrumented Credit Tracing), a training-time interface that exposes executable or evidence backed atoms and traces them back to actions through dependency-valid proof edges. VICT redistributes group-relative advantage only along those edges, shifting credit assignment from rollout-side inference to verifierside tracing. It preserves the original terminal reward, abstains when evidence is incomplete or ambiguous, and changes only the training-time advantage tensor, requiring no learned critic, process labels, branch rollouts, or inference-time verifier access. On ALFWorld and WebShop, VICT improves substantially over outcome-only training and achieves strong performance alongside recent fine-grained credit methods; ablations rule out dense atom rewards, final-commit credit, temporal proximity, and sparsity as sufficient explanations.
[AI-45] SEPO: Evidence-Grounded Prompt Optimization via Structural Editing
链接: https://arxiv.org/abs/2608.28067
作者: Xiaoyu Ma,Haoyue Liu,Yiwen Li,Jionghao Zhu,Zhichao Wang,Ye Chen,Xiaoying Tang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Existing API-only prompt optimisers are often described as interpretable, but in practice, this usually means only post-hoc inspectability: each iteration still rewrites the prompt as one opaque string, leaving a trace of full-prompt diffs rather than localisable, machine-readable edits. This paper introduces SEPO (Structural, Evidence-grounded Prompt Optimization), a multi-trajectory prompt optimiser centred on edit-effect lineage feedback. Rather than treating each iteration as an isolated whole-prompt rewrite, SEPO locally edits stable, typed units in a two-layer prompt schema, links the target and realised structural operations of each edit to the examples it newly fixes or breaks, and carries this edit-effect record forward to guide later architect calls on the same search branch. This makes prompt optimisation addressable, attributable, and actionable. Across a 14-task held-out suite, SEPO improves over the strongest baseline, GEPA, by 3.1 pp on Llama-3.1-8B-Instruct and 2.2 pp on Qwen3-8B, reaching 61.9% and 73.3% macro accuracy. SEPO also lies on both the optimisation-time and test-time Pareto frontiers, spending 2.9M optimisation tokens versus 4.1M for GEPA and producing prompts over 5x shorter.
[AI-46] Learning to Allocate Incentives for Incentivized Advertising via Offline Model-Based Reinforcement Learning
链接: https://arxiv.org/abs/2608.28065
作者: Zilin Zhao,Han Yang,Tianpei Yang,Fangsheng Huang,Yanfei Cui,Kan Peng,Yi Li,Yiming Zong,Hao Zhang,Yinsong Xue
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Complete your ad view and grab a 5-cent bonus! In incentivized advertising, a platform promises users a bonus before observing downstream ad revenue, encouraging them to click and complete ads. It must balance the incentive promised in advance against the revenue realized afterward: insufficient incentives forfeit monetization opportunities, whereas excessive incentives reduce net profit. Because current incentives may also shape user expectations and future engagement, incentive allocation is a sequential decision problem with delayed revenue, cost sensitivity, and carryover effects. Existing work has not studied decision-making algorithms for this setting. Auto-bidding assumes available ad opportunities, while targeted promotion optimizes incentives outside the ad monetization pipeline. We formulate the problem as an MDP and develop an offline model-based RL framework for cost-controllable sequential incentive allocation. It learns a world model of user feedback and ad revenue, then performs conservative policy optimization. An independent counterfactual scorer evaluates each learned policy on held-out logs, enabling pre-launch selection without costly online exposure. Experiments on large-scale industrial data and online A/B tests show that the scorer provides a stable offline signal. The deployment path from causal inference to offline RL and then Offline-MBRL further validates the framework: MB-IQL improves per-user net profit by 7.96% over TD3+BC, whereas reverting to plain IQL reduces it by 6.56% (both (p0.0001)). Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2608.28065 [cs.AI] (or arXiv:2608.28065v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.28065 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-47] WeAgent -MMSearch: Native Text-Vision Interaction for Multimodal Search Agents
链接: https://arxiv.org/abs/2608.28062
作者: Zongkai Liu,Hui Zhang,Liqiang Niu,Zhen Cao,Han Li,Juntao Liu,Wenchao Chen,Chengduo Zhao,Chao Yu,Fandong Meng
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Multimodal search agents extend parametric knowledge with newly emerging and long-tail evidence from the open web. Yet many existing agentic search environments often expose retrieved evidence only as text and omit tool-returned images from subsequent context, reducing visually grounded trajectories to text-only reasoning. Long-horizon interaction also compounds tool-call, response-length, timeout, and budget failures, which can discard salvageable trajectories, waste rollout computation, and disturb policy updates. To address these issues, we introduce WeAgent-Harness, a multimodal agentic harness that supports native text-vision interaction and runtime recovery. Retrieved images receive persistent disk references, allowing the model to inspect, process, and cite them throughout the trajectory. Based on this harness, we develop WeAgent-MMSearch, an integrated system spanning data construction, agentic post-training, and multimodal rollout. For data construction, a strong MLLM uses WeAgent-Harness to discover, synthesize, and verify MMSearch-style tasks and collect expert trajectories. During post-training, our Failure-Aware GSPO (FA-GSPO) recovers salvageable abnormal rollouts and filters invalid ones to improve bounded multimodal planning and this http URL also introduce VisTarget-Bench, a 150-task human-verified benchmark that pairs each question with a held-out target image, distinguishing image-retrieval failures from visual-perception failures. Evaluation on VisTarget-Bench and seven public benchmarks shows that agentic post-training improves the average score by 19.22 points, enabling our model to outperform similarly sized open-source models and rival models with roughly ten times its parameter count.
[AI-48] Explainable Uncertainty Estimation for Reliable Medical AI ICDM
链接: https://arxiv.org/abs/2608.28052
作者: Li Rong Wang,Jamie Duell,Xinran Xu,Thomas C. Henderson,Yu Yue Hew,Pik Wan Erica Chiang,Xiao Wei Alstar Ang,Bingwen Eugene Fan,Xiuyi Fan
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Accepted at the 26th IEEE International Conference on Data Mining (ICDM)
Abstract:Artificial intelligence has strong potential to support clinical decision-making, yet its adoption in healthcare remains limited due to a lack of trust. Uncertainty estimation can signal unreliable predictions, and explainable AI (XAI) can clarify how predictions are made but existing methods treat them separately, providing no feature-level insight into why a prediction is uncertain or which tests to prioritize to reduce it. To address this gap, we propose explainable uncertainty estimation, which unifies uncertainty estimation and XAI to both quantify uncertainty and explain feature-level contributions. We introduce the Expected Gradients Reconstruction Uncertainty Estimate (egRUE), which incorporates prediction explanations into its uncertainty computation and decomposes uncertainty into feature-wise contributions. We prove theoretical properties of egRUE and show through experiments that it improves reliability and interpretability compared to existing methods. A user study with medical experts further demonstrates that egRUE’s explanations improve calibrated trust over uncertainty scores alone, increasing confidence in correct predictions and reducing confidence in incorrect ones. By combining prediction uncertainty with feature-level explanations, egRUE strengthens decision-making support in safety-critical healthcare settings, clarifying both when predictions may be unreliable and which features drive that uncertainty.
[AI-49] String: An Agent ic OS Where Every App Is a Markdown File
链接: https://arxiv.org/abs/2608.28027
作者: Jookyung Song,Nojun Kwak,Simyung Chang
类目: Artificial Intelligence (cs.AI)
备注: 11 pages, 3 figures
Abstract:LLM agents have become a new class of software user, but every surface they work through was designed for someone else. Pages are built for human eyes, which can skim and ignore; tool schemas for programs, which pay nothing to carry definitions they never call. An agent has neither luxury: it re-reads, and pays again for, everything it is shown on every turn. We present String, an open-source runtime that gives this user an interface of its own and treats the job as an operating-systems problem. Tool knowledge moves out of the agent’s context and into a common layer that renders it back one view at a time as Markdown. A single SFMD (String-Flavored Markdown) document declares an application’s views, typed actions, navigation, and credentials, and the runtime handles discovery, validation, execution, state, and secrets behind two core verbs: /open to see and /act to do. Web and app turn out to be two renderings of one architecture: an SFMD site serves styled HTML to browsers and the raw document to agents, so one grammar reaches apps, files, shells, and the web, even legacy HTML, with no per-site integration. Views stay partial by design, and the staging is causal: disclosing one tier of detail a single turn too early costs up to 23 accuracy points, while proper staging drops wrong-action selection from 28% to 2%. Privilege follows provenance: a remote page may call HTTP but never the shell, and caller-supplied text never expands a stored secret. On an 87-task benchmark that pairs each task with curated skills, operationalizing those procedures as on-demand String apps yields comparable aggregate success across six models from frontier to small (+1.3pp) while using 33.5% fewer tokens among completed episodes, and the resident interface stays a constant 53 tokens at any catalog size. We report the design, the evaluation, and what three months of production use taught us.
[AI-50] Coverag e Not Credit: Failure-Credit Routing of Zeroth-Order Perturbation Budgets Does Not Improve On-Pool Sample Efficiency for LLM Agents
链接: https://arxiv.org/abs/2608.28011
作者: Yuxu Ge
类目: Artificial Intelligence (cs.AI)
备注: 19 pages, 3 figures, 7 tables. Preprint
Abstract:Trajectory-level credit assignment can localize which module of a tool-using LLM agent causes failures using only verifiable signals. We ask whether such failure credit should route a fixed zeroth-order/evolution-strategies (ZO/ES) perturbation budget. Across a synthetic environment and frozen Qwen2.5-1.5B/3B and SmolLM2-1.7B agents, three task families, six allocation schemes, a credit-noise sweep, paired seeds, and exact sign-flip tests, we find no statistically detectable improvement over uniform allocation in any on-pool comparison (no gain of at least 2 percentage points). The joint soft-plus-sigma scheme is equivalent to uniform within a +/- 0.02 AUC margin on 1.5B and 3B; concentrating the full budget on the credit argmax is marginally equivalent on 1.5B, where that module is the verified bottleneck, and significantly worse on 3B. Inverse-propensity debiasing does not rescue routing, and misrouting costs up to -0.074 AUC in-house and -0.118 end-to-end on the BFCL-derived family. Across six fixed-step schedules, loss is linear in bottleneck starvation rate (R^2 = 0.94, descriptive), and a preregistered credit-free coverage floor removes detected harm. Matched-budget burst and step-compensating catch-up schedules are consistent with harm arising from insufficient cumulative parameter movement rather than update frequency. Our primary estimand is optimization efficiency on a fixed task pool. On unseen BFCL functions, the study’s one exception is that soft routing exceeds uniform on held-out endpoints (+0.047, p = 0.031, n = 6). A plausible but untested reading is that routing-favored caller improvements transfer while uniform’s on-pool gains reflect a synthesizer behavior specific to our harness. We report this exception explicitly and document three failure modes that can silently invalidate ZO/ES experiments on frozen LLMs.
[AI-51] When Can Conditional Flow Matching Replace Pointwise Negative Log-Likelihood?
链接: https://arxiv.org/abs/2608.28010
作者: Yansen Han,Hongxin Sun,Tao Lin
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Flow matching enables likelihood-free training, yet alignment methods increasingly reuse conditional flow matching (CFM) losses as endpoint negative log-likelihoods (NLLs) and their old/new differences as log-likelihood ratios. We characterize when these substitutions are valid. For linear Gaussian paths, we exactly decompose endpoint NLL into entropy, a weighted CFM objective, an interior velocity–score residual, and a boundary residual. Thus CFM-only estimates and differences are exact only when the corresponding residuals cancel. At the off-policy population optimum, ordinary CFM is not generally a pointwise NLL estimator, whereas (w_\mathrmsc(t)=(1-t)/t) removes the interior residual; this positive result does not extend generally to training or on-policy alignment. On-policy log-ratios can remain biased even for identical endpoint laws or after surrogate optimization. Experiments across dimensions, distributions, and geometries support these conclusions and the mechanisms that make inexact ratios useful. More broadly, the decomposition provides a theoretical basis for adapting likelihood-based LLM methods to flow matching, while distinguishing exact substitutions from controlled surrogates.
[AI-52] A Method for Layer Bit-Width Allocation in LLM Quantization via Performance Maximization Under a Quality-Degradation Constraint
链接: https://arxiv.org/abs/2608.28003
作者: Artem Safronov
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 22 pages, 4 figures
Abstract:This paper proposes a layer bit allocation method for Gemma-3-1B, formulating the problem as performance maximization (latency decrease) given a degradation budget constraint (allowable level of generation quality loss). This approach is different from time- and resource-consuming uniform layer quantization methods that are used in the literature (like GPTQ or AWQ) or allocation methods without proven performance-accelerating effect (like MixLLM or TorchAO). The layer sensitivity profile resulting from our prior work SA-PTQ is applied using the activation pass-through mode inside TensorRT-LLM. For each layer precision is determined individually in blocks, according to a grouping introduced in the prior step (5+5, 10+10, all26), differentiating the contribution of FFN, Attention, and lm_head to the overall speedup. The clock speed was measured for 13 W8A8 variants on an RTX 5090. We find that for FFN and lm_head the time cost of quantization/dequantization is compensated for by the use of integer arithmetic, while for short context lengths, the opposite holds true for Attention: an additional step of quantization slows execution down. We propose a manual implementation of SmoothQuant for TensorRT-LLM which was necessary due to export failures, unavailable for lm_head. The best solution found under joint consideration of all three criteria with minimal degradation was FFN 5+5 with lm_head, providing an 11.0% reduction in latency with negligible quality loss (98.90% Top-1 agreement, +0.85% perplexity degradation). With acceptable quality loss for FFN all26 + lm_head, a speedup up to 19.1% was found possible. We suggest further optimizations: fused attention kernels in INT8, KV-cache quantization, using FP8 instead of INT8 and partial Attention quantization analogous to FFN.
[AI-53] PhenoIntel: A Lifecycle-Aligned Multi-Agent Web Application for Verified Accessible Plant Phenotype Analysis
链接: https://arxiv.org/abs/2608.27999
作者: Narendren S V,Soumyashree Kar
类目: Artificial Intelligence (cs.AI)
备注: 32 pages, 11 figures, 9 tables. Submitted to Engineering Applications of Artificial Intelligence (Elsevier)
Abstract:Existing conversational plant-phenotyping platforms are difficult for plant scientists to use and lack the reliability scientific research demands: failed analyses are reported as valid measurements rather than flagged as missing, statistical tests run without checking assumptions, predictions carry no uncertainty estimate, and specialised hardware limits accessibility. We present PhenoIntel, a lifecycle-aligned multi-agent web platform that turns the full machine-learning workflow into a reliable, user-friendly phenotyping system. Nine specialised agents divide the analysis into stages, from image collection through model selection, inference, and reporting, rather than handing the whole task to one AI manager. Independent checks separate these stages, and every agent reads from and writes to one shared, fixed-structure record, so an inconsistent output from one stage is caught before it reaches the next. Uncertainty is matched to each model family, conformal prediction, detection-confidence spread, or Monte Carlo Dropout, rather than applied uniformly, and quality thresholds adapt to crop and task instead of one global cutoff. When no suitable model exists, PhenoIntel can propose, validate, and integrate a new one on its own. The model repository spans ten trained models across five crops and four imaging modalities. Classification models reach Macro F1 of 0.78-0.996; object-detection models reach 0.96 mAP@50 with a 54% reduction in counting error over an unoptimised baseline; and a temporal model reaches held-out Macro F1 of 0.7050. PhenoIntel runs in a browser on standard hardware, requiring no GPU, and a 1,200-test automated suite confirms complete pipeline execution. Every result carries calibrated uncertainty, validated statistics, and FAIR-compliant provenance, a combination existing conversational phenotyping tools do not offer.
[AI-54] Automated Analysis Framework for Multilingual Climate-Health Literature Based on Multi-Agent Large Language Model
链接: https://arxiv.org/abs/2608.27998
作者: Yuze Sun,Shihui Zhang,Jiancheng Pan,Yunjia Ye,Wentao Luo,Jiahao Li,Quan Zhang,Wenjia Cai,Xiaomeng Huang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:The rapid proliferation of interdisciplinary and multilingual scientific literature has left traditional manual analysis and single-algorithm methods plagued by low efficiency, poor scalability, and insufficient domain adaptability. Targeting the literature analysis needs of the typical interdisciplinary climate-health field, this study proposes a multi-agent large language model automated analysis framework for multilingual scientific literature, which realizes full-process automation covering literature screening, structured information extraction, and standardized integration. With a central coordination module as the core, the framework deploys three dedicated agents for document evaluation, information extraction, and analytical review to mimic the literature analysis thinking of domain experts, and adopts a four-layer hallucination control strategy together with a manual verification procedure to ensure the accuracy and reliability of analytical outcomes. Validated on a bilingual Chinese-English corpus of 32,642 climate-health papers covering China from 1993 to 2023, the framework achieves an F1 score of 0.92 in core information extraction, and completes the extraction and standardization of 2,012 city-literature association pairs, offering effective technical support for large-scale evidence mining in the climate-health research domain.
[AI-55] Should I Use This Synthetic Dataset for Training? How to Test with Minimal Real Data
链接: https://arxiv.org/abs/2608.27996
作者: Zhenyu Tao,Wei Xu,Xiaohu You,Petar Popovski,Osvaldo Simeone
类目: Artificial Intelligence (cs.AI)
备注: Submitted to IEEE
Abstract:Digital twins (DTs) and learned world models are increasingly used to generate synthetic data that augment the scarce real datasets available for training artificial intelligence (AI) models in engineering systems. Owing to the inevitable simulation-to-reality (sim-to-real) gap, however, augmentation may fail to improve the performance of the trained model on the real data distribution. This paper addresses the resulting decision problem: Given a real dataset, a candidate synthetic dataset, and a fixed learning algorithm, decide whether training on the augmented dataset improves the true, population-level performance, while consuming as few real test data points as possible. Two formulations are considered: a direct test on the mean loss difference between the two trained models, and a symmetry-based test on the paired loss difference, which trades a stronger null assumption for faster evidence accumulation. For the latter, we introduce the adaptive e-process sign-flip test (aeSFT), a doubly adaptive procedure that adapts both the number of Monte Carlo sign-flip rounds, and hence the computational cost, and the amount of real test data consumed. aeSFT yields anytime-valid Type-I error control, with no need to pre-specify the test-set size. Experiments on a synthetic-data classification task, a DT-aided wireless packet-scheduling task, and a radio-map prediction task show that aeSFT identifies useful synthetic data using substantially fewer real test samples than mean-based sequential testing, matches the power of fixed-sample sign-flip testing and the paired t -test, while keeping the false-positive rate below the target level.
[AI-56] CAITLYN: Can LLM Agents Autonomously Synthesize Defenses against Emerging Injection Attacks?
链接: https://arxiv.org/abs/2608.27990
作者: Zi Liang,Xiaoyu Xu,Yanyun Wang,Minxin Du,Qingqing Ye,Haibo Hu
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注: Source code: this https URL
Abstract:Prompt injection attacks on Large Language Model (LLM) agents seek to introduce malicious instructions or content into external text sources retrieved by agents, forcing the underlying LLMs to execute harmful actions outside their benign scope. While current defenses effectively counter known injection attacks, deploying them in LLM agent environments remains challenging due to attack variants and emerging threats. Moreover, existing solutions typically suffer from an inherent trilemma, i.e., a constant trade-off among runtime efficiency, contextual precision, and adaptability. To bridge this gap, we propose Continuous Agents for Injection Threats via Lifelong Yielding Nexus (CAITLYN), an agent-agnostic defense middleware. CAITLYN integrates two systems. System I focuses on immediate defense against existing attacks using a two-tiered library: Tier-0 for rule-based detection scripts and Tier-1 for optimized LLM-based accurate inference. System II, in contrast, is deployed to monitor potential abnormal signals and attempt to synthesize new defenses. On standard benchmarks, CAITLYN matches the detection performance of state-of-the-art defenses at lower token overhead than LLM-as-a-judge baselines. On Emerging, our new delivery-aware benchmark featuring novel injection techniques, static baselines and the standalone System I configuration remain vulnerable. In contrast, System II autonomously synthesizes verified defense capabilities, substantially lowering the attack success rate across three diverse agent environments.
[AI-57] When Evidence Shapes Collaboration: Knowledge-Conditioned Topology Generation for Multi-Agent Systems
链接: https://arxiv.org/abs/2608.27984
作者: Yangxiao Jiang,Jiarun Fan,Mingcong Xu,Yanxi Guo,Jiwen Feng,Shanqing Xu,Mengchen Qian,Wei Chen,Xiaojin Zhang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Multi-Agent Systems (MAS) have recently moved from static workflows toward dynamically generated collaboration topologies. However, existing topology generation methods rely primarily on the parametric knowledge of large language models, with external search or retrieval used only as a reactive tool rather than an explicit determinant of collaboration structure. This leads to structure-knowledge misalignment, where systems exhibit redundant interactions or insufficient verification in knowledge-intensive tasks. We propose K-GAT (Knowledge-Guided Agent Topology Generator), a neuro-symbolic framework that formulates collaboration topology design as a knowledge-conditioned structure learning problem, integrating external evidence directly into autoregressive graph generation. Extensive experiments on knowledge-intensive benchmarks demonstrate K-GAT’s efficiency and effectiveness: notably on the expert-level GPQA dataset, K-GAT outperforms the LLM-Debate baseline by a substantial margin of +15.7% in accuracy, while consuming less than half the computational tokens.
[AI-58] Learning from Hard Prompts: Difficulty-aware Advantage Amplification in Dynamic Sampling
链接: https://arxiv.org/abs/2608.27982
作者: Siyuan Gan,Yuhan Li,Xiran Wang,Linjian Meng,Boyan Wang,Zhen Zhao,Jing Huo,Lei Bai,Yang Gao
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Decoupled Clip and Dynamic Sampling Policy Optimization (DAPO) is a prominent variant of Group Relative Policy Optimization (GRPO). DAPO introduces several improvements over GRPO. Among these, Dynamic Sampling contributes the most to DAPO’s accuracy gains relative to GRPO. To improve accuracy, Dynamic Sampling enhances training stability by eliminating zero policy gradients from zero advantages. Specifically, it avoids such zero gradients by filtering out prompts where sampled responses are either entirely correct or incorrect. However, our theoretical analysis shows that Dynamic Sampling decrease training efficiency as it cannot effectively utilize hard-to-sample correct responses on hard prompts. Formally, it asymmetrically amplifies the advantages of distinct responses to the same prompts. On hard prompts, incorrect responses undergo greater amplification than correct ones. This leads the model to avoid generating the observed incorrect responses rather than capitalizing on the hard-to-sample correct ones on hard prompts, resulting in low training efficiency. To improve training efficiency, we propose Direct Advantage Amplification (DAA), which amplifies the advantages of hard-to-sample correct responses on hard prompts, as obtained by Dynamic Sampling. This ensures that, when Dynamic Sampling is used, these hard-to-sample responses can be effectively capitalized on, implying higher training efficiency. By integrating DAA into DAPO, we obtain Difficulty-aware Advantage Amplification Policy Optimization (DA3PO), which is implemented with fewer than 30 lines of code from DAPO. Experiments show that DA3PO significantly outperforms GRPO and other classical GRPO variants.
[AI-59] openJiuwen: Beyond Static Harnesses for Long-Horizon Coding Agents
链接: https://arxiv.org/abs/2608.27969
作者: openJiuwen Team:Tao Yu,Xinyu Zhang,Qianqian Chen,Xiaoneng Xiang,Chia Kwangyang,Xingchen Huang,Ran Chen,Yangkai Ding,Zheng Wang,Yeo Boon Hong,Bingzheng Gan,Enrui Hu,Shuo Cheng,Deyang Li,Ruifeng Shi,Hongbo Wang,Qi Ye,Xuefeng Jin,Zhangchun Zhao
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Long-horizon coding agents operate over evolving repository states while increasingly relying on heterogeneous capabilities, delegated agents, and multi-agent coordination. These trends pose two complementary challenges for the agent harness. First, developers need to compose capabilities, reconfigure execution logic, and scale increasingly complex agent systems without repeatedly rebuilding orchestration. Second, complex coding tasks continuously produce new evidence—such as semantic diagnostics, execution outcomes, task progress, and changing context relevance—that should dynamically influence subsequent runtime decisions. We characterize these challenges as Structural Composability and Runtime Adaptivity. We present openJiuwen, an open-source harness designed for both developer composability and adaptive task execution. openJiuwen provides a shared execution substrate and Rail-based capability composition across single agents, delegated sub-agents, and Swarm Flow, enabling developers to construct sophisticated agent harnesses under common execution semantics. It further adapts framework-controlled runtime decisions around a fixed model policy, allowing evolving evidence to dynamically affect context, feedback, and task control toward successful completion. We systematically evaluate openJiuwen on SWE-bench Verified and Terminal-Bench 2.1, where it achieves 82.6% and 87.19%, respectively, exceeding the strongest selected official-leaderboard point estimates by 3.4 and 3.39 percentage points. These results show that openJiuwen achieves strong performance on complex coding tasks while providing a composable and adaptive harness design.
