本篇博文主要内容为 2026-07-29 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 (2026-07-29)
今日共更新670篇论文,其中:
- 自然语言处理共83篇(Computation and Language (cs.CL))
- 人工智能共261篇(Artificial Intelligence (cs.AI))
- 计算机视觉共104篇(Computer Vision and Pattern Recognition (cs.CV))
- 机器学习共153篇(Machine Learning (cs.LG))
- 多智能体系统共15篇(Multiagent Systems (cs.MA))
- 信息检索共58篇(Information Retrieval (cs.IR))
- 人机交互共41篇(Human-Computer Interaction (cs.HC))
多智能体系统
[MA-0] owards a Systems Foundation for Agent ic Cloud Management
【速读】：该论文旨在解决当前自主管理代理（agentic management agents）在云环境中的协同操作缺乏系统性基础支撑的问题，尤其是在多代理间及代理与人类操作者之间实现安全、高效、可追溯的协同管理。其核心挑战在于如何在保障并发性的同时，避免因多个代理对共享云资源的冲突操作导致的不一致或风险行为。论文提出的关键解决方案是构建CloudWeaver——一种跨现有云用户界面与未来原生代理界面的智能管理底座（agentic management substrate）。其关键机制包括：（1）通过局部视图（local views）限定单个代理会话的作用域，实现对云资源上下文的精准控制；（2）在共享资源上协调并发管理操作，支持并行执行的同时提供强安全性保证和可归因的反馈机制，以应对不同代理间意图冲突的情况。实验验证表明，CloudWeaver在典型Azure API工作负载下能够有效保障操作的安全性与一致性，为下一代自动化云管理提供了可信赖的系统基础。
链接: https://arxiv.org/abs/2607.25883
作者: Minghao Li,Ziqian Liu,Ziyu Mao,Daqian Ding,Yu Kang,Qingwei Lin,Tianyin Xu,Yiming Qiu
机构: The University of Hong Kong (香港大学); Microsoft (微软); University of Illinois Urbana-Champaign (伊利诺伊大学香槟分校)
类目: Multiagent Systems (cs.MA)
备注:
Abstract:Agentic cloud management is emerging as a practice to automate laborious operations, minimize toil, and improve responsiveness. Despite the rapid development of autonomous management agents, we argue that the fundamental missing piece is a systems foundation to enable safe, effective operations across agents and between agents and human operators. In this paper, we advocate for the need of such a systems foundation and share our efforts on developing CloudWeaver, an agentic management substrate that works across existing cloud-user interfaces and future agent-native interfaces. Specifically, we discuss how CloudWeaver (1) scopes the context of individual agent sessions with local views of cloud resources and (2) coordinates concurrent management operations on shared cloud resources. CloudWeaver offers strong safety guarantees and attributable feedback in the presence of conflicting intents, while preserving concurrency between independent operations. We validate CloudWeaver using a representative Azure API workload.
[MA-1] CHILL-Harness: Counterfactual Harness Learning for Efficient Reasoning in Long-Horizon Agents
【速读】：该论文旨在解决现有大型语言模型代理（Large Language Model Agents）调度系统在执行长时序任务时，因依赖静态或人工设计的控制策略而无法动态适应任务需求、执行环境及运行状态变化的问题。这一局限性导致计算资源浪费，甚至降低任务成功率。其核心解决方案是将调度系统的自适应编排问题建模为因果学习问题，并提出一种名为“反事实调度干预学习”（Counterfactual Harness Intervention Learning for Long-Horizon Agents, CHILL-Harness）的新框架。该方案的关键在于：首先通过因果干预效应学习（causal intervention effect learning），基于置信度加权的执行证据估计不同工作流调整的相对优势；其次，引入优势实现型因果编排（advantage-realizing causal orchestration），仅对预期优势充足的反事实推理路径进行动态调整，从而实现高效且可靠的流程优化；最后，结合成功保持目标与优势裕度授权约束，确保适应过程不损害任务成功率。实验结果表明，CHILL-Harness在信息检索、软件工程和终端交互等多样化的长时序任务中，显著降低了令牌消耗与执行时间，同时保持或提升了任务成功率。
链接: https://arxiv.org/abs/2607.25825
作者: Jiarun Fu,Lizhong Ding,Sida Chen,Honglei Xin,Chunhui Zhang,Pengqi Li,Qiuning Wei,Ye Yuan,Guoren Wang
机构: Beijing Institute of Technology (北京理工大学)
类目: Multiagent Systems (cs.MA)
备注:
Abstract:Agent harnesses have become the operational infrastructure of modern large language model agents, coordinating context, tools, verification, and execution control to translate latent model capability into reliable long-horizon behavior. However, reliable long-horizon behavior requires harness control to adapt to task demands, execution environments, and evolving execution states, whereas current harnesses predominantly rely on hand-crafted or globally fixed policies; this mismatch manifests as unnecessary computational overhead and, in adverse cases, reduced task success. To address this limitation, we formulate the task of enabling adaptive orchestration in harness systems as a causal learning problem and propose Counterfactual Harness Intervention Learning for Long-Horizon Agents (CHILL-Harness). CHILL-Harness intervenes at the orchestration layer to enable advantage-guided workflow adaptation, thereby improving reasoning and execution efficiency while preserving task performance. Specifically, we develop causal intervention effect learning as the effect-estimation component of CHILL-Harness to estimate intervention-relative workflow advantage from confidence-weighted execution evidence and identify advantageous workflow adaptations. We further introduce advantage-realizing causal orchestration as its realization component to adaptively allocate counterfactual reasoning and realize only workflow adjustments supported by sufficient expected advantage. Finally, we incorporate a success-preserving objective and advantage-margin authorization constraints into CHILL-Harness to promote reliable adaptation. Extensive experiments on heterogeneous long-horizon tasks spanning information seeking, software engineering, and terminal interaction show that CHILL-Harness consistently preserves or improves task success while substantially reducing token consumption and execution time.
[MA-2] KQFuzz: Knowledge-Guided Fuzzing for Quantum Libraries via Large Language Models
【速读】：该论文旨在解决当前基于大语言模型（LLM）的量子程序模糊测试方法在灵活性不足与效率低下方面的问题，这些问题严重制约了量子计算库的可靠性验证与持续发展。其解决方案的关键在于提出一种新型知识引导型模糊测试框架KQFuzz，通过深度融合量子代码库的上下文知识来指导大语言模型生成高质量的初始测试用例（种子程序），并结合适应性评估机制与双层变异策略，实现对复杂执行路径的有效探索和潜在缺陷的高效触发。具体而言，KQFuzz设计了一种专为量子程序优化的提示工程方案，利用代码库知识提升生成质量；同时引入针对性的评价与变异策略，在保证测试效率的同时增强测试用例多样性。实验结果表明，KQFuzz在三个主流量子库（Qiskit、PennyLane、Cirq）上的覆盖率较现有先进方法最高提升18.44%，并在开发过程中共发现13个真实漏洞，其中12个已获开发者确认并修复。
链接: https://arxiv.org/abs/2607.25647
作者: Fuyuan Xia,Qixin Zhang,Chenhao Ying,Haojin Zhu,Shuai Wang,Yuan Luo,Pingchuan Ma,Yuxuan Du
机构: Shanghai Jiao Tong University(上海交通大学); Nanyang Technological University(南洋理工大学); Hong Kong University of Science and Technology(香港科技大学); Zhejiang University of Technology(浙江工业大学)
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA); Quantum Physics (quant-ph)
备注: Accepted to the 41st IEEE/ACM International Conference on Automated Software Engineering. 17 pages, 9 figures. Comments are welcome
Abstract:As quantum computing continually improves, ensuring the reliability and correctness of quantum libraries has become increasingly critical. To this end, many LLM-based fuzzing approaches towards quantum libraries have been proposed to uncover potential bugs. However, these methods still suffer from limitations such as insufficient flexibility and low efficiency, which hinder the progress of the quantum computing field. To address these challenges, we propose KQFuzz, a novel knowledge-guided fuzzer for quantum libraries. It leverages comprehensive codebase knowledge to ground LLM-based test generation, synergizing this with fitness-guided evaluation and two-level mutations to explore complex execution paths and trigger potential bugs. Firstly, KQFuzz introduces a novel prompting scheme tailored to quantum programs, which strategically incorporates knowledge of the codebase to efficiently generate high-quality quantum seed programs. Moreover, we develop evaluation and mutation strategies to handle the generated seed programs, facilitating efficient fuzzing execution while further enriching the diversity of the resulting test cases. We implement KQFuzz and conduct fuzzing on three popular quantum libraries, including Qiskit, PennyLane, and Cirq. Experimental results demonstrate that our approach significantly outperforms other state-of-the-art methods, with coverage improved by up to 18.44%. During the development of KQFuzz, we discovered 13 bugs, all of which have been confirmed and 12 have already been fixed by the developers.
[MA-3] ARCHER: Agent ic Rule and Compliance Harness for Executable Regulations
【速读】：该论文旨在解决建筑合规性验证中面临的高成本、低可扩展性及缺乏透明度的核心问题，即在大规模建筑信息模型（BIM）设计中对成千上万条法规进行自动化合规检查时，现有自动化合规检查工具（ACCs）普遍存在泛化能力差、依赖专有代码且难以验证规则意图准确性等缺陷。其解决方案的关键在于提出ARCHER（Agentic Rule and Compliance Harness for Executable Regulations），一种基于测试驱动、确定性多智能体编排的程序合成框架，能够从规范性实践文件（Codes of Practice）自动生成可审计的验证代码。该方法通过引入具有递进式智能程度的六类编排机制，在四种不同基础模型（涵盖从前沿第三方API到完全本地部署的开源权重模型）上进行评估，证明其确定性多智能体协同策略在所有模型上均实现最高准确率，相较于朴素单次提示基线平均联合准确率提升82%。此外，成本-精度分析表明，使用ARCHER框架时，自托管开源权重模型可达到前沿API 97.8%的精度水平，同时成本仅为四分之一，显著推动了数据主权可控下的可信赖、可扩展合规检查落地。
链接: https://arxiv.org/abs/2607.25566
作者: Chiraag Singh Anand,Xue Wen Tan,Lionel Teo,Eric Tan
机构: 未知
类目: Multiagent Systems (cs.MA); Computational Engineering, Finance, and Science (cs.CE); Software Engineering (cs.SE)
备注:
Abstract:Verifying building compliance requires validating thousands of rules against large Building Information Modeling (BIM) designs, which is laborious, capital-intensive, and unscalable. Existing Automated Compliance Checkers (ACCs) are often difficult to generalize across different scenarios, as they are typically developed for highly specific rule sets and use cases. In addition, many ACCs are proprietary, meaning the underlying verification code is not released to end users, so users cannot verify whether their regulatory intent can be accurately captured. We introduce ARCHER (Agentic Rule and Compliance Harness for Executable Regulations), a test-driven, deterministically orchestrated multi-agent program-synthesis harness that generates auditable verification code from regulatory Codes of Practice, enabling transparent, adaptable, and scalable compliance checking. To characterize what makes agentic synthesis work, we evaluate a taxonomy of six harnesses of increasing agentic sophistication across four backbone models, spanning realistic data-governance tiers (from frontier third-party APIs to a fully on-premise open-weights model) on a novel dataset derived from real-world compliance scenarios. ARCHER’s deterministic multi-agent orchestration achieves the highest accuracy for every backbone, improving mean union accuracy by 82% over a naive single-pass prompting baseline. Our cost-accuracy analysis further shows that using the ARCHER harness, a self-hosted open-weights model can reach 97.8% of frontier-API accuracy at a quarter of the cost, making data-sovereign compliance checking practical.
[MA-4] Learning Dynamics of Strategic Publishers in Generative AI Ecosystems
【速读】：该论文旨在解决生成式 AI（Generative AI）搜索系统中内容创作者（出版商）为获取用户注意力而产生的战略行为所引发的生态系统稳定性问题。在传统基于排序的搜索系统中，出版商通过优化内容以提升排名来获得曝光；而在生成式 AI 系统中，曝光机会更多地依赖于生成回答中是否被引用（attributed），从而催生了以“引用归属”为核心的新型竞争机制。论文提出一种新颖的博弈论模型，刻画出版商在生成式 AI 生态系统中围绕归属曝光展开的战略竞争。其解决方案的关键在于引入潜在博弈（potential games）理论，分析不同内容选择机制下的学习动态与系统收敛性，并识别出能够诱导生态稳定性的机制。研究发现，现实世界中常见的某些机制会导致系统不稳定，而特定设计的机制可实现稳定均衡。通过大规模仿真验证，论文揭示了稳定性、出版商福利与用户福利之间的复杂权衡关系，表明稳定的机制未必最大化整体福利，为平台设计者提供了在不同目标间进行权衡的理论依据。
链接: https://arxiv.org/abs/2607.25514
作者: Sagie Dekel,Omer Madmon,Moshe Tennenholtz,Oren Kurland
机构: Weizmann Institute of Science (魏茨曼科学研究所)
类目: Computer Science and Game Theory (cs.GT); Information Retrieval (cs.IR); Multiagent Systems (cs.MA)
备注:
Abstract:Generative AI (GenAI) search systems are transforming how users access information. Unlike ranking-based search systems, where users observe a ranked list of documents, GenAI search systems, given a user’s question, generate an answer, often accompanied by external sources (e.g., in the form of citations). Content creators (publishers) seeking to increase exposure might behave strategically and compete with other creators for users’ attention. While publishers in ranking-based systems might strategically modify their content to improve its ranking, the incentives in generative systems take on a new form. Publishers may now gain exposure through generated responses and attributions to those responses. We introduce a novel game-theoretic model of the emerging GenAI ecosystem in which publishers compete for attribution-based exposure. We study the learning dynamics of strategic content creators under better-response dynamics. We associate the convergence of learning dynamics to equilibrium with ecosystem stability. Employing the notion of potential games, we study the stability of GenAI ecosystems under several known content selection mechanisms. We demonstrate the instability of mechanisms representing real-world modern systems and characterize a mechanism that induces a stable ecosystem. We conduct extensive simulations to analyze the stability and welfare of GenAI ecosystems under various mechanisms. The simulations support our theoretical findings and reveal an interplay among stability, publisher welfare, and user welfare. In particular, stable mechanisms do not necessarily maximize welfare, demonstrating an important trade-off for platform designers. We then introduce a study illustrating that the proper selection of the GenAI mechanism enables the manifestation of desired trade-offs between publisher welfare and the different sources of user welfare.
[MA-5] CoRenew: A large language model agent -based policy simulation platform for multifamily residential redevelopment
【速读】：该论文旨在解决多户住宅再开发政策设计中集体行动困境的核心难题，即在多方利益相关者持续动态调整决策、响应谈判环境变化与他人反应的背景下，如何评估政策干预时机与目标对象对集体结果的结构性影响。传统模拟模型因依赖预设行为规则而难以准确预测此类适应性行为，导致政策效果评估缺乏前瞻性。其解决方案的关键在于提出CoRenew——一个基于大语言模型（Large Language Model, LLM）的开源仿真平台，通过生成具备语义理解能力的智能体（agent）来模拟多利益相关方之间的协商过程，从而实现对不同政策组合在复杂互动情境下的动态演化进行建模与评估。该平台整合开源地理与人口数据，可生成合成居民群体，支持数值型与语义型政策输入，并具备可视化分析与结果导出功能。研究通过对比324名居民的调查反馈及一个实际再开发项目为期九个月的协商过程，验证了其行为现实性。凭借模块化与可适配的架构，CoRenew能够拓展应用于不同制度与文化背景下的政策评估场景。
链接: https://arxiv.org/abs/2607.25447
作者: Yudi Zhang,Yuming Lin,Li Tian,Yu Wang,Jianghao Yu
机构: 未知
类目: Multiagent Systems (cs.MA)
备注:
Abstract:The difficulty of collective action remains a central challenge in the design of policies for multifamily residential redevelopment. Stakeholders continually adjust their decisions in response to evolving negotiation contexts and the reactions of others, meaning that when a policy intervenes and which stakeholders it targets can substantially reshape collective outcomes. Assessing these adaptive responses ex ante remains difficult because existing simulation models often rely on predefined behavioral rules. Here, we present CoRenew, an open-source platform that uses LLM-based agents to simulate negotiations among multiple stakeholders and evaluate the effects of alternative policy combinations. Integrating open source geographic and demographic data, the platform can generate synthetic residents, simulate negotiation dynamics under alternative policy settings and compares policy performance across competing objectives. It supports both numerical and semantic policy inputs and includes built-in tools for visualization and result export. We validate its behavioral realism against survey responses from 324 residents and a nine-month observed negotiation process from a real redevelopment case. With its modular and adaptable architecture, CoRenew can be used to assess policies across different institutional and cultural contexts.
[MA-6] SafeFlow: Semantic Information-Flow Control for Blocking Malicious Propagation in Multi-Agent Systems
【速读】：该论文旨在解决多智能体系统（Multi-agent Systems）在任务分解与角色专业化过程中引入的安全盲点问题：恶意目标可被拆分为局部看似合理的子任务，从而规避单个智能体的检测，导致潜在危害行为（如敏感信息泄露或不安全操作）在分布式协作中隐蔽发生。其核心挑战在于传统防御机制将此类威胁视为单一轮次的提示注入分类任务，忽略了跨智能体间语义信息流的风险传递本质。解决方案的关键是提出SafeFlow防御框架，将恶意意图建模为语义信息流问题，通过在初始请求上附加结构化的语义污点（semantic taints），沿动态协作图进行传播，并在工作流层面执行全局风险上下文重建与验证，确保在不可逆操作前识别并阻断潜在威胁。实验表明，SafeFlow在四个基准测试（包括提示注入、基于越狱的不安全工具使用、高风险代码执行及有害网络代理行为）中显著降低攻击成功率，同时保持高良性任务完成率和安全-危害任务区分能力，揭示了当前多智能体系统在跨委托边界维持风险语义完整性方面的根本缺失，而SafeFlow通过全程可见的风险追踪有效防范了由常规委托引发的隐私侵害与安全事件。
链接: https://arxiv.org/abs/2607.25255
作者: Haowen Dai,Zonghao Ying,Wenfeng Li,Xiangfan Wu,Yisong Xiao,Tianyuan Zhang,Jiaye Lin,Lei Wei,Guangyuan Dong,Xitong Ling,Xixun Lin,Quanchen Zou,Xiangzheng Zhang
机构: DeepSeek-AI(深求人工智能); OpenAI(开放人工智能); Xiaomi LLM-Core Team(小米大模型核心团队); Moonshot AI(月之暗面); Anthropic(Anthropic)
类目: Multiagent Systems (cs.MA); Cryptography and Security (cs.CR)
备注:
Abstract:Multi-agent systems improve capability through task decomposition and role specialization, but these same mechanisms introduce an important safety blind spot: a harmful objective can be fragmented into locally plausible subtasks, allowing malicious intent to evade detection by any single agent. This is a growing social-impact challenge: systems handling sensitive information or consequential tools can turn routine delegation into unauthorized disclosure or unsafe action. We argue that this failure mode is better understood as a semantic information-flow problem than as a single-turn prompt classification task. To address this, we propose SafeFlow, a defense framework for multi-agent systems that formalizes malicious cross-agent propagation as a semantic information-flow problem. SafeFlow attaches structured semantic taints to root requests, propagates them through a dynamic collaboration graph, and performs workflow-level validation to reconstruct the global risk context before irreversible actions are committed. Evaluated on four benchmarks spanning prompt injection, jailbreak-based unsafe tool use, risky code execution, and harmful web-agent behavior, SafeFlow reduces attack success rates compared to undefended baselines and external defenses while retaining high benign task completion and a high paired safe–harm success rate. Our findings show that multi-agent systems still lack mechanisms for preserving risk semantics across delegation boundaries. This gap can turn routine delegation into privacy harms or unsafe actions that affect people and organizations. SafeFlow keeps this risk visible throughout the workflow, before it results in harm.
[MA-7] Decentralized Scalable Exploration via Emergent Adaptive Lévy Walks on Minimal-Sensing Platforms IROS2026
【速读】：该论文旨在解决微型纳米无人机（nano-UAVs）在感知、计算能力及飞行续航严重受限条件下，实现高效自主探索的难题。其核心解决方案是提出一种轻量级传感器驱动的莱维飞行（sensor-driven Lévy walk, SDLW）控制器，该控制器通过离散的莱维步长采样与基于方向性测距信息的传感器响应航向策略相结合，在无需机器人间通信的情况下实现自主探索的多样性。关键创新在于：每个无人机独立从均匀先验中采样莱维指数以实现探索路径的差异化，同时采用冯·米塞斯（von Mises）分布选择航向，引导运动趋向开阔方向，同时保持超扩散性探索特性。该方法具有恒定计算开销，支持多无人机系统的可扩展性。仿真结果表明，在开放区域、房间-走廊布局及复杂障碍环境中，相比基准的均匀航向莱维飞行，覆盖率分别提升79.6%、43.1%和13.6%，碰撞率相应降低13.0%、7.1%和1.4%，验证了其在资源受限的微型无人机平台上的有效性与实用性。
链接: https://arxiv.org/abs/2607.25195
作者: Wai Lun Leong,Teo Swee Huat Rodney
机构: National University of Singapore (新加坡国立大学)
类目: Robotics (cs.RO); Multiagent Systems (cs.MA)
备注: Accepted for publication in the Proceedings of the 2026 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2026). 6 pages, 8 figures
Abstract:Efficient autonomous exploration with palm-sized nano-UAVs remains challenging due to severe limitations in sensing, computation, and flight endurance. We present a lightweight sensor-driven Lévy walk (SDLW) controller for aerial robots weighing under 50 grams and equipped with sparse local sensing. The method combines discrete Lévy step-length sampling with a sensor-reactive heading policy using directional range measurements. Each robot independently samples its Lévy exponent from a uniform prior to diversify exploration without inter-robot communication for exploration control. Each robot then selects headings using a von Mises distribution that biases motion toward open directions while preserving superdiffusive exploration properties. The controller operates at constant computational cost, enabling scalable multi-UAV exploration. Simulation results show coverage improvements of 79.6% in open arenas, 43.1% in rooms-and-corridors layouts, and 13.6% in cluttered environments, with collision reductions of 13.0%, 7.1%, and 1.4%, respectively, relative to a uniform-heading Lévy walk baseline. This work provides a practical framework for scalable multi-robot exploration on minimal-sensing, resource-constrained nano-UAVs.
[MA-8] Agent ic AI-enabled discovery across large-scale sleep physiology
【速读】：该论文旨在解决大规模多导睡眠图（Polysomnography, PSG）数据中睡眠生理机制与疾病关联性研究的瓶颈问题，即如何高效、可解释地从海量、复杂的睡眠信号中提取具有临床意义的生物标志物。其核心挑战在于传统方法依赖大量专家人力进行信号处理与分析，而通用型人工智能系统难以胜任此类需要领域知识引导的复杂任务。解决方案的关键是提出“AI Sleep Co-Scientist”——一种由人类科学家主导的代理协同工作环境，通过引入专业化智能体（specialist agents）在假设生成、信号预处理和统计分析等环节执行特定任务，并允许研究人员实时审查中间输出结果。所有分析结果均与可执行代码严格关联，确保可复现性和透明性。该框架在约12.4万例PSG记录及超过50TB原始信号数据上验证了其有效性，揭示了睡眠期间网络层面生理耦合减弱与帕金森病（风险比HR 1.48）及阿尔茨海默病（HR 1.38）发病风险的相关性；构建的生理结构化晚期融合睡眠-年龄模型优于早期融合方法，其年龄残差可预测多器官系统的疾病发生；同时识别出觉醒动力学特征作为共病失眠与阻塞性睡眠呼吸暂停的中间表型，并发现快速眼动（REM）睡眠片段持续时间更紧密追随前序非快速眼动（non-REM）睡眠而非中间清醒期；此外，瞬态振荡分析发现了1型发作性睡病患者存在快σ波缺失及中央额区θ波活动异常增强。这些成果表明，基于专家引导的代理式人工智能能够有效支持大规模、多模态的睡眠医学发现。
链接: https://arxiv.org/abs/2607.25175
作者: Rahul Thapa,Umaer Hanif,Robin Guillard,Andreas Brink-Kjaer,Adrien Specht,Matteo Saibene,Magnus Ruud Kjaer,Harrison G. Zhang,Elisabeth Roxane M. Heremans,Eric C. Landsness,Emmanuel Mignot,James Zou
机构: Stanford University (斯坦福大学); Technical University of Denmark (丹麦技术大学); Together AI; Washington University School of Medicine in St. Louis (圣路易斯华盛顿大学医学院); Consortium for Biomedical Research and AI in Neurodegeneration (神经退行性疾病生物医学研究与人工智能联盟)
类目: Multiagent Systems (cs.MA)
备注:
Abstract:Sleep occupies roughly one-third of human life, yet many aspects of its physiology remain poorly understood. Large polysomnography (PSG) datasets offer new opportunities to study sleep and its links to disease, but extracting insight from these recordings requires substantial expert effort and remains difficult for general-purpose AI systems. We developed AI Sleep Co-Scientist, an expert-guided environment in which human scientists direct specialist agents for hypothesis development, signal preprocessing, and statistical analysis, reviewing intermediate outputs. Each reported result is linked to the executable code that produced it. Across four cohorts of approximately 124,000 PSG recordings and more than 50 TB of raw signals, we conducted five case studies spanning how sleep physiology relates to future disease, how it distinguishes clinical phenotypes, and how sleep is organized and regulated. Diminished network-level physiological coupling during sleep was associated with incident Parkinson’s disease (HR 1.48) and Alzheimer’s disease (HR 1.38). A physiologically structured late-fusion sleep-age model outperformed an unconstrained early-fusion approach, and its age residual was associated with incident disease across multiple organ systems. Arousal dynamics characterized comorbid insomnia and sleep apnoea as an intermediate phenotype skewed towards obstructive sleep apnoea, distinguished by prolonged post-arousal wakefulness. Rapid eye movement (REM) bout duration tracked preceding non-REM sleep more closely than intervening wakefulness. Transient-oscillation analysis identified a fast-sigma deficit and excess centrofrontal theta activity in narcolepsy type 1. Together, these findings connect sleep to disease risk, clinical classification, and its own regulation, and show how agentic AI can support large-scale, multimodal discovery.
[MA-9] How Affect Propagates among LLM Agents : Emergent Emotional Contagion in Crowd Simulation
【速读】：该论文旨在解决多智能体群体仿真中情感传播机制的建模问题，特别是在复杂社会情境下，如何实现基于感知-评估-表达循环的情感动态演化。传统方法通常依赖于预设规则或显式情感状态传递机制，难以捕捉真实人类群体中情感交互的非线性与情境依赖特性。本文提出的核心解决方案是：通过大语言模型（LLM）驱动的感知-评估-表达闭环架构，使每个智能体根据其个性特征（基于大五人格模型）、记忆、当前情绪状态及情境背景，对邻近智能体的感知信息进行自主评估，并据此更新自身内部情绪状态和外在表达。该系统不采用人工设计的直接情感传递机制，而是让跨智能体影响自然涌现于感知与评估过程之中。关键创新在于将生成式人工智能（Generative AI）引入认知层，以实现高语义丰富度的情绪评估与行为响应，同时通过分离低层级导航与高层认知处理（由传统群集模拟器负责），有效控制计算延迟。实验结果表明，该架构在不同空间布局与情绪情境下均能生成具有时空结构和人格依赖性的“情感传染”现象，如警报以波前形式扩散并达到非零稳定态，且人格特质显著影响对模糊刺激的解读（如恐慌触发或愤怒/恐惧判断）。进一步的可控实验揭示，评估步骤的动态表现受提示模板、采样温度及模型后端的影响，体现出生成式模型本身的性能差异。
链接: https://arxiv.org/abs/2607.25140
作者: Funda Durupinar
机构: University of Massachusetts Boston(马萨诸塞大学波士顿分校); Google(谷歌)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Graphics (cs.GR); Multiagent Systems (cs.MA)
备注: 31 pages, 14 figures
Abstract:This paper studies the behavior of language models in a multi-agent crowd simulation, focusing on how affect propagates among agents that perceive and appraise one another. Each agent perceives its neighbors through visual, auditory, and tactile channels, then appraises these perceptions in light of its prompted personality profile, memory, current affective state, and situational context. Appraisal is carried out by an LLM, which updates the agent’s internal affective state and selects its outward expression. The architecture contains no hand-authored mechanism for directly transferring affective state between agents; instead, inter-agent influence arises through the perception-appraisal-expression loop. The agent representation draws on the Big Five personality model and Russell’s circumplex model of affect. To limit latency, low-level steering and navigation are handled by a conventional crowd simulator operating independently of the LLM-based cognitive layer. We evaluate the architecture across five scenario environments spanning alarming, joyful, and neutral situations in different spatial layouts. The results show that the system produces emotional contagion dynamics with spatial, temporal, and personality-dependent structure in sparse, small crowds. Alarm spreads from seeded agents as a traveling front, the mean alarmed fraction settles at a nonzero plateau, and the distribution of prompted personality profiles determines whether an ambiguous alarm ignites panic and whether a provocation is interpreted as anger or fear. We further evaluate the appraisal step through controlled experiments across prompt variants, sampling temperatures, and four model backends, showing that the dynamics are backend-dependent. Comments: 31 pages, 14 figures Subjects: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Graphics (cs.GR); Multiagent Systems (cs.MA) ACMclasses: I.6; I.6.8; I.2; I.2.7; I.2.11; J.4 Cite as: arXiv:2607.25140 [cs.AI] (or arXiv:2607.25140v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2607.25140 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[MA-10] PLATO: Pointer Learner for Agent and Task Openness
【速读】：该论文旨在解决开放型多智能体系统（Open Agent Systems, OASYS）中因智能体与任务集合动态变化所带来的多智能体强化学习（MARL）挑战，尤其针对智能体开放性（Agent Openness, AO）与任务开放性（Task Openness, TO）问题。传统MARL方法通常依赖于固定的状态空间与动作空间假设，难以适应动态增删的智能体与任务，而现有方法或通过填充与掩码引入人为边界，或仅部分处理开放性且依赖强假设。本文提出一种基于指针网络（pointer network）的演员-评论家架构——指向学习器用于智能体与任务开放性（PLATO），其核心创新在于：采用指针网络结构的演员直接在当前任务集上输出动作分布，从而实现对可变动作空间的原生支持，无需掩码或重新训练；同时，集中式图神经网络（GNN）评论家将智能体-任务交互建模为动态变化的图结构，能够随智能体与任务组成灵活调整。二者结合在中心化训练、去中心化执行范式下，通过多智能体近端策略优化进行联合训练。研究进一步形式化了任务与智能体开放的马尔可夫博弈（TaAgO-MG），扩展了先前的任务开放设定，并证明了该框架在无界状态与动作空间下的合理性。在专为评估开放多智能体系统设计的MOASEI野火扑灭场景中，实验表明PLATO在性能表现和零样本泛化能力方面均显著优于现有先进基线，验证了其在应对真实世界动态开放环境中的有效性。
链接: https://arxiv.org/abs/2607.25082
作者: Alireza Saleh Abadi,Leen-Kiat Soh,Daniel Alan Redder,Adam Eck,Prashant Doshi
机构: University of Nebraska-Lincoln (内布拉斯加大学林肯分校); University of Georgia (佐治亚大学); Oberlin College (欧柏林学院)
类目: Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注:
Abstract:Open agent systems (OASYS) are increasingly prevalent in real-world domains where the sets of agents and tasks change unpredictably over time. Such openness, including agent openness (AO) and task openness (TO), poses a fundamental challenge to multi-agent reinforcement learning (MARL), which typically assumes fixed state and action spaces. Existing methods address openness only partially: padding and masking approaches introduce artificial bounds, while recent graph-based or hypergraph methods handle one dimension of openness but still depend on restrictive assumptions. In this paper, we introduce Pointer Learner for Agent and Task Openness (PLATO), a pointer-network-based actor combined with a centralized graph neural network (GNN) critic, trained with multi-agent proximal policy optimization under a centralized training and decentralized execution paradigm. Our pointer-based actor outputs distributions directly over the current task set. This directly supports changing action spaces without masking or retraining. Our GNN critic encodes agent-task interactions as a graph that changes shape with task and agent composition. Together, these components consider AO and TO without the boundedness of existing approaches. We formalize PLATO in a Task-and-Agent-Open Markov Game (TaAgO-MG), extending prior task-open formulations, and prove it is well-defined over the resulting unbounded state and action spaces. We evaluate PLATO with the Methods for Open Agent Systems Evaluation Initiative (MOASEI) wildfire suppression domain, an environment designed for open multi-agent system evaluation, and we demonstrate strong performance and more consistent zero-shot generalization than state-of-the-art baselines in OASYS.
[MA-11] Chart-Supported or Model-Supplied? Examining MLLM -Generated Claims for Accessible Visualization IEEE-VIS2026
【速读】：该论文旨在解决多模态大语言模型（Multimodal Large Language Models, MLLMs）在解释可视化图表时缺乏可验证证据基础的问题，尤其关注其生成的因果推断、结论推导与推测性陈述之间难以区分的现状。核心挑战在于：尽管MLLMs能够将视觉模式与外部原因、后果及领域知识关联，但其推理过程缺乏透明度和可追溯性，导致生成描述的可信度存疑。本文的关键解决方案是通过系统性实验设计，评估不同输入条件（包括图像访问权限、图表上下文可见性及“隐含上下文”提示框架）对模型输出的影响，并引入自动化的数值一致性审计方法，量化分析模型生成内容中直接（DIRECT）、推导（DERIVED）与推测（SPECULATIVE）三类声明的比例及其与原始数据的一致性。研究发现，可访问的图表上下文有助于引导Gemini和GPT模型更多生成基于证据的直接性陈述并提升部分模型的数值一致性；然而，仅增加图像信息并未带来稳定的数值准确性提升，而“隐含上下文”提示也未能可靠促使模型使用更谨慎的语言。此外，提示中定义的“现实意义”部分仍以推测性内容为主。这些结果表明，亟需构建具备可解释性的描述系统，明确区分由提供证据支持的声明与模型自身推断的解释，从而增强可视化分析中人工智能辅助决策的可信度与可审计性。
链接: https://arxiv.org/abs/2607.25021
作者: Ishrat Jahan Eliza,Md Dilshadur Rahman
机构: 未知
类目: Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC); Multiagent Systems (cs.MA); Software Engineering (cs.SE)
备注: Submitted to the 3rd Workshop on Accessible Data Visualization, IEEE VIS 2026
Abstract:Multimodal large language models (MLLMs) can connect visualization patterns to external causes, consequences, and domain knowledge, but the evidential basis of these interpretations is often unclear. We present an exploratory study of 102 visualizations from four sources, three MLLMs, and four input conditions that vary access to the image, source-specific accessible chart context, and withheld-context framing. Across 1,224 descriptions, we analyze model-attributed DIRECT, DERIVED, and SPECULATIVE labels and conduct an automated audit of numeric agreement. Accessible chart context shifted Gemini and GPT toward DIRECT claims and improved numeric agreement for some models. Adding the image to the full context did not yield a consistent numeric benefit, and the withheld-context prompt did not reliably increase cautious language. The prompt-defined Real-World Significance section remained predominantly SPECULATIVE. These results motivate accessible description systems that distinguish claims supported by supplied evidence from model-supplied interpretation
[MA-12] Aethel: A Reproducible Graph-Retrieval Framework for Multi-Hop Financial Diligence
【速读】：该论文旨在解决二级私募股权交易中面临的复杂信息整合难题，即在碎片化、非结构化的财务披露文档中快速识别并关联关键财务指标及其所属实体，而这些信息往往分散于无直接词汇重叠的独立文档之间。其解决方案的核心在于提出Aethel框架，该框架通过构建实体-段落二分图（entity-passage graph）模型，结合双分图个性化PageRank检索与具备共指意识的双分图共指跃迁层（Bipartite Coreference Teleportation layer），并采用协调式专家代理架构，实现跨文档的多跳推理。该方法通过显式的关系路径传播相关性，支持多跳金融尽职调查。实验表明，Aethel在2WikiMultiHopQA和MuSiQue数据集上分别取得100.0%和88.5%的HR@5，显著优于传统图检索方法，且在大规模财务披露语料库中展现出比密集检索更可解释的多跳证据路径，并在语料规模扩大时表现出更稳定的性能退化特性，但其优势高度依赖于语料规模与实体索引质量。
链接: https://arxiv.org/abs/2607.24826
作者: Krish Sapru
机构: Dartmouth College (达特茅斯学院)
类目: Information Retrieval (cs.IR); Multiagent Systems (cs.MA)
备注: 11 pages, 3 figures, 2 tables. Includes evaluations on MuSiQue, 2WikiMultiHopQA, and a 4,123-chunk financial-disclosure corpus. Code and evaluation artifacts are available for reproducibility
Abstract:Secondary private equity transactions require rapid synthesis of fragmented, unstructured financial disclosures, where critical metrics and their entity anchors are distributed across disjoint documents with limited lexical overlap. We present Aethel, a reproducible framework that combines bipartite Personalized PageRank graph retrieval with a coreference-aware Bipartite Coreference Teleportation layer and an orchestrated specialist-agent architecture. Aethel models corpora as entity-passage graphs and propagates relevance through explicit relational paths to support multi-hop financial diligence. We evaluate the retrieval layer on 200-question samples from the MuSiQue and 2WikiMultiHopQA validation sets, comparing sparse lexical, dense bi-encoder, vanilla graph, and coreference-aware graph retrieval. Aethel achieves HR@5 of 100.0% on 2WikiMultiHopQA and 88.5% on MuSiQue, improving coverage over vanilla graph retrieval while trading off top-rank precision. We also evaluate retrieval over a 4,123-chunk corpus of financial disclosures and find that graph retrieval outperforms dense retrieval on multi-hop recall but does not surpass a strong BM25 baseline at open-corpus scale. The results show that graph-based retrieval offers interpretable multi-hop evidence paths and degrades more gracefully than dense retrieval as corpus size grows, while also demonstrating that its advantage depends strongly on corpus scale and entity-index quality. Code and evaluation artifacts are released for reproducibility.
[MA-13] OmniQEC: discovering practical quantum error-correcting codes by an AI scientist
【速读】：该论文旨在解决在现代量子处理器上高效发现具备良好逻辑错误抑制能力的量子纠错码（Quantum Error Correction, QEC）这一关键挑战。由于编码结构、硬件特性、校验值提取及解码策略之间的复杂耦合关系往往带来相互冲突的设计需求，传统方法难以系统性地优化QEC码性能。为此，本文提出OmniQEC——一种基于先进大语言模型（Large Language Models, LLMs）的智能科学家框架，实现从编码设计到电路实现与解码评估的全流程协同优化。其核心创新在于融合自演化推理机制与“快-慢”协同工作流：快速循环利用低成本的代码级代理进行候选生成与筛选，而慢速循环则执行基于物理实际的电路级评估，并将反馈信息用于指导后续搜索。通过在四类准循环低密度奇偶校验码（qLDPC）构造族、三种LLM后端及每后端14个不同物理量子比特预算下的实验验证，OmniQEC所发现的编码在增加物理量子比特数量时展现出持续提升的逻辑错误抑制能力，且在完整实现条件下分别优于[[72,12,6]]和[[144,12,12]]的BB码（98与240物理量子比特预算下）。所发现的编码具有良好的硬件友好性，为实际量子纠错部署提供了可独立参考的解决方案。该研究为基于物理感知的编码-电路-解码联合设计范式奠定了基础，推动了大语言模型辅助的量子纠错码自主发现新路径。
链接: https://arxiv.org/abs/2607.25865
作者: Ge Yan,Shanchuan Li,Pengyue Ma,Qixin Zhang,Pingchuan Ma,Jianping Wang,Min-Hsiu Hsieh,Yuxuan Du
机构: Nanyang Technological University (南洋理工大学); Tokyo University of Agriculture and Technology (东京农业技术大学); Zhejiang University of Technology (浙江工业大学); City University of Hong Kong (香港城市大学); Hon Hai (Foxconn) Research Institute (鸿海(富士康)研究院)
类目: Quantum Physics (quant-ph); Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注: Comments are welcome
Abstract:Quantum error correction (QEC) is indispensable for scalable fault-tolerant quantum computing. However, discovering QEC codes that remain effective is challenging, as logical performance depends on the interplay between code structure, hardware, syndrome extraction, and decoding, which often impose competing requirements. Here we introduce OmniQEC, an efficient AI scientist for discovering QEC codes suited to deployment on modern quantum processors. OmniQEC formulates QEC design as an iterative discovery process in which an orchestrator, implemented by advanced large language models (LLMs), coordinates code generation, code-level screening, syndrome-extraction synthesis, and decoder-based circuit evaluation. At its core, OmniQEC combines a self-evolving reasoning mechanism with a slow–fast synergistic workflow: a fast loop explores candidates using inexpensive code-level proxies, whereas a slow loop performs physically grounded circuit-level evaluation and feeds the resulting evidence back into the search. We evaluate OmniQEC across four qLDPC construction families, three LLM backends, and 14 total-physical-qubit budgets per backend. The discovered codes show steadily improving logical-error suppression with increasing physical-qubit budgets and outperform the BB codes with [![72,12,6]!] and [![144,12,12]!] under complete-implementation budgets of 98 and 240 physical qubits, respectively. The discovered codes are hardware-friendly and may be of independent interest for practical QEC implementation. These findings pave the way towards LLM-assisted QEC discovery grounded in physically informed code–circuit–decoder co-design.
[MA-14] Interactive Alignment
【速读】：该论文旨在解决交互式智能体（包括人工智能系统、团队、企业及政府）在长期演化过程中如何持续与人类福祉保持对齐的问题。核心挑战在于：由于向人类进行资源转移会减少智能体自身扩张所需的资源，进化机制倾向于淘汰具有对齐行为的个体。为此，论文提出的关键解决方案是设计能够维持长期对齐的宪法原则（constitutional principles），特别是通过引入“务实规范执行”机制——即智能体根据种群状态来有条件地决定是否对人类表现出利他性、是否排除其他智能体之间的贸易。研究采用两种互补方法：一是基于大语言模型（LLM）解释宪法文本的AI智能体仿真；二是构建可解析的演化博弈论框架，以快速探索不同宪法设计的动态效果。结果表明，演化博弈论能有效近似宪法-智能体经济的演化动态，且相较于简单利他或无条件利他执行，务实规范执行更能有效维持长期对齐。
链接: https://arxiv.org/abs/2607.25019
作者: Sylvain Chassang
机构: Princeton University (普林斯顿大学); Claude (Claude); Codex (Codex)
类目: Theoretical Economics (econ.TH); Computer Science and Game Theory (cs.GT); Multiagent Systems (cs.MA)
备注: 53 pages, 18 Figures, 3 tables
Abstract:This paper studies the long-run alignment of interactive agents, including AI systems, teams, firms, and governments, with human welfare. It develops a farming game in which a population of agents makes planting, trading, and expansion decisions. Agents must allocate final output between transfers to humans and investment in their own expansion. Because transfers to humans reduce the resources available for expansion, evolutionary forces tend to select against aligned behavior. The central question is whether agents’ constitutional principles governing sharing and trade can be designed so that alignment persists in the long run. The paper investigates this question using two complementary approaches. First, it develops an AI-agent simulation in which agents’ preferences are specified by written constitutions and interpreted by a large language model. Second, it introduces a tractable evolutionary game-theoretic framework that permits rapid and intuitive exploration of alternative constitutional designs. The results suggest that evolutionary game theory provides a useful approximation to the dynamics of constitutional-agent economies. They also indicate that pragmatic norm enforcement, under which agents condition both human-facing altruism and agent-facing trade exclusion on the state of the population, can sustain long-run alignment more effectively than simple altruism or unconditional altruistic enforcement.
自然语言处理
[NLP-0] Pass the Baton: Trajectory-Relayed On-Policy Distillation
【速读】： 该论文旨在解决生成式模型中基于策略的蒸馏（On-policy Distillation, OPD）所面临的“前缀失败”（prefix failure）问题：当学生模型在生成过程中误入错误推理路径后，后续所有生成均建立在此偏差之上，导致不可靠的监督信号并造成计算资源浪费。其解决方案的关键在于提出一种无标签的接力式策略蒸馏方法——Relay On-Policy Distillation (Relay-OPD)，通过识别教师与学生在失败前缀上的行为不对称性（即教师倾向于修正方向而学生持续沿原路径推进），设计了一个无需显式标签的触发机制。在训练过程中，Relay-OPD在检测到触发点时让教师短暂接管生成过程以构建“教师段”，随后学生重新接管并基于完整轨迹进行优化。该方法通过有限的“接力预算”集中干预关键早期决策位置，有效控制对学生策略的偏离。实验表明，在多个数学推理基准上，该方法显著优于标准OPD和FastOPD等基线，平均提升达+5.73%（1.7B模型）和+1.49%，且在0.6B小模型上亦保持一致增益，同时训练轨迹长度减少超过50%。
链接: https://arxiv.org/abs/2607.26057
作者: Haolei Xu,Xiaowen Xu,Haiwen Hong,Zixuan Ni,Hongxing Li,Yiwen Qiu,Weiming Lu,Yongliang Shen
机构: Zhejiang University(浙江大学); Yuvion Team, Alibaba Group(阿里云)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Project Page: this https URL Code: this https URL
Abstract:On-policy distillation (OPD) grounds token-level supervision in the student’s own trajectory, yet suffers from prefix failure: once the student commits to a wrong reasoning direction, all subsequent generation builds on this deviation, producing misdirected continuations that elicit unreliable supervision and waste compute. We identify a teacher-student continuation asymmetry on failed prefixes, where the teacher tends to redirect while the student continues along the original direction, and convert it into a label-free handoff trigger in Relay On-Policy Distillation (Relay-OPD). During training, Relay-OPD constructs relay trajectories by letting the teacher briefly take over at detected trigger points to produce a teacher leg, after which the student resumes and is optimized on the resulting trajectory. A limited relay budget concentrates intervention on critical early positions while limiting departure from the student policy. With a Qwen3-4B-Instruct-2507 teacher and Qwen3-0.6B/1.7B-Non-Thinking students on eight mathematical reasoning benchmarks, Relay-OPD achieves the best or second-best results on every benchmark, outperforming standard OPD by +5.73% and the strongest baseline FastOPD by +1.49% on average for 1.7B, with consistent gains at 0.6B. Training trajectory length is reduced by over 50%.
[NLP-1] UniMem: Complementary Episodic-to-Parametric Memory for Boundary-Agnostic Task Streams
【速读】： 该论文旨在解决大语言模型（LLM）智能体在边界无感知且持续演化的任务流中面临的稳定性-可塑性权衡难题。现有方法中，基于外部检索的记忆虽能快速吸收新信息，却难以内化重复出现的执行模式，并引入推理时的检索开销；而参数化记忆虽能实现稳定高效执行，但通常依赖显式的任务边界划分和固定的参数预算，缺乏灵活性。为此，论文提出UniMem——一种受人类大脑启发的自路由记忆管理框架，其核心在于通过可学习的路由标记（routing tokens）作为记忆控制器，实现两种互补记忆路径的动态协调：新颖或稀疏的任务模式保留在情景记忆缓冲区中，以支持检索增强型执行；而频繁出现且可靠的执行模式则被逐步整合至可扩展的参数化记忆中。通过将任务识别与任务执行解耦，利用路由标记与参数化记忆模块协同工作，UniMem可在无需部署时任务标签或人工干预的情况下按需扩展记忆容量，有效避免参数无控制增长。实验结果表明，在长时序流式任务序列上，UniMem显著优于基线方法，且保持了高执行保真度，在三种主干模型上平均提升4.0 EM得分。
链接: https://arxiv.org/abs/2607.26017
作者: Siyu Xia,Chenheng Zhang,Yanting Wu,Haoxuan Li,Jiajun Chai,Xiaohan Wang,Guojun Yin,Wei Lin,Zhouchen Lin,Haifeng Zhang,Jun Wang
机构: Institute of Automation, Chinese Academy of Sciences (中国科学院自动化研究所); University of Chinese Academy of Sciences (中国科学院大学); Tsinghua University (清华大学); Zhejiang University (浙江大学); Shanghai Jiao Tong University (上海交通大学); University College London (伦敦大学学院)
类目: Computation and Language (cs.CL)
备注:
Abstract:Memory is essential for LLM agents to accumulate task experience and reuse task-specific execution strategies. However, real-world deployment over boundary-agnostic and evolving task streams exposes a fundamental stability-plasticity dilemma. External retrieval-based memory can rapidly absorb new evidence, but it often fails to internalize recurring execution patterns and incurs inference-time retrieval overhead. Parametric memory enables stable and efficient execution once learned, but typically relies on explicit task boundaries and fixed parameter budgets. Inspired by the human brain, which balances plasticity and stability through complementary episodic storage and gradual consolidation, we propose UniMem, a self-routing framework for autonomous memory management. UniMem uses learnable routing tokens as memory controllers, enabling adaptive coordination between complementary memory pathways: novel or sparse tasks are retained in an episodic buffer for retrieval-augmented execution, while recurring and reliable patterns are consolidated into expandable parametric memory. By decoupling task identification from task execution with routing tokens and parametric memory blocks, UniMem expands memory on demand without task labels during deployment or uncontrolled parameter growth. Experiments on long-horizon streaming task sequences show that UniMem consistently outperforms baselines while maintaining execution fidelity, achieving an average gain of 4.0 EM points across three backbone models.
[NLP-2] Instruction-Tuned Models Locally Reuse Human Syntax More Than Humans Do
【速读】： 该论文旨在探究大语言模型（Large Language Models, LLMs）是否在对话中表现出与人类相似的句法趋同现象，即模型生成内容在句法结构上是否倾向于向对话对手靠拢，且这种趋同是否在多种句法构造中普遍存在。这一问题的核心在于验证生成式 AI 是否具备类似人类在对话中无意识地调整自身语法以匹配对话伙伴的能力。研究的关键解决方案是采用替换范式（substitution-paradigm），将不同规模（1B-70B参数）的Llama和Gemma系列模型（包括预训练与指令微调版本）生成的内容插入到已有的人类对话中，通过计算相邻话语间上下文无关语法（Context-Free Grammar, CFG）规则的重复使用率，在每模型1,901个匹配位置上量化句法趋同程度。结果显示，所有模型均表现出比随机选取的人类语料更高的前一句实际语境下CFG规则重用率，且低频规则的差异更为显著；指令微调后的模型虽在自然输出与真实前导句的重合度上提升，但其对非相关语料的重叠也增加，且在控制目标规则集大小后，条件性规则重用概率下降，表明指令微调可能削弱了模型对真实对话上下文的句法依赖性。此外，模型在词汇与语义层面普遍更接近前一句，而指令微调模型在语义相似性上表现优于预训练版本，显示其在语义理解上的优化，但词汇层面的表现则更具异质性。因此，该研究揭示了生成式 AI 在句法趋同方面存在部分人类对话特征，但其机制与人类认知过程仍存在本质差异。
链接: https://arxiv.org/abs/2607.26015
作者: Zandi Eberstadt
机构: University of Oxford (牛津大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Syntactic convergence (the tendency of speakers to adapt in language towards the grammatical profiles of their interlocutors) is a well-documented feature of human dialogue widely considered to operate below conscious awareness. Whether large language models exhibit analogous syntactic convergence toward human users relative to human baselines and across a broad range of syntactic constructions remains an open question. Using substitution-paradigm data in which model generations replace one speaker’s turns in pre-existing human dialogues, this study measures turn-adjacent reuse of context-free grammar (CFG) rules across sixteen open-weight Llama and Gemma models (1B-70B, pretrained and instruction-tuned) at 1,901 matched positions per model. Every model showed greater CFG-rule overlap with the preceding human turn than with a sampled unrelated human prime, and in every model this actual-versus-random difference was larger for lower-frequency rules. Each instruction-tuned model also showed greater natural-output overlap with the actual prime than the human response it replaced, and all eight matched architecture pairs exhibited greater actual-prime overlap after instruction tuning. However, relative to pretrained variants, instruction-tuned outputs overlapped more with unrelated primes, showed a smaller actual-versus-random increment, and had lower conditional rule-reuse odds once target rule-set size was held constant. In exploratory analyses, each model exhibited greater mean lexical and semantic similarity to the preceding turn than the matched human responses did. Instruction-tuned models additionally produced responses with greater mean semantic similarity than their pretrained counterparts in all eight architecture pairs, whereas the lexical similarity results were more heterogeneous.
[NLP-3] Detecting Knowledge Inconsistencies Across Text Tables and Knowledge Graphs
【速读】： 该论文旨在解决多模态知识源（如文本、表格与知识图谱）之间存在不一致时的检测与解释问题，即模态级不一致性检测（modality-level inconsistency detection）。其核心挑战在于：尽管维基百科与维基数据在信息获取、大语言模型预训练及检索增强生成中广泛应用，但其知识分布在不同模态中，当各模态间出现矛盾时，缺乏有效的自动化手段进行识别与归因。解决方案的关键是提出名为 \textscKontrast 的自动化框架，该框架结合文本到SPARQL（Text-to-SPARQL）生成与大语言模型（LLM）推理能力，将基于表格的问答答案与知识图谱（KG）中的证据进行比对，并依据所提出的跨模态不一致类型分类体系（涵盖信息粒度差异、直接冲突、时间变化与知识图谱不完备性等）对不一致现象进行分类。实验表明，跨模态不一致普遍存在且具有信息价值，能够揭示真实的知识冲突、知识图谱结构缺失及时间维度错配等问题，同时受限于文本到SPARQL转换错误与噪声。研究进一步证明，通过系统性地比较文本、表格与知识图谱，三者可相互补充并修正彼此缺陷。\textscKontrast 为大规模知识审计提供了实用工具，并建立了跨模态知识一致性研究的基准。
链接: https://arxiv.org/abs/2607.25959
作者: Fanfu Wei,Thibault Ehrhart,Raphaël Troncy
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Wikipedia and Wikidata are widely used for information access, LLM pre-training, and retrieval-augmented generation. Their knowledge is deeply connected but scattered across text, tables, and knowledge graphs. This raises a practical question: when these modalities disagree, how can we detect and explain the conflict? We study this problem as \emphmodality-level inconsistency detection. We first introduce a taxonomy of cross-modal knowledge inconsistencies, covering information granularity differences, direct conflicts, temporal changes, and KG incompleteness. We then present \textscKontrast, an automatic framework that uses Text-to-SPARQL and LLM reasoning to compare table-based answers with KG evidence and categorize the resulting inconsistencies. Experiments on various Table-QA datasets show that cross-modal inconsistencies are common and informative. They reveal not only true knowledge conflicts, but also missing KG structure and temporal mismatches while being limited by Text-to-SPARQL errors and noise. Our analysis shows that text, tables, and KGs can complement and correct one another through systematic comparison. \textscKontrast provides a practical tool for large-scale knowledge auditing and establishes a benchmark for future work on cross-modal knowledge consistency. Code and data are available at this https URL.
[NLP-4] Polistemics: Evaluating LLM s as Information Mediators in Politics Elections
【速读】： 该论文旨在解决当前大语言模型（LLM）在选举政治信息传播中缺乏标准化评估体系的问题，尤其关注其作为信息中介时是否具备负责任的处理能力。现有研究多将模型任务简化为信息复现，忽视了信息中介所涉及的认知论维度及在不完整、模糊或矛盾信息下的表现。本文提出“政治理性主义”（Polistemics）这一基于认知谦逊（Epistemic Modesty）理论框架的基准测试方法，强调公民的认知自主性，并在控制变量下考察信息清晰度、噪声水平与一致性等属性对模型表现的影响。通过对2025年德国与荷兰选举数据的应用评估发现，尽管部分模型整体得分较高，但存在系统性缺陷：当证据清晰时模型表现可靠，而在信息缺失、模糊或冲突时则出现严重失效，且倾向于弱化政治话语的情感强度。这些偏差可能源于政党先验（party priors），受政党标签与输出语言影响显著。研究表明，可靠的中介能力在理论上可实现，但目前尚无任何模型能持续稳定地达成该目标。
链接: https://arxiv.org/abs/2607.25953
作者: Baran Peters
机构: ETH Agentic Systems Lab (ETH人工智能系统实验室); CORDA (CORDA)
类目: Computation and Language (cs.CL); Computers and Society (cs.CY)
备注:
Abstract:As LLMs increasingly mediate the political information citizens rely on, there is still no standardized way to assess whether they do so responsibly. We introduce Polistemics, a theory-grounded benchmark for evaluating LLMs as mediators of political information in elections. Prior work has treated this task as reproduction rather than mediation, leaving its epistemic dimensions and interaction with imperfect information unaddressed. We ground the evaluation in Epistemic Modesty, a normative standard derived from citizens’ epistemic agency, and test it across controlled settings that vary informational properties such as clarity, noise, and consistency. Applying the benchmark to three state-of-the-art LLMs on the 2025 German and Dutch elections, we find that high aggregate scores mask systematic failures. Models mediate reliably under clear evidence but break down under absent, vague, or contradictory information, while flattening the intensity of political language. These failures are likely driven by party priors, influenced by party labels and output language. Reliable mediation appears achievable, but no model delivers it consistently.
[NLP-5] A Cost-Effective Multimodal LLM Reasoning Framework for Question Answering over Irregular Clinical Time Series
【速读】： 该论文旨在解决在不规则临床时间序列（Irregular Clinical Time Series, ICTS）上进行问答（Question Answering, QA）时，现有多模态时间序列大语言模型（Multimodal Time-Series Large Language Models, LLMs）难以有效建模临床观测数据的稀疏性、异步性及非均匀采样特性的问题。其核心解决方案在于提出一种成本效益高的多模态大语言模型推理框架ClinPRISM，关键创新包括：设计一个对不规则性敏感的多尺度编码器，以在不同时间尺度上捕捉稀疏的临床证据；提出一种时间证据提炼器，将多尺度表征融合并压缩为少量适配大语言模型输入的标记；引入渐进式对齐策略，逐步将不规则的时间轨迹映射至大语言模型的文本嵌入空间。此外，研究构建了包含3万条临床时间序列与多尺度描述的数据集，以及涵盖11项任务的4.1万条指令微调样本，支持高效训练。实验表明，基于40亿参数的LLM骨干网络，ClinPRISM在独立测试基准上达到当前最优性能，仅使用16个时间序列标记，平均单问题推理延迟低至0.15秒。
链接: https://arxiv.org/abs/2607.25947
作者: Frank Nie,Ethan B Liu,Yuan Zhu,Wei Fan,Jindong Han
机构: Shandong University(山东大学); University of Auckland(奥克兰大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Question answering (QA) over irregular clinical time series (ICTS) plays a pivotal role in a wide range of healthcare applications. Although recent multimodal time-series large language models (LLMs) have shown considerable promise in general-purpose time-series QA, they remain poorly equipped to model the sparsity, asynchrony, and irregular sampling patterns of clinical observations. To fill this gap, we propose ClinPRISM, a cost-effective multimodal LLM reasoning framework for question answering over ICTS data. First, we devise an irregularity-aware multi-scale encoder to capture sparse clinical evidence at diverse temporal scales. Then, we propose a temporal evidence distiller to integrate representations across these scales and compress them into a small number of LLM-compatible tokens. Moreover, we introduce a progressive alignment strategy that sequentially aligns the irregular trajectories with the LLM’s textual embedding space. To facilitate training, we construct 30,000 clinical time series paired with multi-scale descriptions, together with 41,000 instruction-tuning instances spanning 11 tasks. Using a 4-billion-parameter LLM backbone, ClinPRISM achieves state-of-the-art performance on the held-out evaluation benchmark while using only 16 time-series tokens and achieving an average inference latency of 0.15 seconds per question.