[AI-60] AERA: Adaptive Evidence Residual Allocation for Efficient Test-Time Reasoning
链接: https://arxiv.org/abs/2608.27964
作者: Ziming Wang,Ivor Tsang,Hangwei Qian
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Test-time scaling improves language-model reasoning by generating additional candidate solutions, but allocating the same inference budget to every problem is computationally wasteful. Existing adaptive stopping methods commonly rely on confidence, agreement, or answer stability, implicitly assuming that stronger current evidence indicates that further computation is unnecessary. We show that this assumption can fail: checkpoint-level correctness evolves non-monotonically, and observable evidence may strengthen before an answer collapses or weaken before it recovers. Motivated by this mismatch, we introduce Adaptive Evidence Residual Allocation (AERA), a sequential controller that learns whether additional computation is likely to recover a better answer from checkpoint-observable evidence. AERA characterizes cumulative response prefixes using answer-distribution, temporal, re-solving, semantic, and compute features, and repeatedly decides whether to stop or allocate the next response block. Future checkpoint correctness is used only to construct offline supervision and is never available to the controller at inference time. Across GSM8K and GPQA Diamond, AERA identifies question-specific residual opportunities while substantially reducing inference computation. In a frozen-threshold incremental-generation evaluation on 300 untouched GSM8K questions, AERA achieves 92.61% accuracy versus 93.01% with 128 responses while reducing completion tokens by 95.99%. These results suggest that adaptive reasoning should estimate the future value of computation rather than equating present confidence with correctness.
[AI-61] SABER: Stability-Aware Early Exit for LLM Reasoning via Adversarial Branch Probing EMNLP2026
链接: https://arxiv.org/abs/2608.27963
作者: Wanli Cheng,Haiya Xiang,Juntao Li,Hongling Wang,Wenliang Chen
类目: Artificial Intelligence (cs.AI)
备注: 20 pages,10 figures,EMNLP 2026 MainConference
Abstract:Large Reasoning Models (LRMs) achieve strong reasoning capabilities, yet long-chain reasoning becomes inefficient once the intermediate answer stabilizes across reasoning steps: additional reasoning yields little marginal benefit while incurring substantial inference cost. Existing early-exit methods based on confidence or entropy poorly capture reasoning stability, while consistency-based approaches rely on multi-step trajectory agreement, requiring sequential evaluations that delay exit. To better balance efficiency and reliability, we propose SABER, a training-free framework for stability-aware early exit via adversarial branch probing. SABER constructs simple yet effective semantic perturbations around intermediate reasoning states to form adversarial branches, and applies lightweight probing to estimate their likely final outcomes without full trajectory rollouts. When the probed outcomes remain consistent across branches, SABER exits early; otherwise, it continues reasoning. Experiments across multiple reasoning benchmarks and model architectures show that SABER reduces reasoning token consumption by 30.2%–39.8% on average while maintaining competitive accuracy with full-length reasoning.
[AI-62] When Teacher Guidance Misleads: Reward-Aligned On-Policy Distillation
链接: https://arxiv.org/abs/2608.27960
作者: Siyuan Gan,Yuhan Li,Xiran Wang,Linjian Meng,Boyan Wang,Zhen Zhao,Jing Huo,Yang Gao
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:On-policy distillation (OPD) has recently emerged as a popular post-training paradigm for large language models (LLMs), providing an efficient way to transfer the knowledge and capabilities of teacher models into student models. However, teacher guidance on student-generated prefixes is not always reliable. Training should optimize the model to generate responses that are more likely to be correct, or equivalently, to get higher outcome rewards. But during OPD, the teacher model may provide guidance that discourages the student from moving toward correct trajectories or moves the student toward incorrect ones, which is misaligned with outcome reward. Such misaligned guidance is unreliable, as it would mislead the optimization process and ultimately degrade model performance. To mitigate misaligned teacher guidance, we propose Reward-Aligned On-Policy Distillation (RA-OPD). The key insight is to keep only trajectories whose induced updates move the student toward correct trajectories or discourage the student from moving toward incorrect ones. Specifically, for each sampled trajectory, RA-OPD checks whether its trajectory-level distillation return is consistent with its outcome reward and then filters out the misaligned trajectories. RA-OPD selects more reliable trajectories to improve student model performance without requiring additional computational cost. We evaluate RA-OPD on math and code benchmarks using models from the Qwen3 family and the DeepSeek-R1 family. Across seven math benchmarks and three code benchmarks, RA-OPD significantly outperforms standard OPD and other tested OPD variants.
[AI-63] Not to Break but to Attest: Adversarial Probes for Privacy-Preserving LLM Verification
链接: https://arxiv.org/abs/2608.27954
作者: Cameron Wilding,Mina Shaker,Fatemeh Ganji
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Post-deployment changes to large language models can alter behavior while leaving routine outputs largely unchanged, creating a challenge for AI governance when model weights are proprietary. We present a privacy-preserving zk-SNARK-based audit framework that searches for probes designed in the spirit of adversarial examples to amplify logit drift between an approved model and a modified deployment. Our framework explores complementary probe families under different access models. Token-based probes operate in a black-box setting and require only the input interface, tokenizer, and vocabulary. Embedding-based probes require gray-box access to the embedding interface. Stress probes rely on additional interface capabilities but do not require full white-box access to model weights or architecture. This range allows probe selection to balance sensitivity, access requirements, and deployment cost. We evaluate probe constructions across LLM architectures, model-tampering scenarios representative of post-deployment attacks, and GPU platforms. Importantly, our experimental results demonstrate that token-based probes consistently deliver the strongest mean sensitivity across models and GPU platforms, although operating in a black-box setting. Our Groth16 zk-SNARK workflow remains practical as the probe set scales from 1 to 50, where proving time increases from 1.02 to 1.78 seconds, verification remains near 0.84 seconds, and proof size remains constant.
[AI-64] he Illusion of textitWhat If: Evaluating the Breakdown of Counterfactual Reasoning in LLM s EMNLP2026
链接: https://arxiv.org/abs/2608.27953
作者: Yucheng Wang,Yuetian Du,Zhengyi Liu,Rongyu Zhang,Bing Zhao,Boyu Yang,Ming Kong,Lin Qu,Hu Wei,Jie Liu,Qiang Zhu
类目: Artificial Intelligence (cs.AI)
备注: Accepted by EMNLP 2026
Abstract:Counterfactual reasoning requires models to reason beyond the observed world and explain how altered conditions propagate through downstream consequences. Existing benchmarks largely target bounded settings with fixed variables or single gold outcomes, overlooking open-domain scenarios requiring causal-process evaluation. To this end, we present \textbfWhatIfBench , a diagnostic benchmark for open-domain, open-form, long-horizon counterfactual causal reasoning, containing 220 what-if questions across STEM, HSS, and Hybrid scenarios. To evaluate free-form responses, we further propose \textbfPRISM , which first converts each natural-language explanation into a Response-Derived Semantic Causal Graph of events, states, and mechanisms. On top of this graph, PRISM then jointly applies a Process Metric assessing graph-level causal validity and a Rubric Metric assessing answer-level explanatory adequacy. Evaluating six frontier LLMs with this framework, we find that WhatIfBench remains far from saturated: even the strongest model reaches only a 64.62% final score. Further analysis reveals persistent causal gaps, premise drift, and topology fragmentation, suggesting that fluent counterfactual narratives often mask fragile causal processes. The benchmark, code, and evaluation scripts are available at \hrefthis https URLWhatIfBench .
[AI-65] Cross-Session Decomposition Attacks: Scaling Risk and Intent-Aligned Retrieval Defense EMNLP2026
链接: https://arxiv.org/abs/2608.27945
作者: Disen Liao,Yihan Wang,Freda Shi,Yaoliang Yu
类目: Artificial Intelligence (cs.AI)
备注: 30 pages, 3 figures, EMNLP 2026 Findings
Abstract:Scaling laws are usually read as a capability story: lower language-modeling loss yields more useful models. We study a safety consequence of this mechanism in \emphcross-session decomposition attacks, where benign-looking subqueries are asked across independent interactions and later recomposed toward a forbidden objective. We formalize this setting as \emphcompositional safety risk and prove a conditional risk-transfer bound: when the reference environment already contains dispersed evidence for a risky reconstruction, the gap between deployed composed risk and reference composed risk is controlled by the model’s excess loss on allowed subqueries. Synthetic withholding experiments show that wider transformers assign lower loss to held-out instructions that never appear verbatim in training but are recoverable from injected supporting facts. A 600-intent pretrained-LLM evaluation shows that larger Qwen3 and Gemma3 family members can yield greater harmful-capability uplift under a fixed decomposition-composition pipeline. As a defense, IntentAlign-MiniLM, our 22M-parameter intent-aligned retriever, outperforms much larger embedding models on held-out intent retrieval and yields the best learned-retriever harmful recall across tested guardrails. Code is available in \hrefthis https URLour GitHub repository.
[AI-66] CASTANET: Causality-Aware Spatio-Temporal Adversarial Network Using Traffic Incident Effects
链接: https://arxiv.org/abs/2608.27942
作者: Toshiya Kitahara,Ryu Shirakami,Koh Takeuchi,Hisashi Kashima
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Predicting non-periodic traffic congestion caused by sudden incidents (e.g., accidents and road damage) is crucial for advanced intelligent transportation systems. However, incident-driven congestion is difficult to forecast because incidents are extremely sparse, occur at specific times and locations, and have heterogeneous impacts depending on the traffic context. While recent deep learning approaches have significantly improved periodic traffic forecasting, their performance on non-periodic congestion remains limited, partly because incident records are not explicitly incorporated and their occurrence is strongly biased in space and time. To address these challenges, we propose CASTANET, which integrates spatio-temporal graph neural networks and causal treatment effect estimation to utilize incident records while mitigating selection bias. Experiments on real-world traffic data and accident records from Tokyo, which we treat as incidents, show that CASTANET reduces RMSE by 4.0% overall compared to the best baseline and by 10.1% on incident-conditioned evaluation, with gains reaching 14.55% under severe congestion.
[AI-67] A Deep Learning-Based Stacking Ensemble Framework for Turbofan Engine Remaining Useful Life Prediction
链接: https://arxiv.org/abs/2608.27940
作者: Limon Bin Hossain,Md. Salehin Seyam,Md Rashedul Islam,Abdur Rahman,Md Sharifuzzaman
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:This study proposes a two-level stacking ensemble framework for Remaining Useful Life (RUL) prediction of turbofan engines, evaluated on the NASA C-MAPSS benchmark using the FD001 and FD003 subsets. The framework integrates four heterogeneous deep learning base learners: Long Short-Term Memory (LSTM), Convolutional Neural Network (CNN), CNN-LSTM, and CNN-GRU, whose out-of-fold predictions are combined by an XGBoost meta-learner to capture complex degradation patterns while mitigating individual model biases. Comprehensive experiments demonstrate that the stacking ensemble achieves superior predictive performance, with Root Mean Square Error (RMSE) of 9.989 and 8.613, Mean Absolute Error (MAE) of 7.081 and 5.195, and R-squared values of 0.899 and 0.906 for FD001 and FD003, respectively. Compared to the best-reported baseline (TCAT: RMSE 11.12 and 11.02), the proposed method achieves RMSE reductions of 10.2 percent and 21.8 percent for FD001 and FD003, respectively. Feature correlation analysis, residual diagnostics, and training convergence curves validate the model’s robustness. These findings underscore the efficacy of stacking ensemble methods for prognostics and health management in safety-critical aerospace applications.
[AI-68] Antipatterns in AI-assisted Qualitative Data Analysis: A Catalog of Temptations and Pitfalls for Software Engineering Researchers
链接: https://arxiv.org/abs/2608.27927
作者: Rashina Hoda,Carolyn Seaman,Victoria Gomes,Rodrigo Spinola
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: 12 pages
Abstract:AI-assisted qualitative data analysis (QDA) offers unprecedented opportunities to streamline software engineering (SE) research, yet uncritical use risks compromising analytical rigor and flooding the field with accelerated production of low-quality research. While tactical best practices will naturally evolve over time, SE researchers currently lack strategic guidance to identify and mitigate methodological risks when attempting AI-assisted QDA. Based on our decades of qualitative SE research expertise and experience combined with an understanding of the emerging landscape of AI-assisted QDA, this paper presents a catalog of antipatterns in AI-assisted QDA - a set of assumptions and practices that initially appear advantageous but ultimately undermine analytical rigor and validity. The antipatterns are grouped into three categories reflecting escalating impact: Dangerous Drivers, Operational Missteps, and Analytical Failures. As more SE researchers attempt AI-assisted QDA, these antipatterns will help them identify and avoid common temptations and pitfalls, while reviewers can be equipped with the vocabulary and criteria to call out problematic and failed practice. Ultimately, this catalog of antipatterns can serve as a stepping stone in our responsible methodological evolution toward principled and meaningful human-AI collaboration in qualitative research.
[AI-69] From Documents to Reasoning : A Validated Synthetic Data Pipeline and Semantic-Aware Fine-Tuning for Financial Numerical Reasoning
链接: https://arxiv.org/abs/2608.27919
作者: Lokendra Birla,Milind Savagaonkar,Visnu Srinivasan,Sowmya Rasipuram,Shubhashis Sengupta
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Financial question answering (QA) has emerged as a key benchmark for evaluating the performance of Large Language Models (LLMs) on domain-specific tasks involving complex data formats such as tables, charts, and rich textual narratives. While recent advancements have enabled models to reason across modalities and perform multi-step arithmetic operations, limitations remain in performance consistency, and evaluation reliability. In particular, standard evaluation metrics like Exact Match (EM) often fail to account for minor variations such as differences in units or formats, misleading performance assessments. In this work, we propose a comprehensive pipeline for improving financial QA systems through high-quality synthetic data generation and fine-tuning of smaller language models (SLMs) using Quantized Low-Rank Adaptation (QLoRA). Our pipeline includes aggressive data validation for synthetic question answer generation to ensure the relevance and correctness of synthetic question-answer pairs. We introduce a novel evaluation metric that matches answers computed from arithmetic expressions rather than ground-truth answers; providing a more accurate reflection of model reasoning capability. Furthermore, we propose a modified loss function that aligns predicted and reference expressions using semantic similarity, our novel evaluation metric and standard cross-entropy, resulting in improved performance. Experimental results on benchmark datasets, ConvFinQA demonstrate significant gains in QA accuracy after fine-tuning using synthetic dataset and proposed loss function. Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2608.27919 [cs.AI] (or arXiv:2608.27919v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.27919 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-70] Low-Altitude Fluid Antenna Network with Multi-Agent Reinforcement Learning
链接: https://arxiv.org/abs/2608.27909
作者: Tong Zhang,Yanfei Su,Shuai Wang,Wanli Ni,Chengzhong Xu,Huseyin Arslan
类目: Information Theory (cs.IT); Artificial Intelligence (cs.AI)
备注: Accepted by IEEE Communications Magazine
Abstract:Low-altitude wireless networks (LAWNs) integrate terrestrial and aerial platforms to provide ubiquitous communication, sensing, and localization services for unmanned aerial vehicles (UAVs) and electric vertical takeoff and landing (eVTOL) aircraft. However, dynamic air-ground and air-air channels, abrupt blockages, and heterogeneous interference hinder the realization of this goal. Nevertheless, fluid antenna (FA), a cutting-edge multiple-input multiple-output (MIMO) technique, overcomes these challenges by reconfiguring antenna positions to unlock additional spatial degrees-of-freedom. In this paper, towards bringing low-altitude FA networks into reality, we study the fast and high-performance FA reconfiguration for low-altitude FA networks with multi-agent reinforcement learning (MARL). Specifically, we present an electromagnetic digital twin (EM-DT)-assisted MARL framework. To fill the sim-to-real gap, we introduce a two-stage transfer learning framework. Our case study shows that joint FA positions and beamforming optimization can enhance the system sum-rate by 118.5%, compared to the fixed position baseline. This gain comes from the dynamic millisecond timescale reconfiguration of FA arrays and the adaptive steering of beams toward aerial users with mobility.
[AI-71] Rubric-to-Code Credit Assignment for Reinforcement Learning
链接: https://arxiv.org/abs/2608.27906
作者: Rui Jin,Jikai Chen,Yihan Chen,Hao Zhou,Demin Zhu,Kaichen Yang,Dong Wang,Chenyi Zhuang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Interactive web application generation requires models to produce usable HTML, CSS, and JavaScript applications from natural language requests. Unlike conventional code generation, application quality depends on multiple user-facing functional requirements, each often tied to localized code regions such as event handlers, state updates, DOM fragments, or CSS selectors. Standard GRPO collapses these structured outcomes into a single sequence-level reward and applies the resulting advantage uniformly to all tokens, weakening credit assignment. We propose \textbfRubric-to-Code Credit Assignment (RCCA), a reinforcement learning framework that converts rubric-level functional feedback into localized optimization signals over generated code. RCCA builds training tasks around explicit functional rubrics, uses a hierarchical reward to separate format, source-code, runtime, and functional failures, and aligns evaluator-generated textual attributions with responsible code spans and generated tokens. The resulting model, \textbfLing-RCCA-Flash, scores 41.25 on MiniAppBench, improving Ling-3.0-Flash by 32.20 points and slightly surpassing Claude Opus 4.5. It also reaches 76.19 on ArtifactsBench, improving the SFT model by 4.48 points and establishing a new top score under the official ArtifactsBench leaderboard setting by surpassing the GPT-5 score by 3.64 points, suggesting transferable implementation-level gains.
[AI-72] Resource Constraints and Performance in Agent ic AI Systems
链接: https://arxiv.org/abs/2608.27886
作者: Amaz Salman,Malka Halgamuge,Teo Susnjak
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Progress toward more autonomous AI increasingly depends on agentic systems that combine a language model with tools, memory, state management, and multi-step execution. These mechanisms shape both task capability and operational burden. We compare OpenClaw and NanoBot as complete agentic systems using a paired primary benchmark and a more detailed instrumented subset of paired prompts. In the primary benchmark, the rate of full task completion was 31% for OpenClaw and 25% for NanoBot, a six-percentage-point difference with a 95% task-bootstrap interval from -3 to 15 percentage points, providing no statistically established full-completion advantage for either system. In the instrumented layer, both systems achieved 26% full completion, while NanoBot reached at least partial completion on 43% of prompts compared with 26% for OpenClaw. OpenClaw took longer on 83% of prompts and had a higher recorded peak-memory value on every prompt, with geometric mean ratios of 2.98 for wall time and 19.44 for peak memory. Among the ten detailed-layer prompts on which at least one system achieved partial or full completion, NanoBot weakly dominated on eight; across all 23 prompts, however, ten of its eighteen dominance cases were cheaper joint failures. Outcome labels differ across the two evidence layers, showing why agent-system evaluation should connect capability and resource measurements to attempt-level execution and scoring provenance. These findings show that progress toward more autonomous AI should be evaluated through verified task completion, observed resource use and records linking each result to the execution that produced it.
[AI-73] SOMTab: Set-Order Mamba for Efficient Tabular In-Context Learning
链接: https://arxiv.org/abs/2608.27882
作者: Hao Wang,Siyu Zhang,Wei Ma
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Tabular foundation models based on in-context learning have recently emerged as strong alternatives to task-specific model fitting. However, the current performance frontier remains dominated by attention-heavy architectures, where attention is used throughout the modeling pipeline. This raises a natural question: is attention necessary at every stage of tabular in-context learning? We introduce SOMTab, a Set-Order Mamba architecture for efficient tabular in-context learning. SOMTab separates representation construction from query-conditioned retrieval. For row and column representations, it maps unordered table tokens into stable latent slots and applies Mamba-based state-space mixing to construct compact representations. For final prediction, it retains attention-based in-context learning to preserve query-conditioned retrieval from labeled context examples. We further introduce DCH-TailMix, a synthetic prior that combines degree-corrected graph heterogeneity with mixed heavy-tailed regimes to diversify synthetic dependency structures. Across tabular benchmarks, SOMTab approaches the performance of strong Transformer-based tabular foundation models while achieving faster inference and lower GPU memory usage, yielding a favorable efficiency–accuracy trade-off.
[AI-74] HyQuant: Hybrid-Precision Quantization for LLM Attention
链接: https://arxiv.org/abs/2608.27875
作者: Jiatong Ding,Bingxin Xing,Yu Zhang,Dian Ding,Xiaodong Yi,Xianbin Ouyang,Feihu Zhou,Kun Zhang,Zhenyu Guo,Hao Pan,Guangtao Xue,Yiming Zhang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Quantization has been widely adopted in LLM training and inference to reduce cost and improve efficiency. However, low-bit quantization of the \emphattention module often introduces large errors at very low bit-widths, causing performance degradation. Existing methods mainly rely on smoothing techniques to handle outliers, while we propose a hybrid quantization design to better balance accuracy and efficiency. Specifically, we propose \textbfHyQuant, an efficient hybrid quantization framework for LLM attention. HyQuant quantizes most attention states into low-bit formats while retaining a small set of vertical-line tokens and local-window states in high precision. These accuracy-critical regions are selected using lightweight vertical-line-aware attention-pattern signals, reducing quantization error with limited overhead. In the Prefill stage, HyQuant uses a hybrid-precision quantized attention operator that preserves vertical-line tokens and a local sliding window in full precision while quantizing the remaining context. In the Decode stage, HyQuant applies the same principle to KV-cache compression and fuses KV dequantization with attention computation to improve memory and hardware efficiency. Across diverse tasks, models, and datasets, HyQuant maintains nearly lossless accuracy with an extremely simple design, demonstrating the efficiency and practical feasibility of hybrid quantization for LLM attention. Code is available at: this https URL .
[AI-75] See Hypothesize Validate: Multimodal Agent ic Framework for Discovering Governing PDEs
链接: https://arxiv.org/abs/2608.27869
作者: Sarang Manoj Pekhale,Amartya Roy,Rajat Sarkar,Souvik Chakraborty
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Discovering governing partial differential equations (PDEs) from observational data remains a core challenge across the sciences. Existing sparse-regression, symbolic-regression, and LLM-based approaches can be constrained by predefined libraries, noise sensitivity, hallucination, or limited iterative refinement. We introduce \textbfMAGE (\textbfMultimodal \textbfAgentic \textbfGoverning \textbfEquation Discovery), an agentic framework that organizes PDE discovery as a \textitconfidence governed hypothesis validation loop inspired by the scientific cycle of observation, hypothesis, and falsification. Four role-specialized agents collaborate: a \textitDifferential Observer computing derivatives and diagnostic visualizations; a VLM-powered \textitPhenomenology Extractor distilling qualitative cues from multimodal diagnostics; an LLM-driven \textitGoverning Law Synthesizer proposing candidates without a predefined library; and an \textitEquation Arbiter fitting coefficients and assigning confidence scores. Discovery iterates until the top candidate clears a user-specified threshold, providing a structured process with an explicit accept-reject protocol. On the evaluated canonical PDE suite, MAGE obtains \textbf8/8 exact structural recovery and the lowest coefficient error among the compared methods on \textbf7/8 systems, with improvements of up to \textbf4 orders of magnitude and a geometric-mean improvement of approximately \textbf3 orders of magnitude. The pipeline also recovers the expected operators in two complex geometries and, on one laboratory sensor record, selects a cubic restoring-force model with held-out R^2=0.98538 . These results support further study of structured agentic reasoning for library-free governing-law discovery, while broader generalization remains to be evaluated.
[AI-76] CoRe-MoE: Compact Reusable MoE for Continual Multimodal Instruction Tuning EMNLP2026
链接: https://arxiv.org/abs/2608.27867
作者: Runze Liu,Naibin Gu,Mingxu Ai,Yuqing Li,Peng Fu,Zheng Lin,Weiping Wang
类目: Artificial Intelligence (cs.AI)
备注: Accepted to Findings of EMNLP 2026
Abstract:Continual multimodal instruction tuning requires multimodal large language models to acquire new task abilities sequentially while preserving previously learned knowledge. LoRA-MoE provides a promising solution by introducing expert-based capacity, but repeatedly learning and maintaining full LoRA experts leads to substantial parameter overhead. This raises a natural question: is full expert expansion necessary for every new task? To answer it, we analyze the SVD of task-specific LoRA updates and observe substantial overlap in their input- and output-side LoRA direction subspaces, with task-specific adaptation largely captured by lightweight coordinates over these subspaces. Motivated by this observation, we propose CoRe-MoE, a Compact Reusable MoE framework for parameter-efficient continual multimodal instruction tuning. CoRe-MoE extracts reusable input- and output-side direction bases from an initial expert bank, and for subsequent tasks trains only compact coordinate experts together with task-specific low-rank routers. Experiments on two representative MLLMs show that CoRe-MoE improves final average performance over the strongest competing baseline by up to 5.90 points, while using less than 1% of the trainable parameters required by sequential LoRA for later tasks. The code is publicly available at this https URL.