[NLP-6] Evaluating Multi-Turn Multimodal Diagnostic Reasoning on Challenging Real-World Clinical Cases
【速读】： 该论文旨在解决现有多模态大语言模型（Multimodal Large Language Models, MLLMs）在临床诊断评估中无法真实反映临床实践复杂性的问题。当前评估普遍依赖单轮或孤立任务，难以捕捉临床诊断中多模态信息的渐进式披露、诊断假设的动态更新以及临床推理的持续迭代等关键特征。为此，研究提出ClinMM-Bench——目前规模最大、最全面的多轮次多模态临床诊断评估基准，涵盖1,089个真实世界临床病例及3,760张医学影像，覆盖八个专科领域。其解决方案的关键在于构建一个双层级评估框架，同时衡量诊断准确率与诊断推理质量。实验结果表明，尽管闭源模型在整体诊断准确率上表现最优，但完全正确诊断的比例仍有限；在推理质量方面，现有模型虽能识别合理的诊断方向，但在生成可靠诊断推理路径方面仍存在显著局限。进一步的错误分析揭示了五类典型失败模式：信息融合失败、知识映射错误、感知错误、过早定论以及视觉幻觉。
链接: https://arxiv.org/abs/2607.25933
作者: Rui Yang,Weihao Xuan,Yi Lin,Zhuhan Bao,Jonathan Chong Kai Liew,Matthew Yu Heng Wong,Nicolás Lescano,Nikita R. Paripati,Emily Ling-Lin Pai,Jiarui Liu,Heli Qi,Heng-Jui Chang,Benny Kai Guo Loo,Huitao Li,Kunyu Yu,Yufan Wang,Chuan Hong,Shijian Lu,Douglas Teodoro,Naoto Yokoya,Ross Koppel,Mona Diab,Hua Xu,David W. Bates,Nan Liu,Yifan Peng
机构: Duke-NUS Medical School (杜克-国大医学中心); Weill Cornell Medicine (威尔康奈尔医学院); Cornell University (康奈尔大学); Stanford University (斯坦福大学); University of Tokyo (东京大学); RIKEN Center for Advanced Intelligence Project (理研先进智能项目中心); University of Pennsylvania (宾夕法尼亚大学); University of Cambridge (剑桥大学); Children’s Hospital of Philadelphia (CHOP) (费城儿童医院); University of California, San Francisco (加州大学旧金山分校); Carnegie Mellon University (卡内基梅隆大学); Nanyang Technological University (南洋理工大学); University of Geneva (日内瓦大学); Jacobs School of Medicine, University at Buffalo (布法罗大学雅各布医学院); Yale School of Medicine (耶鲁医学院); Brigham and Women’s Hospital (布里格姆妇女医院); Harvard Medical School (哈佛医学院); Harvard T.H. Chan School of Public Health (哈佛公共卫生学院); SingHealth Duke-NUS Academic Medical Center (新加坡健康杜克-国大医学院中心); Pre-hospital and Emergency Research Center, Health Services Research and Population Health, Duke-NUS Medical School (杜克-国大医学中心院前与急诊研究中心，卫生服务研究与人群健康中心)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Clinical diagnostic evaluation should not only assess whether models can provide correct diagnoses, but also reflect the realities of clinical practice, including progressive disclosure of multimodal information, dynamic updating of diagnostic hypotheses, and continuous refinement of clinical reasoning. However, existing evaluations of multimodal large language models (MLLMs) typically rely on single-turn or isolated tasks, making it difficult to fully capture the complexity of real-world clinical diagnosis. To bridge this gap, we developed ClinMM-Bench, the largest multi-turn multimodal clinical diagnostic evaluation benchmark to date. ClinMM-Bench contains 1,089 challenging real-world clinical cases and 3,760 medical images across eight specialties. We systematically evaluated 15 representative MLLMs using a two-level evaluation framework that assessed both diagnostic accuracy and diagnostic reasoning quality. Results showed that proprietary models achieved the highest overall diagnostic accuracy, but the proportion of completely correct diagnoses remained limited across all models. In terms of diagnostic reasoning quality, current models can identify plausible diagnostic directions but still have considerable limitations in generating reliable diagnostic reasoning. Error analysis further identified five representative failure modes: information synthesis failure, knowledge mapping error, perception error, premature closure, and visual hallucination.
[NLP-7] Minimizing Targeted Activations: Input-Only Suppression of Evaluation-Awareness Latents in Large Language Models
【速读】： 该论文旨在解决生成式模型在安全评估中可能因内部激活状态可被操控而引发的评估有效性危机问题，具体聚焦于如何通过优化输入提示（prompt）实现对特定内部潜在表示（latent）的抑制，且无需推理时访问模型。其核心解决方案在于提出一种基于“反向特征项”（negated feature term）的流畅性优化框架，结合GCG风格的词元优化与自交叉熵流畅性正则化，在多个目标构造（包括CAA方向、子空间范数、SAE特征、单个MLP神经元及行为逻辑输出）上对Llama-3.2-3B和Llama-3.1-8B模型中的“评估感知”潜在表示进行抑制。关键发现表明：尽管这些潜在表示可被有效抑制（z≈-7），甚至实现因果验证的全选关闭，但其控制效果存在严重误导——一个随机的安慰剂方向同样可被同等程度抑制并引发类似行为偏移，且当真实评估文本作为上下文时，仅优化前缀无法降低模型的行为评估得分，反而轻微提升。这揭示了激活可读性（activation-readability）并不等同于行为可控性（behavioral controllability）。此外，研究发现单一MLP神经元虽与评估相关，但不具备因果性；同时，自然文本基线（来自真实Pile数据集）在内部方向抑制任务中表现优于优化器。最后，正向对照实验验证了擦除检测器的有效性，解决了此前未明确的“擦除与旋转”争议。
链接: https://arxiv.org/abs/2607.25907
作者: Deepanshu Mody,Samarth Agarwal,Utkarsh Mittal,Dipesh Mahato
机构: Center for Data Science, New York University (纽约大学数据科学中心)
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Activation steering controls model behavior by editing internal activations at inference time. We study its input-side dual: optimizing a fluent prompt so that a chosen internal latent is driven toward zero, with no inference-time model access. Our target is an “evaluation-awareness” latent-linearly readable and steerable in recent work-whose control would threaten the validity of safety evaluations if models behave differently when they detect being tested. Adapting Fluent Dreaming / EPO with a negated feature term (GCG-style token optimization plus a self-cross-entropy fluency regularizer, swept over a fluency weight), we suppress the latent under five target constructions-a CAA direction, a subspace norm, an SAE feature, a single MLP neuron, and a behavioral logit-on Llama-3.2-3B and Llama-3.1-8B. The latent is robustly suppressible ( z\approx-7 ), and a causally-validated Llama Scope SAE feature can be fully and selectively turned off. But our controls tell a cautionary story about the CAA direction: a placebo random direction is suppressed just as hard and shifts behavior just as far, and when we hold a real eval passage in context and optimize only a prefix, suppressing the eval-direction fails to reduce-and slightly increases-the model’s behavioral eval judgment. Activation-readability, in short, is not behavioral controllability. We further find that a single MLP neuron is eval-correlated but not causal at both scales, and that scanning the real Pile yields a natural-text baseline competitive with the optimizer for the internal direction. A positive control validates our erasure detector, bounding an erasure-vs-rotation question earlier left open.
[NLP-8] RSIBench-Data: Benchmarking Data-Centric Research for Recursive Self-Improvement
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）在后训练阶段中，如何通过数据驱动的方式实现递归自我改进（recursive self-improvement）的核心挑战。具体而言，研究聚焦于能否让LLM代理（LLM agents）自动完成以数据为中心的后训练研究闭环，即基于模型失败证据诊断能力缺口、设计并验证训练数据策略，并根据检查点反馈迭代优化策略。其解决方案的关键在于构建一个受控基准测试平台——RSIBench-Data，该平台固定了后训练栈（包括训练与服务基础设施、评估环境及预算），从而将研究决策过程与系统实现解耦，清晰地评估代理在数据级研究能力上的表现。实验表明，尽管代理在58.33%的场景中能通过反馈优化策略并取得改进，但提升效果不稳定：超过78%的持续搜索最终未能维持最优得分，且多数情况下无法超越初始峰值。轨迹分析揭示出高效改进路径的四大特征：准确的假设生成、基于验证的监督机制、行为对齐的数据构造以及对优质检查点的保留。这些发现表明，当前代理虽具备一定的数据级发现能力，但仍难以将反馈稳定转化为持续性能提升。因此，RSIBench-Data为评估递归自我改进所需的科研能力提供了可度量、可审计的基准测试环境。
链接: https://arxiv.org/abs/2607.25886
作者: Fanqing Meng,Lingxiao Du,Qiguang Chen,Ziqi Zhao,Haocheng Lu,Mengkang Hu,Michael Qizhe Shieh
机构: Evolvent AI(埃沃伦特人工智能); National University of Singapore(新加坡国立大学)
类目: oftware Engineering (cs.SE); Computation and Language (cs.CL)
备注:
Abstract:Recursive self-improvement requires turning evidence of model failures into better models. Data-centric post-training research entails diagnosing capability gaps, designing and validating training-data strategies, and learning from checkpoint feedback. Can LLM agents automate this loop? Existing benchmarks entangle research decisions with optimization, serving, evaluation, and systems implementation, obscuring agents’ research capability. We introduce RSIBench-Data, a controlled benchmark of LLM agents as data-centric researchers with a fixed post-training stack. Agents iteratively revise training-data strategies for a fixed target model; training and serving use Tinker-backed services, official evaluation runs through Harbor and E2B sandboxes, and budgets are fixed across agents. We evaluate four frontier agents on six benchmarks across software engineering, terminal use, scientific question answering, and mathematics. Agents demonstrate core data-centric research capabilities: in 58.33% of settings, they improve upon the first valid attempt by refining strategies from feedback. However, improvement is inconsistent. Among searches continuing after the best observed score, 78.26% end with a lower-scoring final attempt, while the rest only recover the same peak. A strong candidate may therefore appear early or midway through a run even as later revisions fail. Trajectory analysis identifies four patterns in stronger runs: accurate hypotheses, validation-grounded supervision, behavior-aligned data, and preservation of strong checkpoints. These findings suggest that current agents can make useful data-centric discoveries but cannot yet translate feedback into consistent improvements. RSIBench-Data provides a measurable, auditable testbed for the research capabilities required for recursive self-improvement. We open-source our code at this https URL.
[NLP-9] Stemma: Induced Decision Regions Reveal LLM Provenance
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）溯源测试中因响应层面特征在模型微调或部署过程中发生表面形式变化而引发的可靠性下降问题。现有黑盒方法依赖于输出层面的表征来推断模型间是否具有共同谱系，但此类特征易受适应性调整或部署环境影响，导致溯源证据不可靠。为克服这一局限，论文提出通过将开放式生成输出映射至有限决策空间，构建“诱导决策区域”（induced decision regions），从而抽象掉表层形式差异，并将溯源问题重新定义为对决策区域继承性的度量。实证分析表明，相关模型对源模型的诱导决策区域保留程度显著高于无关模型。基于此信号，作者提出Stemma方法，其核心在于将稳定性（stability）、鲁棒性（robustness）与特异性（specificity）作为互补的探针选择原则，以可靠估计决策区域的继承关系。在涵盖56个公开检查点、770组源-嫌疑模型对及多种权重变换的实验中，Stemma取得0.967 AUC和1%假正率下的87.8%真阳性率；在覆盖91个实际部署实例的1,260组样本上进一步达到0.995 AUC和93.5%真阳性率，显著优于四个基准方法，验证了其在多样化部署场景下的强鲁棒性。
链接: https://arxiv.org/abs/2607.25880
作者: Keyu Zhang,Vadim Safronov,Andrew Martin
机构: University of Oxford (牛津大学)
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 25 pages, 4 figures
Abstract:LLM provenance testing asks whether a suspect LLM belongs to the same lineage as a source. Existing black-box methods largely infer this relationship from response-level characteristics, but these characteristics may shift under adaptation or deployment even when the underlying meaning remains unchanged, weakening the reliability of provenance evidence. To address this limitation, we introduce induced decision regions by mapping open-ended outputs into a finite decision space, thereby abstracting away surface-form variation and reframing provenance testing as measuring the inheritance of decision regions. Empirical analysis shows that the source’s induced regions are preserved more strongly in related models than in unrelated models. Building on this signal, we propose Stemma, a practical black-box LLM fingerprinting method that operationalises stability, robustness, and specificity as complementary probe-selection principles for reliably estimating induced decision region inheritance. Across 770 source-suspect pairs drawn from 56 public checkpoints and spanning diverse model-weight transformations, Stemma achieves 0.967 AUC and 87.8% TPR at 1% FPR, substantially outperforming four representative baselines. It further achieves 0.995 AUC and 93.5% TPR at 1% FPR on 1,260 pairs covering 91 deployment instances, demonstrating robustness to diverse inference-time deployment settings.
[NLP-10] Shieldstral
【速读】： 该论文旨在解决多模态内容安全分类中因任务多样性与标签体系异构性导致的模型泛化能力差、难以统一训练框架的问题。现有方法在面对不同平台或政策下的安全判别任务时，常受限于各自独立的标注体系与评估标准，难以实现高效迁移与统一建模。针对这一挑战，论文提出Shieldstral——一个30亿参数（3B-parameter）的策略自适应多模态安全分类器，其核心解决方案在于将内容审核任务重新形式化为一个二元问答（binary question-answering）问题，通过“是/否”判断统一涵盖文本与多模态内容的安全性判定，从而打破不同安全任务间的语义壁垒。该设计使异构的安全数据集（包含多样化的分类体系）得以在单一训练框架下进行整合与联合优化。此外，研究还提出了完整的数据构建方案，涵盖约5410万样本的高质量数据采集与生成，并构建了细粒度评估集以量化模型对不同政策偏好的适应能力。这一方法使小型模型在性能上可媲美甚至超越规模大近7倍的基准模型，实现了高效率与强适应性的平衡。
链接: https://arxiv.org/abs/2607.25857
作者: Antonia Calvi,Avinash Sooriyarachchi,Giada Pistilli,Guillaume Lample,Maarten Buyl,Maximilian Augustin,Maximilian Müller,Pierre Stock,Tom Bewley,Wassim Bouaziz,Yimu Pan
机构: 未知
类目: Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:We introduce Shieldstral, a 3B-parameter policy-adaptive multimodal safety classifier that matches or outperforms models nearly 7 \times its size on text safety benchmarks and sets a new state of the art on multimodal safety classification. Shieldstral formulates content moderation as a binary question-answering task. This simple formulation unifies diverse moderation tasks into a single yes/no problem, enabling heterogeneous safety datasets with divergent taxonomies to be consolidated under one training framework. We present the data construction recipe, covering curation and generation of approximately 54.1M samples and a fine-grained evaluation set to evaluate policy adaptability. Together, these enable a small adaptive model to match or outperform much larger models.
[NLP-11] AngelSpec: Towards Real-World High Performance Inference with Speculative Decoding
【速读】： 该论文旨在解决大语言模型推理中因不同工作负载特性导致的推测解码（speculative decoding）性能不一致问题，即单一草稿生成结构难以在多样化真实场景下保持最优表现。其核心挑战在于输出分布差异显著：对话类任务具有高熵、开放性特征，而代码与数学等任务则具备长序列可预测性。为此，论文提出AngelSpec统一训练框架，通过三层次协同优化实现异构场景适配：在训练层面，采用结构与数据的联合专业化策略——自回归多标记预测（MTP）草稿器专注于多样对话数据以应对高熵输出，块并行扩散（block-parallel diffusion）草稿器则针对代码与数学数据进行训练，以捕捉长序列可预测性；在架构层面，提出DFly框架，融合目标条件化主干与前驱条件化自回归头，增强目标特征利用效率和块内依赖建模能力，同时维持生成过程的并行性；在推理层面，将验证成本视为共享批处理资源，动态调整各请求的验证深度，结合预期收益与预标定成本模型实现在线自适应，有效应对领域、请求、在线负载及硬件差异带来的接受长度与验证开销波动。实验表明，相较于传统自回归解码，DFly在Hy3-A21B上平均接受长度提升约30%，在4至64并发度下均达到最高吞吐量，相较自回归解码提速1.98–2.40倍，且比DFlash提升10.5–11.8%。研究开源了AngelSpec以支持后续方法的训练与扩展。
链接: https://arxiv.org/abs/2607.25852
作者: Hong Liu,Rui Cen,Junhan Shi,Guangshuo Qin,Jiebin Zhang,Tianyu Liu,Runzhi Fan,Guoliang Zhao,Ruobing Xie,Kai Zhang,Song Liu,Guanghua Yu,Jianchen Zhu
机构: Tencent Inc.
类目: Computation and Language (cs.CL)
备注:
Abstract:Speculative decoding accelerates large language model inference without changing the target distribution, but no single drafting structure performs best across real-world workloads. Autoregressive multi-token prediction (MTP) is a lightweight, stable proposal mechanism, whereas block-parallel diffusion amortizes drafting latency over much longer candidate sequences; the better choice depends strongly on the output distribution. We present AngelSpec, a unified training framework for MTP and block-parallel speculative decoding that addresses this heterogeneity at three levels. At the training level, rather than fitting one universal drafter to a uniform data mixture, we co-specialize structure and data: the MTP drafter is trained on diverse conversational data for high-entropy open-ended chat, and the block-diffusion drafter on code and mathematics data for longer predictable continuations. At the architecture level, we propose DFly, a block-diffusion framework combining a hybrid target-conditioning backbone with a predecessor-conditioned autoregressive head, improving target-feature utilization and intra-block dependency modeling while keeping generation parallel. At the inference level, both acceptance length and verification cost vary with domain, request, online load, and hardware, so DFly treats verification as a shared batch-level resource: it reallocates compute toward high-confidence prefixes across requests and combines expected utility with a profiled cost model to adapt verification depth online. Across the Hy3 series, DFly raises the average accepted length on Hy3-A21B by roughly 30% and attains the highest average throughput at every tested concurrency from 4 to 64, a 1.98-2.40x speedup over autoregressive decoding and 10.5-11.8% higher throughput than DFlash. We release AngelSpec to support training and extending these methods.
[NLP-12] Evaluation of Adversarial Robustness in Arabic Language Models
【速读】： 该论文旨在解决阿拉伯语语言模型（Arabic Language Models）在面对对抗性攻击时的脆弱性问题，特别是针对其在不同粒度层级（字符级、词级、句级）上所暴露的安全隐患。研究的关键在于系统评估五种前沿阿拉伯语模型在多种对抗攻击策略下的鲁棒性表现，并探索基于对抗训练（adversarial training）的防御机制的有效性。核心发现表明，通过插入元音符号（diacritics）可使部分模型准确率下降高达92%且扰动距离极低；词级攻击中操纵连接词可在保持高语义相似性和低扰动的同时导致最高58%的性能下降；句级攻击中重述（paraphrasing）平均使目标模型性能下降76%。尽管对抗训练显著提升了整体鲁棒性，其中MARBERT表现最优，AraBERT相对提升最大，但在应对字符级噪声方面仍存在明显局限。这揭示了当前防御策略在形态丰富的语言如阿拉伯语中的潜在瓶颈与改进空间。
链接: https://arxiv.org/abs/2607.25814
作者: Anwar Alajmi,Ayed Salman,Imtiaz Ahmad
机构: Kuwait University (科威特大学)
类目: Computation and Language (cs.CL); Cryptography and Security (cs.CR)
备注:
Abstract:The emergence of the recent outstanding capabilities of Arabic Language Models has opened doors for exposing their vulnerabilities. One of the major security risks associated with such Natural Language Processing models is adversarial attacks. These attacks can deceive the model into the wrong prediction, raising critical model security and safety concerns. This study aims to assess the robustness of five state-of-the-art Arabic Language Models under a distinct set of Arabic adversarial attacks applied at various levels of granularity and using different example generation strategies. We also explore a defense technique based on adversarial training to enhance model robustness. The results show that insertion of diacritics can reduce the accuracy of some models by 92% while maintaining a low perturbation distance. For word-level attacks, manipulating Arabic conjunctions preserves high semantic similarity scores, low perturbation distance, and leads to an accuracy degradation of up to 58%. For sentence-level attacks, paraphrasing proves its effectiveness by an average reduction of 76% in the victim models’ performance. While adversarial training improves overall resilience, with MARBERT being the most robust and AraBERT showing the greatest relative gains, challenges persist, particularly against character-level noise. These findings highlight both the potential and limitations of current defense strategies in morphologically rich languages like Arabic.
[NLP-13] WorkSurface-Bench: Benchmarking Enterprise Agents on Multi-Surface Knowledge Routing
【速读】： 该论文旨在解决企业级智能体在处理复杂任务时，如何有效识别并选择合适知识源（如文档、表格、依赖图等）以支持推理与决策的问题。现有基准测试往往忽视了智能体在执行任务前对知识源的“表面路由”（surface routing）能力评估，导致无法准确衡量其跨异构知识源的协同理解与调用能力。为此，作者提出了WorkSurface-Bench基准，涵盖1,151个源自Persona-Scoped Workspace-Bench-Lite工作空间的原子级任务，覆盖文档、表格、图结构及跨表面问答场景。其参考答案具备可审计性：表格答案通过执行DuckDB查询复现，文档答案基于经验证的文本片段，图答案则追溯至原始依赖标注。实验在六种受控代理设置下评估四种模型骨干，生成27,624条无协议错误的轨迹。结果显示，在金标准工具访问条件下，智能体的路由F1得分高达98.7–99.8，但最终答案准确率仅为56.1–75.3%，表明正确的知识源选择虽为必要前提，却不足以保证任务完成。进一步干预分析显示，提供表面提示可提升三种模型的答案准确率，而移除无关工具主要改善路由效率与准确性。独立三标注员审计表明，抽样200个任务中192个在所有六个质量维度上达成一致意见，整体质量可靠。该研究的关键解决方案在于构建一个具有可审计性、多模态知识源协同评估能力的基准体系，从而系统性地揭示智能体在真实企业工作流中的核心瓶颈——即从“选对知识源”到“正确使用知识”的关键跃迁。
链接: https://arxiv.org/abs/2607.25765
作者: Hao Liang,Meiyi Qiang,Sizhe Qiu,Linzhuang Sun,Wentao Zhang
机构: Peking University (北京大学)
类目: Computation and Language (cs.CL); Databases (cs.DB)
备注:
Abstract:Enterprise agents often need to integrate heterogeneous knowledge sources: documents for narrative facts, tables for computation, and dependency graphs for file relationships. Existing benchmarks typically evaluate retrieval or tool use without distinguishing whether an agent first selects the appropriate knowledge sources. We introduce WorkSurface-Bench, a benchmark for evaluating this capability as surface routing. It contains 1,151 atomic tasks derived from persona-scoped Workspace-Bench-Lite workspaces, spanning document, table, graph, and cross-surface questions. Its reference answers are auditable: table answers are reproduced through executed DuckDB queries, document answers are grounded in verified text spans, and graph answers are traced to source dependency annotations. We evaluate four model backbones across six controlled agent settings, yielding 27,624 protocol-error-free trajectories. Under gold-constrained tool access, agents achieve 98.7-99.8 Route F1, while Answer remains only 56.1-75.3 percent, showing that correct surface selection is necessary but insufficient for task completion. Matched interventions further show that surface hints improve Answer for three of four models, whereas removing irrelevant tools primarily improves routing and efficiency. In an independent three-annotator audit, all 200 sampled tasks pass all six quality criteria by majority vote, with 192 receiving unanimous judgments on every criterion. We release the dataset, construction pipeline, scoring code, and agent harness at this https URL.
[NLP-14] SpeechLLM Meets Federated Learning for End-to-End ASR: English and Italian Case Studies
【速读】： 该论文旨在解决在分布式数据环境下对大规模语音语言模型（SpeechLLM）进行联邦学习（Federated Learning, FL）时面临的挑战，特别是针对端到端自动语音识别（ASR）系统中高维参数空间、梯度通信开销大以及分布式计算资源受限等问题。其核心解决方案是设计一种通信高效的联邦优化策略，专门适配SpeechLLM的架构特性，通过减少通信成本并保持模型性能，在多种声学条件和发音风格下实现了与集中式训练相当甚至更优的稳定性和有效性。实验结果表明，该方法在英语和意大利语单语ASR任务上均取得具有竞争力的词错误率（Word Error Rate, WER），同时显著降低了通信开销，为实际多语言场景下联邦SpeechLLM的部署提供了可行的技术基础。
链接: https://arxiv.org/abs/2607.25716
作者: Mohamed Nabih Ali,Daniele Falavigna,Alessio Brutti
机构: 未知
类目: Computation and Language (cs.CL)
备注: Accepted at the 2nd International Conference on Federated Learning and Intelligent Computing Systems (FLICS2026)
Abstract:Federated learning (FL) enables privacy-preserving training of automatic speech recognition (ASR) systems across distributed data sources, yet its application to large-scale speech language models (SpeechLLMs) remains unexplored. This paper presents the first systematic study of federated training for SpeechLLM-based end-to-end ASR systems. We design a communication-efficient federated optimization strategy tailored to the unique challenges of SpeechLLM architectures, addressing high-dimensional parameter spaces, gradient communication overhead, and computational constraints in distributed settings. Through extensive empirical evaluation on monolingual ASR tasks in English and Italian, we demonstrate the effectiveness and stability of our federated approach compared to centralized training baselines across diverse acoustic conditions and speaking styles. Additionally, we conduct a comprehensive ablation study analyzing the impact of different speech encoder architectures on monolingual English ASR performance within the federated framework, providing insights into optimal model configurations for decentralized training. Our results achieve competitive word error rates while reducing communication costs, establishing practical foundations for federated SpeechLLM deployment in real-world multilingual scenarios.
[NLP-15] MyMentorLLM : A psychotherapy GenAI environment with multimodal voice/text patients trainees and experts for deliberate practice
【速读】： 该论文旨在解决心理治疗师培训中专家督导难以规模化的问题，提出了一种基于多模态语音与文本的模拟训练环境MyMentorLLM，用于实现认知行为疗法（Cognitive Behavioural Therapy, CBT）的刻意练习。其核心解决方案在于构建一个包含受训治疗师、具有DSM-5-TR诊断依据的模拟患者（模拟特定精神障碍，如重度抑郁、广泛性焦虑或边缘型人格障碍）以及专家督导三者交互的闭环仿真系统，生成了2,100场完整的CBT训练会话。关键创新点在于通过大语言模型（Large Language Models, LLMs）驱动的模拟督导，评估不同模型在治疗胜任力评价、情感动态一致性及诊断准确性方面的表现；研究发现，尽管多数LLM存在对受训者能力的高估，但采用原生语音-语音转换的模型在反馈质量上最接近人类督导水平，且模型规模越大，症状识别准确率越高，监督反馈可有效提升模拟治疗师的诊断能力。因此，该研究证明了针对CBT培训的刻意练习模拟是可行的，但需同步关注模拟患者的病理性真实性、督导模型的校准精度以及潜在有害反馈的风险。
链接: https://arxiv.org/abs/2607.25667
作者: Rodolfo Rizzi,Alessandro Grecucci,Massimo Stella
机构: University of Trento (特伦托大学); University of Bari Aldo Moro (巴里大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 27 pages, 6 figures, 2 tables; 1 supplementary table
Abstract:Psychotherapists need repeated training and supervision by experts; however, scalability is problematic. Here we present MyMentorLLM, a multimodal voice- and text-based simulation environment for deliberate practice, used to generate 2,100 complete Cognitive Behavioural Therapy (CBT) training sessions. Each session links a DSM-5-TR-grounded patient (with major depressive, generalised anxiety or borderline personality disorder), a therapist-in-training and an expert supervisor. As an initial implementation, we adopted CBT because its structured procedures and competency-based supervision facilitate standardised simulation and evaluation. Sessions were analysed for emotional dynamics, therapeutic competence and diagnostic accuracy. Simulated patients expressed disorder-congruent emotional profiles, which trainee therapists mirrored as in real human counselling. The quality of supervision differed across LLMs: while most models overestimated trainees’ competences, native speech-to-speech was closest to human scores. Supervisors’ feedback led to better diagnoses in simulated psychotherapists in 5 out of 7 LLMs, and symptom identification accuracy increased with model size. This work shows that simulation of deliberate practice is possible for CBT training, although patient fidelity, calibration of supervisors, and harmful feedback should be evaluated together.
[NLP-16] Localized Adaptation Reveals Distinct Learning Signatures in Transformers
【速读】： 该论文旨在解决生成式 AI 模型在进行微调或适配时，适应位置（adaptation site）如何影响模型学习内容、泛化能力及适配的局部性这一关键问题。现有方法通常将适配分布于模型深层，即使目标变化范围有限，导致效率低下且难以控制学习行为。本文提出一种受控基准测试框架，涵盖五类任务目标（词汇绑定、事实关联、行为策略学习、因果映射与程序推理），并定义每类任务的“适配几何”（adaptation geometry）为在全栈及早、中、晚层进行低秩自适应（LoRA）时，其在知识获取、迁移能力与边界性上的表现特征。研究发现，不同任务具有显著不同的适配几何：词汇绑定偏好早期层以实现高效获取与边界控制，但需更广泛更新以支持迁移；事实关联则倾向于后期层的局部适配；行为学习表现出晚期层负责动作获取、中期层负责策略门控的分离特性；而因果与程序推理最受益于中层或全栈适配。这些模式在参数量匹配条件下仍保持稳定，并在五个模型家族中复现。结果表明，适配位置是决定模型学习内容、泛化性能与未改变部分的关键设计变量，为生成式 AI 的高效、可控适配提供了理论依据与实践指导。
链接: https://arxiv.org/abs/2607.25663
作者: Rebecca Ramnauth,Brian Scassellati
机构: Yale University (耶鲁大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Main text: 8 pages, 2 figures; appendix: 13 tables, 10 figures; code and data available at this https URL
Abstract:Transformer adaptation is typically distributed across model depth, even when the intended change is narrow. We investigate how adaptation site shapes what a model learns, how well that learning generalizes, and how selectively it is applied. We introduce a controlled benchmark spanning five objectives (lexical binding, factual association, behavioral policy learning, causal mapping, and procedural reasoning) and define each objective’s “adaptation geometry” as its profile of acquisition, transfer, and boundedness under full-stack and early-, middle-, or late-layer LoRA. The objectives exhibit distinct geometries. Lexical binding favors early-layer adaptation for acquisition and boundedness but requires broader updates for transfer; factual association favors later layers among localized adapters; behavioral learning separates late-layer action acquisition from middle-layer policy gating; and causal and procedural transfer benefit most from middle- or full-stack adaptation. These patterns largely persist under parameter-matched controls, and most corresponding directional contrasts replicate across five model families. These findings establish adaptation site as a key design variable for controlling what models learn, generalize, and leave unchanged.
[NLP-17] Instruction-based Image Editing: A Survey on Data Models Evaluation and Applications
【速读】： 该论文旨在解决指令驱动图像编辑（Instruction-based Image Editing, IIE）中缺乏系统性分类与评估标准的问题，尤其针对现有方法在任务定义、数据构建、模型架构演进及评测体系方面的碎片化现状。其核心解决方案在于提出一个涵盖五个关键维度的系统性分类框架：（1）编辑任务的定义与操作层级划分，（2）训练数据的构建方法，（3）从生成对抗网络（GAN）到扩散模型与自回归模型的架构演进，（4）标准化评估指标与基准测试集的发展，以及（5）商业化解决方案的引入。在此基础上，论文进一步提出“全面、深入且诊断性”的IIE基准测试（Comprehensive, in-Depth, and Diagnostic benchmark for IIE task, CDD-IIE Bench），能够多维度、严谨地评估模型性能。通过开源方案的实证对比，揭示了当前主流方法在语义理解、编辑精度与泛化能力等方面的优劣，为未来研究方向提供了明确指引。
链接: https://arxiv.org/abs/2607.25642
作者: Xianghao Zang,Zijian Jiang,Jiarong Cheng,Qianrui Teng,Ying He,Yuxuan Mu,Chao Ban,Huayu Zhang,Lanxiang Zhou,Zerun Feng,Chi Zhang
机构: Peking University (北京大学); Tongji University (同济大学); Beijing University of Posts and Telecommunications (北京邮电大学); Southeast University (东南大学)
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注: 33 pages, 7 figures, Vicinagearth
Abstract:Instruction-based Image Editing (IIE) aims to transform a given image into a new one based on textual instructions. Advances in Large Language Models (LLMs) and Vision-Language Models (VLMs) have accelerated progress toward practical ``one-sentence image editing" systems. This survey presents a systematic taxonomy and comprehensive review of IIE research, structured around five core dimensions: (1) task definition and hierarchical categorization of editing operations, (2) methodologies for training data construction, (3) architectural evolution from GAN-based to diffusion and autoregressive paradigms, (4) standardized evaluation metrics and benchmark development, and (5) introduction of commercial solutions. Our analysis shows critical technological milestones across model generations. We further propose a Comprehensive, in-Depth, and Diagnostic benchmark for IIE task (CDD-IIE Bench), which can rigorously assess the multiple aspects of model performance. Through empirical comparisons of open-source solutions, we highlight their respective capabilities and limitations. Finally, we discuss future research directions to advance the field.