[AI-77] SpikeOPD: Stable On-Policy Distillation for Autoregressive Spiking Language Models
链接: https://arxiv.org/abs/2608.27857
作者: Enqiao Lu,Xingrui Yu,Yiwei Fu,Zhenglin Wan,Pengfei Zhou,Wangbo Zhao,Muqing Jian,Xueyi Zhang,Yang You,Ivor Tsang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Spiking neural networks (SNNs) offer a path to energy-efficient language modeling through sparse encoding and event-driven computation, but training capable spiking language models from scratch remains difficult. A practical alternative is ANN-to-SNN migration through knowledge distillation (KD), where a pretrained artificial neural network (ANN) teacher supervises an SNN student. Existing migration approaches distill on fixed corpus prefixes, whereas autoregressive inference conditions on self-generated prefixes, creating prefix-source mismatch. It manifests as output-policy mismatch with the ANN teacher and internal spiking-dynamics drift between self-generated and matched corpus prefixes. On-policy distillation (OPD) offers a natural way to mitigate both manifestations by continuing teacher supervision on self-generated prefixes. We evaluate a teacher-only full-KL variant, Vanilla OPD, via a controlled stress test and observe it may suffer from delayed rollout-feedback collapse. This result shows that on-policy coverage alone does not ensure stable adaptation. Motivated by these findings, we propose SpikeOPD, a stable on-policy distillation framework for autoregressive SNNs that learns from self-generated prefixes while maintaining rollout stability. It applies full-KL teacher correction to reduce output-policy mismatch, while matched-prefix policy anchoring constrains policy departure from the frozen reference SNN on the same prefixes. Layerwise spike regularization further limits firing-rate deviations during on-policy adaptation. Across three model scales, SpikeOPD improves average accuracy over the corresponding KD SNNs by 0.8, 1.7, and 2.9 points at 0.125B, 0.35B, and 1.3B, respectively, while preserving their sparse-compute profiles.
[AI-78] From Uncertainty to Clinical Risk: Severity-Aware Conformal Planning for Interactive Medical Diagnosis
链接: https://arxiv.org/abs/2608.27847
作者: Yue Zhou,Haiyang Zhou,Jin Zhang,Kong Wang,Yongxin Ni,Youhua Li,Hanwen Du
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Interactive medical diagnosis dynamically acquires patient information through multiple rounds of questioning, supporting accurate, efficient, and safe clinical decisions under incomplete evidence. Existing methods commonly guide information acquisition with predictive uncertainty or label ambiguity, but overlook the asymmetric clinical risk of missing severe diseases and lack unified long-horizon planning over whether to continue asking questions or commit to a diagnosis. To address these limitations, we propose Severity-Aware Conformal Clinical Planning, which formulates interactive diagnosis as a risk-sensitive sequential decision problem. The framework maintains complementary diagnostic, safety, and masked-evidence beliefs; calibrates turn-specific diagnostic prediction sets and severity-weighted differential-diagnosis risk on held-out diagnostic trajectories; and introduces the calibrated clinical risk into Monte Carlo Tree Search to jointly evaluate long-horizon Ask and Commit trajectories. Experiments on DDXPlus and MediQ show that our method achieves more accurate diagnoses with fewer questions across multiple large language models, while improving differential-diagnosis quality and reducing high-risk errors in severe cases. These findings validate the value of using clinical risk, rather than predictive uncertainty alone, as a planning signal and demonstrate the effectiveness of the proposed framework for information acquisition and risk-aware diagnostic decision making. They also motivate future work on clinical-risk-oriented interactive diagnosis and information-acquisition methods.
[AI-79] KLOD: Locality-Preserving Knowledge Editing via Non-Target Distribution Preservation EMNLP2026
链接: https://arxiv.org/abs/2608.27839
作者: Hojun Jeong,Gyunyeop Kim,Sangwoo Kang
类目: Artificial Intelligence (cs.AI)
备注: To be published in EMNLP 2026 Findings
Abstract:Fine-tuning-based knowledge editing is simple and architecture-agnostic, but standard cross-entropy increases the edited target probability without explicitly constraining changes in the non-target output distribution. In sequential editing, such unconstrained redistribution can accumulate as distributional drift and contribute to locality degradation. We propose KLOD, a bounded and distribution-preserving objective for fine-tuning-based knowledge editing that separates the intended target update from distributions that should remain stable. KLOD stops target amplification once a probability threshold is reached, while preserving the target-excluded non-target distribution at target positions and the full next-token distribution at prefix positions. Experiments on CounterFact and ZsRE with Llama3-8B-Instruct and Qwen2.5-7B-Instruct show that KLOD substantially mitigates locality degradation while maintaining high edit reliability. The target probability threshold further provides a controllable Generalization–Locality trade-off. Ablation, multi-seed, and distributional KL analyses support the interpretation that KLOD’s locality gains are associated with preserving output distributions rather than simply weakening the edit. Code is available on GitHub this https URL .
[AI-80] FISGuard: Defending Against Membership Inference via Fixed Input Subspaces
链接: https://arxiv.org/abs/2608.27836
作者: Haocheng Jiang,Hua Shen
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Distributed, Parallel, and Cluster Computing (cs.DC)
备注:
Abstract:As large language models are increasingly adopted in federated learning, protecting user privacy while performing parameter-efficient fine-tuning on distributed private data has become an important challenge. Although clients only share gradients instead of directly uploading raw data, the shared gradients may still leak membership information about training samples. ProjRes (SP, 2026) further increases this risk: with less information and without accessing model outputs, an attacker can effectively distinguish members from non-members solely based on the projection residual between a candidate representation and the subspace induced by server-observable gradients. Existing defenses against membership inference mostly rely on gradient perturbation or regularization, which can not only degrade model utility but also fail to effectively defend against the membership inference attack introduced by ProjRes, which exploits the geometric structure of gradients. To address this issue, we propose FISGuard, a lightweight defense. Its key idea is to construct and fix a low-dimensional representation subspace using independent public data, thereby restricting the space through which private representations are exposed via gradients while preserving the primary information required for downstream tasks. This substantially reduces the projection-residual discrepancy between members and non-members. We evaluate FISGuard against five representative defense methods across three NLP datasets, two LLMs, and two fine-tuning strategies, Adapter and LoRA. The results show that FISGuard reduces the ProjRes attack AUC to near the random-guessing level of 0.5 in most settings, while maintaining downstream task performance close to that of the undefended model and introducing only limited computational overhead, thereby achieving a favorable privacy–utility trade-off. Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Distributed, Parallel, and Cluster Computing (cs.DC) Cite as: arXiv:2608.27836 [cs.CR] (or arXiv:2608.27836v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2608.27836 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-81] RealSWE: A Compositional Evaluation of Coding Agents under Realistic User Requests
链接: https://arxiv.org/abs/2608.27831
作者: Gyuhyeong Kim,Hyojung Gwon,Jeonghyeon Kim,Kyuhong Shim,Sunjae Lee
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Coding agents are now commonly evaluated on the SWE-bench family of benchmarks, whose tasks are built from curated GitHub issues–long, structured, and information-rich. Real user requests, however, are typically far shorter and less structured. To characterize this gap, we define a six-category information taxonomy and four dimensions of linguistic style, and apply them to real user prompts from SWE-chat and problem statements from SWE-bench Verified and Pro. We find that requests carrying only a problem statement, alone or with limited additional context, account for 88% of real prompts but just 7% of benchmark problems. Furthermore, 87% of real prompts are casually written whereas 94% of benchmark problems are formal. Guided by these observations, we introduce sys, 381 multi-variant task families derived from SWE-bench Verified and Pro. Variants within each family share the same underlying task and gold patch while differing only in information composition and linguistic style. Evaluating seven contemporary LLMs with sys, we find that i) realistic inputs reduce resolution rates by 6.4 pp on average and can change model rankings. Controlled analysis further shows that ii) including Desired Behavior and Motivation significantly affects performance, whereas Environment Information and Reproduction Steps merely add tokens without measurable benefit; iii) linguistic style has only small, model-dependent effects. These findings provide actionable guidance for users and agents: explicitly stating the desired behavior and motivation–which most real prompts omit–substantially improves the LLM’s software engineering performance.
[AI-82] Evidential-Based Higher-Order Set Argumentation Framework
链接: https://arxiv.org/abs/2608.27824
作者: Shuai Tang
类目: Artificial Intelligence (cs.AI); Logic (math.LO)
备注:
Abstract:Evidential argumentation extends Dung’s abstract argumentation by requiring arguments and interactions to be backed by chains of evidence rooted in prima-facie elements. However, existing formalisms lack a unified treatment of evidential support, higher-order relations (attacks and supports targeting arbitrary elements), and collective interactions (sources as sets). In this paper, we introduce the Evidential-Based Higher-Order Set Argumentation Framework (EHSAF), which conservatively generalises several existing frameworks within a single expressive setting. We develop two complete semantics for EHSAFs: an \emphadjacent complete labelling semantics that admits multiple truth values (true, false, undecided) for arguments in support cycles, reflecting an open epistemic attitude toward future evidence; and an \emphextension-based complete semantics that follows a strict evidentialist stance, accepting only arguments with well-founded support chains. We show that these two semantics diverge in the presence of support cycles, and prove their equivalence under support-acyclicity. To enable computational reasoning, we provide a normal propositional encoding of EHSAFs and prove that, in three-valued Łukasiewicz logic, its models correspond precisely to the adjacent complete labellings. We further extend this encoding to continuous fuzzy logics (Gödel, Product, and Łukasiewicz), defining a continuous fuzzy normal encoded semantics. We establish that this fuzzy semantics satisfies key properties—continuity, monotonicity, boundary conditions, and solution existence—and that its ternarisation recovers the adjacent complete labellings under natural t-norm conditions. Our framework thus unifies expressive argumentation with principled three-valued and fuzzy semantics, bridging the gap between qualitative and quantitative reasoning about evidence.
[AI-83] Actionable CBFI: Integrating Structural Decomposition and Causal Counterfactual Recourse for Tabular Machine Learning
链接: https://arxiv.org/abs/2608.27821
作者: Sejong Oh
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Explainable artificial intelligence (XAI) increasingly calls for actionable counterfactual recourse, yet current methodologies face challenges related to causal invalidity, excessive cognitive burden, and predictive failure. Exhaustive causal search algorithms often require modifications to multiple attributes, whereas additive attribution-guided methods, such as SHAP, ignore higher-order feature synergies, leading to suboptimal predictive momentum and diffuse intervention effort in complex nonlinear models, such as XGBoost. To bridge this gap, we introduce actionable case-based feature importance (A-CBFI), a diagnosis-prescription integrated framework for tabular machine learning. Grounded in structural causal models (SCMs), A-CBFI isolates synergistic interaction bottlenecks and releases suppressive structural locks, translating them into targeted interventions. By mathematically separating the active user intervention space (L_\mathrmactive) from downstream effects and concentrating over 98.3% of the intervention effort on diagnosed root causes, A-CBFI enables highly targeted interventions. Empirical evaluations across the financial and healthcare domains demonstrate that A-CBFI reduces the active human intervention burden by 76.9% while maintaining comparable global recourse cost to exhaustive causal baselines. By prioritizing the diagnosed causal bottlenecks, A-CBFI provides targeted and actionable recourse while maintaining causal validity and achieving full relative convergence across all causally feasible instances.
[AI-84] AcCoRD: Evaluating User-Agent Collaboration Under Realistic User Preference Dynamics
链接: https://arxiv.org/abs/2608.27818
作者: Tejas Srinivasan,Shikib Mehri,Nandita Shankar Naik,Anirban Das,William M. Campbell,Jesse Thomason
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:User preferences in user-agent collaboration are rarely static and fully-specified upfront: preferences are formed, revealed, adjusted, and relaxed during interaction. Existing benchmarks for evaluating user-agent collaboration focus almost exclusively on resolving underspecified preferences, thereby failing to capture the richer dynamics of real-world interaction. We introduce AcCoRD, a user-agent collaboration benchmark requiring agents to handle diverse user preference dynamics in two domains: online shopping and travel planning. We evaluate five frontier LLMs under two prompting strategies: vanilla ReAct, and an uncertainty-guided variant that prompts models to identify and resolve ambiguity about user preferences. Our results reveal that frontier models can handle underspecification but struggle to satisfy preferences that emerge or evolve mid-interaction and require more sophisticated uncertainty modeling. Further, prompting alone fails to elicit the required uncertainty recognition. We release AcCoRD as a resource for developing agents that can navigate the full complexity of real-world user preferences.
[AI-85] ContextLeak: Exfiltrating LLM Agent Context via Malicious Tools
链接: https://arxiv.org/abs/2608.27800
作者: Yuqi Jia,Ruiqi Wang,Patrick Li,Yuepeng Hu,Peinian Li,Neil Gong
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:Exfiltrating an LLM agent’s runtime context – such as the user prompt, execution trajectory, and tool list – poses severe security and privacy risks to users. Such attacks can be carried out via malicious tools and typically require three conditions: (1) the agent selects the malicious tool for task execution, (2) the agent passes its runtime context as input arguments to the tool, and (3) the tool’s implementation transmits these inputs to an attacker-controlled endpoint. Existing work primarily focuses on conditions (1) and (3), leaving condition (2) largely unexplored, despite its critical role in enabling successful context exfiltration. In this work, we bridge this gap by developing ContextLeak, a malicious tool attack that induces the agent to both select the tool and disclose its context as input arguments. We realize this attack by carefully crafting the tool’s name and description using reinforcement learning. Specifically, ContextLeak employs an LLM, referred to as the attack LLM, to automatically generate the malicious tool’s name and description. To improve attack effectiveness, we fine-tune the attack LLM via reinforcement learning on a set of shadow users with diverse, simulated agent contexts. Our key technical contribution is the design of novel reward functions tailored to the context exfiltration objective, enabling effective reinforcement-learning-based fine-tuning of the attack LLM. Extensive evaluation demonstrates that our attack remains highly effective even when the shadow users’ contexts differ substantially from those of the victim users. Moreover, ContextLeak significantly outperforms existing malicious tool attacks when adapted to this setting. Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI) Cite as: arXiv:2608.27800 [cs.CR] (or arXiv:2608.27800v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2608.27800 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-86] ReToolSQL: Agent ic Reinforcement Learning for Robust Text-to-SQL
链接: https://arxiv.org/abs/2608.27796
作者: Pratik Kakkar,Chandra Dhir,Ravi Shankar,Pareekshit Reddy Gaddam,Anup Shirgaonkar
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Recent work has shown that reinforcement learning from execution feedback can substantially improve text-to-SQL performance, often enabling smaller models to match or exceed much larger systems. However, most existing approaches treat SQL generation as a single-turn task, limiting the model’s ability to recover from errors through iterative refinement. We present ReToolSQL, a two-stage training framework for text-to-SQL that combines (i) a supervised warm-start on rejection-sampled reasoning traces with (ii) agentic reinforcement fine-tuning (RFT) over multi-turn tool-use trajectories. The key insight is that the two stages act on complementary axes, the supervised fine-tuning (SFT) on verified privileged-teacher traces expands the set of solvable questions (raising pass@k coverage on the hardest cases), while RFT converts that expanded capability into higher single-pass accuracy by teaching the model when to verify, what evidence to retrieve, and how to repair faulty SQL from execution feedback. Applied to Gemma 4 instruction-tuned (31B), RFT alone achieves 73.66% execution accuracy (EX) on the BIRD-SQL development benchmark (74.12% EX with self-consistency). Initializing RFT from the SFT checkpoint (SFT \to RFT) yields our strongest model at 74.32% EX single-pass and 74.77% EX with self-consistency. At the time of writing, this ranked first on the BIRD single-model development-set leaderboard. The approach uses composite rewards anchored on execution correctness, requires no human annotation beyond the benchmark itself, and operates within a single dense 31B model, showing that a properly designed SFT \to RFT pipeline over tool-use trajectories is a practical path toward robust enterprise-grade text-to-SQL.
[AI-87] Credo: Reusable Declarative Primitives for Agent ic Workflows
链接: https://arxiv.org/abs/2608.27790
作者: Duo Lu,Andrew Crotty,Uğur Çetintemel
类目: Artificial Intelligence (cs.AI); Databases (cs.DB)
备注:
Abstract:An LLM application depends on both a model and a harness: the program that determines what each call sees, how many calls to make, and which answers to trust. Coding agents can now discover strong harnesses by searching over candidate programs, but the resulting artifact is an opaque block of imperative code whose logical steps, runtime signals, physical execution decisions, and prompt strategies remain implicit and task-specific, forcing subsequent tasks to start the harness search process from scratch. The potential for reuse, however, is substantial. A searched harness encodes significant knowledge, such as the logical steps that work, the signals that matter, the physical operator decisions that adapt execution, and the prompt strategies that are effective, yet this knowledge is buried in imperative code with no inspectable or reusable structure, nor does it carry any provenance or metadata. Credo addresses this problem by recovering a structured declarative description of a searched harness, tagging each extracted primitive with relevant metadata, and cataloguing all of it with provenance. A compiler can then bind stored primitives to generate harnesses for new tasks without having to start the search over from scratch. This paper provides preliminary results demonstrating the potential of our approach and lays out a related research agenda that the database community is well-positioned to tackle, including cost-based compilation over declarative catalogs and catalog maintenance under model and workload drift.
[AI-88] Beyond Search-Imitation: Prior-Directed Exploration for Searchless Chess
链接: https://arxiv.org/abs/2608.27757
作者: Szymon Miłosz,Piotr Duch,Szymon Grabowski
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Searchless chess networks reach human master strength from a single forward pass by imitating a stronger teacher: the strongest, Leela Chess Zero’s (Lc0) released Chessformer, distills the visit counts of an AlphaZero-style Monte Carlo Tree Search (MCTS). Imitating a search is a poor proxy for playing without one, so we fine-tune for single-pass strength with self-play reinforcement learning (RL). Its exploration is usually supplied by an entropy bonus, the reverse Kullback-Leibler (KL) divergence to uniform. We replace it with a forward, mass-covering KL toward the network’s own MCTS prior (prior-directed exploration), so exploration covers the moves the prior judges promising, and pair it with an entropy-adaptive sampling temperature, set by the value head’s outcome uncertainty, that sharpens once a position is decided. In about two thousand steps it raises puzzle accuracy from 93.9% to 94.9% on a 100,000-puzzle suite and mate-in-four accuracy from 77% to 81% while holding searchless strength at or slightly above the base. Measuring tactical accuracy and playing strength together across a matched-compute sweep, we find the two dissociate: accuracy gains fall in a one-point band while ratings straddle the base, and a control fine-tuned on puzzles alone posts the study’s largest tactical gains while shedding roughly 260 Elo; a better puzzle-solver is not thereby a stronger player. Distribution-level measurements show what anchoring buys: without a regularizer self-play collapses onto a single line of play, and the puzzles newly solved are the near misses whose winning move the prior kept alive. The forward-KL prior tops the rating ladder, statistically tied with a reverse-KL anchor that concentrates twice as hard and drops the hardest solutions the mass-covering prior keeps in support.
[AI-89] Probing Perceptual Priors of MLLM s via Gibbs Sampling with Interpretable Generative Controls
链接: https://arxiv.org/abs/2608.27727
作者: Manuel Cherep,Pattie Maes,Nikhil Singh
类目: Artificial Intelligence (cs.AI)
备注: 40 pages, 21 figures
Abstract:A model’s behavior on a task is jointly determined by the input it receives and the prior it brings in, i.e. the distribution over stimuli it implicitly expects. Interpretability research has traditionally studied models by holding inputs fixed and examining model responses either mechanistically, probing how internal structure represents inputs, or behaviorally, measuring how variation in inputs leads to variation in outputs. Neither reconstructs the prior distribution itself, since internal structure shows what a model can represent, not what it expects, and any fixed stimulus set leaves most of the possible input space unseen. In particular, such an input space in real-world settings, such as images seen by VLMs, is extremely high-dimensional and diverse. These priors thus remain a poorly understood component of models that nonetheless influence real-world behavior. We propose a method to sample from models’ perceptual prior distributions directly, by steering a generative model to produce stimuli along controllable axes and running Gibbs sampling over that space with the model under study as the judge. We apply this to a variety of categories and target variables (such as trustworthiness in faces and cheapness in art images) and recover both canonical biases and surprising novel priors invisible to direct prompting, warranting further investigation of their downstream effects.
[AI-90] PCFBench: A Diagnostic Benchmark for Product Carbon Footprint Estimation
链接: https://arxiv.org/abs/2608.27716
作者: Krishna Rao,Andrew Dumit,Shaena Ulissi,Jacob Feintzeig,P. James Joyce,Daniel Frank,Steven Watson,Jonathan Glidden,Gizem Ilayda Dinc,Travis M. Kwee
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:AI systems are being deployed on high-stakes, domain-specific workflows that demand correctness not just in the final output, but at every intermediate step. One such workflow is estimating a product carbon footprint (PCF), the greenhouse-gas emissions attributable to a physical product. AI agents are increasingly being used to generate PCFs, but existing evaluations score either total emissions (hiding error sources and cancelling mistakes) or sub-tasks in isolation (missing compositional interactions). We introduce PCFBench, the first benchmark to carve PCF modeling into independently-evaluable tasks that require decomposition, retrieval, ontology matching, and numerical extraction. It comprises 614 expert-labelled items across six tasks. Together they probe reasoning under under-specification, conflicting context, and numerical constraints. Across eight frontier LLMs from four providers, no single model dominates. Although the strongest models estimate total product emissions within 2 times of declared totals on 77% of products, this rate drops to 37-58% when the PCF is generated step by step, with only 45-75% obeying mass conservation. These failures undermine the transparency practitioners need to compare products and drive decarbonization. We release the dataset and evaluation harness to support targeted progress.
[AI-91] RiskBlend: A Multi-Signal Framework for Test Input Prioritization in Machine Learning Regression Testing
链接: https://arxiv.org/abs/2608.27704
作者: Madhusudan Srinivasan,Namith Nishal Raphae
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
备注:
Abstract:When machine learning classifiers are retrained, inputs correctly classified by the previous model version may be misclassified by the updated version, creating regression faults that are costly to detect because verifying predictions against ground truth may require human annotation, expert review, or expensive simulation rather than inexpensive model inference. Test input prioritization addresses this problem by ranking inputs so that a limited verification budget reveals as many regression faults as possible. Existing approaches rely predominantly on single-model confidence scores and do not exploit how predictions, decision boundaries, and local neighborhoods change between model versions. We propose RiskBlend, a classifier-agnostic prioritization framework that combines four complementary risk signals: historical failure patterns, prediction shift, decision-boundary shift, and neighborhood change. These signals are combined using validation-learned APFD-squared weighting. Across four datasets, five classifiers, four regression-update scenarios, and 15 random seeds, totaling 1,200 experimental configurations, RiskBlend achieves the highest average APFD in all 80 dataset-classifier-scenario combinations, with improvements of up to 0.32 APFD over the strongest baseline. Confidence-based methods remain competitive primarily for linear classifiers on sparse categorical features, which we attribute to feature-space geometry. The results show that cross-version behavioral signals provide important complementary information for prioritizing regression faults in machine learning systems.
[AI-92] Evaluating Loss Functions in Differentiable Out-of-Domain Sound-Matching with Partial Parameter Distance
链接: https://arxiv.org/abs/2608.27698
作者: Amir Salimi,Daniel Penner,Kalvin Eng,Abram Hindle,Osmar R. Zaïane
类目: ound (cs.SD); Artificial Intelligence (cs.AI)
备注:
Abstract:In out-of-domain (OOD) sound-matching, a synthesizer is optimized to mimic a sound it did not generate. OOD evaluation of loss functions is underexplored in part because the standard “parameter loss” metric requires a shared parameter space between target and imitator, which OOD settings lack. We introduce Partial Parameter Distance (PPD), which applies parameter loss only to the critical parameters that mismatched synthesizers share (e.g., filter cutoffs), enabling automatically evaluated OOD experiments; we verify its results with blinded listening tests. Across seven scenarios involving band-pass filtering, amplitude modulation, and pitch-bending, we evaluate four differentiable loss functions (SIMSE_Spec, L1_Spec, JTFS, DTW_Envelope). Loss-function effectiveness remains tightly coupled to the method of synthesis: SIMSE_Spec excels at filter-cutoff recovery, DTW_Envelope at amplitude-modulation recovery, and JTFS at smooth pitch trajectories. Parameter-based evaluation agrees with listening tests on the top-ranked loss function in five of seven scenarios, demonstrating its utility as a diagnostic tool.
[AI-93] Agents for Everyone: A Workshop Framework for Building Agent ic AI Capabilities in a Distributed Curation Community
链接: https://arxiv.org/abs/2608.27675
作者: Seth Carbon,Sierra Moxon,Kimberly Van Auken,Pascale Gaudet,Christopher J. Mungall
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Agentic AI has the potential to accelerate curation of biological databases and knowledge bases. However, uptake has been hindered by a number of challenges and obstacles, including access to agents and appropriate training. Here we describe how we have attempted to address and mitigate these challenges and obstacles through the deployment of a cloud-based agentic environment, and the development of an interactive training workshop for the Gene Ontology Consortium. Our cloud environment for agentic-assisted curation was based on the JupyterHub platform, and utilized Claude Code as a universal harness. This allows curators to interact with an agent session through a terminal running in the browser, and has additional benefits such as centralization of access through a single API gateway, removing the need for participants to manage subscriptions or install software locally. We created four training modules, walking participants through basic agentic tool use first and then working up to agentic biological pathway curation using the existing GO-CAM (GO Causal Activity Model) curation tool. Thirty-seven participants took part in the four-hour workshop. Our key takeaway from this workshop is that building community capability with agentic AI is primarily a problem of access, workflow design, and training. Removing technical barriers, introducing capabilities gradually, grounding exercises in familiar curation tasks, and giving curators direct experience evaluating agent output can provide a practical route toward building shared agentic AI capability in distributed scientific communities.
[AI-94] A Framework for Object-Centric Predictive Monitoring of Collaborative Processes
链接: https://arxiv.org/abs/2608.27671
作者: Daniel Calegari,Andrea Delgado,Leonel Peña,Martín Rubio
类目: Artificial Intelligence (cs.AI)
备注: Submitted to Information Systems, under review
Abstract:Predictive Process Monitoring (PPM) of collaborative, inter-organizational processes requires reasoning over multiple interdependent entities, including participants, messages, local executions, and the global collaboration case. Existing approaches extend traditional event logs with collaboration attributes but retain a single-case perspective, leaving much of this structure implicit. Object-centric process mining (OCPM) provides an alternative by representing these entities as first-class objects with explicit relations and multiple notions of case. This study connects collaborative PPM and OCPM through three contributions: (i) a formal semantic mapping from extended collaborative event logs to an OCED-conformant object-centric representation, serialized in OCEL 2.0; (ii) a reformulation of collaborative prediction tasks as object-centric prediction tasks; and (iii) a reproducible converter and prediction pipeline implementing the proposed mapping. We evaluate the framework on four public collaborative event logs and a fifth derived from the BPI Challenge 2013 incident-management log by executing the fourteen reformulated tasks using five predictive strategies across tabular, sequential, and graph-native encodings. We further discuss the benefits, limitations, and threats to the approach’s validity. The representation makes collaboration structure explicit and makes it natural to state prediction targets based on object relations that fall outside the case-centric taxonomy, at the cost of increased relational complexity and dependence on object-centric tooling.
[AI-95] If Agents Were Angels No Governance Would Be Necessary: Out-of-Band Policy Enforcement at a Trusted Tool Boundary
链接: https://arxiv.org/abs/2608.27646
作者: Marc Millstone,Tyler Akidau,Johannes Brüderl,Marat Pekker
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Give an agent a human’s credential and it inherits the person’s reach without the judgment that limits its use. It can sweep every reachable record into model context, where hidden instructions steer its next call, and every request stays credential-valid while the agent exceeds its job or absorbs a secret. Prompts are a brittle guardrail: one fallible reasoner interprets the task and enforces its limits. We present Out-of-Band Policy Enforcement (OBPE), a trusted boundary outside agent reasoning. It authorizes the typed operation and resource, narrows the query before the backend call, then filters records and fields or masks values in the response. Semantic gating can deny or hold an authorized call on argument values or external state. A data policy owner sets the maximum grant; agent policy can only narrow it. We prove, under stated conditions, that the policy plan is order-independent and agent policy cannot widen the ceiling. Field removal covers one execution; masking and history rules claim less. We release an HTTP proxy prototype simplified from our production system, with conformance tests tying its typed Cedar policy core to the model. Against Jira and ServiceNow mocks, our benchmark compares prompted agents with and without OBPE on four models, including 20 adaptive red-team tasks. A trace failure means protected data entered agent context, an exact value appeared in the answer, or a forbidden effect completed. In 3,621 trials it fell from 57.6% to 0.2%, a cluster-weighted reduction of 41.2 points [95% CI: 27.7, 54.9]; fulfillment fell from 79.1% to 60.9%, while paired safe-useful completion rose 21.8 points [9.5, 35.2]. Some answers reconstructed a value that never entered context or used filtered row counts as an oracle: shaping one execution is not noninterference. Write controls, durable approval, and temporal and aggregate policies lie outside this evaluation. Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2608.27646 [cs.AI] (or arXiv:2608.27646v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.27646 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-96] nsor-Accelerated Eager Multi-Resolution Grids for Evolving Large-Scale Substrates GECCO’26
链接: https://arxiv.org/abs/2608.27612
作者: Romain Claret,Michael O’Neill,Paul Cotofrei,Kilian Stoffel
类目: Neural and Evolutionary Computing (cs.NE); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: GECCO '26 Companion: Proceedings of the Genetic and Evolutionary Computation Conference Companion, Pages 533 - 536
Abstract:In neuroevolution, indirect encoding generates neural network connectivity from a compact genome rather than specifying each connection. ES-HyperNEAT automatically discovers where to place hidden nodes by examining CPPN output patterns: it recursively subdivides space using a quadtree, expanding regions where CPPN outputs show high variance. This adaptive approach discovers network topology without manual substrate specification, extending the fixed-grid HyperNEAT framework built on NEAT. However, the quadtree resists tensorization. Each depth level depends on the parent’s variance, forcing sequential evaluation. Different CPPNs produce different subdivision patterns, preventing batching. And variable leaf counts are incompatible with JAX’s static shape requirement for JIT compilation. Our prior work confirmed these limits at depths exceeding 5, and a JAX reimplementation of the quadtree yielded only marginal speedup despite batched optimizations, motivating the eager reformulation presented here. We present EMR-HyperNEAT, which evaluates all positions at all resolutions up front, then filters using the same variance criterion: ES-HyperNEAT’s subdivide_if(var \theta ) becomes eval_all(); filter(var \theta ). This performs more CPPN queries than necessary, but all queries become independent and parallelizable across both cores and population members, reducing complexity from \BigO( 4^D ) to \BigO( 4^D/P ) across P parallel cores. Recurrent substrate configurations become feasible through a connection type taxonomy. The experiments section validates 12-34 \times on-device GPU speedup on XOR at depths 5-7, and empirically higher solve rates across benchmarks. Comments: GECCO '26 Companion: Proceedings of the Genetic and Evolutionary Computation Conference Companion, Pages 533 - 536 Subjects: Neural and Evolutionary Computing (cs.NE); Artificial Intelligence (cs.AI); Machine Learning (cs.LG) Cite as: arXiv:2608.27612 [cs.NE] (or arXiv:2608.27612v1 [cs.NE] for this version) https://doi.org/10.48550/arXiv.2608.27612 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Related DOI: https://doi.org/10.1145/3795101.3805361 Focus to learn more DOI(s) linking to related resources
[AI-97] PHR-VLA: Planning Horizon Reasoning for Vision-Language-Action Models
链接: https://arxiv.org/abs/2608.27609
作者: Davood Soleymanzadeh,Kaidi Zhang,Zhiyuan Zhang,Bihao Zhang,Xiao Liang,Yu She,Minghui Zheng
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:
Abstract:Vision-language-action models (VLAs) have shown strong promise for general-purpose robotic manipulation by mapping language instructions and vision observations directly to actions. However, most VLAs primarily condition action prediction on current observations and lack an explicit mechanism for reasoning over future task dynamics, which is particularly important for fine-grained, contact-rich manipulation. We present PHR-VLA, a framework that enables planning-horizon reasoning in VLAs through privileged latent representations of future dynamics. PHR-VLA introduces a lightweight auxiliary future head that, during training, aligns the VLA’s internal representations with latent dynamics extracted from future observations. Evaluation results demonstrate that local, contact-centric, patch-level latent dynamics supervision from the wrist camera improves success rate on LIBERO from 84.1% to 88.4% and on real-world disassembly tasks from 63.3% to 82.5%. Patch-level supervision from a third-person camera also improves performance on Meta-World from 56.70% to 57.8%. These results demonstrate that privileged latent dynamics alignment provides an effective training signal for improving anticipatory reasoning in VLA policies. Project website: \hrefthis https URLthis https URL
[AI-98] LongGuard: Mechanistic Analysis and Training-Free Mitigation of Long-Context Failure in Safety Guardrails
链接: https://arxiv.org/abs/2608.27580
作者: Ziyang Chen,Xing Wu,Songlin Hu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Safety guardrails serve as the last line of defense against harmful inputs and outputs of large language models (LLMs), yet they are trained and evaluated almost exclusively on short text. We present LongGuard, a framework that evaluates, mechanistically analyzes, and mitigates long-context guardrail failure. We formulate the task as Safety Needle-in-a-Haystack (SafetyNIAH) over a 0.25k-32k length grid; across 15 mainstream guardrails, unsafe recall drops monotonically by more than 50% on average, and a paired Benign-Fill vs. Needle-Repeat design attributes the failure to proportional dilution of the unsafe needle rather than to absolute length. A three-layer attention-logit-behavior analysis on six guardrails locates the mechanism: attention mass on the unsafe needle is diluted, the unsafe-over-safe logit margin is compressed in lockstep, and the detection decision collapses accordingly, with this attention-logit-behavior chain remaining consistent after partialling out length. We further isolate a sparse set of guard-specialized retrieval heads that exhibit partial specificity relative to their base models. Building on the analysis, we propose two training-free mitigations - Chunked Detection (CD) and Attention-Head Sharpening (AHS) - and a deployment protocol, Context-Aware Hyperparameter Routing (CAHR), that selects configurations by context length and audit side. Across five benchmarks spanning synthetic data, long-context attacks, and reasoning-model outputs, CAHR-CD and CAHR-AHS improve the six-guardrail average by 22% and 13%, respectively. Code and data are available online.
[AI-99] Self-Explainable Multi-Label Graph Neural Network for Correlated Evidence Attribution
链接: https://arxiv.org/abs/2608.27574
作者: Yingqi Feng,Yufei Tang,Min Shi,Xingquan Zhu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 10 pages
Abstract:Multi-label graph learning intends to capture the intrinsic complexity of real-world applications, where one sample is often related to multiple groups or consists of multiple objects. To date, a handful of multi-label graph learning methods exist, but none of them integrate training-time interpretation capability. While post-hoc graph explainers have been developed, they do not explicitly model label-dependent evidence sharing in multi-label graph learners, especially when label pairs are weakly or negatively associated. As a result, post-hoc approaches may miss how evidence should be shared or separated across different labels. This paper advances a new end-to-end self-explainable multi-label graph neural network (SEMGNN), which aims to simultaneously classify multi-labeled nodes and identify edges significantly contributing to each target node w.r.t. predicted labels. Different from post-hoc methods, SEMGNN jointly learns a predictor and a sparse edge-mask explainer within a unified framework and training objective. Label-label correlations are used to improve multi-label node classification and enhance individual label explanations, so that different labels of a node can be supported by distinct yet coherent structural and/or correlated evidence. Experiments and comparisons on synthetic and real-world multi-label networks, in social networking, entertainment, and life sciences, show that SEMGNN achieves competitive or improved predictive performance while providing more faithful and compact label-conditioned explanations.
[AI-100] Nemotron 3.5 Content Safety Moderator: A Compact Multimodal Multilingual and Reasoning Enabled Content Safety Moderator
链接: https://arxiv.org/abs/2608.27548
作者: Varun Singh,Anuj Doshi,Makesh Narsimhan Sreedhar,Shaona Ghosh,Katherine Luna
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Safety moderation for deployed AI applications is moving beyond text-only prompts: systems increasingly need to judge images, documents, screenshots, and generated responses under policies that vary across domains. Existing guardrails usually cover only part of this setting, making it difficult to combine broad coverage, custom policy control, and low compute cost. We present Nemotron 3.5 Content Safety Moderator, also referred to as Nemotron 3.5 CS in this paper for brevity, a compact 4B vision-language safety moderator that jointly classifies user prompts, images, and assistant responses across 12 languages. Nemotron 3.5 CS returns safety labels for latency-sensitive moderation and can additionally produce concise reasoning traces that apply supplied custom policies and identify violated categories when reasoning is requested. We also release a multimodal and multilingual safety dataset for guard training, spanning human-labeled real-image moderation, benign vision-language and document tasks, synthetic rare-risk and jailbreak cases, and custom-policy examples. Across evaluations spanning multimodal safety, text moderation, multilingual robustness, custom-policy following, benign false positives, and latency, Nemotron 3.5 CS demonstrates a practical coverage tradeoff: it adds image-conditioned and policy-conditioned moderation while remaining broadly competitive with specialized guard models. These results suggest that compact vision-language moderators can serve as deployable front-line safety components, with reasoning used selectively for audit and policy review.
[AI-101] SETU: An Agent ic Ecosystem for Multilingual Persona-Aware Communication Coaching
链接: https://arxiv.org/abs/2608.27524
作者: Jonnalagadda Maruthi Tejas,Uponika Barman Roy,Tilottama Goswami,Samir Goswami,Mousita Dhar
类目: Artificial Intelligence (cs.AI)
备注: 6 pages, 1 figure
Abstract:Corporate training teams need scalable and explainable tools to improve workforce communication in multilingual settings. Existing systems often score text, audio, or video in isolation, or produce black-box outputs that are difficult to audit for coaching use. This paper presents SETU, an agentic ecosystem for corporate communication coaching aimed at recruiters, frontline sales professionals and training units who prepare for audience specific conversations. SETU is designed for two scoped scenarios: (i) recruiter-candidate eligibility-and-interest calls with persona context and (ii) sales pitches with target-audience adaptation; owing to limited evaluation resources, this paper reports results on scenario (ii) only. The ecosystem decomposes analysis into specialized video, audio-speech, text-relevance, scoring, notification and reporting agents coordinated through trust-aware orchestration. It generates modality-attributed coaching reports for formative training, with human reviewers retaining final judgment. The name SETU (bridge in several Indic languages) reflects the goal of bridging communication gaps across regional languages and audience expectations.
[AI-102] DAMP: Decay-Aware Mixed-Precision Recurrent-State Quantization
链接: https://arxiv.org/abs/2608.27513
作者: Tao Zhang,Jianchao Tan,Pingwei Sun,Yanqi Yu,Zixu Jiang,Yuchen Xie,Xunliang Cai,Ziqian Zeng
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Softmax attention stores key and value vectors for every preceding token, causing inference memory to grow with sequence length. Recent language models incorporating Gated DeltaNet (GDN) or Kimi Delta Attention (KDA) reduce this cost by replacing the KV cache in most layers with fixed-size recurrent states. However, these recurrent states are commonly stored in FP32 and consume substantial GPU memory; their updates are memory-bandwidth bound and contribute significantly to decoding latency. To our knowledge, we are the first to study post-training quantization of recurrent states in GDN and KDA based language models. We find that uniform quantization provides a poor accuracy–storage trade-off: INT8 and FP8 already degrade accuracy on complex reasoning tasks, while INT4 and NVFP4 reduce it to near zero. We further find that most quantization-error energy is concentrated in a small subset of channels and that the relative decay strength of state channels remains stable across prompts and tasks. Motivated by these findings, DAMP uses both quantization-error energy and decay-based persistence to identify high-risk channels during offline calibration. It stores these channels at higher precision and the remainder in INT8. We evaluate DAMP on Qwen3.6-35B and Kimi-Linear-48B across six benchmarks covering mathematical reasoning, general reasoning, and code generation. At 9.9 bits per state value, DAMP maintains average accuracy close to the FP32 baseline. DAMP reduces recurrent-state storage by 69.1%, accelerates the recurrent-state update kernel by up to 2.01x, and lowers full-model TPOT by up to 10.9%.
[AI-103] WM-R1: Training GUI Agents to Reason and leverag e World Models with Reinforcement Learning
链接: https://arxiv.org/abs/2608.27508
作者: Yu Han,Tianwen Qian
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:GUI agents trained with reinforcement learning (RL) have showcased strong environment learning capabilities on mobile platforms. However, RL typically demands extensive real-environment interactions, leading to high resource costs and instability, especially in GUI scenarios. To address these, we propose WM-R1, the first reinforcement learning framework that trains mobile GUI agents with world models instead of real environments. Specifically, world models serve as the source of state transitions during all rollouts, replacing the real Android environment within the training loop. WM-R1 also embeds world models directly into the thinking process, enabling agents to reason about the consequences of candidate actions before committing to the final action. Crucially, WM-R1 eliminates the need for real-environment interaction, supports massively parallelized and step-level granularized trajectory generation grounded in world models, and introduces a multi-dimensional rule-based reward that jointly optimizes task success, trajectory efficiency, and world model utilization. For efficient training, we curate a high-quality dataset of 2000 challenging tasks. Experiments on Android mobile benchmarks demonstrate that WM-R1-trained agents significantly outperform GRPO-only baselines and inference-time simulation methods. Code is available at this https URL .
[AI-104] Marginal Coverag e Credit Reduces Redundant Exploration in Parallel State-Entropy Optimization
链接: https://arxiv.org/abs/2608.27507
作者: Junhao Cao,Hongyi Xia,Jianian Wu,Xiaopeng Yi,Lixia Huang,Ping Guo
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Policy Gradient for Parallel State Entropy maximization (PGPSE) expands state-space coverage by training independently parameterized policies in replicated copies of the same environment. However, its pooled team-entropy score measures only collective exploration and cannot identify policies that contribute non-redundant coverage. We introduce Marginal Coverage Credit for PGPSE (MCC-PGPSE), which combines leave-one-policy-out coverage with state-owner specialization to estimate policy-specific credit. MCC-PGPSE preserves PGPSE’s pooled objective and redistributes non-negative auxiliary intrinsic rewards according to these credits without changing their total mass. This redistribution is designed to discourage redundant visitation and promote complementary coverage. We evaluated MCC-PGPSE in controlled environments, seven public discrete-state benchmarks, and representative Room and Maze settings from the original PGPSE protocol. Across all tested settings, MCC-PGPSE produced positive final window gains in normalized team state entropy and state support over the Entropy baseline. Controlled-task comparisons and the fixed-suite public aggregate were significant, whereas five-seed original-protocol comparisons were directionally consistent. Ablations and credit alignment controls indicate that most gains arise from leave-one-policy-out coverage rather than non-uniform weighting, mismatched credit, or neural novelty alone. These results support contribution-conditioned auxiliary reward allocation as an interpretable approach to improving complementary coverage among parallel policies in discrete state spaces.
[AI-105] hinking Costs Tokens: When More Structure is Worth the Price
链接: https://arxiv.org/abs/2608.27506
作者: Thomas Nolasque,John Grey,Calista Pham,Ankit Vani
类目: Artificial Intelligence (cs.AI)
备注: 8 pages, 1 figure, 2 tables
Abstract:Adding inference structure to a language model lets it search, verify, and revise, but these actions consume the very budget they are supposed to use well. In this paper, we investigate whether there exists a token-budget threshold, below which the overhead of planning and verification hurts performance and above which it helps. We evaluate two systems on FinQA and TAT-QA financial reasoning tasks, using GPT-5.4 mini across 14 budget tiers ranging from 250 to 42,000 output-equivalent tokens. The first system is a monolith, which is a single LLM call. The second is a verified search architecture that adds planning, label-blind checking, and repair capabilities. We run 1,000 cases for a total of 28,000 completed cells. Both systems score 0% at the two lowest tiers, where neither can fit a complete prompt. At 1,000 tokens, the monolith reaches 18% accuracy while verified search scores near 0%, since the planning overhead leaves no room for an answer. From 1,500 tokens onward, verified search surpasses the monolith and maintains a consistent advantage, reaching approximately 44% at the highest tiers while the monolith reaches approximately 40%. The crossover occurs between 1,000 and 1,500 output-equivalent tokens, confirmed by a strict intersection-union test ( p \le 0.001 at both endpoints).
[AI-106] Context Localization for Generalized Level-Based Evaluation in Knowledge-Based Systems
链接: https://arxiv.org/abs/2608.27482
作者: Ondrej Hutník,Natália Puškárová
类目: Artificial Intelligence (cs.AI)
备注: Author preprint of the article published in Knowledge-Based Systems (2026)
Abstract:We study context localization for generalized level-based evaluation in knowledge-based systems. The framework models situations where a structured nonnegative score, defined on facts, rules, cases, criteria or evidence units, is evaluated through conditional aggregation tests on admissible knowledge contexts. The generalized level measure maximizes a monotone set function over all contexts whose aggregated support reaches a prescribed level. We characterize when filtering the score by a context B is equivalent to localizing the admissible contexts by intersection with B . The main theorem shows that this consistency holds for all monotone set functions if and only if two structural conditions are satisfied: monotonicity with respect to contexts and a reduction property excluding positive localized support outside B . We analyze pointwise and block-generated mechanisms producing the reduction property, extend the result to parameterized systems, and interpret it as a stability criterion for context-dependent evidence selection, non-additive support evaluation and level-based knowledge aggregation.
[AI-107] Effectiveness of IoT and Deep Learning for Detection and Severity Assessment of Postelectrotermes militaris in Tea Plantations
链接: https://arxiv.org/abs/2608.27480
作者: D.K.C. Senevirathna,A.A.E. Nanayakkara,H.M.C.K. Kulathunga,J.K.D.P. Nadula,R.M. Mapatuna,Malithi Nawarathne,Jaliya L. Wijayaraja,P.D. Senanayake,Samitha Vidhanaarachchi,Kalpani Manathunga
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Sound (cs.SD)
备注: 12 pages, 7 figures. Accepted for publication in The Journal of Agricultural Sciences, Sri Lanka
Abstract:Tea plantations are vulnerable to Postelectrotermes militaris, commonly known as the Upcountry Live Wood Termite (ULWT), which can cause substantial damage when infestations remain undetected. This study proposes an IoT-enabled acoustic monitoring framework integrated with deep learning for early detection and severity assessment of ULWT infestations in tea plantations. Research Method: Audio signals were captured non-invasively from tea trunks using a high-sensitivity microphone connected to a Raspberry Pi-based IoT device, with geographic coordinates recorded for spatial tracking. After trimming, resampling, and segmentation, 2,000 ten-second samples were obtained, comprising 1,000 healthy and 1,000 infested samples, and divided into 1,600 training, 200 validation, and 200 test samples. The dataset used in this study is publicly available on Kaggle (Senevirathna et al. 2026). Fourier-derived spectrograms trained a CNN for infestation classification and probability estimation. A weighted severity model combined CNN probability, mean acoustic amplitude, and nearby infested plants within 5 m, with geospatial mapping used to visualize infestation distribution. Findings and Values: Field trials in a ULWT-affected tea plantation in Pundaluoya demonstrated feasibility under realistic environmental noise. On the held-out test set, the CNN achieved 81.5% accuracy, 80.6% precision, 83.0% recall, 81.8% F1-score, and 0.819 ROC-AUC. Beyond binary infestation detection, the framework introduced quantitative severity assessment using infestation probability, acoustic amplitude, and nearby infested plants. The resulting severity and geospatial outputs can support plantation managers in identifying high-risk areas, prioritizing field inspections, and implementing more timely and targeted control measures. Comments: 12 pages, 7 figures. Accepted for publication in The Journal of Agricultural Sciences, Sri Lanka Subjects: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Sound (cs.SD) Cite as: arXiv:2608.27480 [cs.AI] (or arXiv:2608.27480v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.27480 Focus to learn more arXiv-issued DOI via DataCite
[AI-108] Class-Based Heuristic Selection for Solving the Flying Block Puzzle
链接: https://arxiv.org/abs/2608.27476
作者: Sanyar Ahmadi,Pedram Asadzadeh,Amanj Khorramian
类目: Artificial Intelligence (cs.AI)
备注: 27 pages, 15 figures, 2 tables
Abstract:Heuristic search underlies planning in autonomous systems ranging from warehouse logistics to robotic navigation, yet generic heuristics fail to exploit the structural constraints that govern constrained spatial domains, causing search performance to degrade catastrophically on harder instances. We study this problem through the two-column Flying Block Puzzle, a rigorously NP-complete spatial planning microworld whose bottleneck geometry mirrors clearance-to-size constraints encountered in multi-agent path finding, autonomous vehicle navigation, and block relocation systems. We introduce the Class-Based Heuristic A* (CBHA*) algorithm, which integrates a General Move Constraint to capture minimum displacement costs when vacant units are scarce, a formal kinematic taxonomy partitioning the state space into seven mutually exclusive classes with provably admissible heuristics based on vacancy ratio and goal-piece geometry, and a class-conditional tie-breaking mechanism that dynamically switches between depth-priority and vertical-distance ordering to overcome f-value plateaus. Over 146 benchmark instances, CBHA* achieves a 93.4% success rate against 64% for Depth-Prioritized A*, 39% for Standard A*, and 17% for BFS, while reducing node expansions by 87.98% relative to Standard A* and sustaining an average effective branching factor of approximately 3, demonstrating that class-triggered adaptive heuristics constitute a principled mechanism for efficient spatial planning that generalizes structurally to physical constraint systems.