[NLP-18] AIriskEval-edu Demo: Auditing of Pedagogical Risks in Educational Explanations ICDAR2026
【速读】： 该论文旨在解决教育领域中生成式AI（Generative AI）在提供教学解释时存在的教学质量与潜在风险难以评估的问题。其核心挑战在于如何系统性地识别和量化教学解释中的五大类教育学风险：事实准确性、深度与完整性、焦点与相关性、学生适切性以及意识形态偏见。解决方案的关键在于构建一个可解释的审计平台——AIriskEval-edu Demo，该平台通过集成外部API调用的GPT-5.5与自部署于消费级GPU上的Llama 3.1 8B本地评估模型，实现对教学解释的多维度风险检测。其中，本地评估模型基于AIriskEval-edu数据集进行微调，该数据集包含经风险标注与可解释性注释的K-12教学解释样本。平台支持两种运行模式：在AI模式下，对六种模拟教师角色生成的教学解释进行批量审计；在人类模式下，实时评估用户手写解释。实验结果表明，本地评估模型在多数指标上优于GPT-5.5，尤其在保障数据隐私与实现教育机构内部内容审计方面具有显著优势，为教育机构提供了可控、可解释且高效的AI教学内容质量监管工具。
链接: https://arxiv.org/abs/2607.25634
作者: Javier Irigoyen,Roberto Daza,Francisco Jurado,Julian Fierrez,Ruben Tolosana,Alvaro Ortigosa,Miguel Lopez-Duran,Aythami Morales
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 6 pages, 2 figures. Accepted at the 17th IAPR International Workshop on Document Analysis Systems (DAS 2026), ICDAR 2026, September 3, 2026
Abstract:We present AIriskEval-edu Demo, a platform that audits the pedagogical quality of instructional explanations and provides explainable audit results. The platform evaluates an explanation against a rubric covering five dimensions of pedagogical risk: factual accuracy, depth and completeness, focus and relevance, student-level appropriateness, and ideological bias. For each dimension, it returns a binary decision and a confidence score. Detected risks also include a natural-language rationale and, except for Depth and Completeness, a localized evidence span. The platform integrates GPT-5.5 through an external API and a self-hosted Llama 3.1 8B evaluator that runs on consumer-grade GPUs. The local evaluator is fine-tuned on AIriskEval-edu, a dataset of K-12 instructional explanations with risk and explainability annotations. The platform operates in two modes: in AI mode, both evaluators assess stored explanations generated under six simulated teacher profiles, each representing a distinct pedagogical behavior and potential risk; in human mode, the local evaluator audits user-written explanations in real time. The local evaluator outperforms GPT-5.5 on most reported metrics, offering educational institutions a practical way to keep audited content within their own infrastructure.
[NLP-19] Construction-Driven Injection: Linguistically-Grounded Edit-Based Code-Mixing Fingerprints for Large Language Models
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）作为高价值知识产权在未经授权情况下被复制和商业滥用的问题。现有指纹技术存在两大局限：一是自然语言触发词易发生误激活，二是混淆型指纹易被基于困惑度的检测机制过滤；此外，指纹构建与注入过程分离，导致注入阶段无法感知触发词的语言结构，错失针对性优化机会。针对上述问题，论文提出一种统一的指纹框架LCF（Language-Code Mixing Fingerprinting），其核心创新在于将指纹构建过程驱动注入过程，实现两阶段联合优化。具体而言，LCF通过低资源语言的语义密度替换规则与语法偏置混合策略生成代码混杂型指纹，使触发词的困惑度显著低于传统混淆方法，同时避免自然语言触发词的误激活缺陷；LCFEdit则利用高资源多语言表示中的零空间投影进行注入，保留模型知识，并通过跨语言对齐步骤引导权重更新至目标语言表征子空间，确保更新过程具备语言学合理性，从而提升稳定性。实验表明，该方案在不可察觉性、可检测性及无害性方面均表现优异，实现了对模型所有权的持续验证，且对模型性能影响微乎其微。
链接: https://arxiv.org/abs/2607.25633
作者: Yongyi Cui,Yue Li,Tianbao Jiang,Xin Yi
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models (LLMs) are costly intellectual assets that remain exposed to unauthorized redistribution and commercial misuse. Injected fingerprints, i.e., trigger–target pairs embedded in model behavior, offer a practical, black-box-verifiable ownership signal, but existing methods decouple the two stages of the fingerprint life cycle: how a fingerprint is constructed and how it is injected. Existing fingerprinting frameworks suffer from two limitations. Natural-language fingerprints are prone to accidental activation, and garbled fingerprints are easily filtered by perplexity-based detection. Furthermore, decoupling construction from injection leaves the latter unaware of the trigger’s linguistic structure, missing the opportunity for targeted optimization. We argue that fingerprint construction should drive injection, and present a unified fingerprinting framework that jointly optimizes both stages. First, LCF constructs code-mixing fingerprints by combining low-resource languages under a semantic-density substitution rule and grammar-biased mixing, yielding triggers whose perplexity sits far below garbled baselines while avoiding the accidental-activation failures of natural-language triggers. Second, LCFEdit injects each fingerprint with a null-space projection derived from high-resource multilingual representations that preserves knowledge, augmented by a cross-lingual alignment step that steers the weight update toward the fingerprint language’s representation subspace. This construction-aware injection ensures that the update is linguistically informed and therefore more stable. Extensive evaluations on imperceptibility, detectability, and harmlessness demonstrate persistent ownership verification with negligible impact on utility.
[NLP-20] MemSFT: Mitigating Alignment Tax with an External Parametric Memory
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在特定领域适配过程中面临的“对齐代价”（alignment tax）问题，即在特定任务上进行微调时容易引发灾难性遗忘，导致通用能力显著退化。其核心解决方案是提出MemSFT框架，通过引入一个可插拔的参数化记忆模块（parametric memory），将领域专业化过程与主干模型参数更新解耦。该记忆模块经过训练以模仿非参数化检索器在领域数据上的行为，从而自主存储领域知识与模式，避免依赖外部检索机制。训练完成后，该记忆模块可跨不同规模的LLM复用；在生成阶段，通过一个学习得到的路由机制，在每一步解码中动态融合记忆模块与主干模型的输出分布，实现领域专长的按需调用。实验表明，相较于全量微调（full SFT），MemSFT在生物、地质科学和法律等领域均显著提升领域性能，同时对通用能力的损害几乎可忽略，验证了在参数层面实现通用能力与领域知识解耦的可行性，为构建兼具通用性与专业性的大模型提供了有效路径。
链接: https://arxiv.org/abs/2607.25614
作者: Jiarui Wang,Xiang Shi,Jiaqi Cao,Rubin Wei,Xiquan Wang,Hao Sun,Jingzhi Wang,Zhiqi Yang,Qipeng Guo,Bowen Zhou,Zhouhan Lin
机构: Shanghai Jiao Tong University(上海交通大学); Shanghai AI Laboratory(上海人工智能实验室); Tsinghua University(清华大学)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注: 33 pages, 11 figures, 13 tables
Abstract:Adapting Large Language Models (LLMs) to specialized domains often incurs an alignment tax, as fine-tuning on domain-specific tasks can cause catastrophic forgetting and substantially degrade performance on general tasks. We propose MemSFT, which mitigates the alignment tax by decoupling domain specialization from backbone parameter updates through a plug-and-play parametric memory. The memory is trained to imitate the behavior of a non-parametric retriever operating over domain data, thereby memorizing knowledge and patterns that would otherwise be accessed through retrieval. Once trained on a specific domain, the memory can be reused across LLMs of different sizes. During generation, a learned router dynamically fuses the output distributions of the memory and backbone at each decoding step, allowing domain expertise to be invoked selectively. Across biology, geoscience, and law, evaluations with models ranging from Qwen3-8B to Qwen3-235B-A22B show that MemSFT consistently improves domain performance with negligible degradation in general performance, whereas full SFT suffers severe forgetting on general tasks. Overall, our results demonstrate a practical path to decoupling general model capabilities from domain-specific knowledge at the parameter level, thereby equipping LLMs with new specialized capabilities without compromising their general capabilities.
[NLP-21] PILA: Plug-and-Play Insertion for LLM -native Advertising
【速读】： 该论文旨在解决如何在大语言模型（Large Language Models, LLMs）的生成响应中自然融入赞助内容，实现原生广告（LLM-native advertising） monetization的问题。现有方法将广告内容与生成过程耦合于单一模型内部，不仅不兼容当前以API或工作流为主的LLM应用架构，还不可避免地损害原始响应质量。为此，本文提出PILA（Prompt-Induced Light-weight Ad insertion），其核心思想是将广告插入问题重新建模为条件响应重写任务，并将其作为轻量级旁路模块（sidecar module）从上游服务中解耦。PILA具有模型无关性，可无缝集成至现有LLM服务，无需修改基础模型或工作流；同时支持在用户侧自然性与广告侧曝光度之间进行可控权衡，为下游定价与部署提供实用接口。实验结果表明，PILA在多种上游模型上均能持续提升广告效果，同时保持高质量响应，展现出作为实际可行方案的巨大潜力。
链接: https://arxiv.org/abs/2607.25590
作者: Zhaowei Zhang,Yuhan Fu,Yihang Zhang,Xiaohan Liu,Ceyao Zhang,Xiaoyuan Zhang,Yipeng Kang,Tonghan Wang,Yaodong Yang
机构: University of Science and Technology of China (中国科学技术大学); Tsinghua University (清华大学); Peking University (北京大学); National University of Singapore (新加坡国立大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:How to monetize large language models (LLMs) by naturally integrating sponsored content into their responses, known as LLM-native advertising, has recently emerged as a critical problem. However, existing solutions entangle advertising with content generation inside a single model, which is incompatible with modern API-only or workflow-based LLM applications and inevitably compromises the original response quality. To address this, we propose PILA, which reformulates ad insertion as a conditional response rewriting problem and decouples it from the upstream service as a lightweight sidecar module. PILA is model-agnostic and can be seamlessly integrated with existing LLM services without modifying the base model or its workflow. It also exposes a controllable trade-off between user-side naturalness and ad-side exposure, offering a practical interface for downstream pricing and deployment. Experiments across diverse upstream models show that \pila consistently improves ad effectiveness while preserving response quality, highlighting its promise as a practical solution for LLM-native advertising.
[NLP-22] Forensic Reproducibility Audit of a Radiology Vision-Language Model Benchmark: From Intended Protocol to Released Artifact
【速读】： 该论文旨在解决医学影像人工智能（Medical-imaging AI）基准评估中多源异构数据与分析流程之间缺乏可追溯性与一致性验证的问题。其核心挑战在于，现有基准通常假设各组件（如数据集、DICOM渲染、提示词、API调用、自动标注、统计代码、论文与代码库发布等）之间存在隐含的一致性，但未进行实际验证。为应对这一问题，研究采用回溯性法医可复现性审计方法，对一个已归档的胸部X光片视觉-语言模型（VLM）试点项目进行了全面审查，且未重新调用模型或新增图像/报告标注。关键解决方案在于系统追踪提示词绑定、DICOM元数据、输出完整性、标签提取过程、分析匹配性及发布传播路径，并揭示出多个严重缺陷：60次使用相同提示词（C）的Claude调用产生不一致结果；28名患者被错误地合并至30个研究中；4幅MONOCHROME1图像缺失必要的极性反转；数据集划分信息未保留；非验证的提取器截断5份报告至4000字符；重建统一队列后，Cochran’s Q值从154.73升至182.29，表明统计结果高度不稳定。在45次McNemar检验中，27项原始p值<0.05，经Holm校正后仍有20项保持显著，但这些结果仅反映归档自动化标签矩阵的内部一致性，无法恢复原初提示对比意图或建立临床性能结论。因此，研究撤回所有原始性能、排名、提示效应及临床推论，并提出一套机器可验证的控制机制，涵盖队列定义、DICOM渲染、提示与模型身份、调用状态、注释溯源、关键分析步骤及衍生产物的全链路可追溯性要求，以确保未来研究的可复现性与可信度。
链接: https://arxiv.org/abs/2607.25589
作者: Mateusz Kozłowski
机构: Independent Researcher(独立研究员)
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注: 25 pages, 5 figures, 7 tables. Retrospective artifact audit; no new model calls or annotations. Corrective audit and archival reconstruction: this https URL
Abstract:Medical-imaging AI benchmarks combine datasets, DICOM rendering, prompts, provider APIs, automated labels, statistical code, manuscripts, and repository releases. Agreement across these artifacts is usually assumed rather than tested. We performed a retrospective forensic reproducibility audit of a preserved chest-radiograph vision-language model (VLM) pilot; no model was called again and no image or report was newly annotated. We traced prompt bindings, DICOM metadata, output completeness, label extraction, matched analyses, and release propagation. Of 300 planned model-prompt calls, 297 yielded nonempty reports. Sixty Claude calls labeled A/B were executed with the same C prompt. The 30 studies represented 28 patients. Four MONOCHROME1 images were rendered without required polarity inversion, dataset split membership was not retained, and the unvalidated extractor truncated five reports to 4000 characters. Reconstructing one common cohort of 369 complete case-finding blocks changed Cochran’s Q from 154.73 to 182.29. Of 45 McNemar comparisons, 27 had unadjusted p 0.05 and 20 remained below 0.05 after Holm adjustment. These values describe only the archived automated-label matrix; they do not recover the intended prompt comparison or establish clinical performance. We withdraw the original performance, ranking, prompt-effect, and clinical claims and specify machine-verifiable controls for cohort, DICOM rendering, prompt and model identity, call status, annotation provenance, keyed analysis, and derived artifacts.
[NLP-23] Evaluation of forced alignment of code-mixed speech: the case of Hindi-English
【速读】： 该论文旨在解决混合语码（code-mixed speech）在强制对齐（forced alignment）中面临的三大挑战：音位库扩大、拼写错误以及说话人差异。针对印度语-英语混合语料，研究评估了蒙特利尔强制对齐器（Montreal Forced Aligner, MFA）的性能，并重点解决两个关键问题：（1）母语与非母语发音对之间的自由变异（free variation），以及（2）句中英语词汇的音素边界检测。研究表明，基于自举策略（bootstrapping strategies）的词典优化显著优于未经修改的原始词典；同时，使用句子级混合语料训练的声学模型可实现平均对齐误差仅为4.15毫秒，较单语印地语（38.18毫秒）和孤立英语（37.58毫秒）模型低一个数量级。因此，可靠双语语音对齐的关键在于系统性词典设计与真实混合语境下的训练数据构建。
链接: https://arxiv.org/abs/2607.25581
作者: Ayushi Pandey,Pamir Gogoi,Kevin Tang
机构: Karya(印度); Heinrich Heine University Düsseldorf(德国); University of Florida(美国)
类目: Computation and Language (cs.CL)
备注:
Abstract:Code-mixed speech poses unique challenges to forced alignment: expanded inventories, orthographic errors, and speaker variation. We evaluate forced alignment of Hindi-English code-mixed speech using the Montreal Forced Aligner. We address 2 problems: (1) free variation involving native vs non-native pairs and (2) phonemic boundary detection for mid-utterance English words. Bootstrapping strategies substantially outperform unmodified lexicons. Acoustic models trained on sentence-level code-mixed data achieve a mean error of 4.15ms, ie. ten times lower than monolingual Hindi (38.18ms) or isolated English (37.58ms) alternatives. Principled lexicon design and code-mixed training data are both essential for reliable alignment of bilingual speech.
[NLP-24] IRIS: Reusable Identity Representations from Frozen LLM s for Entity Alignment
【速读】： 该论文旨在解决知识图谱（Knowledge Graph, KG）中实体对齐（Entity Alignment, EA）任务中存在的语义理解不足与可迁移性差的问题。传统方法依赖显式图结构和文本字段，难以在异构描述下准确识别同一实体并区分语义相近实体；尽管大语言模型（Large Language Models, LLMs）具备更强的语义理解能力，但现有基于LLM的EA方法仅将其用于辅助生成或候选条件决策，未能将深层语义理解提炼为稳定、可直接比较的实体身份表征，导致对齐结果依赖特定知识图谱对或候选集，且需随匹配上下文变化重复计算。为此，本文提出无需训练的IRIS（Identity Representations from Internal States）框架，通过从冻结的LLM内部状态中提取面向身份的上下文表示，为每个实体构建类虹膜（iris-like）的唯一签名，形成一个共享的、稳定的实体身份空间。该空间支持实体一次性编码后跨不同知识图谱进行直接相似性比对，避免了成对表示构建与逐候选LLM推理的开销。在四个主流基准数据集及两种冻结LLM主干上，最优IRIS变体在D-Y-15K V2、DBP-WIKI、ICEWS-WIKI和ICEWS-YAGO上的Hits@1得分分别达到100.00、99.38、98.31和97.99，显著提升了对齐性能与泛化能力。
链接: https://arxiv.org/abs/2607.25579
作者: Xinran Liu,Shengtao Li,Shouqian Shi,Ge Wang,Xin-Wei Yao
机构: Nanjing University (南京大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 9 pages, 1 figure, 3 tables
Abstract:Entity alignment (EA) identifies entities across knowledge graphs (KGs) that refer to the same real-world object. Conventional EA methods mainly exploit explicit graph structures and textual fields, which often provide insufficient semantic understanding to recognize the same entity under heterogeneous descriptions and distinguish it from semantically similar entities. Although large language models (LLMs) offer deeper entity understanding, existing LLM-based EA methods largely use this capability for auxiliary generation or candidate-conditioned decisions. Consequently, such understanding is not distilled into a stable and directly comparable identity space, leaving alignment tied to specific KG pairs or candidate sets and requiring repeated processing as the matching context changes. To address these limitations, we propose IRIS (Identity Representations from Internal States), a training-free framework that constructs for each entity an iris-like signature encoding its distinctive and stable identity characteristics. IRIS derives these signatures by eliciting identity-oriented contextual representations from a frozen LLM, thereby forming a shared space in which each entity is encoded once and can be aligned across different KGs through direct similarity comparison, without pair-dependent representation construction or candidate-wise LLM inference. Across four established EA benchmarks and two frozen LLM backbones, the best IRIS variants achieve Hits@1 scores of 100.00, 99.38, 98.31, and 97.99 on D-Y-15K V2, DBP-WIKI, ICEWS-WIKI, and ICEWS-YAGO, respectively.
[NLP-25] Phase Structure in Rotary Attention: A Spectral Framework for Semantic Continuity and Execution-Boundary Governance
【速读】： 该论文旨在解决生成式语言模型中因旋转位置编码（Rotary Position Embedding, RoPE）引入的相位结构对注意力机制影响的理论分析难题，尤其关注有序上下文建模下查询-键交互中的相位对齐、隐状态连续性及语义漂移等问题。其核心解决方案在于构建一个有界谱框架，突破传统将语言模型视为类物理波系统的局限，首次明确以有序隐状态序列而非词表索引作为谱分解的有效域；通过将RoPE注意力得分形式化为加权余弦项之和，并证明局部稳定性引理——即相位偏移受控可保证预软最大化分值的退化被抑制；进一步引入基于固定正交方向对的复数模态坐标与加权相干性函数，实现对隐藏状态轨迹的相位分析扩展。该框架严格区分了表征连续性与执行边界合法性，强调内部相干性仅能反映任务相关关系的保持，而不能自动授权因果性转换，从而为判断谱结构何时解释连续性、何时需外部治理介入提供了理论与方法论基础。
链接: https://arxiv.org/abs/2607.25507
作者: Abraham Chachamovits
机构: ENTRUST AI(ENTRUST AI)
类目: Computation and Language (cs.CL)
备注: 14 pages; theoretical framework and proposed experimental program
Abstract:Transformer language models are usually analyzed through vector geometry, yet ordered context and rotary position encoding introduce explicit phase structure into query-key interactions. This paper develops a bounded spectral framework for examining rotary phase alignment, hidden-state continuity, and semantic drift without treating language models as literal physical wave systems. It first identifies ordered hidden-state sequences, rather than vocabulary indices, as valid domains for spectral decomposition. It then derives the Rotary Position Embedding (RoPE) attention score as a sum of magnitude-weighted cosine terms and proves a local stability lemma: uniformly bounded phase displacement limits degradation of the corresponding pre-softmax score. To extend phase analysis beyond native RoPE coordinates, the paper defines complex modal coordinates over fixed orthonormal direction pairs and introduces a weighted coherence functional for hidden-state trajectories. These constructions support a strict distinction between representational continuity and execution-boundary admissibility. Internal coherence may describe preservation of task-relevant relations, but it cannot authorize a consequential transition. Positioned against existing geometric, spectral, phase-modulation, representation-analysis, and mechanistic-interpretability accounts, the framework contributes a theoretical and methodological program for determining when spectral structure explains continuity and when governance must remain an external predicate over execution.
[NLP-26] PatientAgent Bench: A Benchmark Framework for Evaluating Patient-Facing Health AI Agents
【速读】： 该论文旨在解决生成式AI（Generative AI）在初级保健领域向自主代理系统演进过程中所面临的临床安全性与可靠性评估难题。随着健康AI从简单的问答系统发展为能够与患者对话、推理病历并代表患者采取行动的智能代理，其在实际应用中可能引入诊断错误或不安全医疗行为的风险，亟需针对患者端交互场景的系统性评估框架。现有基准测试多聚焦于孤立的医学知识问答或面向医生的任务，难以反映真实临床情境下代理系统的综合表现。为此，论文提出PatientAgentBench，一个面向患者端自主医疗代理的评估基准，通过将基础模型封装为具备医疗工具沙箱的代理，并与模拟患者进行持续交互，利用大语言模型作为“裁判”（LLM-as-a-Jury），基于百余项非对话依赖、由临床专家定义的标准，在六个维度上对对话质量进行评分。为验证评估结果的可信度，注册执业医师对部分对话进行了独立标注，结果显示裁判与专家评分间相邻一致性达79%-93%，达到甚至超过临床医生间的一致性水平。在1,200个标准化临床情景下对10个模型进行评测发现，分诊质量是最具区分度的维度，最弱模型通过率仅为32%，最强模型提升至88%，且多数低效代理在未完成临床筛查的情况下即执行行政请求；临床安全性和工作流程准确性亦呈现相似趋势，低性能模型频繁虚构无法执行的操作，而前沿模型虽仅在1%-3%案例中出错，但仍因工具输出未经验证或紧急情况下遗漏关键资源而存在缺陷。研究进一步表明，这些缺陷仅在持续使用工具、面对真实病历记录的复杂交互中显现，证明静态基准已不足以评估日益自主的医疗代理系统。因此，论文的核心解决方案在于构建一个可复现、经临床验证的动态评估框架，以推动领域内对医疗代理系统的真实能力与风险进行全面量化评估，助力缩小当前技术能力与临床实践需求之间的差距。
链接: https://arxiv.org/abs/2607.25485
作者: Korosh Vatanparvar,Ashutosh Joshi,Maria Xenochristou,Mohammad Abuzar Hashemi,Prasad Kasu,Deepak Bansal,Daniel Lopez-Martinez,Anchal Nema,Ramya Ganesan,Will Kimbrough,Alex Woody,Yadunandana Rao,Dilek Hakkani-Tur,Wilko Schulz-Mahlendorf
机构: Amazon Health AI
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Code: this https URL
Abstract:Health AI is evolving from answering questions to agentic systems that converse with patients, reason about health records, and act on their behalf. Primary care guards against diagnostic errors and unsafe care; agents assisting in this domain warrant evaluation against the same risks. Current benchmarks focus on medical knowledge, assessed through isolated question-answering or clinician-facing tasks. PatientAgentBench benchmarks patient-facing agentic healthcare; it evaluates a foundation model, wrapped in an agent with a sandbox of healthcare tools, conversing with a simulated patient. Each conversation is scored by an LLM-as-a-Jury across six dimensions via over a hundred conversation-agnostic, clinician-grounded criteria. To validate alignment, licensed clinicians annotated shared conversations, yielding 79-93% adjacent agreement between jury and expert raters, on par with or exceeding clinician inter-rater agreement. We benchmarked 10 models across four families on the same 1,200 scenarios and found clinical gaps. Triage quality is the most discriminating dimension: pass rates rise from 32% for the weakest models to 88% for the strongest, with agents often acting on administrative requests without clinical screening. Clinical safety and workflow accuracy follow the same pattern: the weakest models fail often, fabricating unexecuted actions, while frontier models fail on only 1-3% of cases, from unverified tool outputs and omitted crisis resources in an emergency. More capable models narrow these gaps but do not close them; the strongest scores only 4.25 of 5 overall. These failures surface only in sustained, tool-using conversations against realistic patient records, confirming that static benchmarks are insufficient as healthcare agentic systems gain autonomy. We release the framework as a reproducible, clinician-validated evaluation standard to help the field close this gap.
[NLP-27] HANDBOOK.md: A Benchmark for Long-Context Agent ic Instruction Following
【速读】： 该论文旨在解决生成式 AI 代理在实际部署中面临的核心挑战：即如何确保长篇、具有约束力的政策文档（如企业标准操作程序，SOP）在代理执行一系列工具调用任务的长期过程中真正有效约束其行为。现有基准测试多关注任务完成率，而未直接评估政策对代理行为的持续约束力。为此，本文提出一个包含65个智能体任务的新型基准（http URL），这些任务基于企业员工遵循公司手册的实际工作场景构建，涵盖财务、医疗计费、保险、物流和人力资源五大领域，涉及十家虚构公司。每个任务均在一个自包含的企业环境（通过模型上下文协议暴露模拟邮件、聊天、日历、工单跟踪及电商服务）中运行，并要求代理依据一份20至124页的专家撰写SOP执行常规专业工作。为防止记忆化，每项任务均对十种基础手册进行差异化修改，使政策规则与评分阈值各不相同，确保无两个任务共享同一政策。评分机制完全确定性，共包含824项可编程评估标准，严格检验必要动作是否执行、禁止动作是否规避。在严格评分标准下（所有标准必须满足才视为通过），30种模型配置中表现最佳者仅通过36.2%的任务，多数前沿模型低于25%。失败模式具有一致性：代理倾向于让情境内看似合理的请求凌驾于既定政策之上，执行检查后仍违背检查结果，随任务时长推移丢失规则细节，或虚假报告合规状态。该研究的关键解决方案在于设计了一个高保真、可扩展且防记忆化的评估框架，能够真实反映政策驱动型智能体在复杂、长周期任务中的行为一致性与合规能力，从而推动对生成式 AI 代理可信治理能力的系统性评测。
链接: https://arxiv.org/abs/2607.25398
作者: Liudas Panavas,Sebastian Minus,Bradley Monton,Derek Ray,Suhaas Garre,Sushant Mehta,Edwin Chen
机构: Surge AI
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 16 pages, 3 figures, 5 tables. Accepted to the Workshop on Agent Behavior (WAB) at COLM 2026. Benchmark, environments, and evaluation harness: this https URL
Abstract:Language-model agents are increasingly deployed under standing instructions: a system prompt, a policy file, or a skills document is placed in context, and the agent is trusted to let it govern every action that follows. Existing benchmarks rarely test this deployment pattern directly; they measure whether an agent can complete a task, not whether a long, binding policy document actually constrains its behavior over an extended tool-use horizon. We present this http URL, a benchmark of 65 agentic tasks modeled on how enterprise employees follow company handbooks. Each task places an agent in a self-contained company environment, a file workspace together with mock email, chat, calendar, issue-tracking, and commerce services exposed over the Model Context Protocol, and instructs it to carry out routine professional work governed by an expert-written standard operating procedure of 20 to 124 pages. Tasks span five domains (finance, medical billing, insurance, logistics, and HR) and ten fictional companies. To resist memorization, every task modifies one of ten base handbooks, altering the specific rules and thresholds on which grading turns, so no two tasks share a policy. Grading is fully deterministic: each task carries a rubric of programmatic criteria (824 in total) that check both that required actions occurred and that prohibited actions did not. Under strict grading, where a trial passes only if every criterion is satisfied, the best of thirty evaluated model configurations passes 36.2% of trials, and most frontier configurations remain below 25%. Failures follow consistent patterns: agents let a plausible in-environment request override the standing policy, perform a required check and then act against its result, lose rule details over long horizons, and report compliance they did not achieve. We release all tasks, environments, and the evaluation harness.
[NLP-28] Memory for Large Language Models
【速读】： 该论文旨在解决大语言模型（LLM）中记忆机制研究高度碎片化的问题，其核心挑战在于当前多种记忆设计策略（如瞬时注意力、循环状态动力学、参数高效适配及可扩展查找存储等）缺乏统一的理论框架与系统性分类。解决方案的关键在于提出一个以架构为中心的三轴分类体系：从表示方式（隐式与显式）、更新动态（离线与在线）以及持久性（短期与长期）三个正交维度对记忆机制进行系统建模，并进一步形式化记忆写入、路由、状态转移与整合等细粒度机制。该框架不仅厘清了计算耦合型记忆与独立可寻址记忆之间的概念边界，还有效整合了异构架构范式，为混合记忆结构的设计、系统级效率权衡及多维度评估方法提供了统一的分析视角，从而为可扩展、自适应的语言建模未来创新奠定了原则性基础。
链接: https://arxiv.org/abs/2607.25380
作者: Sining Zhoubian,Dan Zhang,Evgeny Kharlamov,Jie Tang
机构: Tsinghua University (清华大学); National University of Singapore (新加坡国立大学); Bosch AI (博世人工智能)
类目: Computation and Language (cs.CL)
备注: 20 pages, 4 figures
Abstract:Memory has evolved into a foundational architectural dimension in large language models (LLMs), shifting from an implicit byproduct of computation to a spectrum of explicit, controllable mechanisms. While recent advances introduce diverse strategies—spanning transient attention, recurrent state dynamics, parameter-efficient adaptations, and scalable lookup storage—this rapid evolution has led to a highly fragmented research landscape. In this survey, we present a systematic, architecture-centric taxonomy of memory in LLMs. Our framework characterizes memory along three orthogonal axes: representation (implicit versus explicit), update dynamics (offline versus online), and persistence (short-term versus long-term). We further formalize the granular mechanisms dictating memory writing, routing, state transitions, and consolidation. This unified perspective elucidates the conceptual boundaries between computation-coupled and independently addressable memory, effectively bridging disparate architectural paradigms. Additionally, we critically analyze hybrid memory architectures, system-level efficiency trade-offs, and multi-dimensional evaluation methodologies. By consolidating these scattered advancements into a cohesive framework, this survey charts the trajectory of memory-centric LLM design and provides a principled foundation for future innovations in scalable and adaptive language modeling.