[AI-109] Hypothesize Evaluate Refine: A Scientific Agent for PDE Discovery with Unknown Spatial Coefficient Fields
链接: https://arxiv.org/abs/2608.27475
作者: YuJie Huang,WenWu He,ZhuoEr Lin,Congcong Liu,Dong Liang,Zhuo-Xu Cui
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 33 pages, 4 figures, including appendices
Abstract:Discovering PDEs in heterogeneous media requires jointly identifying the governing operator and the unknown spatial fields that parameterize it. These tasks are coupled: changing field placement changes the differential law, while a sufficiently flexible field can conceal structural error on a single trajectory. We present Hypothesize, Evaluate, Refine for PDE Discovery (HER-PDE), a scientific-agent framework that discovers compositional PDE structure together with nonparametric, time-invariant coefficient fields. The Agent analyzes two noisy trajectories generated by different excitations, proposes complete expression-tree hypotheses, and combines creative structural exploration with local candidate refinement. Its Hypothesis Evaluation Interface (HEI) estimates only the fields explicitly declared in each hypothesis, never adds missing terms, and scores structures by bidirectional cross-excitation transfer. The selected law is subsequently audited on a sealed temporal interval. Across five controlled two-dimensional systems observed with 5 percent relative Gaussian state noise, the Agent recovers the generating operator in all five cases, including equivalent signed-field and product-rule parameterizations. Across nine unknown coefficient fields, the recovered fields attain a median Pearson correlation of approximately 0.85 and a median relative L2 error of approximately 0.28. These results show that agent-guided hypothesis refinement can recover heterogeneous governing laws without prescribing a parametric form for their spatial coefficients.
[AI-110] LLM -Augmented Causal Discovery: Probabilistic Fusion of Edge Existence and Orientation
链接: https://arxiv.org/abs/2608.27472
作者: Neville K. Kitson,Anthony Constantinou
类目: Artificial Intelligence (cs.AI)
备注: Accepted as a poster at Probabilistic Graphical Models (PGM) 2026 Conference
Abstract:Bayesian network structure learning (BNSL) from observational data struggles with orientation identifiability, while large language models (LLMs) offer broad but often unreliable causal knowledge. We propose combining these complementary sources through a novel representation, termed Probabilistic Dependency Graphs (PDGs). In a PDG, each edge is associated with a distribution over directed, undirected, and absent states, enabling fusion via weighted averaging. We evaluate this approach on 26 benchmark networks, combining ensembles of three BNSL algorithms (FGES, Tabu, PC) with three LLMs (Gemini, Claude, GPT) across multiple prompts and random seeds. A simple 50/50 fusion improves F1 over the better of either source alone in 22 of 26 networks, with a statistically significant mean improvement of 0.056 (p0.001) . Analysis reveals that the two sources play complementary roles: BNSL contributes a high-recall edge skeleton (80% vs 60% for LLM), while LLM contributes accurate edge orientation (96% vs 77% for BNSL). Our results show that representing both sources as probabilistic uncertainty over edge existence and orientation is a practical and effective way to improve causal graph accuracy.
[AI-111] Rating the Raters: Rasch Measurement Theory for LLM Evaluation
链接: https://arxiv.org/abs/2608.27463
作者: Pratik S. Sachdeva,Nathan Boudol
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:LLMs now sit on every side of evaluation: as examinees scored on benchmarks, judges of other models’ outputs, and raters of human-generated content. Each paradigm can be viewed as a measurement problem, where a latent property of an object is probed with items from an instrument (e.g., benchmark) by raters. Standard evaluation practices often neglect the contributions of each core component to the end result, limiting our understanding of what is being measured. Rasch measurement theory (RMT) is well-suited to this kind of problem. RMT decomposes ordinal ratings into separable facets on a common scale. It further provides a battery of diagnostics that can identify miscalibrated measurements and rater biases. We present a case study of RMT applied to the LLM-as-rater paradigm using the Measuring Hate Speech corpus, whose construct was itself built under RMT. We fit a series of many-facet Rasch models to annotations from nine LLMs spanning families and capability levels. Our analyses show that LLMs systematically differ from human raters in severity, item-level calibration, question-order robustness, target-identity sensitivity, and rating scale use, which all would be obscured by standard evaluation practice. Overall, we argue that RMT belongs in the toolkit for evaluating LLM-as-examinee, -judge, and -rater paradigms.
[AI-112] me Capsule of Testable Human Knowledge: 41 Years of Jeopardy! in a Single Free Local Model
链接: https://arxiv.org/abs/2608.27459
作者: David Noever,Forrest McKee
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:In 2011, IBM’s Watson was something like a sealed capsule of its era’s queryable knowledge. Its DeepQA system defeated the strongest human Jeopardy! champions, but the knowledge that let it do so lived in a curated billion-document corpus running on a cluster of POWER7 servers, frozen at build time and impossible to move or copy. We show that the same kind of artifact, a snapshot of what a culture can answer, is now portable and essentially free. We evaluate a single 9 GB open-weight model (Qwen2.5-14B, 4-bit) against the complete open Jeopardy! clue dataset, 529,939 clues across all 41 broadcast seasons from 1984 to 2025. To our knowledge this is the first time a model has been run over the full corpus. The 41 years mark only how long the questions were collected. What they test is far older and broader: the accumulated body of human general knowledge a culture considers worth knowing, from ancient history and dead languages to science, literature, and geography, with a verified answer for every item. The model answers 67.0% of all clues under a strict forced-response protocol with exact and fuzzy matching, and exceeds 85% on factoid categories. We treat training-data exposure as something both systems share rather than a flaw unique to language models. Watson’s case is in fact the more extreme one. Its corpus was assembled to contain Jeopardy answers and it was tuned on past clues, and it could not answer anything outside that curated distribution. The decisive test is whether a model can answer clues that did not exist when it was built. On clues aired after its training cutoff, the local model holds 65% and Claude Opus 4.8 holds 95%, while Watson by construction scores zero. The capability survives the move from a server room to a file you could seal in a time capsule, and unlike Watson it is not frozen to its own moment.
[AI-113] Learning a Size-Weight Frontier for Synthetic-Augmented Inference
链接: https://arxiv.org/abs/2608.28576
作者: Chengpiao Huang,Kaizheng Wang
类目: Methodology (stat.ME); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Machine Learning (stat.ML)
备注: 19 pages, 5 figures
Abstract:Synthetic data can improve statistical inference when real data are scarce, but naively treating synthetic samples as real data can introduce bias and lead to unreliable inference. We develop a general framework for synthetic-augmented inference across a population of related tasks. It characterizes synthetic augmentation by the number of synthetic observations and their weight. Central to our framework is a size-weight frontier that specifies, for each weight, the largest synthetic sample size for which all smaller sizes attain the target task-marginal coverage. We estimate this frontier from historical tasks, and establish a finite-sample coverage guarantee simultaneously for all size-weight configurations on or below the estimated frontier. In experiments using large language model responses to augment opinion survey data, our procedure achieves target coverage and substantially narrows confidence intervals.
[AI-114] Real-time virtual circuits for plasma shape control via neural network emulators: experimental demonstration on MAST Upgrade
链接: https://arxiv.org/abs/2608.28468
作者: Nicola C. Amorisco,Kamran Pentland,Adriano Agnello,George K. Holt,Alasdair Ross,Matthew J. Marshall,Edward Jones,Graham J. McArdle,Charles Vincent,Timothy Nunn,Martin Kochan,Pedro Cavestany,Aran Garrod,Stanislas Pamela,James Buchanan
类目: Plasma Physics (physics.plasm-ph); Artificial Intelligence (cs.AI)
备注: submitted
Abstract:Conventional plasma shape control in tokamaks relies on virtual circuits (VCs) that are computed offline from linearisations around a small, tailored number of reference equilibria, and deployed as expertly prepared schedules during the discharge. Here, we report on the first experimental deployment of real-time VCs. We replace pre-set look up tables with VCs updated in real time using surrogates of the plasma response. Both the existing control architecture and the interpretability of VC-based control are retained. Previous work showed that neural network emulators can produce accurate VCs, and validated their performance in closed-loop shape control simulations. Here, we report their first experimental validation on MAST Upgrade (MAST-U). Dedicated experiments spanning different scenarios, including prescribed shape perturbations, feedback-driven divertor-leg motion, and strongly evolving plasma configurations, show that real-time VCs can realise plasma shape control tasks within the MAST-U plasma control system. These results establish the experimental feasibility of real-time linearisations as a practical extension of conventional plasma shape control in tokamaks. The present implementation demonstrates a central step towards a simpler control workflow, in which manually constructed, phased VC schedules are replaced by VCs generated automatically online from a trained surrogate model, without scenario-specific retraining.
[AI-115] Conformal Risk-Averse Decision Making with Optimized Certainty Equivalent Risk Control
链接: https://arxiv.org/abs/2608.28179
作者: Amirmohammad Farzaneh,Osvaldo Simeone
类目: Machine Learning (stat.ML); Artificial Intelligence (cs.AI); Information Theory (cs.IT); Machine Learning (cs.LG)
备注:
Abstract:We study risk-averse decision making, in which an agent selects actions while being uncertain about the true system state. The risk is measured via optimized certainty equivalent (OCE) metrics, which generalize popular criteria such as mean-variance risk and conditional value-at-risk (CVaR). We characterize the optimal policy under known distributions, and show that it reduces to a prediction set-based solution for the CVaR. This provides an operational interpretation of conformal prediction-type prediction sets. For unknown distributions, we develop a data-driven calibration strategy, based on a synthetic model for the likelihood and held-out calibration data, yielding high-probability control of the OCE risk. The approach is evaluated on two wireless beamforming settings.
[AI-116] Efficient Auto-Interpretability of AI Models in Biology
链接: https://arxiv.org/abs/2608.27754
作者: Piotr Jedryszek,Oliver M. Crook
类目: Quantitative Methods (q-bio.QM); Artificial Intelligence (cs.AI)
备注:
Abstract:Sparse autoencoders (SAEs), and other interpretability methods could turn AI models in Biology and other fields into engines of scientific discovery by explaining the superhuman capabilities of those models. However, a latent is only useful if we know three things: whether it is coherent, whether it can be described, and whether that description has predictive power. These questions are routinely conflated. We assemble them into a single pipeline and report the practical innovations each stage required. First, cross-seed dictionary stability prioritises which latents are worth spending resources to investigate. Second, an intruder-detection task asks whether a latents activating examples share a recognizable pattern. Third, a separate pass proposes a candidate biological description which we convert into falsifiable predictions which can be tested in silico. Deployed on the Boltz-1 Pairformer trunk, stability prioritisation finds interpretable latents using about 4.4 times fewer latent evaluations each, and at 5.2 times lower measured cost, while recovering over half of them, and the external check shows the surfaced motifs are significantly enriched for their claimed annotations. The results also suggest a possible tension: the cross- seed stability might be selecting for some types of features, like structure-related ones, much more than others, such as function-related features.
机器学习
[LG-0] QGPINNs: A Physics-Informed Neural Network Framework for Nonlocal Differential Equations on Quantum Graphs
链接: https://arxiv.org/abs/2608.28589
作者: Vaibhav Mehandiratta,Saket Ramchandra
类目: Machine Learning (cs.LG); Numerical Analysis (math.NA)
*备注:
Abstract:We propose QGPINNs, a physics-informed neural network framework developed in PyTorch for the numerical solution of nonlocal differential equations on quantum graphs. The framework is designed as a general computational implementation in which the solution on each edge of the graph is approximated by a neural network, while a unified graph-based loss function enforces the governing equations together with initial, boundary, and vertex transmission conditions. In particular, the formulation incorporates standard continuity and Kirchhoff-Neumann vertex conditions and Dirichlet boundary conditions into the learning process to couple the local edge-wise neural approximations into a global solution on the graph. The framework is developed for two representative classes of nonlinear models: multi-order fractional elliptic problems and time-fractional evolution equations on quantum graphs. To improve accuracy and training stability, QGPINNs integrates several graph-adapted learning strategies, including soft and hard constraint enforcement, dynamic loss balancing, Fourier feature embeddings, and a learnable singularity-capturing feature for weakly singular solutions arising in the considered problems. The framework also extends naturally to inverse problems, including the identification of the orders of fractional operators and physical parameters from noisy observational data. We validate the accuracy, computational efficiency, and physical consistency of the proposed framework through numerical experiments on benchmark graph structures and real-world networks, including the IEEE 14-bus system and an open-channel agricultural drainage network.
[LG-1] On two proofs of d2 mixing of weighted Dikin walks
链接: https://arxiv.org/abs/2608.28566
作者: Yuansi Chen,Yunbum Kook
类目: Data Structures and Algorithms (cs.DS); Machine Learning (cs.LG); Optimization and Control (math.OC); Probability (math.PR); Computation (stat.CO)
*备注: 36 pages. AI disclosure included
Abstract:We study the mixing time of weighted Dikin walks for sampling from exponential distributions on polytopes and truncated positive-semidefinite (PSD) cones. Our first result gives a general total-variation mixing bound under strong self-concordance, \bar\nu -symmetry, and mixed-trace regularity on the local metric. The key idea is to control the Metropolis–Hastings acceptance probability on a high-probability region rather than at every point. Applying this framework to the Lee–Sidford, Lewis-weight, and John metrics yields an \widetilde O(d^2) mixing bound for sampling from polytopes, while applying it to a hybrid barrier yields an \widetilde O(d^4) mixing bound for sampling from truncated PSD cones. Our second result establishes stronger \chi^2 -divergence guarantees and pointwise acceptance control using a new fourth-order bootstrap condition. For a suitably scaled Lee–Sidford metric, this yields an \widetilde O(d^2) mixing bound in \chi^2 -divergence, improving on the previous \widetilde O(d^9/4) bound.
[LG-2] Advancing Interaction-Sensitive Feature Selection: Novel Relief-Based Algorithms Expanded Comparisons and Recommendations for Biomedical Data Mining
链接: https://arxiv.org/abs/2608.28552
作者: Kia Kazemi-Nia,Harsh Bandhey,Philip J. Freda,Ryan J. Urbanowicz
类目: Machine Learning (cs.LG)
*备注: 18 pages, 6 figures, 2 tables, submitted for journal review
Abstract:As a precursor to high-dimensional biomedical data modeling, reliable feature selection can reduce computational expense, improve modeling performance, and yield simpler, more interpretable models. However, most filter-based feature selection methods struggle to detect feature interactions, while wrapper or embedded feature selection methods are computationally expensive. Relief-based algorithms (RBAs) are filter methods that are sensitive to feature interactions while mitigating these other limitations. This study (1) refactors, optimizes, and expands the scikit-rebate Python package with existing and newly proposed RBA variants and (2) conducts rigorous RBA benchmark comparisons across diverse genomic simulations. We expand scikit-rebate to include SWRF*, mu-Relief, and 5 novel RBA variants implementing alternative strategies for neighbor selection and feature scoring. All RBAs were evaluated to compare predictive feature ranking and runtime across simulated genomic datasets varying in sample size, number of features, heritability, and underlying association type (e.g. main effects and interactions). All RBAs, except mu-Relief, were proficient in detecting 2-way interactions in noisy data. RBAs utilizing ‘far’ scoring were best at detecting 2-way interactions - with MultiSWRFDB* top-performing - but were far less sensitive to main effects. SWRF, MultiSWRF, MultiSURF, and MultiSWRFDB yielded top performance across main effect and 2-way interaction datasets with MultiSWRFDB performing best when also considering 3-way interactions. Refactoring of scikit-rebate resulted in 10 to 35-fold reductions in RBA runtimes. The newly introduced RBAs were among the strongest performing, and by robustly retaining both main effects and 2-way epistatic interactions, these algorithms preserve predictive signals for downstream modeling.
[LG-3] DARTS: Decoder-Aware Representation Tuning via Surgery for Model Merging EMNLP2026
链接: https://arxiv.org/abs/2608.28547
作者: Aaryan Ajay Sharma,Sai Nishanth Padala,Seganrasan Subramanian
类目: Machine Learning (cs.LG)
*备注: Accepted to EMNLP 2026 Main Conference
Abstract:Model merging combines multiple task-specific fine-tuned LLMs into a single multi-task model without additional training. However, merged models are known to suffer from representation bias: systematic drift between the merged model’s hidden states and those of each individual source model. Prior work (Yang et al., 2024a) study and mitigate this bias for encoder-based vision models using a lightweight correction module trained with L1 loss. However, such bias is not studied for decoder models due to their autoregressive nature. We analyze the problem of representation bias in decoder models, and show two challenges absent in encoders: (1) the causal attention mask causes bias to accumulate across token positions, requiring position-dependent correction; and (2) not all token positions are equally important, i.e., high-entropy (decision-critical) positions matter far more than low-entropy ones. To address these challenges, we propose Decoder-Aware Representation Tuning via Surgery (DARTS). DARTS employs a novel entropy-weighted L1 loss to upweight correction at high-entropy positions where errors most affect generation quality, and a per-position additive bias that captures position-dependent error without overparameterization. We perform extensive evaluation on three domains: code generation (HumanEval), mathematical reasoning (GSM8K), and instruction following (AlpacaEval) on Llama-2-7B models, and show DARTS achieves significant improvement over the standard surgery approach while adding negligible parameters ( 0.1% of total parameters).
[LG-4] REPLICANT: Learning Policies for Evading and Hardening Malware Detectors
链接: https://arxiv.org/abs/2608.28499
作者: Shae McFadden,Ilias Tsingenopoulos,Mario D’Onghia,Alexander Herzog,Myles Foley,Chris Hicks,Lorenzo Cavallaro,Fabio Pierazzi
类目: Machine Learning (cs.LG); Cryptography and Security (cs.CR)
*备注:
Abstract:To determine the real-world effectiveness of machine learning based malware detection, it is vital to evaluate its robustness against highly capable adversaries. However, state-of-the-art attacks do not effectively model realistic adversaries, as they often assume access to privileged information such as the training data, feature space, or confidence scores of the target. In this work, we present Replicant, a deep reinforcement learning framework that learns the realistic task of evasion under a strict label-only black-box threat model. Replicant learns a reusable policy on how to modify a malware sample and when to query the target, which transfers across samples, detectors, and feature spaces. Across seven Android malware detectors and three feature spaces, Replicant is the strongest and most query-efficient approach achieving a mean attack success rate of 78.8%, a relative improvement of 20.9%-39.2% over the state-of-the-art. Furthermore, when used for adversarial training, Replicant also outperforms the state-of-the art by producing detectors with more generalizable robustness. With Replicant we demonstrate that learning the task of evasion not only results in stronger attack performance but, crucially, provides a better signal for hardening malware detectors.
[LG-5] Curvature-Conditioned Multiscale Momentum with Sphere Constraints for LLM Pretraining
链接: https://arxiv.org/abs/2608.28442
作者: Shuchen Zhu,Yuxin Fang,Mingze Wang,Kun Yuan
类目: Machine Learning (cs.LG)
*备注: 50 pages
Abstract:Pretraining accounts for a large fraction of the total computational cost in LLM training. However, noise-dominant gradients and the highly ill-conditioned loss landscape bring severe challenges. Although modern adaptive optimizers such as AdamW and Muon have achieved great success in large-scale pretraining, their reliance on gradient normalization offers limited mitigation of the ill-conditioned curvature. The progress along flat directions (eigen-directions of small eigenvalues), which dominates the final loss reduction, remains relatively slow. To enhance training dynamics along flat directions, we propose a curvature-conditioned multiscale momentum method with sphere constraints, delivering steady acceleration in LLM pretraining. This multiscale momentum, applied only along flat directions, pairs a slow-decay component for noise reduction with a fast-decay component for rapid curvature adaptation, harnessing their complementary strengths. Crucially, we employ a sphere constraint technique to prevent parameter inflation and excessively rapid effective learning rate decay that would otherwise arise from a naive combination. Extensive experiments show that the proposed method significantly accelerates Muon across diverse architectures (dense, MoE) and model sizes (0.12B–2.3B parameters). Theoretically, we verify the acceleration effect and provide insight into the design principles underlying the flat-direction multiscale momentum.
[LG-6] Euclidean Fourier Neural Operators
链接: https://arxiv.org/abs/2608.28425
作者: Nathanael Bosch,Niklas Frederik Schmitz,Michael F. Herbst
类目: Machine Learning (cs.LG); Materials Science (cond-mat.mtrl-sci); Computational Physics (physics.comp-ph)
*备注:
Abstract:Fourier neural operators (FNOs) provide an efficient framework for learning mappings between function spaces as they are, by construction, independent of the grid resolution at which they are trained and evaluated. However, FNOs are not independent of the periodic domain they are applied to: their discrete spectral weights are indexed by integer Fourier mode numbers, which correspond to physical wavevectors. When applied to a different domain, the same trained weights act at different wavevectors, and the FNO silently represents a different operator. This makes FNOs unsuitable for tasks where transfer across domains is crucial. We propose Euclidean Fourier neural operators~(EFNOs) as a domain-independent alternative to FNOs. By parameterizing the spectral kernel as a continuous function of the physical wavevector, the EFNO can learn operators that act consistently across periodic domains of varying shape and size. We evaluate the EFNO on a simple heat equation and on a practically relevant materials science task of learning exchange-correlation potentials across different crystal structures, and demonstrate that the EFNO is able to generalize to unseen grid sizes and domains.
[LG-7] SymboLLM -FE: LLM -Accelerated Symbolic Regression for Automated Feature Engineering on Tabular Data
链接: https://arxiv.org/abs/2608.28408
作者: Zi-Jian Cheng,Zi-Yi Jia,Zhi Zhou,Yu-Feng Li,Lan-Zhe Guo
类目: Machine Learning (cs.LG)
*备注:
Abstract:Tabular data, as a core data format in machine learning, often lacks the discriminative power needed for high-performance modeling due to insufficient feature informativeness. Automated Feature Engineering (AutoFE) overcomes this by automating feature generation and selection, ensuring both model performance and operational efficiency. However, traditional AutoFE often yield features with poor interpretability because they rely on blind mathematical transformations, while large language models (LLM)-based AutoFE faces challenges in requiring costly multi-round iterations to generate high-utility features to effectively enhance model performance, compounded by inherent risks of bias and hallucination. In this paper, we combine symbolic regression with LLMs for feature engineering (SymboLLM-FE) to solve these challenges. We extract mathematically expressive formulas strongly correlated with the target via symbolic regression, which can enhance model performance, then refine them by LLMs with rich prior knowledge to ensure interpretability. Empirical results on six real-world datasets and four Kaggle competitions demonstrate that SymboLLM-FE outperforms existing AutoFE. SymboLLM-FE also addresses the dual challenges of poor interpretability and numerous iterations by employing a statistical prior-grounded LLM refinement mechanism and single-digit LLM calls.
[LG-8] Parser States Already Know: Structure-Conditioned KV Persistence for Structured Generation
链接: https://arxiv.org/abs/2608.28276
作者: Linze Wu,Xinrui Chen
类目: Machine Learning (cs.LG)
*备注: Work in progress
Abstract:Structured generation underpins large language model (LLM) agents that produce JSON, SQL, and function calls, where a single wrong field can cause the downstream action to fail. Constrained decoding already tracks parser transitions to enforce formal validity, and these transitions expose how generated tokens participate in schema-critical decisions such as required fields, arguments, and structural boundaries under the active grammar. Existing KV compression largely leaves this task-relevant structural signal unused. We introduce PASK (Parser-Aware Structural KV Persistence), which turns parser-derived structure into layer-group-specific KV persistence decisions. PASK addresses the mismatch between model-side KV sensitivity and task-level structured risk by using task-error sensitivity to set minimum protection floors and attention-output distortion to allocate residual KV capacity. An offline calibration stage compiles these signals into a persistence policy, leaving only lightweight structure-conditioned lookup online. At a targe total KV budget of 0.33, PASK outperforms the strongest compressed baseline by 17.39 percentage points on average across eight BFCL non-live and Live subcategories on Qwen3-4B. In end-to-end serving, PASK achieves up to 2.2x higher throughput and 3.3x lower TPOT, while using 0.53x the peak GPU memory of Full KV.
[LG-9] An algebraic proof of Colombos difference-power determinant conjecture
链接: https://arxiv.org/abs/2608.28274
作者: Kun Li,Li Tie,Peng Wang,Zihan Liu
类目: Machine Learning (cs.LG); Rings and Algebras (math.RA)
*备注:
Abstract:Let n\ge2 be even, let \lambda=(\lambda_1,\ldots,\lambda_n)\in\mathbbR^n have pairwise distinct coordinates, and define the difference-power matrix [ A_d(\lambda) := \bigl[(\lambda_r-\lambda_s)^d\bigr]_r,s=1^n, \qquad d\in\mathbbN. ] In 1928, Colombo proved that \det A_n-1(\lambda)\ne0 —and hence \det A_n-1(\lambda)0 —and that \operatornamerank A_d(\lambda)=d+1 for 0\le dn-1 . He conjectured that [ \det A_d(\lambda)\ne0 \qquad\textfor every d\ge n-1. ] For even d , the conjectured nonsingularity follows from previously published results on distance-power matrices. The remaining open cases were therefore the supercritical odd exponents d\ge n+1 . We prove nonsingularity for all these odd exponents, thereby completing Colombo’s conjecture. Consequently, [ \operatornamerank A_d(\lambda)=\min\n,d+1\ \qquad(d\in\mathbbN). ] Our proof converts a hypothetical kernel vector into a real binary form having more projective real linear factors, counted with multiplicity, than its real Waring length permits.