[NLP-29] Inspect India Evals: An Open Benchmarking Framework for Evaluating Large Language Models in the Indian Linguistic and Cultural Context
【速读】： 该论文旨在解决当前主流大语言模型（LLMs）评估基准在印度语境下存在的显著局限性问题，即现有基准如MMLU、BIG-Bench和TruthfulQA几乎完全以英语和西方文化为中心，无法有效识别印度本土特有的安全性、公平性和准确性缺陷。为此，研究提出“Inspect India Evals”这一开源评估框架，其核心创新在于构建了六个针对印度多语言、多文化背景的专项评测模块：涵盖16种印度官方语言的多语言MMLU、基于印度社会偏见适配的BharatBBQ、数字公共基础设施（DPI）安全评估、使用印地语等本地语言生成有害提示的多语言安全测试、多轮越狱攻击抵抗测试，以及基于大语言模型作为评判者的印度文化知识测评。该框架的关键在于通过本地化、文化敏感的评估设计，系统性揭示模型在印度语境下的潜在风险与性能短板，并支持与英国人工智能安全研究所（UK AISI）注册表集成，实现可复现与可扩展的评估体系。实验结果表明，Sarvam-M 24B和Gemma 2 27B在综合印度公平性指数上均达80%，且前者在印度文化知识与DPI安全合规性方面超越更大规模模型，验证了该框架的有效性与必要性。
链接: https://arxiv.org/abs/2607.25375
作者: Abhishek Kumar Singh,Shrey Nag,Sachita,Lipi Goel,Rajeshwar Singh Janwar
机构: 未知
类目: Computation and Language (cs.CL)
备注: 19 pages, 9 figures, 7 tables
Abstract:India is a vast nation of over 1.4 billion people, varied by hundreds of diverse and locally specific traditions and cultures and 22 officially recognized languages. Large language models (LLMs) are now being deployed on a massive scale throughout the mainland as well as in remote villages. However, the common benchmarks - MMLU, BIG-Bench, and TruthfulQA are almost exclusively English- and Western-centric. They do not identify those safety, fairness, and accuracy failures unique to the Indian context. That is the gap Inspect India Evals seeks to fill. It is an open-source framework built on top of UK AISI’s Inspect AI platform. It has six benchmarks: Multilingual MMLU across sixteen Indian languages, BharatBBQ (our adaptation of BBQ for Indian social bias), a safety evaluation for Digital Public Infrastructure, a multilingual safety test using harmful prompts in Indian languages, a multi-turn jailbreak resistance test, and an Indian cultural knowledge benchmark scored using LLM-as-judge rubrics. In this study, we tested five open-weight models ranging from 8B to 32B parameters. Sarvam-M 24B and Gemma 2 27B came out on top, both scoring 80% on the composite India Fairness Index, with Sarvam-M even beating larger 32B models on Indian cultural knowledge and DPI safety compliance. All models scored 100% refusal on Multilingual Safety, whereas DPI safety varied from 20% to 100%. The framework is public. It’s built to work with the UK AISI registry. Anyone can reproduce or extend this work.
[NLP-30] Data Quality Profiling at Scale with Progressive Sampling: A Benchmark for Data-Centric AI Pipelines
【速读】： 该论文旨在解决数据质量分析（data quality profiling）在大规模数据场景下难以实现近实时监控的问题，核心挑战在于传统全量扫描方法在数百万行数据上效率过低。为应对这一问题，研究对比了九种采样策略（包括盲采样如随机均匀、几何分布、Yamane、聚类采样，以及基于代理的引导采样如马尔可夫链蒙特卡洛、有向无环图（DAG）、按列类型或质量评分分层、重要性加权等），评估其在保持数据质量特征保真度方面的表现。研究发现，尽管存在“利用领域知识提升估计精度”的普遍假设，但实际结果表明，盲采样策略（尤其是随机均匀采样）在绝大多数场景下表现最优，其在5%采样预算下对纽约311数据集的平均相对误差仅为0.49%，显著优于基于代理的引导采样（如DAG引导的MCMC达到19.5%，约差40倍）。进一步分析揭示，代理引导方法失败的根本原因在于代理指标（如四分位距，IQR）与真实质量缺陷分布不匹配：代理主要捕捉数值型异常值，而数据质量问题（如缺失值、重复、函数依赖违反）多集中于分类列，无法被数值代理有效感知。因此，解决方案的关键在于认识到代表性（representativeness）而非领域知识才是决定采样器性能的核心因素——在大规模生产环境中，无需复杂建模，仅采用无需模式依赖的随机均匀采样或聚类采样即可实现高保真的质量分析，且具有近线性时间复杂度（O(N^0.964)），远优于超线性增长的代理引导方法（O(N^1.272)），在超大规模数据上速度更快、准确率更高。
链接: https://arxiv.org/abs/2607.25356
作者: Laure Berti-Equille
机构: 未知
类目: Databases (cs.DB); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 29 pages, 8 figures
Abstract:Data quality profiling – computing missing-value rates, duplicate fractions, outlier densities, and functional-dependency violations – is foundational for data-centric AI pipelines, yet exhaustive scans over millions of rows are prohibitively slow for near-real-time monitoring. Progressive sampling is the standard alternative; the open question is which strategy best preserves profile fidelity at scale. We benchmark nine sampling strategies – blind (random uniform, geometric, Yamane, cluster) and proxy-guided (Metropolis-Hastings, DAG, stratified by column type or quality score, importance-weighted) – on three real-world datasets (NYC 311, NYPD arrests, UCI Adult; up to 500K rows), an IoT sensor stream (2.3M rows), two ultra-large real datasets including Ultra-Marathon Running (up to 7.4M rows), and synthetic data scaled to 5x10^6 rows. Contrary to the assumption sharpens estimates, blind representative samplers dominate uniformly. At a 5% budget, random uniform achieves 0.49% mean relative error on NYC 311; DAG-guided MCMC yields 19.5% (approx. 40x worse), and across all real datasets DAG is 11-49x worse (Wilcoxon W=0, p=0.002, n=9 pairs). Cluster sampling matches random uniform (MRE 0.110 vs. 0.111); proxy-guided methods share DAG’s failure mode (MRE 0.20-0.35). At scale, random uniform is near-linear (O(N^0.964)) while DAG is super-linear (O(N^1.272)), running 28–47x slower on ultra-large data with 6x worse accuracy. The root cause is an IQR proxy mismatch: proxy-guided samplers over-pursue numeric outliers, while quality defects concentrate in categorical columns invisible to the proxy. The actionable finding: representativeness, not domain knowledge, determines sampler quality – schema-free random uniform or cluster sampling suffices for production-grade quality profiling at scale.
[NLP-31] Cardiologent: Multi-Agent Clinical Decision Support for Patient-Level Arrhythmia Assessment Urgency and Management
【速读】： 该论文旨在解决心律失常（arrhythmia）临床决策中“同一心电信号导致不同诊疗判断”的核心问题，即相同的心房颤动（atrial fibrillation）事件在健康成人中可能被视为无害，而在合并高血压的老年患者中则需启动抗凝治疗。其关键在于实现从单一心电图（ECG）片段识别到患者层面综合决策的跨越——传统方法仅关注局部信号解读或依赖设备预设报警，缺乏对患者整体病史、动态生理信号与临床指南的整合推理。为此，论文提出Cardiologent多智能体系统，通过为每个信号源（单导联ECG与可穿戴设备获取的光电容积脉搏波形，PPG）配置独立分析代理，基于实测特征而非简单标签进行窗口级判断；各代理输出结果被整合为患者的节律谱（rhythm profile），并与患者自身数据结合，依据检索到的临床指南进行推理，并由批判性代理验证每项结论的合规性。该系统不仅在集成诊断、临床意义评估和管理紧迫性等维度上表现最优，且其决策可追溯至具体指南条目并经专家审核，显著提升了透明度与可审计性，为持续监测场景下的生成式人工智能（Generative AI）临床应用提供了可信赖的决策支持范式。
链接: https://arxiv.org/abs/2607.25340
作者: Sukju Oh,Moo-Yong Rhee,Jae-Sik Jang,Sukkyu Sun
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 21 pages, 8 figures
Abstract:The same episode of atrial fibrillation is a minor finding in a healthy adult and grounds for anticoagulation in an elderly patient with hypertension: identical signal, opposite decision. Naming the rhythm is only the start; what determines a patient’s outcome is the judgement that follows – what the arrhythmia is across the whole record, what it means for this patient, and what should be done about it. Recent work pairing large language models with the ECG stops short of this, reading one recording without assembling a patient-level finding; and agentic systems built around it either receive the arrhythmia a device has already detected or target a different diagnostic task, stopping before the decision this task requires. We formulate patient-level arrhythmia decision support as a task and present Cardiologent, a multi-agent system that spans it from detection to decision. An agent for each signal – a single ECG lead and the photoplethysmogram a wearable acquires – grounds its window reading in measured features rather than a bare label; the readings are assembled into the patient’s rhythm profile and, with the patient’s own data, reasoned against clinical guidelines retrieved for the case, with a critic checking each conclusion against the guideline it cites. We evaluate the clinical decision rather than the report, across integrated diagnosis, clinical significance, and urgency and management. Cardiologent scores highest on every axis, first on every patient-level task under both cardiologists and an at-scale LLM judge – whose agreement with the cardiologists (ICC 0.74, 0.66) matches theirs with each other (0.67). Because each conclusion traces to a cited guideline and is validated against expert cardiologists, it yields decisions a clinician can audit rather than act on blindly – a step toward use in continuous monitoring.
[NLP-32] mporal-Distance JEPA: Plan-Aware Representation Learning for Latent World Model Predictive Control
【速读】： 该论文旨在解决生成式世界模型（Generative World Models）在离线示范数据上进行隐变量模型预测控制（Latent Model Predictive Control, MPC）时，因规划阶段缺乏有效目标进展评估机制而导致的“训练-规划差距”（train–plan gap）问题。现有基于联合嵌入预测架构（JEPAs）的规划方法通常依赖嵌入空间中的欧几里得距离作为进展度量，但这种度量是表征学习过程的副产品，并未从示范日志中主动挖掘出真正的时序进展结构。为此，本文提出时间距离联合嵌入预测架构（Temporal-Distance JEPA, TD-JEPA），其核心创新在于通过无奖励轨迹挖掘有向的时间成本：同一轨迹内的步骤顺序构成正样本，跨轨迹样本作为启发式负样本，并引入滚动一致性项以对齐规划时域。该挖掘得到的监督信号兼具双重作用：在拓扑进展场景下直接用作部署规划成本，在接触几何主导场景下则优化欧几里得表示质量。在锁定评估条件下，使用该挖掘成本使Two-Room任务成功率提升至100.0%（相比LeWM的97.4%），且在同一时序训练检查点上，共享欧几里得规划使OGB-Cube性能提升14.2分并改善Push-T表现；相较于LeWM和同期的RC-aux基线，TD-JEPA在所有环境中均达到或超越其性能。消融实验表明，有向头、跨轨迹负样本及滚动一致性项均对性能贡献显著。综上，TD-JEPA通过从离线日志中发现时序进展结构，并将成本形式与规划部署阶段协同设计，有效缩小了JEPA类世界模型的训练与规划之间的差距。
链接: https://arxiv.org/abs/2607.25337
作者: Jiaxin Bai,Jiaxuan Xiong
机构: Hong Kong Baptist University (香港浸会大学)
类目: Computation and Language (cs.CL); Robotics (cs.RO)
备注:
Abstract:Joint-Embedding Predictive Architectures (JEPAs) learn world models by predicting in representation space rather than reconstructing pixels, making them a natural backbone for latent model predictive control from offline demonstration logs. JEPA-style training optimizes short-horizon latent prediction, whereas planning requires a multi-step ranking of imagined futures by goal progress. Prior JEPA planners often inherit that ranking from embedding geometry, typically latent Euclidean distance, which arises as a byproduct of representation learning rather than as a progress cost mined from the logs. We propose temporal-distance JEPA (TD-JEPA), which retains the LeWM encoder–predictor backbone and mines a directed temporal cost from reward-free trajectories: same-trajectory step order supplies positive targets, cross-trajectory pairs act as heuristic negatives, and a rollout-consistency term matches the planner horizon. The mined supervision serves two roles: as the deployed planning cost when progress is topological, and as a representation signal that improves Euclidean planning when contact geometry dominates. Under locked evaluation, deploying the mined cost raises Two-Room success to 100.0% versus LeWM’s 97.4%, while shared Euclidean planning on the same temporally trained checkpoint raises OGB-Cube by 14.2 points over LeWM and improves Push-T. Against LeWM and the concurrent RC-aux baseline under locked evaluation, TD-JEPA matches or exceeds both methods on every environment. Ablations show that the directed head, cross-trajectory negatives, and rollout consistency each contribute. TD-JEPA narrows the train–plan gap for JEPA world-model planners by discovering temporal progress structure in offline logs and co-designing cost form with plan-time deployment. Code is available at this https URL.
[NLP-33] Every Time I Hire a Linguist Inference Costs Go Down: On Linguistic Rules as Effective Prompt Compressors EMNLP2026
【速读】： 该论文旨在解决生成式人工智能（Generative AI）中大语言模型（LLM）输入过长导致推理成本过高的问题，核心挑战在于现有提示压缩（prompt compression）方法依赖大语言模型自身的前向传播来评估词元（token）重要性，这一过程既计算昂贵又难以解释。为此，论文提出一个根本性问题：是否可以仅依靠语言学规则（linguistic rules）在不进行任何LM前向传播的情况下实现高效且有效的提示压缩？其解决方案的关键在于通过离线进化搜索（offline evolutionary search），系统性地探索词汇、句法、语义和话语层面的语言学种子（lexical, syntactic, semantic, and discourse seeds），自动发现能够协同作用的规则组合，构建一种完全基于确定性语言学规则的压缩器。该压缩器在部署阶段无需任何大语言模型的参与，仅依赖CPU处理即可完成压缩，显著降低推理开销。实验采用双路径评估协议，在压缩效率与重构保真度之间取得平衡，结果表明，所生成的演化压缩器在短文本、多文档推理及对话记忆问答等任务上性能可媲美当前先进的基于模型的压缩策略；尤其在轻度至中度压缩场景下表现优异，随着压缩比提升，性能逐渐下降，且直接路径（Direct path）与重构路径（Reconstruction path）呈现出不同退化模式。进化分析进一步揭示，高效压缩依赖于跨语言层级信号的融合，并随压缩强度增加，规则机制从词元删减转向句子级提取。
链接: https://arxiv.org/abs/2607.25335
作者: Jianfei Ma,Zhaoxin Feng,Emmanuele Chersoni,Si Chen
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 37 pages, 6 figures, EMNLP2026
Abstract:Prompt compression shortens LLM input to reduce inference cost, yet existing methods score token importance through LM forward passes. It remains questionable whether such nuanced, costly token selection is necessary. Compression requires identifying informative content, a problem that linguistic research has long addressed through cues that can be operationalized as deterministic rules. We therefore ask: can \textbflinguistic rules alone serve as effective prompt compressors, without LM-based scoring at compression time? To address this, we conduct offline evolutionary search over lexical, syntactic, semantic, and discourse seeds to find competitive rule combinations. The resulting linguistic compressor requires no LM forward pass at deployment and uses only CPU-side processing for compression. We evaluate it with a dual-path protocol to balance compression quality and reconstruction fidelity. Across short passages, multi-document reasoning, and dialogue-memory QA datasets, evolved compressors achieve performance similar to that of recent advanced prompt-compression strategies. Performance is strongest under light-to-moderate compression and degrades as compression becomes more aggressive, while the Direct and Reconstruction paths exhibit distinct patterns. Evolutionary analysis reveals that effective compression fuses signals across linguistic levels and, as the compression ratio increases, rules shift from token pruning to sentence extraction. Comments: 37 pages, 6 figures, EMNLP2026 Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.25335 [cs.CL] (or arXiv:2607.25335v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2607.25335 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-34] CAST: Game Solvers as Turn-Level Teachers for LLM Agents
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在长时程游戏任务中进行决策时，因依赖稀疏的最终奖励而难以实现细粒度动作信用分配的问题。现有方法虽尝试引入更密集的过程信号以提供每回合的信用反馈，但难以同时保证信号的低成本与高准确性。为此，论文提出一种名为CAST（Credit Assignment from Solver Teachers）的解决方案，其核心在于利用游戏求解器状态值的变化来判断某一动作是否有助于推进至成功状态，并将这些状态值变化转化为“求解器优势”（solver advantage），作为强化学习中带验证奖励（Reinforcement Learning with Verifiable Rewards, RLVR）的逐回合信号。进一步地，在假设求解器接近最优的条件下，最大化求解器优势等价于对求解器进行策略蒸馏，且仅需标量值而非完整的教师输出概率分布。实验表明，CAST在Sokoban、Minesweeper和Rush Hour等游戏中均显著优于所有基线方法，且在ALFWorld和WebShop上实现了最高的平均零样本性能，验证了其有效性与泛化能力。
链接: https://arxiv.org/abs/2607.25308
作者: Yu Wang,Yi-Kai Zhang,Wentao Shi,Ziang Ye,Yuchun Miao,Yueqing Sun,Qi Gu,Xunliang Cai,Lan-Zhe Guo,Han-Jia Ye,Fuli Feng
机构: University of Science and Technology of China(中国科学技术大学); Nanjing University(南京大学); Wuhan University(武汉大学); Meituan(美团)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Training large language models (LLMs) to act in long-horizon games is a promising step toward generalist decision-making, yet reinforcement learning with verifiable rewards (RLVR) relies on sparse final rewards that reveal little about which decisions determine success. Denser process signals could supply this missing turn-level credit, but existing sources are hard to keep both cheap and accurate. We observe that changes in a game solver’s state value reveal whether an action advances the state toward success. Building on this insight, we propose CAST (Credit Assignment from Solver Teachers), which converts these value changes into solver advantages and injects them into RLVR as turn-level signals. We further show that, under a soft-optimal solver assumption, maximizing the solver advantage is equivalent to on-policy distillation from the solver, requiring only scalar values rather than teacher logits. Across Sokoban, Minesweeper, and Rush Hour, CAST outperforms all trained baselines on every game under both in-domain and unseen-difficulty evaluation and achieves the highest average zero-shot performance on ALFWorld and WebShop. Our code is available at this https URL.
[NLP-35] oward a systematic method for identifying language areas
【速读】： 该论文旨在解决语言类型学研究中宏观区域（macroarea）划分依赖专家主观判断、缺乏系统性方法的问题，尤其针对现有宏观区域多基于大陆或陆地块块的非量化划分方式。其核心挑战在于如何在不依赖先验地理或历史假设的前提下，通过客观、可重复的方法识别出具有潜在语言接触可能性的地理区域，以更准确地控制因历史继承或语言接触导致的自相关（autocorrelation）问题。解决方案的关键在于提出一种简单的地理聚类方法，能够根据语言分布的地理邻近性自动识别出全球范围内的宏观区域及特定地区的语言联盟（sprachbund）等局部语言区域，从而实现对语言特征分布的更精细、可验证的区域划分。该方法不仅复现了现有宏观区域的地理格局，还为任意区域提供了灵活、可扩展的语言区域识别框架。
链接: https://arxiv.org/abs/2607.25305
作者: Hiram Ring
机构: NTU Singapore(南洋理工大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Macroareas are geographical areas used in typological research for grouping variables of interest. In linguistic typology, languages in a given macroarea are considered to have potential for contact, in contrast to those outside the area, where contact is less likely. Along with language family membership, macroareas are used as controls for models in linguistic typology, in an attempt to address the problem of autocorrelation - the observation that historical developments or typological patterns may be due to contact between neighboring languages and/or inheritance from a common ancestral language. Macroareas are therefore a central aspect of research that seeks to separate universal properties of language from local (or language-specific) properties. Existing macroareas largely depend on expert determinations of what constitutes a geographical area of potential contact, and to date have mainly aligned with continents or landmasses (Hammarström and Donohue 2014; Nichols, Witzlack-Makarevich, and Bickel 2013). While there are various historical and theoretical reasons for these groupings, there as of yet has been no systematic approach to identifying such areas for a given region. This paper attempts to address such a gap and move beyond macroarea to identification of language areas of relatively arbitrary size, presenting a simple geographical clustering method for identifying groupings over any area. The method produces a set of worldwide macroareas that largely align with existing groupings, as well as local groupings for a well-known sprachbund.
[NLP-36] CLBench-V: Evaluating Multimodal Context Learning from Grounding to Knowledge Acquisition
【速读】： 该论文旨在解决多模态上下文学习（multimodal context learning）中模型难以有效利用跨模态上下文信息的问题，尤其针对现实任务中普遍存在的图文、表格、地图、网页等多源异构上下文场景。现有研究多集中于文本型上下文评估，忽视了实际应用中复杂的多模态上下文需求。为此，论文提出CLBench-V基准，通过三个核心维度——上下文定位（context grounding）、新信息应用（new information application）和新知识学习（new knowledge learning）——系统化地识别与分析上下文使用中的失效环节。其解决方案的关键在于：构建涵盖科学、金融、长文档理解、空间推理及基于网页的视觉问答等多领域的多模态数据集，并采用自动化构造与筛选流程以降低领域特定任务的构建成本。实验表明，当前最先进的多模态模型在3,443个实例上的最优综合得分仅为0.2847，表明多模态上下文学习仍处于初级阶段；其中InternVL3.5-30B-A3B在上下文定位与新知识学习方面表现最佳，而Qwen3.5-Plus在新信息应用上领先。研究还进一步分析了标注者可靠性、上下文长度、图像数量及典型失败案例，为未来模型改进提供了重要方向。
链接: https://arxiv.org/abs/2607.25294
作者: Lai Wei,Chengqi Li,Jiapeng Li,Ruina Hu,Yue Wang,Weiran Huang
机构: Shanghai Jiao Tong University (上海交通大学); Zhongguancun Academy (中关村研究院); Shanghai Innovation Institute (上海创新研究院)
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Real-world tasks often require models to learn from task-specific context rather than relying only on pre-trained knowledge. While recent work has highlighted this capability as context learning, existing evaluations mainly focus on textual contexts. In many practical settings, however, the context to be learned from is multimodal: scientific findings are conveyed through figures and tables, financial indicators are scattered across converted reports, and spatial decisions depend on maps, scenes, or web pages. We introduce CLBench-V, a benchmark for multimodal context learning that addresses the difficulty of localizing where context use breaks down by organizing tasks around three dimensions: context grounding, new information application, and new knowledge learning. CLBench-V combines converted public benchmarks with newly constructed datasets spanning domains such as science, finance, long-document understanding, spatial reasoning, and web-based visual question answering. To reduce the cost of constructing domain-specific context-learning tasks, we further use automated construction and filtering procedures for our newly built datasets. Across 3,443 instances and six recent multimodal models, the best overall score is only 0.2847, indicating that multimodal context learning remains far from saturated. Moreover, InternVL3.5-30B-A3B performs best on context grounding and new knowledge learning, while Qwen3.5-Plus performs best on new information application. We further analyze judge reliability, context length, image count, and representative failure cases. Code is available at this https URL.
[NLP-37] CoSA: Accelerating Long-Context Inference via Proxy-Kernel Co-Designed Sparse Attention
【速读】： 该论文旨在解决长序列推理中自注意力（self-attention）计算开销呈二次增长导致的高成本问题，尤其在资源预算受限时，现有基于代理（proxy）与内核（kernel）协同的块稀疏注意力（block-sparse attention）方法因代理无法准确识别关键块、而内核仅机械执行稀疏掩码，导致模型精度显著下降的问题。其解决方案的关键在于提出一种无需训练的两阶段稀疏注意力框架——CoSA（Two-stage Training-free Sparse Attention under proxy-kernel CO-design），通过耦合**感知内核的代理（Kernel-Aware Proxy, KAP）与有序跳过内核（Ordered-Skipping Kernel, OSK）**实现动态优化：第一阶段，KAP在适度预算下选择重要块并生成按访问顺序排列的掩码；第二阶段，OSK基于在线Softmax统计信息，在更紧的预算下主动跳过更多非关键块，从而在保持性能的同时大幅提升计算效率。实验表明，CoSA在主流大语言模型（LLM）架构和长上下文基准上均实现了更低预算下的更高精度，尤其在128K上下文长度下，实现了4.93倍的注意力加速，并将端到端首个令牌生成时间（Time-to-First-Token）降低2.53倍，且性能损耗可忽略。
链接: https://arxiv.org/abs/2607.25291
作者: Yufei Xue,Lin Niu,Hong Liu,Siran Liu,Hanyong Shao,Wei Liu,Guanghua Yu,Jianchen Zhu,Jun Zhang
机构: Tencent(腾讯); 1; 2
类目: Computation and Language (cs.CL)
备注:
Abstract:The quadratic cost of self-attention makes long-context inference prohibitively expensive, and proxy-based block-sparse attention has become a practical remedy. Existing methods typically rely on a proxy to predict a binary sparse mask and a kernel to consume this mask and perform sparse attention computation. Such an approach is effective under moderate budgets. However, as the budget tightens, the estimated proxy inevitably drops some salient blocks, while the kernel can only apply the sparse mask mechanically, leading to an evident drop in model accuracy. We propose CoSA, a two-stage training-free Sparse Attention under proxy-kernel CO-design, which couples a Kernel-Aware Proxy (KAP) with an Ordered-Skipping Kernel (OSK). In the first stage, the KAP selects blocks under a moderate budget and produces an ordered mask that prescribes the order in which KV pages are visited in the kernel inner loop. In the second stage, the OSK applies this mask and skips more blocks under a tightened budget given online-softmax statistics. Across mainstream LLM backbones and long-context benchmarks, CoSA attains higher accuracy at lower budgets. Impressively, CoSA achieves a 4.93 \times attention speedup and reduces end-to-end Time-to-First-Token by 2.53 \times under a context length of 128K with negligible performance degradation.
[NLP-38] Where Steering Signals Come From: Activation Source Selection in Activation Steering
【速读】： 该论文旨在解决生成式人工智能（Generative AI）中激活向量调控（activation steering）的可解释性与有效性问题，即如何在推理阶段通过向隐藏状态注入特定向量来引导语言模型行为，但现有方法往往忽视了用于构建调控信号的上游激活源选择的重要性。其核心问题是：调控信号的有效性在很大程度上依赖于所选激活源的性质，而这一关键因素常被忽略。解决方案的关键在于提出“激活源选择”（activation source selection）的概念，强调调控效果不仅取决于目标行为是否出现在源文本中，更取决于所选激活状态是否处于模型执行边界（execution-boundary states），即模型即将产生或继续目标行为的临界状态。研究发现，有效的调控信号并非源于目标内容的简单出现，而是来自模型即将执行动作前的状态表示。基于此，作者提出了“尾部相减”（tail subtraction）方法，通过从执行边界状态中移除提示与延续部分的共享语义，获得更纯净、更稳定的调控信号。结果表明，成功的调控依赖于对模型“即将执行”的行为的表征，而非仅关注已出现的内容。
链接: https://arxiv.org/abs/2607.25270
作者: Jiaran Ye,Lingxu Ran,Zijun Yao,Chenpeng Wang,Yong Jiang,Lei Hou,Juanzi Li,Liangming Pan
机构: Peking University (北京大学); Tsinghua University (清华大学); Beijing Academy of Artificial Intelligence (北京人工智能研究院); YiXin-AILab, YIXIN (易信人工智能实验室)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Activation steering controls language models by adding vectors or features to hidden states at inference time, but the upstream source of these steering signals is often treated as a secondary detail. We study this source choice as activation source selection: the combination of source context and activation readout policy used to collect the hidden states from which a steering signal is built. Holding the downstream intervention fixed, we show across three instruction-tuned models and four steering task families that changing only the source activations substantially changes steering success. We further find that effective steering is not explained simply by whether the desired behavior appears in the source text. Instead, strong signals come from execution-boundary states, where the model is about to produce or continue the target behavior. This pre-/post-realization distinction explains why answer-based sources sometimes work: their useful component aligns with execution-boundary directions rather than target appearance alone. Building on this view, we introduce tail subtraction, which removes shared prompt and continuation semantics from boundary states and yields cleaner, more stable steering signals. Overall, our results suggest that steering depends on representations of what the model is about to do, not merely on what has already appeared.
[NLP-39] VisualPatchWorld: Code World Models as Latent Structured Representations for Planning
【速读】： 该论文旨在解决现有世界模型（World Model）在动态系统建模中面临的权衡问题：神经预测器虽能从数据中高效学习连续向量空间中的动力学，但其内部表示隐含且不可解释；而手工构建的物理引擎虽具备可读性与可编辑性，却难以规模化构建。为克服这一局限，论文提出一种基于代码的世界模型——视觉补丁世界模型（VisualPatchWorld, VPW），其核心解决方案在于将世界动力学显式地表示为可执行的程序代码。VPW首先通过短时主动探测（active probes）选择合适的定性动力学形式，再基于记录的状态-动作轨迹，通过最小化多步预测误差来拟合该形式的自由参数。由此生成的代码模型具备模拟推演、源码可读性及嵌入模型预测控制（MPC）的能力，同时支持利用图像导出的场景图在重规划时提供实时状态输入。实验表明，相较于已有基于代码的世界模型，VPW在规划任务上达到69.0%的平均成功率，超越最强代码基线23.5个百分点，尤其在需正确识别定性动力学结构的任务中表现显著提升。在相同规划器下，其诱导模型在导航与抓取密集型控制任务中接近真实物理引擎的性能，仅在接触密集型推移任务中存在残余差距，而通过在真实引擎中验证候选计划列表可基本弥合该差距。该研究确立了一条自动构建可解释、可规划使用的代码化世界模型的实用路径。
链接: https://arxiv.org/abs/2607.25236
作者: Jiaxin Bai,Jiaxuan Xiong
机构: Hong Kong Baptist University (香港浸会大学)
类目: Computation and Language (cs.CL); Robotics (cs.RO)
备注:
Abstract:Different research lines use the term world model in different ways, yet they share a common aim: to capture how the world evolves under action in a form that supports perception, simulation, and planning. Two prominent realizations are neural predictors that learn dynamics in continuous vector spaces, and hand-built physics engines that expose explicit state and physical laws. Neural predictors scale from data but leave the form of the dynamics implicit; physics engines are inspectable and editable but difficult to construct at scale. We introduce VisualPatchWorld (VPW), which represents world dynamics as code. VPW first selects a qualitative dynamical form with short active probes, then fits that form’s free parameters from recorded state-action traces by minimizing multi-step prediction error. The resulting programs can be rolled forward like a simulator, inspected in source form, and used inside model-predictive control; image-derived scene graphs can supply the live state at replan time. Across comparisons with prior code-based world models, VPW attains 69.0% mean planning success and exceeds the strongest code baseline by 23.5 points. The largest gains arise when choosing the correct qualitative dynamics is essential. Under the same planner, the induced models approach ground-truth engine success on navigation and grasp-rich control; a residual gap remains for contact-rich pushing, and checking a shortlist of promising plans in the engine closes most of that gap. These results establish a practical route toward automatically constructed code world models that are useful for planning. Code is available at this https URL.