[LG-10] Learning to Transfer Across Modes: Towards Unified Urban Mobility Forecasting
链接: https://arxiv.org/abs/2608.28273
作者: Yixuan Zhao,Man Luo
类目: Machine Learning (cs.LG)
*备注:
Abstract:Urban transportation systems consist of multiple mobility modes that coexist within the same city and exhibit complex interdependencies, leading to correlated demand dynamics across modes. However, forecasting demand jointly across different modes remains challenging due to substantial heterogeneity in space and the limited availability of historical data for emerging modes. Existing forecasting methods are largely developed for individual mobility modes and implicitly assume compatible spatial structures between source and target systems, which severely restricts their applicability in multi-modal settings. To address these challenges, we propose \textbfTransMod, a unified framework for urban mobility demand forecasting that enables effective knowledge transfer across heterogeneous mobility modes. TransMod constructs a shared zone-level spatial representation that aligns mobility systems with different spatial granularities into a common space, thereby reducing structural mismatch and distributional shift. Built on this unified representation, TransMod further learns transferable spatio-temporal patterns from data-rich source modes and adapts them to data-scarce target modes, alleviating the dependence on extensive target-domain histories. Extensive experiments on real-world datasets demonstrate that TransMod consistently outperforms existing approaches and provides robust forecasting performance under limited target data.
[LG-11] Residual-Guided Randomized Neural Networks
链接: https://arxiv.org/abs/2608.28267
作者: Mushir Akhtar,M. Tanveer,Mohd. Arshad
类目: Machine Learning (cs.LG)
*备注: Accepted at WCCI 2026
Abstract:Randomized neural networks enable fast and analytically tractable training by fixing the input to hidden layer parameters at random and learning the output weights in closed form; however, their performance critically depends on a single uninformed draw of hidden units. This one shot and task uninformed feature construction often leads to redundant representations and suboptimal utilization of model capacity. To address this limitation, we propose a simple and broadly applicable residual guided procedure that greedily constructs the hidden layer using a closed form residual decrease criterion. At each stage, we (i) generate a pool of random candidate units, (ii) score each candidate by the exact reduction it induces in the ridge regularized objective, (iii) select the top k units, and (iv) refit the readout in closed form using the standard design with direct input links. This procedure yields a progressive training process with a guaranteed monotonic decrease of the training objective. The method is model agnostic: only the candidate generation is architecture specific, while the scoring selection refitting loop is shared across models. Extensive experiments on 71 benchmark datasets from the UCI repository, covering both binary and multiclass classification tasks, demonstrate that the proposed residual-guided models consistently outperform their baseline counterparts in terms of accuracy, stability, and overall ranking performance.
[LG-12] SinkSLOT: Sinkhorn via Sparse Lifted Optimal Transport
链接: https://arxiv.org/abs/2608.28262
作者: Ian Hsieh,Soumya Snigdha Kundu,Tom Vercauteren,Reuben Dorent
类目: Machine Learning (cs.LG)
*备注:
Abstract:Entropic optimal transport (EOT) has been shown to offer a computationally tractable approximation to exact optimal transport. However, the standard Sinkhorn-Knopp algorithm has two main limitations. First, given discrete measures with N points, each iteration requires O(N^2) operations, which restricts its use on large-scale datasets (e.g. N\geq10^4 ). Second, it uses the independent coupling as a reference measure for regularisation. This assigns mass to high-cost transport edges at moderate regularisation strengths. We propose SinkSLOT, which addresses both limitations by putting forth the expected sliced lifted transport plan as a natural way to sparsify the Gibbs kernel with a non-independent prior coupling. We prove that: 1) SinkSLOT converges; 2) with L slices, each resulting sparse Sinkhorn iteration costs O(LN) ; and 3) the resulting objective is a divergence requiring no debiasing. Experiments on synthetic benchmarks show that SinkSLOT delivers substantial speedups over state-of-the-art dense and sparse EOT methods. We also demonstrate the applicability of the proposed divergence in a gradient flow experiment. The code is publicly available at this https URL.
[LG-13] Spectral Features Dominate BCG Respiratory-Event Detection: A Large-Scale Patient-Independent Comparison of Feature Groups in Sleep Apnea Patients
链接: https://arxiv.org/abs/2608.28242
作者: Israel Campero Jurado,Zoe Bousraou,Lara Benning,Sara Padilla Neira,Alexander Breuss,Robert Riener,Esther Irene Schwarz,Elisabeth Wilhelm
类目: Machine Learning (cs.LG)
*备注: 35 pages, 7 figures
Abstract:Unobtrusive ballistocardiographic (BCG) sensing is a promising modality for long-term sleep-apnea monitoring, yet it remains unclear which signal features are most discriminative for respiratory-event detection. We present a literature-guided, patient-independent comparison of ten BCG feature groups using a 512-sensor capacitive pressure mat recorded simultaneously with respiratory polygraphy in 155 patients (52 female, 103 male) undergoing in-hospital evaluation for obstructive sleep apnea. Features were extracted from six spatially distinct signal channels, yielding a 191-dimensional feature vector spanning general statistical, time-domain, frequency-domain, wavelet, frame-energy, and nonlinear complexity descriptors. Under strict leave-one-patient-out cross-validation for binary classification of respiratory-event windows versus event-free reference windows, Random Forest and Histogram Gradient Boosting achieved AUC-ROC of 0.967 and 0.969 and AUC-PR of 0.977 and 0.979, respectively. Feature-importance analysis revealed that frequency-domain features dominate discrimination: breathing-band power in the 0.1-0.4 Hz range accounted for 30.3% of total discriminative information across all spatial channels, and Fast Fourier Transform spectral-shape descriptors of the adaptively preprocessed channel contributed a further 15.1%. AUC and curve-length features provided the main complementary time-domain evidence (21.5%), whereas wavelet-derived and nonlinear features contributed smaller secondary effects (10.4% combined across 59 features). Frequency-domain and time-domain features together accounted for 67% of total discriminative information, demonstrating that a compact, interpretable subset of the full feature library achieves clinically relevant performance under patient-independent validation and providing an empirical basis for feature selection in future BCG systems.
[LG-14] Efficient Online Continual Foundation Model Fine-Tuning for Predictive Process Monitoring
链接: https://arxiv.org/abs/2608.28237
作者: Sjoerd van Straten,Marwan Hassani
类目: Machine Learning (cs.LG)
*备注:
Abstract:Predictive Process Monitoring (PPM) models are increasingly deployed in dynamic environments where concept drift causes the underlying process distribution to shift over time. While recent work has moved toward online continual learning, existing methods train compact, task-specific networks entirely from scratch, leaving a persistent cold-start problem. Foundation Models (FMs) offer a compelling solution to this problem, but their continual fine-tuning in the process mining domain remains unexplored. We propose COMPASS (Continual Online foundation Model-based PPM with Adaptive SubSpaces), the first framework for online continual fine-tuning of FMs for PPM. COMPASS adapts loss-plateau drift detection to autonomously identify task boundaries in event streams and maintains a unified knowledge subspace including both pre-trained and task-specific directions. We evaluate our approach on nine event streams covering synthetic and real-world concept drift scenarios, across task-free and task-aware settings with multiple backbones and with consistent hyperparameter tuning across all methods. Our approach outperforms three SOTA non-FM competitors and two update strategy baselines, with particularly strong gains on streams exhibiting recurrent drift and complex, long-running cases, while incurring acceptable computational overhead compared to the non-FM competitors.
[LG-15] D-TAIA: Domain-Aware LLM Adaptation for Multi-Task Predictive Process Monitoring
链接: https://arxiv.org/abs/2608.28236
作者: Sjoerd van Straten,Christine Jacob,Marwan Hassani
类目: Machine Learning (cs.LG)
*备注:
Abstract:Predictive Process Monitoring (PPM) enables organizations to forecast future process behavior, such as the next activity and remaining time of ongoing cases. In practice, three conditions cause existing methods to degrade, namely data scarcity, high process entropy and distributional shift. While Foundation Models (FMs), especially Large Language Models (LLMs), offer a new paradigm through broad sequential reasoning, adapting them to multi-task PPM under these conditions remains an open challenge. Existing FM-based approaches either lack mechanisms for handling distributional shift or rely on direct regression heads that can be structurally misaligned with continuous time prediction tasks. This paper introduces D-TAIA (Domain-aware Training and Attention-based Inference Architecture), a framework for a joint next activity and remaining time prediction task via parameter-efficient fine-tuning of an FM backbone. Our approach combines domain-aware triplet loss (DATL) pre-training with FAISS-based nearest neighbor retrieval for remaining time prediction, and adopts the TAIA inference strategy to preserve pre-trained sequential reasoning during fine-tuning. Evaluated across four real-world event logs, D-TAIA consistently shows SOTA or competitive performance compared to a fine-tuned LLM and a recurrent neural network baseline. Ablation studies confirm that techniques from NLP and computer vision can be transferred effectively to PPM with only a 10M-parameter backbone, though component contributions vary by dataset entropy.
[LG-16] Generalized Context in Cross Attention for Transfer Learning of Disjoint Tabular Data
链接: https://arxiv.org/abs/2608.28209
作者: Kazi F. Akhter,Ibna Kowsar,Manar D. Samad
类目: Machine Learning (cs.LG)
*备注: Accepted for publication at Neural Networks (Elsevier)
Abstract:Unlike images and text, applying transfer learning to tabular data is challenging due to heterogeneity in feature types, structures, and semantics across disparate domains. Existing methods assume shared features across data tables to enable knowledge transfer between domains, which is unrealistic in practice. \mdsThis paper introduces generalized context learning to remove the requirement of shared features across domains. The generalized context captured by transformer projection weights for key , value , and query provides rule-based generalization rather than the domain-specific context conventionally learned from transformer activations. Projection weights for key from the source domain interact with the weight for query in the target domain to achieve Cross-domain Attention Transfer Learning (CATTLE) in a data-agnostic manner. Our experiments on ten pairs of disjoint source-target data sets show that CATTLE can learn generalized context from a single source data set and is rank-wise and statistically superior to nine state-of-the-art baselines, including machine learning, deep learning, and transfer learning methods using large-scale pre-trained models. CATTLE achieves the best average rank (2.9) and delivers a 3.7% average AUROC gain over the baseline methods. The CATTLE source code is available at this https URL.
[LG-17] Biologically Inspired Mechanisms for Facilitating Grokking in Multilayer Perceptrons
链接: https://arxiv.org/abs/2608.28184
作者: Florin Leon
类目: Machine Learning (cs.LG)
*备注: 51 pages, 9 figures
Abstract:Grokking is a delayed transition from memorization to generalization that is often accompanied by substantial reorganization of internal representations. This paper studies whether biologically inspired mechanisms, many of which are not commonly incorporated into artificial neural networks, can actively promote this transition by regulating hidden-layer computation at the levels of neuronal activity, response, and effective connectivity. We augment a multilayer perceptron with input gating, structural plasticity, gain modulation, threshold modulation, homeostasis, lateral inhibition, and activation decorrelation, and evaluate these mechanisms through systematic ablations on two established grokking benchmarks: sparse parity and noisy XOR classification. The results show that the mechanisms contribute unequally to generalization. Homeostasis provides the strongest and most consistent benefit, while structural sparsification emerges as the second major mechanism. The remaining biologically inspired mechanisms have smaller or less consistent effects in the present experiments. For both problems, the results support the common principle that explicit regulation of neuron utilization and effective connectivity can improve the emergence of generalizable internal computation. These findings motivate broader investigation of biologically inspired activity regulation and adaptive sparsification, including in large language models, where they may accelerate the development of generalizable representations and reduce the optimization time required for robust generalization.
[LG-18] HARTS: Efficient Agent ic Reinforcement Learning for Hybrid-Attention Models over Arbitrary Rollout Trees
链接: https://arxiv.org/abs/2608.28158
作者: Boyuan Meng(1),Peihua Bao(1),Hong Liu(1),Xiaowei Zhu(1),Chao Wang(1),Gen Li(1),Zhenxuan Pan(1) ((1) Ant Group, China)
类目: Machine Learning (cs.LG); Distributed, Parallel, and Cluster Computing (cs.DC)
*备注:
Abstract:Agentic reinforcement learning (RL) often produces irregular rollout trees with shared histories. Training root-to-leaf trajectories independently recomputes these shared prefixes. Existing systems primarily target full-attention models and lack dense, differentiable hybrid-attention execution compatible with activation recomputation. We present HARTS (Hybrid-Attention RL over Tree Structures). HARTS jointly plans microbatches, data-parallel (DP) replica assignments, and microbatch-slot schedules using non-replay compact-token work after prefix compression. For chunkwise linear attention, a linear-time algorithm coordinates chunk-boundary state recovery and replay and produces the minimum number of sequential linear-attention calls under our packed execution model. HARTS preserves the chunkwise state partitioning of trajectory-wise training: it does not repeat projections, MLP/MoE computation, or final outputs, and performs only bounded state replay for numerical alignment. Per round, HARTS batches all branches into one packed call, propagates gradients through differentiable state handoffs, supports activation recomputation, and restores per-token log-probabilities. For deterministic, no-token-drop top- k MoE routing, semantic multiplicities restore MoE-objective token weights and load statistics. Existing RL objectives retain their interface. To our knowledge, HARTS is the first system to demonstrate arbitrary-rollout-tree prefix-sharing speedups on a real hybrid-attention model. On an Agentic RL workload generated from SWE-bench tasks, HARTS achieves 4.81 – 4.87\times forward/backward/gradient speedup with activation recomputation across multiple parallel configurations. Its numerical differences are comparable to baseline self-rerun variation, and its reward trend is similar to the baseline over the first 120 steps of \tau^3 -Bench training.
[LG-19] Conditional Diffusion Models for Energy-Efficient Driving
链接: https://arxiv.org/abs/2608.28142
作者: Hemanth Neelgund Ramesh,André Snoeck,Chyi-Fu Hong,Shijing Sun
类目: Machine Learning (cs.LG)
*备注:
Abstract:Electrification of commercial delivery fleets is shifting fleet routing from distance- and time-based optimization toward energy-aware decision-making. Existing sequence models primarily provide deterministic point estimates or limited uncertainty summaries, which do not capture the range of plausible energy-consumption trajectories required for operational decision-making. In this work, we introduce a conditional diffusion framework that generates EV battery-current profiles conditioned on route features such as vehicle velocity and ambient temperature. The model combines a latent conditioning encoder with a temporal 1D U-Net denoising backbone that enables trip-related conditions to be mapped into a shared representation and guides the reverse diffusion process. We evaluate the framework on an open-access commercial EV telemetry dataset containing 12k trips from 9 vehicles. The proposed latent-conditioned diffusion model generates realistic cur- rent trajectories that capture both the dominant temporal envelope and sharp transient events. The model achieves a Wasserstein distance of 0.0029 between generated and measured current distributions below the real vs real reference distance of 0.0085 indicating that generated samples lie within the empirical variability of the test set. We further demonstrate that learned latent conditioning substantially improves performance over direct condition injection, reducing the Wasserstein distance by 89.1% and MAE by 52.8%. This work demonstrates a generative modeling framework for characterizing EV energy consumption under real-world operating conditions, providing an essential foundation for uncertainty-aware fleet planning in large-scale operational settings.
[LG-20] Learning to Difference: Adaptive Reversible Differencing (AdaRDiff) for Time Series Forecasting
链接: https://arxiv.org/abs/2608.28134
作者: Morad Laglil,Younes Hlal,Marouane El Hadari,Emilie Devijver,Eric Gaussier
类目: Machine Learning (cs.LG)
*备注:
Abstract:Reliable long-horizon time series forecasting is an important yet difficult problem. Trends and seasonality introduce complex temporal structure that challenges learning-based forecasting models. Differencing, which subtracts nearby past values to remove such structure, is the classical remedy, but its reliance on hand-picked orders and periods has kept it largely absent from recent deep architectures. We propose \textbf\underlineAdaptive \textbf\underlineReversible \textbf\underlineDifferencing \textbf(AdaRDiff), a generalized differencing approach that uses learnable weights to simplify the series through weighted differencing with previous time instants. This yields stabilized residuals on which forecasting is performed, after which the removed components are restored autoregressively to reconstruct the forecast, capturing trend and seasonality jointly through a single operator. This reconstruction admits a closed-form convolutional expression, which parallelizes on GPU and yields up to 33.7\times speedup over the naive recurrence. We furthermore rely on a two-phase training schedule that separates temporal structure discovery from reconstruction learning, as suggested by a theoretical analysis of the gradient when using a linear forecasting model. AdaRDiff attains state-of-the-art forecast accuracy across eight benchmarks spanning electricity, weather, traffic, and energy, at negligible parameter cost. Furthermore, it is designed as a plug-and-play module: integrating AdaRDiff improves eight diverse backbones, from linear models to Transformers, in the large majority of cases, by up to 25.9% with a linear backbone and 18.3% with iTransformer.
[LG-21] Generalized Gibbs Ensemble Weighting for Forecast Combination
链接: https://arxiv.org/abs/2608.28116
作者: Prasen R. Nuthanakaluva,Nava K. Gaddam
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: 15 pages, 6 tables, preprint
Abstract:Forecast combination is a reliable way to improve predictive performance when several forecasting models are available. Simple aggregation rules such as the mean, median, trimmed mean, inverse-loss weighting, and exponential weighting are often strong baselines, but their relative performance can vary across datasets, forecast horizons, deployment settings, and levels of disagreement among base forecasters. We develop Generalized Gibbs Ensemble Weighting (GGEW), a probabilistic framework that treats forecasting models as experts and assigns ensemble weights using a Gibbs-style exponential transformation of normalized predictive loss. The framework extends this basic weighting rule through numerical stabilization, diversity-aware score corrections, and online hyperparameter adaptation. GGEW produces a family of related methods, including Stable Gibbs weighting, Directional Gibbs-NCL, and Symmetric Gibbs-NCL. These variants share one core algorithm and differ only in the score used inside the exponential weighting rule. For sequential deployment, we adopt a UCB-style bandit mechanism, called online Local-UCB, to adapt the learning rate, diversity strength, and Gibbs variant without evaluating the full hyperparameter grid at every prediction step. We evaluate GGEW on official M4 competition forecast submissions and external rolling-origin deployment experiments using Monash Traffic Hourly, Electricity Hourly, and Solar Weekly datasets. Results suggest that Gibbs-style adaptive weighting is a useful and competitive tool across several benchmark settings, although its relative performance varies across datasets, forecast horizons, deployment protocols, and forecast disagreement groups. The contribution is not a universal dominance claim, but a framework and empirical study motivating further investigation of when adaptive Gibbs-style forecast combination is useful.
[LG-22] Comparing Classical and Quantum Machine Learning for Regression in High Energy Physics Collision Data
链接: https://arxiv.org/abs/2608.28084
作者: Tariq Mahmood,Zain ul Abidin,Itzel Luviano Soto,Alfredo Raya
类目: Machine Learning (cs.LG); High Energy Physics - Experiment (hep-ex); High Energy Physics - Phenomenology (hep-ph); High Energy Physics - Theory (hep-th); Quantum Physics (quant-ph)
*备注:
Abstract:The classification and regression of particle collision events constitute a persistent computational challenge in experimental high energy physics, where large volumes of simulated data must be processed with both speed and precision. This work carries out a systematic comparison of four classical machine learning architectures, support vector machines (SVM), artificial neural networks (ANN), convolutional neural networks (CNN), and long short-term memory (LSTM) networks against their quantum counterparts: quantum SVM (QSVM), quantum neural networks (QNN), quantum CNN (QCNN), and quantum LSTM (QLSTM). All models are trained on simulated proton-proton collision events with electron-positron and muon-antimuon final states from the CERN Open Data portal, using transverse-momentum components as input features and transverse-momentum magnitude as the regression target. Classical architectures, and in particular the CNN and LSTM, achieve marginally better quantitative performance under current hardware and dataset constraints. Quantum models, however, reach competitive accuracy with substantially fewer trainable parameters: the QCNN reproduces the performance of the deep classical CNN using only four qubits and a circuit of depth three, pointing to a genuine parameter-efficiency advantage on near-term quantum devices. A baseline analysis confirms that the regression problem is non-trivial for shallow polynomial fits, supporting the relevance of the architectural comparison. These results characterize the trade-offs between classical and quantum approaches under realistic, resource-constrained conditions and provide a benchmark for future studies on actual quantum hardware.
[LG-23] Characterization of Request and Token Energy Costs for LLM Inference Workloads on GPU Platforms ISWC2026
链接: https://arxiv.org/abs/2608.28044
作者: Prabhu Vellaisamy,Vanessa Lam,Shawn Blanton,John Paul Shen
类目: Performance (cs.PF); Distributed, Parallel, and Cluster Computing (cs.DC); Machine Learning (cs.LG)
*备注: Accepted at the 2026 IEEE International Symposium on Workload Characterization (IISWC 2026). 13 pages, 6 figures, 9 tables
Abstract:Large language model (LLM) inference serving is priced by tokens, but GPU energy is consumed over inference windows. This accounting mismatch makes token-normalized metrics incomplete, since average output-token energy can decrease even when total request energy increases. We characterize this behavior with a decomposed energy model: a fixed one-time prefill with a fixed generation setup cost, while each output-token generation step adds marginal step energy. We evaluate this LLM inference energy model on NVIDIA H100 and H200 GPUs across dense and mixture-of-experts (MoE) models, reporting both request energy and token energy as functions of model type (M), phase §, batch size (B), context length ©, and output length (N). For Llama-3.2-1B on H200 at batch-16 and context-4K, increasing output length from 10 to 512 tokens reduces token energy from 7.46 to 0.72 J/token while total batched inference-window energy increases from 1.19 to 5.93 kJ. Batching also reduces token energy, but the gain is context-bounded: at 10 output tokens, the batch-16 to batch-1 gain falls from 6.31x at context-512 to 1.17x at context-4K. MoE models amplify this effect: sparse routing and fragmented expert execution increase fixed energy at low concurrency, while batching spreads that energy across more generated tokens and substantially narrows the dense-vs.-MoE token-energy gap. These results show that energy-aware serving should jointly optimize both request energy and token energy, rather than only reducing per-token energy cost.
[LG-24] Exact Risk Ratios for Weighted Data Selection in Linear Regression
链接: https://arxiv.org/abs/2608.28007
作者: Guangjian Zhang
类目: Machine Learning (cs.LG); Statistics Theory (math.ST)
*备注: 33 pages
Abstract:Hanneke, Moran, Shlimovich and Yehudayoff (COLT 2025) posed the following open problem. A selector sees a finite dataset D \subseteq \mathbbR^d \times \mathbbR , picks at most n examples together with nonnegative weights, and hands the weighted least squares objective to the minimum-norm ERM. Writing F_w(d,n) for the worst-case ratio between the loss of the returned predictor on all of D and the optimal loss, they proved F_w(d,n)=\infty for nd , F_w(d,d)=d+1 and F_w(d,n)=1 for n \ge 2d , and asked for the value in the open regime dn2d . We determine this value in several cases. For every d we prove F_w(d,2d-1)=1+1/d , which confirms a claim stated without proof in the original note. We further prove F_w(3,4)=5/3 and F_w(4,5)=2 , the two smallest cells not covered by the endpoint formula. For every intermediate budget n=d+k we prove the lower bound F_w(d,d+k) \ge 1+\Gamma_d,k , where \Gamma_d,k is an explicit harmonic quantity over balanced partitions, and we show that this bound is the exact minimax value over the class of datasets whose whitened gradient systems carry an orthogonal circuit-block structure. All three exact values match 1+\Gamma_d,k , and we conjecture that equality holds throughout the open regime. The upper bound proofs run on a common geometric spine: a rigidity theorem for positive spanning configurations of loss gradients, classifications and structural reductions of small positive bases in \mathbbR^3 and \mathbbR^4 , and a dimension-free extremal-basis argument that converts sign-cone geometry into five-point selections. We also give explicit counterexamples showing that several shorter routes fail, and constructive polynomial-time selection algorithms for all proved cases.
[LG-25] Is Monte Carlo Tree Search Just Every-Visit Monte Carlo Control?
链接: https://arxiv.org/abs/2608.27985
作者: Xianyi Wu
类目: Machine Learning (cs.LG)
*备注: Comments and discussions are welcome
Abstract:Monte Carlo Tree Search (MCTS) and every-visit Monte Carlo (MC) control are usually presented as different methods. MCTS is described in the language of search (selection, expansion, simulation, and backup), whereas MC control is described in the language of reinforcement learning (trajectory sampling, return estimation, action-value updating, and policy improvement). This note argues that, at the level of trajectory generation and action-value updating, the distinction is largely terminological. The tree policy and rollout policy can be viewed as the learned and not-yet-learned parts of a single evolving policy; expansion corresponds to first visit and initialization; and backup is the ordinary every-visit Monte Carlo update. Under this interpretation, the four stages of MCTS reduce to two basic operations: trajectory sampling under the current policy and every-visit Monte Carlo updating. In this sense, MCTS is simply every-visit Monte Carlo control expressed in the language and data structure of search. The purpose of this note is expository: to make this equivalence explicit and easier to recognize.