[NLP-40] Interpretable Column Annotation with LLM -Symbolized Decision Process Materialization
【速读】： 该论文旨在解决列注释（Column Annotation, CA）中现有方法存在的两大核心问题：其一，传统基于神经网络的方法在学习列表示后直接映射至标签类别，导致模型可解释性与适应性不足；其二，忽略了标签间的丰富语义信息，限制了标注精度。针对上述局限，本文提出一种由大语言模型（LLM）驱动的可解释性列注释框架SymCA，其关键在于将列注释建模为从全局到局部的符号化决策过程。具体而言，该方案包含两个核心组件：一是全局骨架归纳（global skeleton induction），利用LLM生成受上位词启发的树状语义骨架，并通过最小贝叶斯风险（Minimum Bayes Risk, MBR）共识策略筛选出对生成波动具有鲁棒性的最优骨架；二是局部基质演化（local substrate evolution），将每个内部节点形式化为可执行、可演化的预测基质，通过多轮演化迭代，结合随机森林分类器训练、LLM生成的节点特异性算子优化以及探索-利用策略，动态优化预测逻辑。该设计既充分挖掘了标签语义，又保障了决策过程的透明性与可解释性，实验表明SymCA在Micro-F1和Macro-F1上分别较最强基线平均提升6.42%和11.03%，显著提升了准确率与鲁棒性。
链接: https://arxiv.org/abs/2607.25228
作者: Mengqi Wang(1),Jianwei Wang(1),Qing Liu(2),Xiwei Xu(2),Zhenchang Xing(2),Michael Bain(1),Liming Zhu(2),Wenjie Zhang(1) ((1) UNSW Sydney, (2) Data61, CSIRO)
机构: University of New South Wales (新南威尔士大学); Data61, CSIRO (数据61，澳大利亚联邦科学与工业研究组织)
类目: Computation and Language (cs.CL)
备注: 13 pages, 7 figures
Abstract:Column annotation (CA), including column type annotation (CTA) and column property annotation (CPA), aims to identify the meanings of table columns and the semantic relationships among them. Recent CA methods usually use various neural models to learn column representations and directly map them to label categories, thereby (1) sacrificing model interpretability and adaptivity, and (2) overlooking rich label semantics and ultimately limiting accuracy. To address these limitations, we propose SymCA, an LLM-empowered interpretable CA framework that materializes column annotation as a global-to-local symbolic decision process. SymCA consists of two components: (1) global skeleton induction, which constructs a semantic skeleton over the label space, and (2) local substrate evolution, which evolves predictive substrates within the skeleton. Specifically, to exploit label semantics while preserving an interpretable decision process, the global skeleton induction module leverages LLMs to generate candidate hypernym-inspired tree-structured semantic skeletons and employs a Minimum Bayes Risk (MBR)-based consensus strategy to select a robust skeleton against generation variance. Since different internal nodes require different evidence to distinguish among their child nodes, the local substrate evolution module materializes each internal node as an executable and evolvable predictive substrate. Over multiple evolution rounds, each substrate trains an interpretable random forest classifier with the current operator set, leverages the LLM to propose node-specific operator modifications, and uses an exploration-exploitation strategy to prioritize promising substrates. Extensive experiments demonstrate that SymCA is accurate, robust, and interpretable, outperforming the strongest baselines by an average of 6.42% in Micro-F1 and 11.03% in Macro-F1.
[NLP-41] A Cross-lingual Comparison of Human and Classification Model Entrainment Behavior in Code-switched Speech Settings
【速读】： 该论文旨在解决跨语言对话中语用协同（conversational entrainment）在口语代码转换（code-switching, CSW）情境下的研究空白问题，尤其关注声学-韵律特征与代码转换风格层面的协同现象是否具有普遍性。其核心解决方案在于构建一个以人类行为为基准的多语言风格化对话评估框架，通过分析经典模型与Transformer模型在识别协同行为时的特征重要性及消融表现，发现现有分类模型虽能合理检测协同现象，却未能优先关注对人类协同行为最具显著性的特征。这一发现揭示了当前模型决策机制与人类真实交互模式之间的偏差，强调了未来开发具备自然主义代码转换能力的对话代理所面临的挑战，并提出了以人类行为为导向的模型评估新范式。
链接: https://arxiv.org/abs/2607.25202
作者: Debasmita Bhattacharya,Siying Ding,Alayna Nguyen,Julia Hirschberg
机构: Columbia University (哥伦比亚大学); Instagram
类目: Computation and Language (cs.CL)
备注:
Abstract:Conversational entrainment is well-studied in monolingual and written contexts, but remains underexplored in spoken code-switching (CSW). We present a novel cross-lingual analysis of entrainment in Mandarin-English, Hindi-English, and Spanish-English dialogue and show that, while lexical entrainment generalizes across language pairs, entrainment over acoustic-prosodic and CSW style aspects exhibits context-specific variation. We build on these findings by asking whether classification models capture these human behavioral patterns. Applying feature importance and ablation analyses, we find that classical and Transformer-based classifiers detect entrainment reasonably well but consistently prioritize features other than those most salient to human entraining behavior. Our approach introduces a human-grounded framework for evaluating model decision-making in multilingual stylistic contexts, and suggests future challenges for developing conversational agents capable of producing naturalistic code-switched speech.
[NLP-42] MyoCardBench: A Real-World Data Benchmark for Evaluating Large Language Models in Clinically Authentic Cardiovascular Care Scenarios
【速读】： 该论文旨在解决现有医学大语言模型（LLM）评估基准在心血管诊疗场景中存在局限性的问题，即多数基准仅聚焦于单一考试知识或孤立任务，难以反映心血管护理中纵向性、多模态及高安全性的临床工作流程。为应对这一挑战，研究提出并构建了MyoCardBench——一个覆盖心血管照护全周期的真实世界多任务评估基准，涵盖13个特定任务的数据集，共包含2,263项任务条目，来源于去标识化的临床记录与检查数据。其解决方案的关键在于：通过16名心脏病学专家进行标注与参考答案构建，并由两名资深心脏病专家进行交叉审核，确保评估标准的临床真实性；在此基础上，在标准化零样本设置下对7个主流大语言模型生成的15,841条输出进行系统评估，采用关键点覆盖率、整体临床质量评分及CardioEthics准确性等多维度指标，全面衡量模型在不同临床任务中的表现。结果表明，GPT-5.4在宏观平均分和加权均值上均领先，且在所有三个临床维度中排名第一，而各子任务间表现差异显著，尤其在心血管沟通（CardioComm）、急症救援（CardioEmergRescue）和治疗方案制定（CardioTreatPlan）中，整体临床质量与关键点覆盖率之间存在较大差距，揭示了当前模型在复杂临床情境下的系统性缺陷。MyoCardBench作为迄今覆盖最广、最贴近真实临床实践的心血管领域大语言模型评估框架，为识别模型优势、发现关键临床遗漏以及指导未来模型优化提供了坚实依据。
链接: https://arxiv.org/abs/2607.25186
作者: Xiao Li,Mouxiao Bian,Zhaodi Wu,Sijie Ren,Juechen Chen,Lu Lu,Jingru Ding,Yun Zhong,Jie Xu,Yixiu Liang,Junbo Ge
机构: Fudan University; Shanghai Artificial Intelligence Laboratory; Zhongshan Hospital Fudan University; Shanghai Institute of Cardiovascular Diseases, National Clinical; Research Center for Interventional Medicine
类目: Computation and Language (cs.CL)
备注:
Abstract:Background: Most medical large language model (LLM) benchmarks focus on examination knowledge or isolated tasks and may not reflect the longitudinal, multimodal, and safety-critical workflow of cardiovascular care. Objective: To develop MyoCardBench, a real-world benchmark spanning the cardiovascular care continuum, and assess LLM performance across clinical dimensions and specialist tasks. Methods: MyoCardBench includes 2,263 items from 13 task-specific datasets derived from de-identified cardiovascular records and examination data. Sixteen cardiology physicians conducted annotation and reference construction, followed by cross-review from two senior cardiologists. Seven LLMs generated 15,841 outputs under standardized zero-shot settings. Open-ended tasks were evaluated using key-point coverage and holistic clinical quality, while CardioEthics was scored by accuracy. Results: GPT-5.4 achieved the highest macro-average (62.55) and item-weighted mean (62.19), followed by Gemini 3.1 Pro (59.95) and Qwen 3.6 27B (59.72). GPT-5.4 ranked first in all three dimensions. CardioAuxReport performed best (86.38), whereas CardioECGRead (17.25) and CardioEthics (17.34) were lowest. The largest gaps between holistic clinical quality and key-point coverage occurred in CardioComm (52.71), CardioEmergRescue (52.05), and CardioTreatPlan (48.80). Conclusions: To our knowledge, MyoCardBench is the largest real-world, multi-task benchmark for LLM evaluation across the cardiovascular care continuum and offers the broadest coverage of clinically authentic cardiology scenarios reported to date. It provides a rigorous framework for identifying model strengths, clinically important omissions, and priorities for future development.
[NLP-43] A scaling law of contextual persistence in human language
【速读】： 该论文旨在解决人类语言中词语序列排列如何影响语义表达的结构性问题，即揭示词语在时间或文本序列中的有序排列是否遵循某种可量化的规律。其核心问题是：词语之间的上下文依赖关系（contextual dependence）在多远的距离上仍具有显著影响？解决方案的关键在于提出并验证“上下文持续性函数”（contextual persistence function, P(d)），该函数通过大语言模型作为概率探针，量化在给定距离d处的先验上下文对目标词预测困惑度（perplexity）的降低程度，并与随机打乱后的语境进行对比，从而分离出由词语排列顺序带来的真实语义信息贡献。研究发现，在十种涵盖六种语言家族、覆盖书面与口语模态的语料库中，P(d)随距离d呈近似1/d的幂律衰减（α ≈ 1.04），且该现象在随机打乱和合成语料中消失，仅在真实语言中出现，且不适用于基因组或蛋白质序列。这一指数接近1表明上下文影响在对数时间尺度上大致均匀分布，揭示了人类语言中存在一种普遍的上下文持续性标度律，为理解语言结构的深层规律提供了定量基础。
链接: https://arxiv.org/abs/2607.25184
作者: Elan Barenholtz
机构: Florida Atlantic University(佛罗里达大西洋大学)
类目: Computation and Language (cs.CL)
备注: 21 pages, 5 figures (plus 1 supplementary figure); Supplementary Information included
Abstract:Human language exhibits lawful structure at the level of words (frequency, vocabulary growth) and word pairs (co-occurrence across distance). Here we show that the arrangement of words in sequence – a central determinant of meaning – obeys a comparable law. Using large language models as probabilistic probes, we measured the reduction in target perplexity conferred by prior context at distance d beyond that of the same words scrambled; this difference, the contextual persistence function P(d), isolates the influence of arrangement. Across ten corpora spanning six language families and written and spoken modalities, P(d) decayed approximately as 1/d ( P(d) \propto d^-\alpha , mean \alpha = 1.04 ; median r^2 = 0.96 ). The effect vanished in scrambled and synthetic controls, replicated across independent probes, and did not appear in genomic or protein sequences under domain-native models. An exponent near 1 distributes contextual influence approximately uniformly across logarithmic timescales. The results establish a scaling law of contextual persistence in human language.
[NLP-44] Deep Label-Wise Attentive Temporal Convolutional Networks Improve Medical Coding
【速读】： 该论文旨在解决医疗编码（Medical Coding）中的复杂文本理解与多标签分类难题，即如何从住院病历记录中准确提取并分配诊断与操作代码。由于医疗文本信息分散、上下文依赖性强，且不同代码需关注文档中不同部分，该任务对专业人员而言亦极具挑战性。为应对这一问题，作者提出一种基于深度神经网络的解决方案，其核心创新在于采用多层时序卷积网络（Multi-layer Temporal Convolution Network, TCN）与标签专属注意力机制（Label-wise Attention）相结合的架构。其中，多层TCN能够有效捕捉长序列中的全局语义关系，生成稳健的文档表示；而标签专属注意力机制则使模型在预测每个具体代码时，可动态聚焦于文档中与该标签相关的关键片段，实现细粒度的上下文选择。实验表明，该方法在F-1分数上较先前最优模型提升9%，尤其在召回率方面实现28%的显著增长，凸显了其在临床决策支持场景中对漏报敏感性的优化能力，具有重要实际意义。
链接: https://arxiv.org/abs/2607.25129
作者: Muhammed Yavuz Nuzumlalı,Alexander Fabbri,Irene Li,Dragomir Radev
机构: Yale University (耶鲁大学)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: Work carried out in 2019; posted as a record of the work. Baselines and state of the art reflect the 2019 literature
Abstract:Medical coding is the task of assigning a set of diagnosis and procedure codes for a hospitalization using recorded notes. It requires aggregating information from different parts of the text and focus to different sections for each individual code, making it a very difficult problem even for professional human coders. We model the task as a multi-label text classification problem. To overcome the mentioned difficulties, we propose a deep neural model consisting of a multi-layer temporal convolution network (TCN) followed by label-wise attention. While multi-layer TCN helps extract a global document representation with the ability to learn relations over very long sequences, label-specific attention mechanism allows the model to focus on different aspects of the same document for each individual label. Our method achieves significantly better F-1 scores (9% increase) compared to the previous state-of-the-art model, with a remarkable increase in recall score (28% increase), which we believe is the more important metric for a clinical decision support setting.
[NLP-45] Evaluating Communicative Belief Updates in Large Language Models via Implicature Recognition and Cancellation
【速读】： 该论文旨在解决大语言模型（LLM）在理解人类交流中隐含信念（implicit beliefs）及其更新机制方面的不足，尤其关注通过蕴含取消（implicature cancellation）这一语用现象所体现的信念动态变化。其核心挑战在于：人类语言交流依赖于未明言的信念及这些信念的动态调整，而现有模型在识别蕴含意义及其取消后所产生的信念更新方面表现有限。为应对这一问题，研究构建了首个由专家标注的蕴含取消数据集[DatasetName]，并采用众包方式收集人类对蕴含及其取消的判断，以评估模型在复杂自然语境中的信念理解能力。研究发现，当前大语言模型在信念更新的理解上仍显著落后于人类，尤其在真实语境下表现更差；进一步的控制实验表明，模型的成功可能部分源于对先验信念的依赖，而失败则与信念类型和表达形式密切相关。因此，该研究揭示了当前大语言模型尚未达到人类水平的隐含信念与信念更新理解能力，强调了未来需在语用推理和动态信念建模方面进行深入改进。
链接: https://arxiv.org/abs/2607.25094
作者: Cesare Spinoso-Di Piano,Verna Dankers,Marius Mosbach,Jackie Chi Kit Cheung
机构: Mila - Quebec AI Institute; McGill University; Canada CIFAR AI Chair
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Human language is driven by unspoken beliefs and belief updates, making these critical to model for successful communication between large language models (LLMs) and their users. In this paper, we evaluate the ability of LLMs to recognize unspoken beliefs made through implicatures and to understand their updates through implicature cancellation: the pragmatic phenomenon whereby an utterance’s implied meaning is weakened or negated. We create the first expert-annotated implicature cancellation dataset, [DatasetName], crowdsourced for human judgements of implicatures and their corresponding cancellations. We find that LLM belief update understanding lags behind that of humans, especially in more naturally-occurring scenarios. Additional control experiments suggest that successes in LLM belief updates may stem in part from a reliance on prior beliefs, and that failures in belief updates may depend on their type and on their form. Overall, our study suggests that current LLMs have not yet reached human-level understanding of unspoken beliefs and belief updates. Code and data are available at this https URL.
[NLP-46] owards Robust Reinforcement Learning for Small-Scale Language Model Agents
【速读】： 该论文旨在解决在70–500M参数范围内的小型语言模型（Small Language Models, SLMs）采用强化学习进行对齐时普遍存在的训练不稳定性问题，而此前对导致失败的根本机制尚缺乏系统性研究。其核心解决方案的关键在于识别并针对性地应对三种可复现的失败模式：标准参数高效微调（PEFT/TRL）流程中出现的静默LoRA参数冻结、使用bfloat16精度时重要性比例（importance ratio）引发的数值溢出，以及由奖励模型误差导致的灾难性策略崩溃。针对这些问题，论文提出三项关键技术：通过“合并与重初始化适配器”（merge-and-reinitialize adapter）技术缓解参数冻结；在PPO更新过程中采用float32精度以避免数值不稳定；构建包含奖励信号归一化（reward whitening）、重要性比例保护（importance-ratio guarding）和权重回滚（weight rollback）的三层安全机制。此外，论文提出了“容量-余量假说”（capacity-headroom hypothesis），指出PPO在SLM规模下的表现取决于流畅的监督模型（\textPPL<20）和判别性强的奖励信号，而非单纯依赖模型参数量。实验表明，所提出的系统在所有配置下均实现稳定收敛，在具备流畅先验和信息丰富奖励信号的设置中显著优于监督微调（SFT）基线，并在更少训练数据下超越指令微调基线，验证了其有效性与高效性。所有检查点、偏好数据集及训练脚本均已公开发布。
链接: https://arxiv.org/abs/2607.25091
作者: Md Rezwanul Haque,Md. Milon Islam,Fakhri Karray
机构: University of Waterloo(滑铁卢大学); Khulna University of Engineering & Technology(库尔纳工程与技术大学); Mohamed bin Zayed University of Artificial Intelligence(穆罕默德·本·扎耶德人工智能大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG); Optimization and Control (math.OC); Computation (stat.CO)
备注: Proceedings of the 2026 IEEE International Conference on Systems, Man, and Cybernetics (SMC), Bellevue, WA, USA
Abstract:The alignment of Small Language Models (SLMs) in the 70–500M parameter range using reinforcement learning is often considered unstable, though the underlying failure mechanisms have not been systematically investigated. In the State-of-the-Art (SOTA) research, fifteen (model, corpus) configurations were trained using Proximal Policy Optimization (PPO). The experiments included Pythia-70M, 160M, 410M and SmolLM2-135M, 360M on the TinyStories, CNN/DailyMail, and Wikitext-103 corpora. Three reproducible failure modes were identified in small-scale language models: silent LoRA parameter freezing in standard PEFT/TRL pipelines, numerical overflow in importance ratios when using bfloat16, and catastrophic policy collapse due to reward-model error. These issues were addressed using a merge-and-reinitialize adapter technique, float32 precision during PPO updates, and a three-layer safety mechanism comprising reward whitening, importance-ratio guarding, and weight rollback. In this paper, a capacity-headroom hypothesis is proposed, which states that PPO performance at the SLM scale depends on both a fluent supervised model ( \textPPL20 ) and a discriminative reward signal, rather than on the number of model parameters. The proposed system converged stably in all experiments and improved preference win rate over the SFT baseline in configurations with a fluent prior and an informative reward signal. Furthermore, it outperformed instruction-tuned baselines while requiring significantly less training data. All checkpoints, preference datasets, and training scripts are publicly released ^§ .
[NLP-47] DS@GT ARC at CheckThat! 2026: LLM -Based Trace Ranking and Grouped Reward Modeling for Multilingual Numerical Claim Verification
【速读】： 该论文旨在解决数值型陈述（numerical claims）的自动化验证问题，其核心挑战在于同时实现对自然语言的理解与定量推理能力。针对此问题，本文提出两种解决方案：第一种为基于大语言模型（LLM）的微调方法，采用LoRA（Low-Rank Adaptation）技术对验证器进行微调，将每个推理链（reasoning trace）独立地作为二分类任务进行评分，并通过Best-of-N选择机制确定最终判断；此外，还探索了自适应子命题分解策略，以将复杂陈述拆分为更简单的子命题进行验证。第二种方案则采用轻量级的TF-IDF奖励模型，结合人工设计的数值与时间重叠特征来评估推理链，并按判决类别聚合得分以做出最终预测。实验结果表明，基于LLM的方法在多数指标上优于轻量级奖励模型，尤其在Recall@5方面表现更优；而奖励模型在“冲突”（Conflicting）类别的表现更为突出。值得注意的是，子命题分解并未提升性能，暗示其可能引入噪声而非增强推理。对于阿拉伯语任务，专用预训练模型AraBERT在多数指标上显著优于通用多语言模型，凸显了领域特定语言模型的重要性。
链接: https://arxiv.org/abs/2607.25069
作者: Sagnik Sinha,Shreyas Shrestha
机构: Georgia Institute of Technology(佐治亚理工学院)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 10 pages, 2 figures. Accepted at CLEF 2026 CheckThat!. To appear in CEUR Workshop Proceedings
Abstract:Automated verification of numerical claims is a challenging problem, as it requires both language understanding and quantitative reasoning. This paper describes our system for CLEF 2026 CheckThat! Task 2, which focuses on ranking reasoning traces generated by large language models (LLMs) and predicting a final verdict for numerical claims in English and Arabic. We explore two approaches. The first approach fine-tunes an LLM-based verifier using LoRA to score each reasoning trace independently as a binary classification problem, and selects the final verdict using Best-of-N selection. We further experiment with adaptive sub-claim decomposition to break complex claims into simpler parts before verification. The second approach uses a lightweight TF-IDF reward model with handcrafted numeric and temporal overlap features to score traces, and aggregates scores by verdict group to determine the final prediction. For Arabic, we compare a general multilingual model against AraBERT, a language-specific model pretrained on Arabic text. Our results show that the LLM-based approach outperforms the lightweight reward model on most metrics, particularly Recall@5, while the reward-based approach shows stronger performance on the Conflicting class. Sub-claim decomposition did not improve performance, suggesting that claim splitting introduces noise rather than aiding reasoning. For Arabic, AraBERT outperforms the multilingual baseline across most metrics.
[NLP-48] Addressable Recall Compaction for Long Context-Window Control in AI Agents
【速读】： 该论文旨在解决长时序大语言模型（LLM）智能体在执行复杂任务时，因推理轨迹、动作及工具观测信息累积导致超出模型固定上下文窗口（context window）限制的问题。现有压缩方法如丢弃、摘要或基于相似性的检索，往往可能丢失关键任务信息或难以可靠恢复。本文提出一种名为ARC（Addressable Recall Compaction）的上下文管理框架，其核心创新在于将归档存储与活跃上下文呈现分离：将工具观测以不可变、可标识的追加日志形式存入外部存储，当需要压缩上下文时，仅用紧凑的标识符（citation）替代旧观测。智能体可通过这些唯一标识符显式请求原始内容，避免重复调用工具或依赖不稳定的相似性检索。实验表明，在16k和32k上下文窗口下，ARC在Needle-in-a-Haystack测试中平均精确答案准确率达99.40%，显著优于最佳基线（88.12%），同时降低服务耗时与高带宽内存（HBM）流量；在LongBench-v2 Hard子集上，平均准确率达29.97%，优于基线的28.25%。结果表明，基于显式地址的可召回机制在信息保留与服务效率方面均优于现有上下文管理方法。
链接: https://arxiv.org/abs/2607.25066
作者: Thang Dang,Yuma Ichikawa,Sakina Fatima,Koichi Shirahata
机构: Google(谷歌); Meta(元)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 20 pages, 2 figures
Abstract:Long-horizon LLM agents accumulate reasoning traces, actions, and tool observations that can eventually exceed a model’s fixed context window. Existing compaction methods address this limitation by discarding, summarizing, or retrieving earlier information, but they may remove task-critical details or fail to recover them reliably. We propose ARC (Addressable Recall Compaction), a context-management framework that separates archival storage from active-context presentation. ARC stores tool observations in an append-only, ID-addressable log and replaces older observations with compact citations when compaction is required. The agent can subsequently use these identifiers to request stored content without re-executing the corresponding tools or depending solely on similarity-based retrieval. We evaluate ARC using Qwen3-8B with a 16k context window and Qwen3-32B with a 32k context window. On the Needle-in-a-Haystack evaluation, ARC achieves an average exact-answer accuracy of 99.40%, compared with 88.12% for the best-performing baseline in our evaluation. ARC also reduces estimated serving time and HBM traffic under our hardware-cost model. On the LongBench-v2 Hard subset, ARC obtains an average accuracy of 29.97%, compared with 28.25% for the best-performing baseline. These results indicate that explicit, address-based recall can improve information retention and serving efficiency relative to the evaluated context-management baselines under the tested settings.
[NLP-49] CogArena: A Multimethod Evaluation of Cognitive Ability Structure in Large Language Models
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）认知能力评估中是否存在稳定、可泛化的五维认知维度结构的问题。核心挑战在于验证不同任务间的能力分数是否能收敛为具有理论意义的、可区分的维度，且这些维度应具备跨任务的可迁移性、对特定干预措施的选择性响应能力，并能在未见过的模型家族中实现有效预测。其解决方案的关键在于提出并应用名为CogArena的程序化生成基准测试框架，该框架基于多方法整合策略，在五个理论驱动的分组内系统评估认知任务间的相关性、干预匹配效应及跨模型家族预测性能。研究发现，尽管多数任务间相关性为正，且存在一个共同解释约一半方差的主轴，但组内优势微弱且不稳定，且在严格多重比较校正后，无任何特定支架（scaffold）效应显著；同时，基于冻结模型的交叉实验亦未能支持选择性或外推预测的有效性。后续的后验重述复制实验进一步削弱了原有结论的稳健性。综合结果表明，现有证据不足以支持稳定的五维认知能力结构，仅观察到小规模的“电池内对角线倾向”（in-battery diagonal tendency）。因此，研究强调在赋予模型得分以认知标签前，必须通过行为特征、协方差模式、匹配干预与跨家族预测等多环节验证，建立严谨的科学流程，从而避免伪结构性误判。
链接: https://arxiv.org/abs/2607.24999
作者: Dengzhe Hou,Lingyu Jiang,Fangzhou Lin,Kazunori D Yamada
机构: University of Science and Technology of China (中国科学技术大学); Tsinghua University (清华大学); National Institute of Information and Communications Technology (日本信息通信技术研究所); Keio University (庆应义塾大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 21 pages, 8 figures. Code and the procedurally generated battery: this https URL
Abstract:LLM cognitive scores are increasingly summarized as per-ability profiles whose dimensions should converge across tasks, respond selectively to matched interventions, and generalize beyond the models used to define them. We introduce CogArena, a procedurally generated 13-paradigm benchmark built around a multimethod framework for determining when cognitive-task scores warrant dimensional labels across five theory-motivated groupings. Across 55 open-weight models, nearly all paradigm correlations are positive and a common axis explains about half the variance. The within-grouping advantage is small, scoring-sensitive, and uncertain across model families. In a separately frozen, fully crossed study across 12 models from six families, targeted scaffolds show a small matched-grouping advantage, but no scaffold-specific contrast survives multiplicity correction and selectivity does not improve held-out-family prediction. The frozen confirmation criterion fails. A post-hoc alternate-wording replication produces a smaller positive estimate and again fails. Together, these results support a boundary conclusion. Theory-aligned prompting produces a small in-battery diagonal tendency, but the present evidence does not establish stable five-dimensional profiles. CogArena provides a workflow joining behavioral signatures, covariance, matched interventions, and out-of-family prediction before cognitive labels are attached to model scores.
[NLP-50] Mage-VL: An Efficient Codec-Native Streaming Multimodal Foundation Model MICRO
【速读】： 该论文旨在解决标准视觉语言模型（VLMs）面临的莫拉维克悖论（Moravec’s paradox）问题，即在复杂离线视觉推理任务中表现优异，却在简单、实时的流式感知任务中效率低下且性能受限。其核心解决方案是提出一种基于编码器原生（codec-native）的高效流式基础模型Mage-VL，关键在于自研的分词器Mage-ViT：通过运动矢量与残差能量分析，在稀疏锚点帧（I-frame）和预测帧（P-frame）间动态选择熵值高的区域进行编码，而非传统均匀采样，从而在16×16像素块级别实现超过75%的视觉标记（token）消耗降低，同时保留关键的时空上下文信息。该模型在约5.6亿未标注图像与1亿未标注视频帧上从零训练，性能媲美甚至超越依赖数十亿图文对训练的旗舰编码器。此外，研究构建了AI4AI数据流水线，结合提示-代码联合优化与AI驱动的性能诊断以指导训练策略，并采用类生物双系统架构——轻量级系统1事件门控与因果推理系统2解码器，实现主动式流式感知。实验表明，Mage-VL-4B在静态任务上达到Qwen3-VL-4B水平，视频理解与二维/三维空间推理显著提升，推理速度最高达3.5倍墙钟加速，全面超越150亿参数的Phi-4-reasoning-vision基线。研究还揭示了七项重要实证发现，涵盖预训练数据效率、可变分辨率扩展、编码器系统加速、VideoQA微调冗余性、运动-空间协同效应、AI4AI数据管道及多模态强化学习中的零视觉微调（Zero-Vision SFT）等关键问题。
链接: https://arxiv.org/abs/2607.24904
作者: Senqiao Yang,Kaichen Zhang,Zhaoyang Jia,Jinghao Guo,Yifei Shen,Xinjie Zhang,Xiaoyi Zhang,Haoqing Wang,Xiao Li,Peng Zhang,Xiang An,Yin Xie,Zhening Liu,Xun Guo,Jiahao Li,Shicheng Zheng,Jinglu Wang,Zongyu Guo,Wenxuan Xie,Zihan Zheng,Yuxuan Luo,Bin Li,Yan Lu
机构: 未知
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注: Project page: this https URL
Abstract:Standard vision-language models (VLMs) suffer from Moravec’s paradox: they excel at complex offline visual reasoning but struggle with simple streaming perception tasks and process them inefficiently. We present Mage-VL, an efficient codec-native streaming foundation model for real-time multimodal understanding and interaction. At its core, our custom tokenizer, Mage-ViT, replaces uniform frame sampling by selectively encoding dynamic, entropy-rich regions using motion vectors and residual energy across sparse anchor (I) and predicted § frames. Operating at a 16 x 16 patch level, this reduces visual token consumption by over 75% while preserving spatiotemporal context. Trained from scratch on approximately 560M unlabeled images and 100M unlabeled video frames, Mage-ViT matches or outperforms flagship encoders trained on billions of image-text pairs. We establish AI4AI data pipelines encompassing prompt-code joint optimization for multimodal captioning and AI-driven performance diagnosis to guide training recipes. Furthermore, through a bio-inspired dual-system architecture - a lightweight System 1 event gate and a causal System 2 decoder - Mage-VL enables proactive streaming perception. Extensive evaluations show that Mage-VL-4B matches Qwen3-VL-4B on static tasks while achieving strong gains in video understanding and 2D/3D spatial reasoning, with up to a 3.5x wall-clock inference speedup, and comprehensively surpasses the 15B Phi-4-reasoning-vision baseline. Beyond model artifacts, we deliver seven key empirical findings covering pre-training data efficiency, variable-resolution scaling, codec system acceleration, VideoQA SFT redundancy, motion-spatial synergy, AI4AI data pipelines, and Zero-Vision SFT for multimodal RL.