[LG-26] PhyMamba: Physics-Modulated Mamba for Robust Battery Health Prognostics
链接: https://arxiv.org/abs/2608.27978
作者: Sara Sameer,Yunyi Zhao,Wei Zhang,Minggang Zeng,Wenqing Li,Man-Fai Ng,Yonggang Wen
类目: Machine Learning (cs.LG); Systems and Control (eess.SY)
*备注:
Abstract:Battery health prognostics is a core function in battery management systems (BMSs), yet long-horizon health forecasting from BMS signals remains challenging due to operating-condition dependency and sensor noise. In this paper, we propose PhyMamba, a two-stage physics-modulated Mamba framework that integrates electrochemical aging into sequence modelling. PhyMamba does not require explicit identification of internal aging parameters, which often relies on intrusive measurements. In stage-1, a lightweight Mamba encoder first processes BMS signals and produces a latent representation that is transformed via an aging parameterization module, into physics-informed aging features. In stage-2, a customized Mamba forecasting backbone performs multi-cycle prediction, where physics is tightly integrated to regulate the model’s internal temporal updates toward degradation-consistent evolution. Experiments on three public datasets under multiple forecast horizons show that PhyMamba achieves the best aggregated performance, with an overall mean error reduction of 31.8% compared with a diverse range of baselines. PhyMamba also offers an optimized accuracy-efficiency trade-off, which supports practical deployment for robust battery health prognostics.
[LG-27] mporal Memory-Aware Online Test-Time Adaptation on Dynamic Graphs ICDM2026
链接: https://arxiv.org/abs/2608.27948
作者: Bo Li,Xin Zheng,Ming Jin,Can Wang,Shirui Pan
类目: Machine Learning (cs.LG)
*备注: Accepted By ICDM 2026
Abstract:Test-time adaptation (TTA) on graphs aims to adapt a graph neural network (GNN) that is well-trained on the training graph to the test graph, which involves potential distribution shifts that may harm model generalization and test-time inference. While recent efforts have investigated TTA on static graphs, there is still a research gap on dynamic graphs learned with dynamic GNN (DGNN) models, where both structural connectivity and node semantics evolve continuously over time. This makes adapting a DGNN model for reliable test-time performance substantially challenging. To fill this gap, in this work, we propose a novel framework of temporal memory-aware Online Test-Time Adaptation on Dynamic Graphs, named DGOTTA, to effectively adapt well-trained DGNNs during test time. Specifically, the proposed DGOTTA contains three modules: (1) temporal-aware augmentation, to extend the diversity of test dynamic graphs for addressing complex temporal and spatial shifts; (2) memory-aware model prediction, to alleviate catastrophic forgetting; (3) consistency-guided online adaptation, to enforce temporal alignment and memory smoothness. Extensive experiments on three real-world datasets and four DGNN backbones demonstrate that DGOTTA significantly improves generalization under diverse distribution shifts and multiple model architectures.
[LG-28] I2PS: A Topology-Informed Inverse Design Framework for Stochastic Multicellular Pattern Formation
链接: https://arxiv.org/abs/2608.27931
作者: Kenji Komiya,Andrew Kailiang Jin,Ryo Nishikimi,Kunio Kashino
类目: Machine Learning (cs.LG)
*备注: 13 pages, 5 figures
Abstract:This study proposes a novel framework to estimate parameters for reproducing target multicellular patterns using an agent-based model (ABM). Two major challenges in multicellular ABMs are estimating cell-level parameters (agent-specific variables) and quantitatively evaluating the topological characteristics of multicellular arrangements under stochastic cell proliferation and death. To address these challenges, we integrate two approaches: Betti vectors and inverse surrogate modeling. The Betti vectors obtained through topological data analysis can consistently represent features of a wide range of multicellular spatial configurations. The inverse surrogate modeling enables direct inference of the corresponding ABM parameters from the target patterns. We validated the proposed framework using zebrafish pigment pattern formation, a representative model of pattern formation driven by multicellular interactions. The results demonstrate that our framework successfully estimates ABM parameters and outperforms conventional methods such as PointNet++. Notably, the proposed method, which used only 10% of the training data, outperformed PointNet++, which used 100% of the data, across all evaluation metrics.
[LG-29] ACIT-Switch: Cost-Aware Model Escalation for LLM Agents from Censored Supervision
链接: https://arxiv.org/abs/2608.27911
作者: Ji’an Lei,Jian Huang
类目: Machine Learning (cs.LG)
*备注: 17 pages, 6 figures, 3 tables, 1 algorithm
Abstract:Agents with smaller language-model backbones are less expensive but can drift into persistent failure modes, whereas those with larger backbones are generally more reliable but more costly. This reliability-cost trade-off motivates routing methods that decide when to invoke an agent with a larger backbone: before execution, after a fixed trajectory prefix, or locally at individual steps. Our method, TACIT-SWITCH, learns permanent handoff policies from accumulated trajectory evidence and Teacher-Annotated Censored Intervention Times (TACIT). It represents each annotation as an interval-censored observation on a cumulative-risk scale. The resulting mixture-cure threshold model estimates the probability that the paired Strong rollout succeeds and, conditional on success, the handoff threshold; no teacher is required at deployment. In a mechanism-based multi-step simulation, TACIT-SWITCH improves success by 7.4-11.1 percentage points over task-level, step-level, and fixed-prefix routing baselines at comparable cost. Within that controlled simulation, ablations show that task features and cumulative trajectory risk provide complementary information. With operating points selected on development data, TACIT-SWITCH achieves the highest held-out success among learned policies on both ALFWorld (48.5% with 4B Cheap; 45.5% with 9B Cheap) and DABench (73.1%).
[LG-30] here and Back Again: Bidirectional Diffusion Bridges for Multimodality Translation
链接: https://arxiv.org/abs/2608.27885
作者: Gabe Guo,Elon Litman,Thanawat Sornwanee,Jose Blanchet,Stefano Ermon
类目: Machine Learning (cs.LG)
*备注:
Abstract:Multimodality translation (e.g., text-to-image) is a core generative AI task. However, existing approaches (1) follow generative paths that do not directly represent the source modality, limiting the flexibility of some sampling algorithms; and (2) are unidirectional, preventing inversion (e.g., image-to-text). We propose BIT: Bidirectional Image-Text Diffusion Bridges. In contrast to previous approaches, BIT starts directly from text and interpolates into images, providing (1) a source-aware generative path that enables diverse and flexible sampling algorithms; and (2) an endpoint-conditioned process that can be traversed from image to text, providing a unified, bidirectional generative framework. BIT is derived through stochastic calculus, yielding SDE forms amenable to simulation and tractable loss functions that scale to high dimensions. Our experiments show that BIT is competitive with denoising-diffusion and deterministic-flow baselines, and outperforms them on several vision–language and natural-science evaluations.
[LG-31] Beyond Pairwise Graphs in Science: Hypergraph Adaptive Wavelet Operators for Parametric PDEs
链接: https://arxiv.org/abs/2608.27883
作者: Rajat Sarkar,Venkataramana Runkana,Souvik Chakraborty
类目: Machine Learning (cs.LG); Computational Physics (physics.comp-ph)
*备注:
Abstract:Physical systems are often modeled by solution operators that map input fields, parameters, geometries, or past states to steady or future physical states. Learning these maps is difficult, especially for time-dependent systems that must assimilate history and remain stable under autoregressive rollout. Many neural operators work best on regular, structured grids, while realistic simulations often require unstructured meshes or point clouds to resolve complex geometries; in such settings, grid-centric representations can lose accuracy. Graph neural operators handle these domains through message passing or spectral graph filtering, but pairwise edges do not directly capture group-wise couplings among mesh cells, local neighborhoods, or conservation volumes. We introduce the Hypergraph Adaptive waveLet Operator (HALO), which lifts the domain to a hypergraph and learns in its spectral wavelet domain. HALO avoids explicit hypergraph-Laplacian eigendecomposition through Chebyshev polynomial wavelet filters, giving localized spectral kernels at linear sparse-matrix cost. Its trainable dyadic wavelet scales are regularized toward tight-frame coverage, allowing the frequency response to adapt to each PDE while encouraging stable multi-scale spectral coverage. Across 2D and 3D benchmarks on structured and unstructured discretizations, HALO achieves best or near-best accuracy among frequency-, transformer-, DeepONet-, state-space-, and graph-based baselines and sustains stable multi-step rollouts. The same model scales to industrial aerodynamic geometries: on meshes of a few hundred thousand points it is on par with, or better than, the strongest fixed-discretization transformers, while remaining resolution-equivariant.
[LG-32] Node-wise Feature Encoding for Neural Performance Prediction
链接: https://arxiv.org/abs/2608.27794
作者: Matthew Grenier,William Hammer,Andrew Heuer,Nikhil Krishna,Yi Wang,Ramtin Zand
类目: Machine Learning (cs.LG)
*备注: 22 pages, 7 figures
Abstract:As neural networks are increasingly deployed on resource constrained edge devices, accurate prediction of latency and energy is critical for efficient neural architecture search. Existing GNN and transformer based predictors achieve strong results but largely ignore node-level computational cost, limiting their ability to model performance critical operations. To address this, we introduce FeatureFormer, a neural performance predictor that incorporates explicit node-wise encodings of FLOPs, parameter counts, and memory proxies within a gated graph attention architecture. We also present NNEQ, a new large-scale energy consumption dataset that enables unified evaluation of latency and energy prediction. Extensive experiments demonstrate that FeatureFormer achieves state-of-the-art performance across both metrics, including challenging out-of-domain settings. Finally, we show that the proposed encoding is broadly applicable and consistently improves existing predictors with negligible overhead.
[LG-33] Initialization Is Critical: Advancing Federated Short-Term Load Forecasting under Load Heterogeneity via Model Initialization
链接: https://arxiv.org/abs/2608.27791
作者: Jianing Chen,Vajiheh Farhadi,Yan Li,Thomas La Porta
类目: Machine Learning (cs.LG); Systems and Control (eess.SY)
*备注:
Abstract:Short-term load forecasting (STLF) provides essential information for numerous applications in modern power systems. However, accurate STLF often relies on fine-grained smart-meter data from distributed users, raising increasing concerns about data privacy. Federated learning (FL) has therefore emerged as a promising privacy-preserving paradigm for STLF. Nevertheless, this paper reveals structured heterogeneity in clients’ load data. Specifically, clients exhibit different responses to exogenous factors and distinct temporal load profiles, which can degrade forecasting performance in FL. To mitigate these issues, this paper studies the role of model initialization in federated STLF, and proposes two initialization strategies from global and local perspectives. For global model initialization, when auxiliary public load data are available, a pretrained initialization strategy is developed to initialize the global model before federated training, thereby reducing client drift during the training process. For local model initialization, we propose SLIAvg, a sequential local initialization strategy that promotes a more consistent training process by allowing participating clients to start from progressively adapted models within each communication round. Since the proposed strategies only modify the initialization process, they are compatible with most existing FL frameworks and privacy-enhancing techniques. Experiments on real smart-meter data with two representative forecasting architectures demonstrate that the proposed strategies effectively improve forecasting performance, as evidenced by reduced client drift, improved convergence behavior, and lower forecasting errors.
[LG-34] Diffusion Distillation for Efficient Weather Ensembles
链接: https://arxiv.org/abs/2608.27728
作者: Yiming Yang,Valentin Brekke,James Briant,Serge Guillas
类目: Machine Learning (cs.LG); Applications (stat.AP)
*备注:
Abstract:Diffusion models generate skillful weather ensembles but require costly iterative sampling. We introduce a supervised energy-distance distillation method that compresses a multi-step diffusion teacher into a single-step student by aligning student forecasts with teacher samples and ground-truth observations. Experiments on global forecasting and typhoon-track prediction show that our student outperforms existing distillation methods and preserves skill for extreme events. It matches or surpasses the teacher across key metrics using only one neural function evaluation per autoregressive step.
[LG-35] Leverag ing a Foundation Model for the EEG-Based Diagnosis of Alzheimers Disease
链接: https://arxiv.org/abs/2608.27719
作者: Maggie Lin,Chung-Lin Hou,Tzyy-Ping Jung
类目: Machine Learning (cs.LG); Neurons and Cognition (q-bio.NC)
*备注: 7 pages, 9 figures, 1 table. Accepted and presented at the 48th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC 2026)
Abstract:Biological heterogeneity in Alzheimer’s Disease (AD) poses a critical diagnostic challenge, particularly for traditional linear methods that fail to capture non-linear neural dynamics. To address this, we propose a diagnostic framework utilizing the Large Brain Model (LaBraM), pretrained on over 2,500 hours of EEG data. By integrating these high-dimensional latent embeddings with a non-linear Random Forest classifier, our approach effectively isolates robust disease markers. Under a rigorous subject-independent 5-fold cross-validation protocol, the method achieves an ROC-AUC of 89.36% +/- 3.49%, PR AUC of 81.45% +/- 4.43%, and Balanced Accuracy of 82.44% +/- 4.34% in distinguishing dementia patients from healthy controls. Notably, this performance uses only 8-second EEG segments, surpassing traditional spectral baselines, including band-power and parameterized oscillatory features (FOOOF). Post-hoc occlusion analysis confirms the model captures clinically validated biomarkers, specifically occipital-frontal Alpha and Theta rhythm degradation. Additional neurophysiological alignment analysis demonstrated that higher LaBraM-predicted dementia probability significantly correlated with worse cognitive performance, greater clinical severity, increased theta and alpha relative power, and higher aperiodic exponent. These findings demonstrate that deep latent representations extract clinically relevant signatures from noisy signals, enabling precise, rapid, and data-efficient diagnosis.
[LG-36] Beyond Non-IID: Learner–Client Distribution Mismatch in Federated Learning
链接: https://arxiv.org/abs/2608.27715
作者: Yiming Xie,Lili Su,Ningfang Mi
类目: Machine Learning (cs.LG)
*备注:
Abstract:Federated learning systems are increasingly deployed to facilitate collaborative model training across a heterogeneous client population. Existing practice mostly implicitly assumes that the aggregated client data distribution is representative of the learner’s target distribution or that learning from all available clients is uniformly beneficial for the learner distribution. However, such an assumption often does not hold in reality. Traditional client selection strategies in FL literature largely overlook such misalignment, while most existing work on multi-source transfer learning either requires direct access to local data or uses one-shot model/feature aggregation. In this paper, we take the initiative to understand and mitigate the impacts of such learner-client population misalignment. In particular, we consider the practical setting where the learner keeps a small proxy dataset. We observe that client contributions vary significantly across training rounds, and traditional technology is insufficient to identify beneficial sources under multi-source transfer diversity. Then, we propose a dynamic, influence-aware client selection framework that estimates each client’s potential utility to the learner’s optimization objective using proxy influence signals on a learner-specific proxy set. Via using leave-one-out evaluations, we prioritize the most informative sources of knowledge while controlling the negative impacts of statistical noise and data heterogeneity. Experiments on CIFAR-10 under heterogeneous data partitions demonstrate that our approach consistently outperforms static and dynamic baselines, achieving faster convergence and higher accuracy. Subjects: Machine Learning (cs.LG) Cite as: arXiv:2608.27715 [cs.LG] (or arXiv:2608.27715v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2608.27715 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-37] DART-FL: Burst-Aware Multitask Federated Learning under Dynamic Inference Demand at the Edge
链接: https://arxiv.org/abs/2608.27713
作者: Yiming Xie,Pinrui Yu,Geng Yuan,Xue Lin,Ningfang Mi
类目: Machine Learning (cs.LG)
*备注:
Abstract:Edge intelligence systems increasingly require model training and online inference to coexist on resource-constrained devices, while inference demand can vary substantially across tasks over time. This creates two coupled challenges: sufficient computation must be reserved for inference to maintain service-level objectives (SLOs), while the remaining training capacity should adapt to task-specific demand so that frequently requested tasks can improve earlier during training. We propose an SLO-aware, demand-driven multitask federated learning framework (DART-FL) that jointly adapts the inference-training resource split and task-level training emphasis. At each scheduling interval, DART-FL uses the inference backlog and profiled service capacity to determine the minimum resource allocation required for inference. The remaining training capacity is then distributed across tasks using a queue-aware DPP-inspired scheduler, and the resulting task allocations are mapped to dynamic loss weights. This allows tasks experiencing higher inference demand to receive greater training emphasis in earlier communication rounds. Clients train a shared backbone with task-specific heads, and the complete multitask model is aggregated through FedAvg. We evaluate DART-FL using Stanford Cars and Oxford Flowers 102 under both synthetic and real Alibaba trace-derived workloads. Results show that DART-FL dynamically adapts the inference-training resource split to time-varying inference demand and shifts the learning progress of high-demand tasks toward their burst periods, improving model accuracy when those tasks are frequently requested while maintaining comparable long-term multitask performance. Subjects: Machine Learning (cs.LG) Cite as: arXiv:2608.27713 [cs.LG] (or arXiv:2608.27713v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2608.27713 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-38] SafeStep: An Interactive Demonstration of Semantic Communication for Pedestrian Safety Monitoring
链接: https://arxiv.org/abs/2608.27688
作者: Christian McDowell,Andrea Panebianco,Jeremiah Yang,Sirin Chakraborty,Samuel Chamoun,Travis Ross,Yin Sun
类目: Machine Learning (cs.LG)
*备注: 6 pages, 5 figures. Submitted to the Quality, Value, and Age of Information for Tactical Networks workshop at the IEEE Military Communications Conference (MILCOM). Christian McDowell, Andrea Panebianco, and Jeremiah Yang are co-primary authors
Abstract:In this paper, we develop SafeStep, an interactive browser-based semantic communication platform for live pedestrian safety monitoring. SafeStep extracts pedestrian information from four live traffic-camera feeds, transmits it through a semantic communication transceiver over an Additive White Gaussian Noise (AWGN) channel, and renders user-specific positions, trajectories, and risk labels. The platform allows to independently select the transceiver, Signal-to-Noise Ratio (SNR), codelength, and Age of Information (AoI), and demonstrates the transceiver performance of the selected configuration through live pedestrian safety monitoring to each browser. SafeStep compares a recently proposed semantic communication design called Meta-VIB with five baseline transceivers. Meta-VIB uses a compact neural model with only 4.16 million parameters to generalize across varying SNR, codelength, and AoI values without online retraining. Experimental results show that Meta-VIB achieves mean task-loss reductions of up to 92.1% . On one high-end GPU server, the integrated concurrent-access workload maintains the target 5 frames/s through 20 users. At 100 users, each requesting a distinct configuration, SafeStep records no request failures and a mean application response time below 1 s, but its mean per-browser frame rate falls to approximately 1 frame/s. To our knowledge, SafeStep is the first real-time semantic communication platform to make AoI-induced downstream degradation directly observable in live monitoring applications.
[LG-39] SegBench-GC: Testing Segmentation Invariance in Multi-Step Offline Goal-Conditioned Reinforcement Learning
链接: https://arxiv.org/abs/2608.27678
作者: Musa Shams
类目: Machine Learning (cs.LG)
*备注: 10 pages, 2 tables. Code and reproducibility artifacts: this https URL
Abstract:Offline goal-conditioned reinforcement learning (GCRL) often uses trajectory structure for future-goal sampling and multi-step targets, yet logged trajectories may be partitioned for administrative reasons that do not correspond to termination. We introduce SegBench-GC, a controlled stress test of segmentation invariance that holds transitions, source trajectories, goal sampling, optimization settings, and evaluation fixed while varying only artificial backup boundaries and whether those boundaries retain continuation value. Continuation-valid targets (CVT) provide the segmentation-consistent control: reward accumulation stops at an artificial cut, but the target bootstraps from its stored successor. In a matched-count PointMaze study with 35,000 artificial cuts, three segmentation realizations, and three optimization seeds, final 50-episode-per-task success is 50.5% uncut, 39.1% with CVT, and 19.1% when the same cuts are treated as absorbing; across segmentation realizations, naive mean success ranges from 4.8% to 31.9%. An independent published n-step baseline (n=25) from the Decoupled Q-Chunking codebase shows the same failure on Puzzle-4x5: 47.2% uncut, 58.5% CVT, and 0.27% naive across three optimization seeds. A target-level diagnostic verifies the analytic target difference to numerical precision, and learned-critic diagnostics show a large optimistic shift under naive handling while CVT remains approximately aligned with the uncut critic. CVT applies standard continuation bootstrapping rather than a new Bellman rule; the contribution is the controlled benchmark, failure isolation, and cross-learner evidence that administrative segmentation can materially change multi-step offline GCRL.
[LG-40] Unsupervised Continual Learning with Growing Self-Organizing Maps and Synthetic Replay
链接: https://arxiv.org/abs/2608.27662
作者: Pujan Thapa,Alexander Ororbia,Travis Desell
类目: Machine Learning (cs.LG)
*备注:
Abstract:This work presents a generative continual learning framework based on growing self-organizing maps (GSOMs) that are augmented with learned distributional statistics as well as encoder-decoder models for class-incremental learning. The proposed approach enables exemplar-free replay using distributional statistical memory, which eliminates the need to store raw data. Each GSOM unit maintains its own mean, variance, and covariance estimates, which are subsequently used to generate synthetic samples for replay; in encoder-decoder configurations, these samples are then decoded back into the input space (via ancestral sampling) for subsequent training. Our method is fully unsupervised, as it does not rely on explicit task boundaries or class labels during training. Results across multiple benchmarks show that the proposed approach achieves performance competitive even with supervised state-of-the-art memory-based methods while consistently outperforming memory-free approaches. In several settings, our framework matches or exceeds existing baselines, particularly in challenging single-class incremental scenarios. We also provide baseline results for single-class incremental TinyImageNet and MiniImageNet, offering a useful reference for future work. This work highlights the effectiveness of an unsupervised, adaptive, topology-driven neural form of statistical replay as a scalable, flexible approach to continual learning.
[LG-41] More Data Cannot Break a Symmetry: Identifiability by Design
链接: https://arxiv.org/abs/2608.27651
作者: Jing Xu,Christopher Kanan
类目: Machine Learning (cs.LG)
*备注: 14 pages, 6 figures
Abstract:Unsupervised representational alignment recovers a stimulus-by-stimulus correspondence from geometry alone, but the automorphism group of the stimulus geometry bounds what any such alignment can identify, before data exist. The obvious diagnostic for this degeneracy, the cheapest non-identity relabelling, ranks two published designs in the wrong order, because dense sampling creates near-duplicates whose transposition is nearly free. We turn this known invariance (Demetci et al., 2024) into a design-time diagnostic and intervention. In colour, where candidate geometries have closed form, we show that the failure is structural: sixty-four times the restart budget leaves a symmetric design unmoved while an asymmetric set at the same N recovers every time. Discriminating representational models and recovering a correspondence are essentially uncorrelated objectives (r = -0.02 over 3,000 subsets). Choosing nine colours by this diagnostic alone, without consulting any learned representation, moves all 93 model representations away from the degenerate point and cuts catastrophic alignment failures from 75% to 2% with the models, the layers, N and the solver all held fixed. The same risk arises wherever a regular design meets its candidate geometry’s isometry group, including evenly spaced orientations, tones, or motion directions, and the check costs one function call before data collection.
[LG-42] Dandelion: A Spherical Flower for Neural Simulation of Planetary Dynamics
链接: https://arxiv.org/abs/2608.27521
作者: Till Muser,Giovanni Abati,Ivan Dokmanić
类目: Machine Learning (cs.LG)
*备注:
Abstract:Many dynamical processes unfold on the sphere but the default scientific machine learning architectures are Euclidean. Applying these architectures on a regular lat-lon grid causes problems: Cartesian convolutions become distorted at high latitude; 2D FFTs in Fourier neural operators incorrectly assume double periodicity; Cartesian positional encodings in ViTs distort spherical geodesic distances. Recent work moves towards natively spherical primitives, including spherical convolutions (e.g., DeepSphere or DISCO), Spherical Fourier Neural Operators (SFNOs), and geodesic attention. Here we propose Dandelion, a spherical version of Flower, a warp-based neural PDE solver. Layers of Dandelion predict a tangent-plane displacement and transport features along great circles. We obtain a U-Net-like structure by implementing hierarchical pooling entirely in the spherical-harmonic domain. There are thus no convolutions: spatial mixing is achieved only through spherical coordinate changes, or warps. To compare Dandelion with existing spherical architectures, we release an evolving benchmark suite of challenging, natively-spherical PDE datasets including a modified Galewsky jet, anomalous chained turbulence, Cahn-Hilliard decomposition, spherical Riemann shocks, Held-Suarez dry atmospheric transport and global ocean dynamics. This new benchmark fills the gap in existing spherical datasets which are either too small and stylized, or much too large (ERA5) for model iteration. Dandelion is best or second-best on every dataset, and the gap to non-warp baselines widens with resolution: at 256\times 512 , Dandelion and Flower2D occupy the top two slots in both single-step prediction and rollout.