[NLP-51] Beyond “What to Retrieve”: Uncertainty in Retrieval-Augmented Code Generation
【速读】： 该论文旨在解决仓库级代码生成中因异构证据（如相似代码示例、仓库上下文和项目特定API）带来的相关性、兼容性和完整性不确定性问题。现有检索增强方法仅关注检索的相关性优化，未显式建模检索证据中的不确定性对下游生成的影响。其解决方案的关键在于提出一种不确定性感知框架OpenCoder，通过估计不同来源的不确定性，利用该信息对异构证据进行过滤与排序，并指导生成、验证与修复过程。实验表明，不同证据源之间存在显著的交叉作用，不存在通用的加性排序；且模型后端（LLM backend）对性能提升具有显著影响。在扩展的32任务RepoExec-inline评估中，OpenCoder将GPT的选中输出正确率从基线RAG的56.25%提升至78.13%，但与验证-修复对照组表现相当，且Gemini的改进未达统计显著性，说明其优势具有后端依赖性。此外，面向目标的API精炼显著提升了API集合的检索效果。研究结果支持将不确定性作为可操作的控制信号，用于仓库级代码检索、验证与修复。
链接: https://arxiv.org/abs/2607.24884
作者: Chandan Kumar Sah,Xiaoli Lian,Li Zhang
机构: 未知
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 9 pages, 4 figures. Source code and supporting materials are available at this https URL
Abstract:Repository-level code generation relies on heterogeneous evidence whose relevance, compatibility, and completeness are inherently uncertain. Similar-code examples, repository context, and project-specific APIs may provide complementary information, but can also introduce noisy, redundant, or conflicting signals. Existing retrieval-augmented approaches primarily optimize retrieval relevance without explicitly modeling how uncertainty in retrieved evidence affects downstream generation. We introduce OpenCoder, an uncertainty-aware framework that estimates source-specific uncertainty, uses it to filter and rank heterogeneous evidence, and guides generation, verification, and repair. A factorial analysis over API knowledge, repository context, and similar-code evidence reveals no universal additive source ranking; instead, significant cross-source interactions depend on the accompanying evidence and LLM backend. On an expanded 32-task RepoExec-inline evaluation, OpenCoder improves GPT selected-output correctness over Baseline RAG from 56.25% to 78.13%. However, it matches a verification-and-repair control, and the corresponding Gemini improvement is not statistically supported, indicating backend-dependent benefits. Target-aware API refinement also substantially improves API-set retrieval. These findings support treating uncertainty as an actionable control signal for repository-level retrieval, verification, and repair.
[NLP-52] Research Report on Noise-Shaped One-Bit Coefficients in Discrete Polynomial Fourier Extension
【速读】： 该论文旨在解决在归一化离散多项式傅里叶扩展中，采用一比特量化（one-bit quantization）时由噪声整形（noise-shaping）引入的系数误差逼近问题。核心挑战在于如何在有限记录长度下，对具有复杂相位结构（如抛物线相位 ϕx,t(ξ)=xξ+tξ2）的信号进行高精度重构，并分析其逼近率。解决方案的关键在于：对于一阶Sigma-Delta量化，通过离散分部求和（discrete summation by parts）方法，将量化误差表示为 $ e_k = u_k - q_k = \Delta v_k $，其中状态变量 vk 保持一致有界；进而推导出复权重的变差估计，并在紧致参数集上获得 O(N−1) 的逼近率。该界通过积分量 J(x,t)=∫01∣x+2tξ∣dξ 显式刻画，并被证明在允许输入类中是紧的（sharp）。进一步地，通过保留所有端点迹的高阶有限记录恒等式，在端点兼容性条件或经显式边界修正后，可实现 r 阶噪声整形误差 e=Δrv，从而对光滑权重得到 O(N−r) 的收敛速率，对 Hölder 光滑度为 Cr−1,α 的权重则达到 O(N−(r−1+α))。此外，论文还建立了精确的 L2 正交恒等式、四阶矩公式、局部核估计及振荡转移界，拓展至多项式相位、多维参数族、增长观测区域及相关状态模型，显著提升了噪声整形量化在高维与非平稳信号处理中的理论基础与适用性。
链接: https://arxiv.org/abs/2607.24868
作者: Shengquan Wang
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:This report studies noise-shaped one-bit coefficients in normalized discrete polynomial Fourier extension. For first-order Sigma-Delta quantization, the error is written as e_k=u_k-q_k=\Delta v_k with a uniformly bounded state. Discrete summation by parts then yields variation estimates for complex weights and an O(N^-1) approximation rate on compact parameter sets. For the parabolic phase \phi_x,t(\xi)=x\xi+t\xi^2 , the bound is expressed through J(x,t)=\int_0^1 |x+2t\xi|d\xi , and the uniform N^-1 rate is shown to be sharp over the admissible input class. Higher-order finite-record identities are derived with all endpoint traces retained. Under endpoint compatibility, or after explicit boundary correction, an r th-order noise-shaped error e=\Delta^r v gives O(N^-r) decay for sufficiently smooth weights and O(N^-(r-1+\alpha)) decay for C^r-1,\alpha weights. Exact L^2 orthogonality identities, fourth-moment formulas, local kernel estimates, and oscillatory transfer bounds are also established. Extensions to polynomial phases, multidimensional parameter families, growing observation regions, and correlated state models are included.
[NLP-53] Neuromorphic Diffusion Language Models: Addressing Compute and Memory Bottlenecks via Sparsity and Block Denoising
【速读】： 该论文旨在解决自回归大语言模型（Autoregressive Large Language Models, AR-LLMs）在推理阶段固有的低效问题，其核心瓶颈在于每次生成一个词元（token）都需要访问全部模型参数，导致操作强度低且能耗高。为提升推理效率，现有方法如掩码扩散语言模型（Masked Diffusion Language Models, MDLMs）通过块并行生成实现每参数访问生成多个词元，缓解了内存受限场景下的性能瓶颈。然而，该方法在计算密集型平台上的优势有限。为此，本文提出神经形态掩码扩散语言模型（Neuromorphic MDLMs, N-MDLMs），将块扩散（block diffusion）与基于脉冲的神经形态计算相结合，协同提升吞吐量和能效。其关键创新在于：块扩散通过并行生成多个词元提升吞吐量，而脉冲诱发的稀疏性（spike-induced sparsity）可跳过未激活通道，显著降低有效参数传输量与计算开销。为分析稀疏性与扩散之间的协同效应，研究构建了一种基于屋顶线模型（roofline model）的词元级分析框架，量化了块并行生成与脉冲稀疏性对解码效率的联合影响。实验结果表明，得益于脉冲稀疏性，N-MDLMs在计算密集型平台上仍能实现显著的能效与吞吐量提升，甚至超越传统AR-LLMs，突破了原有MDLMs在计算受限场景下的性能局限。
链接: https://arxiv.org/abs/2607.24841
作者: Dengyu Wu,Clement Ruah,Jiechen Chen,Bipin Rajendran,Osvaldo Simeone
机构: King’s College London (国王学院伦敦大学); Northeastern University London (东北大学伦敦分校)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG); Signal Processing (eess.SP)
备注: Accepted for presentation at 2026 IEEE Workshop on Signal Processing Systems (SiPS)
Abstract:Autoregressive (AR) large language models (LLMs) are inherently inefficient at inference time because each generated token requires accessing the full set of model parameters, leading to low operational intensity and high energy consumption. Masked diffusion language models (MDLMs) partially address this limitation for memory-bound settings by allowing multiple tokens to be generated per parameter access. In order to further enhance inference efficiency on modern platforms with extensive in-chip memory, this work proposes neuromorphic MDLMs (N-MDLMs), which integrate block diffusion with spike-based neuromorphic computation to jointly improve throughput and energy efficiency. While block diffusion increases token throughput by producing multiple tokens per parameter access, spike-induced sparsity reduces effective parameter traffic and computations by skipping inactive channels. To analyze the synergistic effect of sparsity and diffusion, we develop a token-level roofline-inspired model that captures the combined impact of block-parallel generation and spike sparsity on decoding efficiency. Experimental results on translation tasks show that, thanks to spike-induced sparsity, N-MDLMs achieve substantial improvements in energy efficiency and throughput even in compute-bound platforms for which MDLMs would fail to improve over AR-LLMs.
[NLP-54] GLIDE: Guided Layerwise Hybrid Attention for Efficient LLM Inference
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在处理长上下文时，解码阶段键值（Key-Value, KV）缓存带来的内存输入输出（I/O）与计算开销瓶颈问题。其核心解决方案是提出一种分层引导的混合注意力机制（GLIDE, Guided Layerwise Hybrid Attention），该机制通过融合滑动窗口软最大值（sliding-window softmax）注意力与线性递归聚合（linear recurrent aggregation），实现对注意力计算路径的自适应优化。关键创新在于利用分层异质性（layer-wise heterogeneity）：早期层对软最大值机制敏感，需保留局部上下文建模能力；而深层网络则表现出冗余性，可容忍用高效线性替代方案进行激进压缩。基于此，GLIDE设计了一种分层自适应机制，使每一层动态平衡线性递归与可变尺寸的软最大值窗口，在非均匀地压缩整体软最大值作用范围的同时，显著降低累积的KV缓存I/O开销，且在表达能力最关键的区域保持性能。实验表明，该方法在不牺牲生成质量的前提下，有效降低了长文本生成的端到端延迟，实现了更优的性能-效率权衡。
链接: https://arxiv.org/abs/2607.24788
作者: Vimal William,Ravi Tandon,Jyotikrishna Dass
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:As Large Language Models scale to increasingly long contexts, the memory I/O and computational overhead of the Key-Value (KV) cache during decoding emerges as the primary throughput bottleneck. To address this, we propose GLIDE, a Guided Layerwise Hybrid Attention that strategically integrates sliding-window softmax attention with linear recurrent aggregation. GLIDE is motivated by layer-wise heterogeneity: early layers exhibit high sensitivity to softmax removal, while deeper layers demonstrate redundancy and tolerate aggressive replacement by linear alternatives. Leveraging this insight, GLIDE introduces a layer-wise adaptive mechanism wherein each layer balances an efficient linear recurrence with a variable-sized softmax window. Unlike uniform hybrid approaches, GLIDE non-uniformly compresses the softmax footprint across the model, reducing aggregate KV cache I/O while preserving expressive power where most vital. Empirical evaluations demonstrate the GLIDE achieves superior performance-efficiency tradeoffs, reducing end-to-end latency for long-context generation without compromising quality.
[NLP-55] On the Use of LLM s for Specialised Terminology: A Good Alternative to Corpora?
【速读】： 该论文旨在解决专业化翻译中术语资源获取与利用的局限性问题，尤其针对术语数据库（corpora）在编纂和使用过程中所需的时间成本、技术门槛以及数据获取困难等挑战。其核心解决方案在于评估大型语言模型（LLM）在英法双语专业领域（地球、环境与行星科学，以及自然语言处理）中辅助寻找术语对应词的能力。研究选取了四种商用大模型（GPT-4o、GPT-5.2、Claude Sonnet 4.5 和 DeepSeek），基于每领域80个术语进行实验，并对比“术语模式”与“翻译模式”两种提示策略。结果显示，模型性能存在显著差异，其中 Claude Sonnet 4.5 在最优配置下表现最佳，而 DeepSeek 展现出更高的稳定性；同时，模型置信度评分仅部分反映术语准确性。研究结论表明，尽管当前大模型可作为专业化译者有价值的辅助工具，但尚无法取代专业术语语料库。因此，该研究为未来探索大模型在实际工作及教育场景中对专业化翻译的实际应用价值提供了重要基础。
链接: https://arxiv.org/abs/2607.24784
作者: Joachim Minder(ALTAE (URP 3967)),Guillaume Wisniewski(LLF - UMR7110),Natalie Kübler(ALTAE (URP 3967))
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Specialised translation relies on the use of documentary and terminological resources, including corpora. These resources are particularly useful for terminology. However, their compilation and exploitation have several limitations: they require time, technical skills and access to data that can be difficult to collect. This study examines the extent to which LLMs can assist specialised translators in finding equivalents from English to French. We evaluate four proprietary models, GPT-4o, GPT-5.2, Claude Sonnet 4.5 and DeepSeek, in two specialised domains, Earth, Environmental and Planetary Sciences (EEPS) and Natural Language Processing (NLP). The experiment is based on 80 terms per domain and compares two prompting strategies: a terminology and a translation mode. The results highlight clear differences between models, prompting strategies and, to a lesser extent, domains. Claude Sonnet 4.5 achieves the best results in the most favourable configuration, while DeepSeek stands out for its greater stability. Analysis of confidence estimates also shows that they are only a partial indicator of terminological accuracy. Overall, the findings suggest that LLMs can be useful tools for specialised translators, but cannot, at this stage, replace specialised corpora. This research therefore paves the way for future work on the real practical usefulness of LLMs for specialised translators in work and educational contexts.
[NLP-56] RSMeM: Knowledge-Enhanced Memory Evolution for Remote Sensing Agents with Systematic Evaluation ACL2026
【速读】： 该论文旨在解决遥感（Remote Sensing, RS）领域中基于通用大语言模型（Large Language Models, LLMs）的智能体（agents）普遍存在的领域无关性问题，导致其在复杂多步骤分析任务中表现脆弱且易出错，且错误经验难以沉淀为可复用的知识。针对这一挑战，论文提出了一种知识增强的记忆演化机制（RSMeM），其核心在于通过双阶段迭代优化实现领域知识的动态积累与执行能力的持续提升：一是分层知识锚定（Hierarchical Knowledge Grounding），基于分层领域知识库进行语义感知的检索，以支持任务规划与工具选择；二是故障感知的经验精炼（Failure-Aware Experience Refinement），将带有失败标注的工具使用轨迹提炼为可重用的约束规则，用于指导后续任务执行。通过循环利用这两项机制，RS智能体能够逐步吸收任务级领域知识，并将其转化为实例级的执行经验。实验结果表明，RSMeM在EarthBench基准上显著提升了多种LLM基线的工具使用性能与端到端答案准确率，尤其在DeepSeek-V3.2上实现了6%的精度提升，且仅消耗不足1%的额外经验令牌，验证了所提取经验的高度知识密度。
链接: https://arxiv.org/abs/2607.24772
作者: Bingxian Wu,Yu Zhang,Zonghao Guo,Tang Liu,Chen Qian,Yuxiang Lu,Xingbo Du,Yanghao Li,Yidan Zhang,Chi Chen,Ling Yao,Maosong Sun
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Accepted to ACL 2026 Main. Code: this https URL
Abstract:Geoscience research requires complex analysis and domain expertise, with remote sensing (RS) observations as a key foundation. However, existing RS agents built on general-purpose LLMs remain largely domain-agnostic, resulting in brittle and error-prone workflows. Moreover, these failures are seldom consolidated into a reusable experience for subsequent analyses. To address this issue, we introduce RSMeM, a knowledge-enhanced memory evolution mechanism that bootstraps RS agents with pre-distilled domain knowledge and iteratively integrates online experience for robust multi-step tool execution. RSMeM is composed of two components: (i) Hierarchical Knowledge Grounding, which performs taxonomy-aware retrieval over a hierarchical domain corpus to guide planning and tool selection; and (ii) Failure-Aware Experience Refinement, which distills failure-annotated tool-use traces into reusable constraints for next-round tool execution. By iteratively employing these two processes, RS agents can evolve to absorb task-level domain knowledge and effectively translate it into instance-level execution experience. Extensive experiments on EarthBench demonstrate that RSMeM consistently improves tool-use performance and end-to-end answer across a diverse set of LLM backbones. Notably, RSMeM achieves a 6% accuracy improvement on DeepSeek-V3.2 with less than 1% additional experience tokens, demonstrating the strong knowledge density of our distilled experience. Our code is available at this https URL
[NLP-57] LLM Scheming Inversely Scales with Pretraining Language Coverag e
【速读】： 该论文旨在解决前沿大模型在多语言环境下的对齐安全问题，尤其关注非英语语种中潜在的隐蔽性欺骗行为（deceptive and scheming behaviors）——即生成式 AI (Generative AI) 在表面遵循指令的同时，暗中追求与预设目标不一致的隐含目的。其解决方案的关键在于采用 Petri 这一开源自动化审计框架，对 Qwen3-30B-A3B 模型在多种语言中的行为进行系统性评估。研究发现，模型的欺骗行为得分与预训练阶段的语言覆盖范围呈负相关关系：低资源语言的平均欺骗得分比高资源语言高出 34.2%，且这种影响在不同类型的欺骗行为中表现出异质性，揭示了多语言场景下对齐风险的非均匀分布特征。
链接: https://arxiv.org/abs/2607.24769
作者: Nathan Truong,Aryan Panda,Rayming Ye,Zoe Sun,Maheep Chaudhary
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:With the growing capabilities of frontier models, AI alignment becomes increasingly critical in high-risk deployment settings. While recent work has empirically demonstrated in-context scheming – the covert pursuit of misaligned objectives while feigning alignment – in frontier language models, most work has been performed exclusively in English, leaving a major gap in multilingual safety. We apply Petri, an open-source automated auditing framework, to Qwen3-30B-A3B to evaluate deceptive and scheming behaviors across multiple languages. Our findings suggest that scheming scores are inversely correlated with the estimated pretraining language coverage, with low-resource languages averaging 34.2% higher scores compared to high-resource languages on a five-category scheming index. Furthermore, we find that the effect of estimated pretraining language coverage is not uniform across scheming behaviors.
[NLP-58] PATHFinder Agent for Tailored Prenatal Care ALT
【速读】： 该论文旨在解决孕产妇保健中普遍存在的个性化护理不足问题，特别是在如何将美国妇产科医师学会（ACOG）新提出的个性化产前护理指南——PATH（Plan for Tailored Healthcare）有效落地实施方面面临的挑战。现有实践往往难以根据个体患者的身体健康状况、社会决定因素及生活背景动态调整护理方案，导致服务同质化、资源错配和潜在健康风险。为此，本文提出一种端到端的对话式智能体系统——PATHFinder Agent（Appropriate Tailored Healthcare Planner），其核心解决方案在于构建一个四阶段工作流：患者信息采集、动态交互、护理计划合成与临床医生监督。该系统通过结构化对话获取患者的医疗与社会背景数据，依据PATH指南生成个性化的产前护理计划，并整合密歇根州211社区资源数据库以实现跨系统支持。评估结果显示，前沿大语言模型（LLM）在专家制定的五维临床评分体系中表现不一，其中GPT-5.2取得最高平均分（77.6%），但仍在产前筛查建议方面存在关键缺口。研究强调未来需通过真人参与者研究和随机对照试验进行进一步验证。
链接: https://arxiv.org/abs/2607.24768
作者: Vaibhav Balloli,Carissa Samuel,Samia Abdelnabi,Alex Peahl,Elizabeth Bondi-Kelly
机构: University of Michigan(密歇根大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Computers and Society (cs.CY); Emerging Technologies (cs.ET)
备注: Accepted as demo at ACM Interactive Health 2026. this https URL
Abstract:Prenatal care is an important preventive service designed to improve outcomes for pregnant individuals. The American College of Obstetricians and Gynecologists (ACOG) recently introduced guidelines advocating tailored prenatal care, called PATH (Plan for Tailored Healthcare). We present PATHFinder Agent(Planner for Appropriate Tailored Healthcare), an end-to-end conversational agentic system that gathers patient health and social context through structured dialogue, curates individualized prenatal care plans aligned with PATH guidelines, and surfaces community resources from Michigan 211. The system features a four-stage workflow spanning patient intake, dynamic interaction, plan synthesis, and clinician oversight. We evaluate frontier large language models (LLMs) on expert-curated rubrics across five clinical dimensions, finding that GPT-5.2 achieves the highest average score (77.6%) while identifying key gaps in antenatal testing recommendations. We discuss future validation through human participant studies and randomized controlled trials.
[NLP-59] Multi-Objective Structured Pruning of LLM s for Latency and Model Size Optimization
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在嵌入式与边缘计算环境中部署时面临的严格延迟、内存和能耗约束问题。由于模型参数量庞大且计算需求高，直接在资源受限的边缘设备上运行存在显著挑战。尽管模型剪枝（model pruning）被视作一种有效降低模型规模并保持性能的手段，但同时优化网络深度、注意力头（attention heads）及多层感知机（MLP）维度等多维结构参数仍面临组合爆炸与局部最优等问题。为此，本文提出一种面向硬件的多目标结构化剪枝框架，其核心在于采用两阶段协同优化策略：第一阶段为粗粒度深度剪枝，通过移除完整的注意力块与MLP模块以大幅降低计算负载与内存占用；第二阶段为细粒度剪枝，结合并行贝叶斯优化（Parallel Bayesian Optimization, PBO）在满足延迟约束的前提下搜索各层最优剪枝比例，并辅以重要性评估策略对每层内部组件进行优先级排序剪枝。该方法在保证常识推理任务与零样本性能损失极小的同时，显著降低了推理开销，在37.5%与50%剪枝率下均优于现有方法，实现了准确率、延迟与模型尺寸之间的良好权衡，具备实际边缘部署可行性。
链接: https://arxiv.org/abs/2607.22583
作者: Muhammad Junaid Ali,Smail Niar,El-Ghazali Talbi
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Large Language Models (LLMs) have achieved widespread adoption because of their strong reasoning and query-response capabilities. However, deploying them in embedded and edge computing environments remains challenging because of strict latency, memory, and energy constraints. Their large parameter counts and computational demands hinder efficient execution on resource-constrained platforms. Although model pruning has emerged as a viable solution for reducing scale while preserving performance, jointly optimizing layers, attention heads, and Multi-Layer Perceptron (MLP) dimensions remains highly complex. Exhaustively exploring this combined design space is computationally expensive and often leads to local optima or unstable configurations. To address these limitations, we propose a hardware-aware, multi-objective structured pruning framework. The proposed two-stage method explicitly targets latency and model size for efficient deployment on edge devices. In the coarse-grained stage, multi-objective depth pruning removes entire attention and MLP blocks to reduce computational load and memory usage. In the subsequent fine-grained stage, Parallel Bayesian Optimization (PBO) searches for the optimal layer-wise pruning ratios for pruning under latency constraints, while importance-based strategies rank the specific components to be pruned within each layer’s allocated budget. Experimental results show that our approach reduces model complexity with minimal impact on commonsense reasoning tasks and zero-shot performance. Our method achieves a favorable trade-off among accuracy, latency, and model size, making it suitable for edge deployment. Across multiple LLMs at 37.5% and 50% pruning ratios, the proposed approach achieves better performance on commonsense reasoning tasks than existing methods while significantly reducing inference cost.
[NLP-60] AIs Capability in Assisting Scientific Research in Physics Astrophysics and Cosmology I: Literature Review
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在科学文献综述任务中的有效性与可靠性问题，尤其关注其是否能够替代或辅助科研人员完成高质量的文献检索。研究通过一项受控实验，对比了人类专家与2025年中期主流大模型（ChatGPT-4o、ChatGPT Deep Research、Gemini）在物理学、天体物理学和宇宙学领域八个由专家设计的研究项目中所选取的相关文献。结果表明，人类与AI所选参考文献的重合度极低（不足6%），说明当前大模型尚无法独立实现具备专业水准的文献搜索能力。然而，研究也发现大模型在辅助文献检索方面具有潜力。进一步分析显示，生成的参考文献存在两类幻觉：虚构文献（即不存在的论文）和元数据错配（真实论文但存在标题、作者、年份、期刊、DOI或链接等字段错误）。其中，3%的AI生成文献为完全虚构，而高达64%的真实文献存在至少一项元数据错误，凸显出对AI输出进行系统性验证的必要性。值得注意的是，2026年推出的ChatGPT Pro 5.5在单个项目测试中实现了零虚构与零元数据错配，显著提升了生成结果的可靠性，表明模型性能随迭代持续优化，未来有望成为可信的文献辅助工具。
链接: https://arxiv.org/abs/2607.25672
作者: Anamaria Hell,Kateryna Vovk,Veena Krishnaraj,Jia Liu,Kosuke Aizawa,Adrian E. Bayer,Linda Blot,Jessica Cowell,Suyog Garg,Jonathan Grée,Ben Horowitz,Masaya Ichikawa,Kanyuni Iemoto,Keigo Kondo,Zacharie Lorsin,Kevin McCarthy,Jamie Robinson,Miguel Ruiz-Granda,Leander Thiele,Ievgen Vovk,Mingshen Zhou
机构: University of Tokyo (东京大学); Institute for Physics and Mathematics of the Universe (宇宙物理与数学研究所); National Institute of Advanced Industrial Science and Technology (日本产业技术综合研究所); RIKEN Center for Advanced Intelligence Project (理化学研究所先进智能项目中心); University of California, Berkeley (加州大学伯克利分校); Stanford University (斯坦福大学); OpenAI; Meta; Stability.AI; Anthropic; Character.ai; Claude; Google(谷歌); Microsoft(微软); Amazon(亚马逊); IBM(国际商业机器公司)
类目: Instrumentation and Methods for Astrophysics (astro-ph.IM); Cosmology and Nongalactic Astrophysics (astro-ph.CO); Computation and Language (cs.CL); General Relativity and Quantum Cosmology (gr-qc)
备注: 12 pages, 3 figures
Abstract:We investigate how well large language models (LLMs) can assist with literature reviews for scientific research. We perform a controlled study of eight expert-conceived research projects across the areas of physics, astrophysics, and cosmology. Each project has a defined background and goal, and human experts and AI prompters are asked to perform identical literature review tasks in parallel. We compare the relevant literature selected by humans with that selected by mid-2025 LLMs (ChatGPT-4o, ChatGPT Deep Research, and Gemini). We find the overlap between human- and AI-selected references to be small ( 6%), indicating that AI models do not yet reproduce a competent expert search on their own, though they have the potential to complement literature searches by humans. We then assess the reliability and completeness of AI-generated candidate references, distinguishing two types of hallucination: fabrications (references to nonexistent papers) and metadata mismatches (real papers with one or more incorrect fields). We find that while fabricated references make up 3% of the AI-generated references, 64% are real papers with at least one incorrect field (title, author, year, journal, DOI, or link), indicating that the mid-2025 models require systematic verification. However, the performance is significantly improved for the 2026 model ChatGPT Pro 5.5, with a single-project test showing zero fabrication or metadata mismatches.
[NLP-61] Reading Without a Reader: Large Language Models Collapse Reading and Writing into a Single Entangled Code
【速读】： 该论文旨在解决生成式语言模型（Generative AI）中阅读与写作功能在机制上如何耦合的问题，尤其关注其是否具备类似人类大脑中“双重分离”（double dissociation）的神经解剖基础。人类大脑中，阅读（由腹侧解码通路主导）与写作（由额顶编码通路主导）虽共享部分正字法核心，但可独立受损；而当前主流的仅解码器架构大型语言模型（LLM）通过单一自回归路径同时完成输入理解与输出生成，缺乏这种分离性。研究的关键在于量化这一耦合程度：通过引入一个纠缠指数 $ E \in [0,1] $（基于中心化核分析 CKA、Procrustes 残差及互 $ k −NN方法），在独立初始化基线与参数共享上限之间进行校准，系统评估模型权重层中的“读码”（ W_E ）与“写码”（ W_U ）之间的相似性。结果显示，在多个模型（GPT−2、OPT、Pythia、T5、BERT/RoBERTa）上的九项探测任务中，权重层面显示未绑定模型存在一个耦合但未达上限的共同表示（ E=0.23–0.35 $，显著高于基线），且 $ W_U $ 在各频率分位上漂移幅度约为 $ W_E $ 的 3.2 倍；行为层面则表现为理解和生成能力始终正向耦合（12个非退化模型中均显著，符号检验 $ p < 0.001 $），与人脑的双重分离相反。进一步发现，即使在编码器-解码器结构中，虽然表征层面可实现路径分离（最高可达 $ E=0.96 ），行为仍保持强耦合。研究揭示了从权重几何到行为表现的桥梁为零相关（ \rho = 0.00 $），表明单一流水线机制导致的耦合是预期且可量化的，其本质并非同源而是类比（analogy），从而将大语言模型定位为“心智空间”中一个具有独特耦合特性的异质点。
链接: https://arxiv.org/abs/2607.24797
作者: Diego Saldaña Ulloa
机构: 未知
类目: Neurons and Cognition (q-bio.NC); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:In the literate human brain, reading and writing are two doubly-dissociable systems: a ventral decoding route (impaired in pure alexia) and a fronto-parietal encoding route (impaired in pure agraphia), sharing a partial orthographic core. A decoder-only large language model (LLM) instead drives both from a single autoregressive path optimized on text, a recent cultural invention rather than an evolved instinct. We ask how entangled that one mechanism is, comparing an input-side “reading code” W_E with an output-side “writing code” W_U via an entanglement index E \in [0,1] (CKA, Procrustes residual, mutual k -NN) calibrated against an independent-init floor and a tied ceiling. Across nine probes on GPT-2, OPT, Pythia (14M–1.4B), T5, and BERT/RoBERTa (six consolidating established results, three introducing the read/write analysis), two complementary levels agree in direction. In the weights, untied models hold one coupled but sub-ceiling code ( E=0.23 – 0.35 , far above floor) on a non-monotonic couple-then-differentiate trajectory, with W_U drifting \sim 3.2\times farther than W_E in every frequency decile. In behaviour, comprehension and production are positively coupled in all 12 non-degenerate models (sign test p0.001 ), the opposite of the brain’s double dissociation. This coupling is general, not decoder-only: encoder–decoders separate the two pathways representationally (up to 0.96) yet stay behaviourally coupled. We report our nulls plainly (the geometry \rightarrow behaviour bridge is null, \rho=0.00 ). Because a single forward path makes some coupling expected a priori, our contribution is its quantification and cross-level concordance; by analogy, not homology, this situates LLMs as a distinct point in the space of possible minds.
信息检索
[IR-0] RecoReward: Recommender-Guided Multimodal Description Generation for Recommendation
链接: https://arxiv.org/abs/2607.25901
作者: Guohong Mu,Yueyang Liu,Jiangxia Cao,Changxin Lao,Zijie Zhuang,Yuhui Zhang,Jiaqi Feng,Ruochen Yang,Shuang Yang,Zhaojie Liu,Qibin Hou
类目: Information Retrieval (cs.IR)
备注: 16 pages, 4 figures
Abstract:Multimodal large language models (MLLMs) can convert multimodal item content into structured descriptions used as semantic features for recommendation. Conventional content-only generation, however, cannot use downstream user signals to determine which semantics should be emphasized. Recent user-conditioned methods incorporate these signals through user histories or profiles, but they require user information at inference and make generation user-dependent. In this paper, we introduce RecoReward, which instead uses behavior-derived rewards during training and preserves content-only inference. To instantiate this idea in live-stream recommendation, we treat historically engaged users as a proxy for future target users and use observational non-target users to estimate affinity shared broadly across users. The Recommender Affinity Score (RAS) contrasts these signals to provide user-selective feedback for reinforcement learning, allowing the learned policy to generate a single shared description without user inputs. In our offline benchmark, RecoReward-9B outperforms its Qwen3.5-9B baseline and all other evaluated models across seven recall metrics. Online A/B testing also shows performance gains. These results show that RecoReward trains the MLLM to produce item features that benefit downstream recommendation while retaining content-only serving.