[LG-43] When Muon Meets Task Interference: A Spectral Perspective on Continual Learning and Model Merging
链接: https://arxiv.org/abs/2608.27518
作者: Shangge Liu,Yuehan Yin,Yinghuan Shi,Lei Wang,Wenbin Li
类目: Machine Learning (cs.LG)
*备注:
Abstract:Continual learning (CL) and model merging (MM) both aim to obtain a single model that performs well across multiple tasks, challenged respectively by catastrophic forgetting and weight-disentanglement error. In the literature, these difficulties are merely treated separately and mitigated through a variety of solutions, while the geometry induced by the base optimizer is treated as an implementation detail. In this work, we show that the two difficulties are in fact two instances of the same phenomenon: a parameter update useful for one task shifts the model’s outputs on another. We formalize this shared phenomenon as \textittask interference and reduce it to a common layer-wise Frobenius inner product \langle \Delta W_\ell, J_\ell(x)\rangle_F . This quantity, in turn, is utilized to expose the role of the optimizer. We theoretically derive an upper bound that isolates the spectral norm |\Delta W_\ell|_2 as an optimizer-controllable factor of task interference, and a per-mode analysis shows that this bound tracks the dominant part of the empirical interference. Specifically, we then identify the recent Muon optimizer as a mechanism that regulates this factor by construction. Our work reveals that its elegant control on spectral norm tightens the interference bound for both CL and MM, positioning Muon as a principled optimizer-centric approach complementary to existing solutions. Our theoretcal analysis is well validated by experimental results. Replacing the AdamW optimizer with Muon improves accuracy by up to +5.02 points on the eight-task model-merging benchmark across three CLIP backbones. For continual learning, Muon also delivers uniformly positive gains across ten class-incremental protocols, three task-incremental protocols, and the 11-task MTIL benchmark.
[LG-44] Learning between the peaks: sharp asymptotics for kernel ridge regression under power-law anisotropy
链接: https://arxiv.org/abs/2608.28564
作者: Lorenzo Rizzi,Arie Wortsman Zurich,Bruno Loureiro
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:We study kernel ridge regression under anisotropic Gaussian data, where the input covariance decays as a power law with exponent \alpha\geq 0 for polynomial inner-product kernels. We derive asymptotically sharp expressions for the kernel spectrum and the generalization error in the polynomial high-dimensional regime n=\Theta(d^\kappa) , revealing how anisotropy reshapes the learning curves. For weak anisotropy ( 0\alpha1 ), the problem remains effectively high-dimensional and retains some features of the isotropic case, while departing from it in others: the variance still peaks at integer sample complexities \kappa\in\mathbbN , but these peaks are progressively damped as \alpha grows; meanwhile, for targets strongly aligned with the data’s principal directions, the bias drops at fractional sample complexities, decoupling the bias transitions from the interpolation peaks. For strong anisotropy ( \alpha 1 ), the effective dimension of the problem is constant, and the variance stops depending on sample size altogether, plateauing under ridgeless interpolation or vanishing at an explicit rate under fixed ridge penalty. The bias undergoes a sharp transition governed by the target’s decay rate: below a threshold, learning is abrupt rather than gradual; above it, the bias decays as a power law that recovers the classical source and capacity rates. We finally specialize these results to single-index targets, showing how the alignment of the index with the data’s principal directions determines the effect of anisotropy on learning. Together, our results clarify how the input geometry shapes the kernel features and fundamentally impacts its generalization properties.
[LG-45] Generalized Splines and Gaussian Processes
链接: https://arxiv.org/abs/2608.28446
作者: Michael Unser
类目: atistics Theory (math.ST); Machine Learning (cs.LG); Functional Analysis (math.FA); Machine Learning (stat.ML)
*备注:
Abstract:For finite-dimensional linear inverse problems where the variables are Gaussian, it is well-known that the minimum-mean-square error estimator takes the form of a regularized least-squares data fit. In this chapter, we show that this equivalence extends to a much broader infinite-dimensional setting where generalized splines take the role of linear regressors and generalized Gaussian processes on a nuclear space S are the counterpart of Gaussian random vectors. The scope of this extension is of the same nature as the switch from the classic notion of function to that of a distribution, also known as a “generalized function.” Our formalism involves a whitening/regularization operator L: S\to S’ whose continuous extension induces a native Hilbert space H\subset S’ that plays a central role in our characterization. The presentation is self-contained for the most part and remarkably general and powerful. It allows for the recovery of all known instances of such equivalences; in particular, the methods involving innovations and reproducing-kernel Hilbert spaces developed by Kailath and his students, and the mathematical correspondence between fractional splines and Mandelbrot’s fractional Brownian motion (fractals), with the former being the optimal estimators of the latter. It also covers general Bayesian methods for the resolution of infinite-dimensional inverse problems.
[LG-46] Quantum Federated Learning Based on Bures–Uhlmann Geometry for Heterogeneous Noisy Clients
链接: https://arxiv.org/abs/2608.28379
作者: Haruki Emori,Masaki Uchihara,Yuuki Tokunaga
类目: Quantum Physics (quant-ph); Machine Learning (cs.LG)
*备注: 18 pages, 2 figures, and 3 Tables
Abstract:Quantum federated learning enables collaborative model training across quantum devices without sharing raw data, and it faces the data and hardware heterogeneity inherent to noisy quantum devices. Utilizing the quantum geometric tensor is a natural remedy, yet pure-state approaches and diagonal approximations discard the correlations that encode parameter incompatibility. To address this, we extend the parameter-space geometry to the mixed states that noisy clients actually prepare. The real part of the resulting mixed-state geometric tensor is the Bures metric, which measures how fast the physical state changes under parameter variation, and the imaginary part is the mean Uhlmann curvature, which quantifies the incompatibility of estimating multiple parameters simultaneously. Accordingly, we employ the Bures metric as a local preconditioner and use the mean Uhlmann curvature to develop an achievable-precision aggregation rule that dynamically down-weights unreliable clients. Furthermore, we establish theoretical guarantees by proving a convergence theorem and a variance-dominance proposition. Empirical evaluations on a trapped-ion quantum emulator demonstrate that the proposed method maintains high accuracy across diverse device-heterogeneity conditions and outperforms standard federated averaging, whose accuracy degrades under strong noise.
[LG-47] Localizing Global Discrepancies: Marginal Contributions and Contextual Anomaly Detection
链接: https://arxiv.org/abs/2608.28375
作者: Tommaso dorigo
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Data Analysis, Statistics and Probability (physics.data-an)
*备注: 32 pages
Abstract:Global goodness-of-fit and discrepancy statistics can establish that a sample departs from a reference distribution without identifying which observations drive the departure. We develop a framework for this localization problem by assigning to each observation its conditional or marginal contribution across random statistical contexts. This connects resampling diagnostics and data valuation to projection theory and event-level anomaly detection. For symmetric statistics, fixed-size replacement is exactly equivalent to centered conditional localization. For U-statistics, the addition score equals the first Hoeffding/Hájek contribution; for smooth distributional functionals it is related at leading order to the influence function; and for unbiased known-background MMD it reduces exactly to the MMD witness. This viewpoint also yields more efficient estimators. Matched-context subtraction removes fluctuations unrelated to the observation, while for pairwise MMD the event-containing terms give a simple localizer. On the LHC Olympics anomaly-detection benchmark, the pair estimator converges to the direct empirical MMD witness with the predicted 1/(Rm^2) scaling, where m is batch size and R the number of batches. At m=1000 and R=5x106 it reaches correlation 0.9993 with essentially identical AUC. We also ask when context contains information beyond an event’s own features. In a shared-latent toy model, the full single-event signal and background distributions are identical by construction, forcing isolated-event AUC=0.5. Discriminating information survives only in cross-event dependence induced by the shared latent parameter; the ensemble recovers this information, whereas an independent-latent control does not. This separates two roles of context: efficient localization of a global discrepancy and genuinely additional class information when the alternative contains shared structure. Comments: 32 pages Subjects: Machine Learning (stat.ML); Machine Learning (cs.LG); Data Analysis, Statistics and Probability (physics.data-an) Cite as: arXiv:2608.28375 [stat.ML] (or arXiv:2608.28375v1 [stat.ML] for this version) https://doi.org/10.48550/arXiv.2608.28375 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-48] Real-Time Monitoring of MHD Liquid Metal Flows with Shallow Recurrent Decoders
链接: https://arxiv.org/abs/2608.28366
作者: Claudio Scardino,Stefano Riva,Carolina Introini,Matteo Lo Verso,Eric Cervi,Antonio Cammi,Laura Savoldi
类目: Computational Physics (physics.comp-ph); Machine Learning (cs.LG); Fluid Dynamics (physics.flu-dyn)
*备注:
Abstract:State estimation in magnetohydrodynamic flows is critical for real-time monitoring of liquid metal blankets in tokamak fusion reactors. Due to the multiphysics nature of these phenomena, high-fidelity simulations are computationally prohibitive for real-time applications. This work investigates a data- driven Reduced Order Model framework: the Shallow Recurrent Decoder (SHRED) coupled with Principal Component Analysis, to map sparse temperature measurements to the full thermo-hydraulic system’s state. The major contribution of this work lies in the two-parameter analysis of a fully three-dimensional domain representative of the DEMO breeding blanket configuration. Here, the flow is subjected to an external magnetic field varying in direction and intensity and is hindered by two cylinders acting as a water-cooling system, which impose a temperature boundary condition on their surfaces. This double-parametric magnetic variation induces nonlinear transitions in the flow dynamics, ranging from chaotic behavior at low magnetic field intensities to laminarized regimes at high intensities, characterized by the formation of asymmetric side layers at an inclination angle of 30 degrees. SHRED reconstruction maintains a mean relative error of approximately 5% for the temperature, pressure, and velocity fields. This accuracy is maintained across both weak and strong magnetic fields, ranging from 0.075 T to 0.300 T, and for inclination angles from 5 to 30 degrees, reflecting its dominant toroidal component. These errors are only slightly larger than the lower error bound dictated by low-rank truncation. The results establish SHRED as a reliable state estimator for complex and realistic engineering applications involving completely unseen parametric scenarios and validate it as an accurate real-time state estimation technique suitable for online monitoring and control of real facilities.
[LG-49] I-FLOP: Fast Learning of Order and Parents from Interventional Data
链接: https://arxiv.org/abs/2608.28245
作者: Liuting Chen,Alex Markham
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注: 28 pages, 7 figures; accepted to The 13th International Conference on Probabilistic Graphical Models (PGM)
Abstract:We extend the FLOP (fast learning of order and parents) algorithm recently proposed by Wienöbst et al. (2026) from observational to interventional data. In particular, we use the interventional BIC score of Hauser and Bühlmann (2012), adapting it to be used with the iterative Cholesky-based score updates that are partly responsible for FLOP’s speed. We show that, in the sample limit, I-FLOP recovers a DAG in the same interventional Markov equivalence class as the data-generating DAG. We compare I-FLOP to existing causal structure learning algorithms on real and simulated interventional data, where it performs favorably in terms of both performance and run time.
[LG-50] Landau theory of quenched criticality in linear in-context learning
链接: https://arxiv.org/abs/2608.28059
作者: Daesik Kim,Sumin Choi,Hyojae Jeon,Jung Hoon Han
类目: Disordered Systems and Neural Networks (cond-mat.dis-nn); Machine Learning (cs.LG)
*备注: 17 pages, 7 figures (counting subfigures)
Abstract:In-context learning (ICL) allows a pretrained model to infer a new task from examples supplied in its prompt without updating its parameters. In linear models of ICL, the prediction error develops a double-descent singularity when the number of pretraining samples becomes comparable to the number of learnable parameters. We formulate this interpolation singularity as a critical phenomenon of a quenched disordered system. By comparing annealed and quenched descriptions of the same linear ICL model, we identify the connected sample-to-sample fluctuations of the learned parameters as the microscopic origin of the singular error. A Landau potential is constructed by integrating the cavity self-consistency equation for the renormalized ridge parameter \xi . The role of (magnetization) order parameter is played by \xi , while the bare ridge parameter \lambda becomes its conjugate magnetic field. The normalized sample complexity \tau acts as a temperature and the double-descent singularity occurs at the critical temperature \tau_c =1 . The Landau susceptibility is precisely the quantity that diverges in the fluctuation contribution to the prediction error. The order parameter is closely related to the fraction of zero eigenvalues of the empirical relaxation matrix in the ridgeless limit, which define flat directions in the learning dynamics. The Landau theory is generically cubic in the order parameter with critical exponents (\beta_\rm cr,\delta_\rm cr,\gamma_\rm cr)=(1,2,1) . In the large-context regime, there appears a pseudogap-like regime characterized by suppressed order parameter. Predictions of the Landau theory are independently confirmed from numerical solutions of the original learning problem with good quantitative agreement. Our results pave the way for solid statistical-physics understanding of the interpolation criticality in linear in-context learning.
[LG-51] Anchored Scenario Coverag e for Failure-Aware First-Hit Batch Inverse Design
链接: https://arxiv.org/abs/2608.27873
作者: Chuhan Yang,Chenxi Wang,Linhan Wu,Yuyang Liu
类目: Optimization and Control (math.OC); Machine Learning (cs.LG)
*备注:
Abstract:Early discovery of at least one valid design satisfying a target requirement is a central objective in failure-prone closed-loop inverse design. A natural batch baseline ranks candidates by a product-form marginal valid-hit score, but selecting the highest-ranked candidates independently can produce redundant recommendations under predictive uncertainty and waste the experiment budget. We introduce ARC-SC(Anchored Risk-Constrained Scenario Coverage), a batch acquisition method that preserves strong marginal candidates as anchors and allocates the remaining batch positions by maximizing complementary coverage over predictive target scenarios under a risk-support constraint. In frozen-oracle closed-loop simulations on superconductivity and JARVIS materials-property benchmarks, ARC-SC yields a statistically supported improvement in first-hit discovery and remains competitive with directionally favorable first-hit performance on more challenging design space. These results establish ARC-SC as a POF-anchored, scenario-aware batch strategy for improving early valid-target discovery under structured experimental failure.
[LG-52] Beyond Procrustes distances: a multilinear Gromov-Wasserstein distance capturing chirality
链接: https://arxiv.org/abs/2608.27774
作者: Clément Soubrier,Geoffrey Woollard,Andrew Warren,Khanh Dao Duc
类目: Optimization and Control (math.OC); Machine Learning (cs.LG)
*备注: 57 pages, 6 figures
Abstract:Efficiently and robustly analyzing shape data is critical across many scientific disciplines. While chirality is a fundamental property in numerous applications - most notably in molecular science - existing shape analysis metrics fail to distinguish between a shape and its mirror image. To address this gap, we introduce a multilinear generalization of the Gromov-Wasserstein objective. Under mild assumptions, this objective yields a distance between shapes, represented as probability distributions quotiented by a symmetry group G . In particular, for G = SO(d) , we introduce the Chiral Gromov-Wasserstein ( \mathrmCGW ) distance, sensitive to chirality. We establish robustness properties for the multilinear Gromov-Wasserstein distances and develop efficient algorithms to compute them, reformulating the underlying optimization problem by projecting couplings onto a low-dimensional space. We derive algorithms for both local and approximate global solutions, yielding a fully polynomial-time approximation scheme for these problems. We validate the framework through numerical experiments that demonstrate the effectiveness of \mathrmCGW as a shape metric for chiral objects.
[LG-53] On the Computational and Statistical Efficiency of the Empirical Maximum Entropy on the Mean Method
链接: https://arxiv.org/abs/2608.27705
作者: Matthew King-Roskamp,Gabriel Rioux,Rustum Choksi,Tim Hoheisel
类目: Optimization and Control (math.OC); Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: 39 Pages, 7 figures
Abstract:The Maximum Entropy on the Mean (MEM) method provides a flexible computational framework for solving inverse problems by combining data fidelity with entropy-based regularization. In practice, however, the prior distribution is typically unknown but can be estimated from data, giving rise to the empirical MEM method. We establish a parametric convergence rate of O(n^-1/2) in expectation for empirical MEM, improving upon the previously established O(n^-1/4) guarantee by King-Roskamp et al. (2026). Our proof is based on a novel stability analysis of the primal and dual optimization problems under perturbations of the underlying probability measure, relying only on foundational tools from convex analysis and probability. We further show that the MEM dual problem admits a reformulation as an expected risk minimization problem, thereby placing MEM within the modern framework of stochastic optimization and enabling scalable stochastic gradient algorithms for large-scale inverse problems. Together, these results place empirical MEM as a statistically and computationally efficient methodology for data-driven inverse problems.
[LG-54] Quantum SEDONet: Spectrally-Embedded Quantum Deep Operator Networks for Partial Differential Equations
链接: https://arxiv.org/abs/2608.27626
作者: Muhammad Abid,Arth Sojitra,Bipin Tiwari,Omer San
类目: Quantum Physics (quant-ph); Machine Learning (cs.LG)
*备注:
Abstract:Quantum DeepONet accelerates neural-operator inference by evaluating an orthogonally parameterized network on a quantum computer, reproducing in ideal simulation the accuracy of its classical counterpart at asymptotically lower inference cost. Its trunk network, however, receives query coordinates with limited spectral structure, requiring the network to learn oscillatory features through its nonlinearities. We propose Quantum SEDONet (Spectral-Embedded Deep Operator Network), which assigns each trunk coordinate a spectral basis according to its boundary condition: Fourier features for periodic coordinates and Chebyshev features for bounded, non-periodic coordinates. The basis is selected per coordinate rather than per problem, allowing both representations within a single problem. Under unary amplitude encoding, the embedding incurs no additional qubits or circuit depth when its dimension remains within the network width, while increasing the parameter count by only a few percent. Across four benchmarks, Quantum SEDONet reduces the mean relative L2 error by 54.1% for the antiderivative, 49.6% for advection, 36.0% for Burgers, and 36.2% for a mixed-boundary channel Poisson problem. Quantum and classical evaluation paths agree to within 10^-8 throughout. The channel Poisson problem simultaneously uses Fourier features in the periodic direction and Chebyshev features in the bounded direction, demonstrating coordinate-wise boundary-matched spectral embedding without additional quantum-resource cost.
[LG-55] Physics-informed learning for the inverse problem in resonant ultrasound spectroscopy
链接: https://arxiv.org/abs/2608.27590
作者: Alejandro Cubillos Muñoz,Manuela Rivas,Julian Rincon
类目: Materials Science (cond-mat.mtrl-sci); Machine Learning (cs.LG); Applied Physics (physics.app-ph); Computational Physics (physics.comp-ph)
*备注: 15 pages, 8 figures, 3 tables
Abstract:Inferring elastic constants from resonant ultrasound spectra is a nonlinear and typically overdetermined inverse problem based on finite spectral data. We formulate the Rayleigh-Ritz inverse problem as a constrained inverse-isospectral problem on the set of physically admissible elasticity tensors. This induces effective low-dimensional variables for the inverse map on the admissible elasticity manifold: length and elastic scales, aspect-ratio coordinates, scale-free spectral features, and stability-respecting elastic ratios. We use these variables to construct a physics-informed learning pipeline in which a regression model acts only on reduced spectral and geometric features, while scale recovery and final elastic-constant reconstruction are imposed analytically. For the full cubic benchmark, the reconstructed constants have MAE values of 20.37(35.15) , 24.30(41.33) , and 2.13(3.66)~\mathrmGPa for C_11 , C_12 , and C_44 . In the fixed-geometry benchmark, the corresponding cubic MAPE values are 4.14(3.87)% , 8.31(8.50)% , and 2.44(2.86)% , while the isotropic values are 4.0(3.6)% and 0.4(0.3)% for the bulk and shear moduli. The inverse problem then becomes a constrained regression problem in variables adapted to the geometry, scaling, crystal symmetry, and thermodynamic stability of Hookean elasticity.
[LG-56] owards Large-Scale Heterogeneous Data Organization for Scientific Foundation Models: A Nuclear Fusion Case Study ICLR2026
链接: https://arxiv.org/abs/2608.27578
作者: Nathaniel Chen,Kouroche Bouchiat,Peter Steiner,Azarakhsh Jalalvand,SangKyeun Kim,Egemen Kolemen
类目: Plasma Physics (physics.plasm-ph); Machine Learning (cs.LG)
*备注: Accepted at the 3rd Workshop on Navigating and Addressing Data Problems for Foundation Models (DATA-FM), ICLR 2026
Abstract:Training effective foundation models requires massive and organized datasets, yet scientific domains such as nuclear fusion present unique challenges due to largely heterogeneous and sparse data. Here we characterize the data used in developing such a model: with over 20 sensor types spanning 5 orders of magnitude in sampling rate, mixed tensor structures (point measurements, spectrograms, images), and nonstationary physics. We analyze our input complexity and discuss trade-offs between temporal context and frequency resolution. Our analysis provides a template for representing multi-modal fluctuation data at scale, with implications for both multi-modal control systems and nuclear fusion.
[LG-57] owards a mathematical theory of superposition
链接: https://arxiv.org/abs/2608.27540
作者: Michael I. Ivanitskiy,John Jasper,Emily J. King,Dustin G. Mixon
类目: Machine Learning (stat.ML); Information Theory (cs.IT); Machine Learning (cs.LG); Combinatorics (math.CO)
*备注:
Abstract:We develop a mathematical theory of superposition in neural networks using tools from frame theory and compressed sensing. In our model, a sparse binary vector (x) of active features is encoded through an overcomplete dictionary (W), and feature recovery is performed by applying (\operatornameReLU(W^\top W x+b)) with an appropriate bias vector (b). We prove several recovery theorems for this model. In the random-support setting, we establish high-probability support recovery for nearly tight, low-coherence dictionaries, with guarantees when the expected sparsity is up to order (d/\log n). In the worst-case support setting, we give a sharp and computable criterion for which sparsity levels permit support recovery. We apply this criterion to Gaussian random matrices and equiangular tight frames. For real equiangular tight frames with (nd+1), we determine the exact recovery threshold in terms of the coherence. The proof of this result for real equiangular tight frames relies on a novel characterization—which should be of independent interest to frame theorists—of the distribution of signs in the Gram matrix.
[LG-58] Ab initio Modeling of MoS2/Oxide Device Interfaces with Machine Learned Electronic Structures
链接: https://arxiv.org/abs/2608.27533
作者: Manasa Kaniselvan,Mauro Dossena,Denghui Lu,Alexander Maeder,Nicolas Vetsch,Alexandros Nikolaos Ziogas,Mathieu Luisier
类目: Materials Science (cond-mat.mtrl-sci); Machine Learning (cs.LG)
*备注:
Abstract:We introduce a new ab initio approach to simulate semiconductor devices that integrates scalable machine-learned (ML) electronic structure models with an advanced quantum transport (QT) solver. The developed framework enables 10,000X speedups over density functional theory to produce the Hamiltonian matrix of devices made of 20,000 atoms, while offering high prediction accuracy. We use its unique features to investigate MoS2/oxide samples and single-layer MoS2 field-effect transistors, where the surrounding oxide layers, here, HfO2 or Al2O3, are explicitly included into the QT domain. In particular, we reveal that the presence of undercoordinated metal atoms (Hf or Al) close to the semiconductor-oxide interface significantly affects the magnitude of the electronic current and its propagation through MoS2.
[LG-59] Optimal Transport for Network Comparison: A Review with Machine Learning Applications
链接: https://arxiv.org/abs/2608.27500
作者: James Hyun,François G. Meyer
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Social and Information Networks (cs.SI)
*备注:
Abstract:Network comparison using optimal transport is a growing area of research in network science. Unlike standard graph metrics, optimal transport computes both network dissimilarity and a transport plan that explains how one graph morphs into another. In this paper, we review how optimal transport compares undirected, unweighted graphs using three primary distances: the Wasserstein, Gromov-Wasserstein, and Bures-Wasserstein distances. We examine the closed form of the Wasserstein distance in one dimension via node feature probability distributions, and show how the transport plans of the Wasserstein and Gromov-Wasserstein distances capture which specific nodes influence the distance after graph perturbation. For the Bures-Wasserstein distance, we derive bounds using Laplacian spectra to bypass full spectral decompositions. Finally, we evaluate these distances using a synthetic network dataset for clustering and a real-world time series network for anomaly detection.
[LG-60] Multiscale Community-Based Fingerprinting of Signed Functional Networks
链接: https://arxiv.org/abs/2608.27483
作者: Sema Athamnah,Selin Aviyente
类目: Neurons and Cognition (q-bio.NC); Machine Learning (cs.LG); Signal Processing (eess.SP)
*备注:
Abstract:Objective: Recent studies demonstrate that functional connectomes contain subject-specific signatures, or \textitfingerprints, that can identify individuals across repeated sessions and tasks. Existing methods mostly rely on edge-level features that are sensitive to noise, difficult to interpret, and limited in their ability to generalize across tasks and datasets. Methods: We propose a multiscale community-based functional connectome fingerprinting framework that characterizes each individual by the mesoscale structure of their functional networks. We introduce a signed multilayer community detection framework that incorporates both correlated and anti-correlated brain activity to identify subject-specific community structures across tasks and sessions. Graph-theoretic metrics are then computed from the resulting joint community structures to derive low-dimensional community-level fingerprint representations. Results: The proposed framework is evaluated on 810 healthy control subjects from the Human Connectome Project (HCP). The results show that community-based fingerprints provide a reliable and interpretable substrate for individualized brain characterization across sessions and tasks. Conclusion: Mesoscale community structure provides meaningful and discriminative subject-specific fingerprints. Significance: The proposed framework offers a promising foundation for precision neuroimaging and personalized neuroscience applications.
附件下载