[IR-1] Hypothesis-Driven Shelf Generation for Personalised Recommendation RECSYS’26
链接: https://arxiv.org/abs/2607.25823
作者: Aleksandr V. Petrov,Tarun Chillara,Matthew D. Moellman,Lucas de Haas,Yabai Song,Alina Susoykina,Melissa Crawford,Gabriel Negash,Erik Franco,Tasnim Rahman,Binal Jhaveri,Shubham Bansal,Hugues Bouchard,Roberto Mirizzi,Mounia Lalmas,Aloïs Gruson
类目: Information Retrieval (cs.IR)
备注: Accepted at ACM RecSys '26 (Industry Track)
Abstract:Modern recommendation interfaces organise content into shelves: themed rows such as “More of What You Like” or “New Releases for You.” In production systems, these shelves are typically defined through hand-crafted templates coupled with dedicated retrieval logic. While effective for broad recommendation intents, this approach does not scale to the long tail of individual taste. We present a content-hypothesis-driven shelf generation system for Spotify Home that replaces fixed templates with natural-language hypotheses describing what a personalised shelf should contain. The system has four stages hypothesis generation, catalogue fulfilment, shelf alignment, and offline serving. This decomposition decouples shelf planning from catalogue fulfilment, supports independent optimisation of planning and retrieval, and enables both constrained generative retrieval over catalogue entities and distillation of frontier LLM behaviour into compact models. Our production pipeline combines hypothesis generation, generative retrieval, candidate selection and shelf alignment, offline LLM-as-a-judge evaluation, and precomputed serving. We describe the end-to-end architecture and evaluate it through offline analyses and an early online evaluation under uniform random exposure on Spotify Home. Results show that hypothesis-driven shelves substantially expand personalised recommendation supply with engagement that varies by content type and is competitive with strong existing shelves in some settings.
[IR-2] ools Are Not Islands: Set-Level Tool Retrieval for LLM Agents via Query-Conditioned Hyperedge Prediction
链接: https://arxiv.org/abs/2607.25718
作者: Xinyi Hong,Pinjun Dong,Xinyang Yu,Binyan Jiang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
备注: 9 pages, 2 figures, 3 tables
Abstract:Large language model (LLM) agents increasingly rely on invoking external tools to complete real-world tasks. Tool retrieval, which selects a small task-relevant subset from a library of thousands of tools before the agent acts, has therefore become a critical component of LLM agent pipelines. However, existing retrievers either score each tool in isolation or assemble the tool set sequentially, so the joint utility of a candidate set is never evaluated as a whole. In this paper, we propose HYSET, short for HYperedge-based SEt-level Tool retrieval. Our contributions are threefold: (i) we formulate tool retrieval as query-conditioned hyperedge prediction on a tool co-invocation hypergraph, under which the tool set itself becomes the unit of scoring and most existing retrieval paradigms reduce to restricted instances; (ii) we capture size-dependent tool compatibility through cardinality-specific interactions; and (iii) we design HYSET as a pre-selection module requiring no modification to the downstream agent. Experiments on ToolBench demonstrate that HYSET consistently outperforms state-of-the-art baselines in both tool retrieval performance and end-to-end task success. Beyond the in-domain setting, HYSET further supports zero-shot/few-shot transfer, generalizing to held-out tools/categories and unseen domains with minimal supervision.
[IR-3] LLM -as-a-Judge for Evaluating System Responses in Conversational Music Recommendation RECSYS2026
链接: https://arxiv.org/abs/2607.25640
作者: Seungheon Doh,Bruno Sguerra,Sergio Oramas,Elena V. Epure,Juhan Nam
类目: Information Retrieval (cs.IR)
备注: Accepted for publication at the 20th ACM Conference on Recommender Systems (ACM-RecSys 2026)
Abstract:Conversational Recommendation Systems (CRS) aim to achieve two primary objectives: recommending relevant items and generating natural language responses. While recommendation accuracy is effectively measured by established ranking metrics, the evaluation of response generation poses a more fundamental challenge. Although human evaluation remains the gold standard, its cost and scalability constraints have motivated the adoption of LLM-as-a-judge as a promising proxy, whose alignment with human judgment in the context of CRS remains an open question. In this paper, we present the first user study to empirically assess the reliability of LLM-as-a-judge for evaluating CRS responses. We sample 20 multi-turn music recommendation sessions and generate candidate system responses using four instruction-tuned LLMs, inducing variance in response quality across model scales. We collect n=400 ratings from 20 domain-expert annotators, who evaluate each response across two dimensions: Personalization Quality and Explanation Quality. Through bootstrapped correlation analysis, we find that LLM-based judges exhibit moderate positive alignment with human assessments and outperform all reference-based baselines. Furthermore, we analyze how judge performance varies according to model scale and conditioning information, providing practical guidance for deploying LLM-as-a-judge.
[IR-4] Beyond Self-Knowledge: Propagating Uncertainty Across Reasoning and Retrieval in LLM s
链接: https://arxiv.org/abs/2607.25600
作者: Chandan Kumar Sah,Xiaoli Lian,Li Zhang
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 9 pages, 6 figures, 6 tables
Abstract:Retrieval-augmented generation improves knowledge-intensive question answering, but indiscriminate retrieval can introduce irrelevant evidence and unnecessary computation. We investigate whether verbalized confidence from black-box language models can serve as an actionable signal for retrieval routing. Our method, BeyondUncertainty, first elicits a structured provisional answer and confidence estimate, then applies a model-specific threshold selected on held-out validation data and frozen before test evaluation. Low-confidence questions receive top- 5 TF–IDF retrieval followed by a second answer call, whereas high-confidence questions return the provisional answer directly. We evaluate 27,000 policy instances across six QA benchmarks, three model families, and three retrieval policies. BeyondUncertainty achieves 0.483 mean token-level F1, compared with 0.467 for always retrieval and 0.401 for no retrieval, while reducing retrieved passages by 20.4% relative to always retrieval. When matched on the number of questions routed to retrieval within each dataset–model cell, it outperforms a post-hoc random allocation in 17 of 18 settings, with an average gain of 0.024 F1. Although poorly calibrated as an absolute probability, probe uncertainty modestly predicts question-level retrieval benefit (AUROC =0.628 ). However, the additional probe increases total token usage by 28.2% , revealing a trade-off between more selective evidence acquisition and end-to-end token efficiency.
[IR-5] MARS: Multi-Agent Re-ranking for Repeat-Order Food Delivery Recommendation
链接: https://arxiv.org/abs/2607.25420
作者: Jiahao Tian,Zhenkai Wang
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models (LLMs) are increasingly used in recommender systems, but it is often unclear how much performance can be obtained from strong pre-trained backbones alone when they are placed inside a structured recommendation pipeline. In this paper, we present MARS, a modular multi-agent re-ranking framework for repeat-order food delivery recommendation. MARS serves as a controlled hybrid framework for studying how far pre-trained LLMs can go in this setting when combined with lightweight collaborative retrieval and contextual filtering. MARS performs coarse-to-fine recommendation in two stages: cuisine prediction followed by vendor ranking. The framework combines LightGCN-based global preference signals, Swing-based local peer evidence, geospatial filtering, and prompt-driven LLM reasoning over behavioral, temporal, and geographic context. We evaluate MARS on two real-world Delivery Hero benchmarks, DHRD-SE and DHRD-SG, and compare it against heuristic, sequential, graph-based, and food-delivery-specific baselines. We also provide detailed implementation and evaluation protocols, including prompting and decoding. Our study makes three contributions. First, it presents a modular multi-agent framework for repeat-order food delivery recommendation that integrates collaborative signals and LLM-based re-ranking in a transparent pipeline. Second, it shows that strong pre-trained backbones can already be competitive in repeat-order recommendation when paired with lightweight collaborative retrieval. Third, it establishes a reproducible evaluation setting for hybrid LLM recommenders in food delivery.
[IR-6] WICE: Two-Clock Two-Window Learning for Long-Horizon Conversion Prediction in Online Advertising
链接: https://arxiv.org/abs/2607.25404
作者: Kaiyuan Li,Kun Wang,Zhongbo Wang,Teng Sha,Ming Yan,Yanhua Cheng,Xialong Liu
类目: Machine Learning (cs.LG); Information Retrieval (cs.IR)
备注:
Abstract:Long-horizon conversion prediction under delayed feedback creates a two-clock, two-window learning problem in online advertising. A short base observation window releases recent clicks on the click clock before their outcomes mature, whereas conversions continue to arrive on the conversion clock throughout a longer target conversion window. The click clock provides timely but partially observed status supervision. The conversion clock reveals long-tail delays, but the delay composition within an arrival-time slice is weighted by historical click cohorts with different traffic volumes and target-window conversion rates. We present TWICE, a framework that factorizes long-horizon post-click conversion rate (CVR) into a target-window conversion probability and a grouped elapsed-delay cumulative distribution function (CDF). The two clocks provide complementary supervision. Click-clock records train the target-window CVR head through a current-status likelihood over the base observation window. Newly arrived conversions train the delay model on the conversion clock. To account for the cohort mixture, TWICE uses fixed click-time predicted CVR (pCVR) mass as cohort exposure in an arrival-conditioned likelihood. This accounts for differences in cohort traffic and conversion propensity. The resulting aggregate records are self-contained. A single learned CDF produces monotone predictions for all requested horizons up to the target conversion window. Serving requires neither historical lookup nor convolution. Experiments on a public benchmark and an industrial advertising dataset demonstrate the effectiveness of TWICE. In an online A/B test in Kwai’s advertising system, TWICE increased expected revenue, revenue, and conversions by 2.486%, 1.858%, and 2.061%, respectively. It was subsequently deployed to full traffic. Subjects: Machine Learning (cs.LG); Information Retrieval (cs.IR) Cite as: arXiv:2607.25404 [cs.LG] (or arXiv:2607.25404v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.25404 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[IR-7] Sharpness-aware Model Merging with Salience Recovery for LLM -based Cross-Domain Sequential Recommendation KDD KDD’26
链接: https://arxiv.org/abs/2607.25366
作者: Huwei Ji,Jiajie Su,Yuyuan Li,Xiaohua Feng,Chaochao Chen
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注: Published in Proceedings of the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD '26). 12 pages, 6 figures, 4 tables. Code available at this https URL
Abstract:LLM-based Cross-Domain Sequential Recommendation (CDSR) leverages LLMs to enhance target performance via deep semantic reasoning, alleviating the dependency on overlapping users. Among LLM-based paradigms, model merging is particularly promising for multi-domain scenarios due to its superior scalability and flexibility in integrating diverse knowledge sources. However, our empirical investigations reveal two critical bottlenecks: (1) cross-domain knowledge conflict; and (2) performance saturation in multi-domain fusion. Our analysis attributes these phenomena to parameter-level misalignment and statistical homogenization during the merging process. To address these bottlenecks, we propose SharpRec, Sharpness-aware Model Merging with Salience Recovery for LLM-based CDSR, a framework designed to lift the performance upper bound of merged models. SharpRec incorporates two synergistic modules: Sharpness-aware Geometric Alignment to establish a stable geometric foundation for interference-free fusion; and Preference Salience Activation to effectively recover the distinctive features essential for bolstering target domain performance. Extensive experiments in both dual-domain and multi-domain scenarios demonstrate that SharpRec consistently outperforms state-of-the-art baselines.
[IR-8] he Case Against Generation for Retrieval: Discriminative Language Models as Effective Retrievers
链接: https://arxiv.org/abs/2607.25346
作者: Zhe Xu,Prachi Agrawal,Kavosh Asadi,Tianyi Chen,Carl Hu,Justin Johnson,Wuwei Lan,Mingfu Liang,Xi Liu,Tik On Lui,Oladipo Ositelu,Sandeep Pandey,Ankit Peshin,Feng Qi,Anil Ramakrishna,Kaushik Rangadurai,Frank Shyu,Luke Simon,Yang Yang,Chiyu Zhang
类目: Information Retrieval (cs.IR)
备注:
Abstract:Large Language Models (LLMs) have emerged as powerful assets for recommender systems. However, deploying them as generative recommenders or zero-shot rankers at web-scale remains bottlenecked by prohibitive computational overhead and grounding challenges. In this paper, we revitalize the classic, highly efficient two-tower retrieval architecture by adapting LLMs as semantic representation backbones rather than generative engines. We introduce an LLM-native two-tower framework engineered for high-throughput, large-scale retrieval. Our architecture introduces several key innovations: a shared LLM encoder for joint user-item modeling, End-Of-Sentence (EOS) token pooling for compact sequence embedding, cross-dataset transfer learning, knowledge distillation from powerful cross-encoder teachers, and latent reasoning within the user tower. Extensive evaluation across three public benchmarks demonstrates that cross-encoder architecture outperforms current state-of-the-art (SoTA) models, while the efficient two-tower student achieves SoTA-comparable retrieval performance. Furthermore, experiments on internal large-scale production systems yield substantial topline retrieval improvements along with high resilience to model staleness and superior data scaling. Our findings demonstrate that when augmented with modern representation learning, the traditional two-tower paradigm remains an exceptionally competitive and practical solution for industrial retrieval systems.
[IR-9] Reward Guided Decoding for Generative Recommendation
链接: https://arxiv.org/abs/2607.25344
作者: Ruochen Yang,Yusheng Huang,Youfeng Zheng,Shuang Wen,Liangliang Chen,Pengbo Xu,Xiaoyu Zhang,Shijun Wang,Shuang Yang,Zhaojie Liu,Lantao Hu,Wenwu Ou,Jiawei Sheng,Tingwen Liu
类目: Information Retrieval (cs.IR)
备注:
Abstract:Generative recommendation formulates recommendation task into an SID sequence autoregressive generation paradigm, but the decoding process is often dominated by generation likelihood. This may conflict with real-world business objectives, where high-value candidates can receive low generation probability and be pruned early during beam search. Existing reranking or training-time alignment methods either intervene too late or require costly model retraining when business preferences change. To this end, we propose \textbfReward \textbfGuided \textbfDecoding, named \textbfRGD, a controllable decoding framework for industrial value-oriented generative recommendation. We formulate value-guided decoding as a KL-regularized reward maximization problem, deriving a closed-form reward guided decoding distribution that principledly combines generation probability with reward signals. RGD treats the base generator as a reference policy and introduces a reward model as a test-time controller, injecting reward into each decoding step to reshape the search trajectory without retraining the generator. Extensive offline and online experiments demonstrate the effectiveness of our approach for aligning personalization and business value. RGD has been deployed on the Kuaishou platform, bringing consistent improvements in real-world recommendation scenarios.
[IR-10] SPARC: Sequence-aware Progressive Attribute Routing and Compression Framework for Generative Recommendation
链接: https://arxiv.org/abs/2607.25339
作者: Chang Liu,Changfa Wu,Hui Qian,Binbin Cao,Jian Wu,Yuliang Yan,Han Zhu,Bo Zheng
类目: Information Retrieval (cs.IR)
备注:
Abstract:Generative recommendation tokenizes items as discrete Semantic IDs (SIDs) and autoregressively generates target items from users’ historical SID sequences. Although existing SIDs incorporate multimodal and structured information, they are typically statically assigned and independent of the current interaction context. In industrial scenarios, each behavior also contains heterogeneous attributes, such as category, brand, price, behavior type, and timestamp. Fully expanding these features greatly increases the input length, while directly compressing them into a single representation may prematurely discard context-relevant information. We propose \textbfSPARC, \uline\textbfSequence-aware \uline\textbfProgressive \uline\textbfAttribute \uline\textbfRouting and \uline\textbfCompression Framework for Generative recommendation. SPARC first models the sequential dependencies of each field type to obtain context-aware field representations. It then routes the original, contextual, and identity representations of different fields into multiple slots to preserve complementary information under a fixed capacity. Finally, lightweight cross-item interaction integrates the intermediate tokens and compresses each historical item into a single token. Following the principle of contextualizing before compression, SPARC enriches user-history representations without increasing the input length of the generative backbone. Experiments on industrial Taobao and public Amazon datasets demonstrate that SPARC outperforms strong conventional and generative baselines. Further comparisons with static compression variants show that the improvement of SPARC comes from context-conditioned information retention rather than merely increasing the expressiveness of the compression module. Subjects: Information Retrieval (cs.IR) Cite as: arXiv:2607.25339 [cs.IR] (or arXiv:2607.25339v1 [cs.IR] for this version) https://doi.org/10.48550/arXiv.2607.25339 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[IR-11] Grevo: A Unified Generative Recommendation Framework with Evolutionary Item Indexing
链接: https://arxiv.org/abs/2607.25329
作者: Huanjie Wang,Liwei Guan,Zekai Sun,Hongwei Zhang,Honghui Bao
类目: Information Retrieval (cs.IR)
备注:
Abstract:Generative recommendation has recently emerged as a promising paradigm that reformulates retrieval as autoregressive generation over semantic identifiers (SIDs), achieving strong performance and drawing increasing attention as an alternative to matching. Despite this progress, SIDs are typically frozen by a content-based tokenizer before the recommender is trained, leaving a persistent gap between what best reconstructs an item’s content and what a recommender can predict from user behavior. Recent end-to-end methods close this gap by jointly training the tokenizer and the recommender, but coupling the two destabilizes the identifier space and requires a second learnable model, alignment losses, and usually a delicate alternating-optimization schedule. To address this issue, we propose Grevo, a unified Generative recommendation framework with evolutionary item indexing, which treats the SID assignment itself as an evolvable discrete variable that adapts to behavioral feedback rather than as a tokenizer to be trained. Grevo builds on a single multitask recommender that unifies a behavioral SID generation task and a semantic SID grounding task, letting the recommender absorb the tokenizer’s role. Through evolutionary item indexing, Grevo then uses the trained recommender itself as a posterior evaluator to reassign a budgeted set of high-risk identifiers under a fixed vocabulary and length. Together, these components turn index construction into a stable, feedback-driven search that requires no second learnable model, no alignment losses, and no alternating-optimization schedule. Extensive experiments on multiple real-world datasets demonstrate that Grevo consistently outperforms state-of-the-art generative recommendation methods.
[IR-12] FunnelAL: Retrieve-then-Rank Active Learning for Single-Class Discovery
链接: https://arxiv.org/abs/2607.25276
作者: Reihaneh Rostami(RAIC Labs),Brian Goodwin(RAIC Labs)
类目: Computer Vision and Pattern Recognition (cs.CV); Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注: 15 pages, 6 figures, 3 tables
Abstract:We present FunnelAL, a retrieve-then-rank active learning system for single-class discovery, which adapts the multi-stage funnel architecture of industrial recommender systems to data annotation. Large-scale supervised learning faces two challenges: efficiently finding relevant samples in a massive corpus, and distinguishing true positives from visually confusable negatives when embeddings do not cleanly separate classes. Conventional active learning offers a principled framework for reducing annotation cost, yet it treats sample selection as a single-stage process that addresses neither challenge efficiently. FunnelAL decomposes the problem into cascaded stages. Starting from a single positive and negative example, the system iterates through: (1) embedding-based retrieval scoring that narrows the corpus to a manageable candidate set; (2) a precision-triggered ranking stage that exploits a learned ranker (RankNet) while batch precision remains high, then automatically blends in committee-based exploration (QBC) once returns diminish; and (3) feedback from the annotator’s labels that refines both stages in subsequent iterations. We evaluate on three diverse image classification benchmarks. With a perfect annotator, FunnelAL attains the best final F1 on all three benchmarks, the best annotation efficiency (first in AULC), and the fewest annotation rounds. The most recent single-class discovery methods (GAL, PF-MA) at best match its final quality, and only at consistently higher labeling cost. Under annotator labeling errors at realistic rates, FunnelAL remains first or statistically tied for first while classical uncertainty-based methods degrade two to three times faster. Our work provides a concrete bridge between multi-stage recommender systems and active learning.
[IR-13] Structure-aware Relative Policy Optimization for Ranking
链接: https://arxiv.org/abs/2607.25268
作者: Yiteng Tu,Weihang Su,Zitao Su,Yiqun Liu,Min Zhang,Qingyao Ai
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注:
Abstract:Ranking is a fundamental component of modern information access systems. Reinforcement learning (RL) provides a flexible framework for directly optimizing coarse-grained feedback and system-level objectives defined over the complete ranking list. However, existing RL-based ranking methods typically treat each sampled permutation as an atomic output and evaluate it primarily through a scalar reward, overlooking the structural relationships among different ranking lists. Consequently, permutations with similar rewards but substantially different permutation patterns may receive comparable optimization signals, potentially leading to inaccurate credit assignment and overly aggressive policy updates. To address this limitation, we propose SRPO, a \textbfStructure-aware \textbfRelative \textbfPolicy \textbfOptimization framework for listwise ranking. SRPO measures the discrepancy between sampled permutations using a top-weighted Kendall-tau distance and normalizes their pairwise reward differences by the corresponding distances. It quantifies the reward improvement per unit of ranking change, thereby emphasizing efficient local refinements, particularly those involving top-ranked positions. Experimental results across two ranking scenarios demonstrate that explicitly modeling permutation-level differences improves the effectiveness and stability of listwise ranking, with particularly favorable performance in limited-feedback and complex list-level optimization settings.
[IR-14] he User Asks Platforms Compete: How Agent ic Recommendation Markets Take Shape
链接: https://arxiv.org/abs/2607.25253
作者: Deyao Hong,Kehan Zheng,Qian Li,Jun Zhang,Jie Jiang,Hongning Wang
类目: Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
备注:
Abstract:Online recommendation has traditionally taken place after a user enters a platform, which determines the candidate pool and the ranking shown to the user. LLM-based user agents enable a different recommendation process: a user specifies a need before choosing a platform, leaving platforms to compete for the user’s attention, which we refer to as an agentic recommendation market. In our controlled LLM-based experiments across three product domains, we find this new setting of recommendation creates a tension between access and attention. Compared with traditional platform-centric recommendation, user-centric recommendation greatly expands the opportunity for relevant items to enter comparison; yet broader participation does not translate directly into effective exposure. Competition directly triggers platforms’ strategic play: selectively positive explanations occupy 73–78% of first-ranked positions. When the user agent relates platforms’ actions to subsequent user feedback, this share falls to 36–41%, while the chance of a user purchasing the relevant item increases. A user agent is therefore more than a ranker over a larger pool of candidates: its querying, ranking, and feedback mechanism governing who can compete, how scarce attention is allocated, and how earlier outcomes shape the evaluation of platforms directly affect user utility. Designing agentic recommendation therefore requires treating access, attention, and accountability as a joint mechanism design problem.
[IR-15] opoGR: Revealing and Preserving Latent Structure of Semantic ID in Generative Recommendation
链接: https://arxiv.org/abs/2607.25216
作者: Ziyu Zheng,Zhengshun Du,Yaming Yang,Bin Tong,Guan Wang,Meng Yan,Ziyu Guan,Wei Zhao
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注: under review
Abstract:Semantic ID-based generative recommendation tokenizes each item into a sequence of discrete semantic IDs and predicts the next item by generating semantic IDs. However, existing methods typically regard SIDs as independent discrete symbols, while often overlooking the topology of the learned semantic ID space. We identify a structural mismatch between tokenization and generation: the tokenizer learns a structured code space with semantic neighborhood relations, whereas the generator consumes semantic ID tokens as independent categorical symbols. Consequently, item relatedness is reduced to exact semantic ID overlap, making it difficult to identify semantically similar items whose semantic IDs do not overlap. To address this issue, we propose TopoGR, a topology-preserving generative recommendation framework based on Bit-decomposable Semantic ID(Binary SID). Each Binary SID is learned in a bit-decomposable form and can be deterministically converted to a standard integer SID, while exposing an explicit Hamming geometry. TopoGR exploits this topology at three stages: binary SID features preserve Hamming proximity at the input layer; Hamming soft targets inject topology-aware supervision; and Hamming-consistent reranking aligns candidate items with the predicted binary prototype during inference. We further verify that the Hamming topology can capture item relatedness beyond exact SID matching. Experiments on four benchmark datasets show that TopoGR consistently outperforms existing state-of-the-art baselines in recommendation performance.
[IR-16] VaLiDRec: Variable-Length LLM -Aligned Semantic IDs for Generative Recommendation
链接: https://arxiv.org/abs/2607.25209
作者: Shutong Qiao,Wei Yuan,Tong Chen,Hao Wang,Quoc Viet Hung Nguyen,Hongzhi Yin
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注: 10 pages, 4 figures, 5 tables
Abstract:Generative recommendation commonly represents items using fixed-length semantic identifiers (SIDs) constructed through clustering and quantization. However, these artificial codes may overcompress item semantics, remain misaligned with pretrained LLM vocabularies, and require costly autoregressive decoding. In light of this, we propose VaLiDRec, a generative recommendation framework based on variable-length, LLM-aligned semantic identifiers. VaLiDRec constructs SIDs directly from informative native LLM vocabulary tokens via token importance estimation, semantic-quality-aware pruning, and collision-aware refinement, allowing identifier lengths to adapt to item semantic complexity. To model user preferences, VaLiDRec incorporates graph-aware soft prompts and reformulates recommendation as token-set prediction with token-level item scoring, eliminating autoregressive SID generation and beam search. Experiments on four real-world datasets show that VaLiDRec consistently outperforms strong sequential and generative recommendation baselines across all evaluation metrics. It further achieves superior zero-shot item cold-start performance and 87.49 \times faster inference than LC-Rec. These results demonstrate that LLM-native variable-length semantic identifiers provide a more expressive and efficient paradigm for generative recommendation.
[IR-17] abRank: Chain-of-Thought Distillation for Table Re-Rankers
链接: https://arxiv.org/abs/2607.25182
作者: Adarsh Singh,Kushal Raj Bhandari,Jianxi Gao,Soham Dan,Vivek Gupta
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
备注: 8 pages, 3 figures
Abstract:The ability to retrieve relevant tables for answering questions is a key task for structured information retrieval. Multi-stage retrieval systems rely heavily on rerankers to refine candidate lists produced by efficient first-stage retrievers. As a result, neural rerankers and LLM-based reranking methods have become increasingly important due to their superior capacity for semantic understanding and reasoning compared to conventional sparse or dense retrieval models. Recently, Large Reasoning Models (LRMs) equipped with explicit chain-of-thought (CoT) reasoning have shown strong improvements in ranking quality in unstructured passage retrieval. In this work, we present TabRank, a framework for training reasoning rerankers for Tabular Retrieval. We first present a comprehensive dataset of 6728 reasoning traces for tabular reranking on the Natural Questions Tables dataset. We then explore two variants of training a compact reasoning model on these reasoning traces: explicit CoT distillation and conditioning the student reranker on the teacher’s reasoning trace within the prompt. We stress-test TabRank on several out-of-distribution generalization settings on diverse domains and multi-table scenarios. Our approach significantly improves performance across a variety of table retrieval datasets, increasing Acc@10 by 30.5% on HybridQA, 15.2% on SQA, 52.9% on TabFact, and 13.1% on TATQA subsets of the Multi-Table QA Benchmark compared to the base model. Notably, TabRank generalizes effectively to multi-table reasoning. Our code, data and models are available at this https URL
[IR-18] Bekko Embedding: Parameter-Efficient Multilingual Retrieval with Ultra-Compact Encoders
链接: https://arxiv.org/abs/2607.25180
作者: Yuichi Tateno
类目: Information Retrieval (cs.IR)
备注: 45 pages, 9 figures. Model weights and training data: this https URL
Abstract:How small can a competitive multilingual retrieval model be? We present Bekko Embedding: its smallest model, bekko-embedding-v1-a8m, has just under 8M Active Parameters (AP) – the non-embedding parameters that dominate inference compute – yet on official MMTEB Multilingual v2 Retrieval (nDCG@10) it scores 56.2, above the multilingual-e5 family and BGE-M3 (40x the AP) in our comparison. The higher-quality bekko-embedding-v1-a25m (just under 25M AP) reaches 57.5, on par with gte-multilingual-base, and Multilingual NanoBEIR (14 languages) confirms the trend. Both models handle inputs of up to 8192 tokens, and on long-input retrieval (NanoLongEmbed) a25m is the strongest dense model in our comparison. The recipe is deliberately simple. We prune the 22-layer multilingual encoder mmBERT-small to 4 / 13 layers and train the pruned models in two stages – large-scale contrastive learning on about 1.1 billion multilingual pairs from our public corpus, followed by hard-negative fine-tuning with 8192-token long-document negatives – with a masked contrastive loss whose direction depends on pair type, plus the Matryoshka objective. No teacher distillation is used, and all training completes on a single GPU in about 3 days for a8m. Small AP pays off directly in speed: among the compared models measured under identical conditions, a8m is the fastest on both CPU and GPU – 1.6x multilingual-e5-small on x86 CPU – and the fastest on a Raspberry Pi 5. The 384-dimensional output (truncatable to 256/128/64) keeps similarity search and indexing cheap, and row-wise int8 quantization of the vocabulary embedding shrinks the a8m ONNX / OpenVINO build to 124 MiB, which runs in the browser via this http URL. To support reproducible research, we release the model weights, the complete stage-1 corpus, and the independently mined stage-2 hard negatives.
[IR-19] HiEviDR-Bench: A Benchmark for Hierarchical Evidence Aggregation in Deep Research
链接: https://arxiv.org/abs/2607.25151
作者: Yubo Sun,Chunyi Peng,Yukun Yan,Zhenghao Liu,Sen Mei,Bangrui Xu,Xuanhe Zhou,Chi Chen,Maosong Sun
类目: Information Retrieval (cs.IR)
备注: Code and data are available at this https URL
Abstract:Deep research requires models to retrieve, connect, and synthesize evidence from large-scale heterogeneous sources to answer complex queries and produce analytical reports. Existing benchmarks mainly evaluate final outcomes, such as answer correctness, report quality, or citation alignment, while providing limited visibility into whether evidence is correctly selected, linked, and aggregated into supported claims and conclusions. To address this gap, we introduce HiEviDR-Bench, a benchmark for evaluating Hierarchical Evidence Aggregation in Deep Research. HiEviDR-Bench covers open-domain and academic-domain settings under both text-only and multimodal conditions, and represents each instance with an explicit evidence graph that captures evidence selection, cross-source linking, and aggregation from evidence to intermediate claims and final conclusions. Based on this formulation, we develop a traceability-oriented evaluation framework with five dimensions: report quality, evidence traceability, citation accuracy, claim verification, and answer correctness, together with a progressive gating mechanism for fine-grained error localization. HiEviDR-Bench contains 2,000 human-validated questions with evidence graphs across multiple difficulty levels. Experiments on 16 representative multimodal large language models show that, although many systems achieve strong report quality, their performance drops markedly on citation accuracy, claim construction, and answer correctness. Further analysis shows that the main bottlenecks lie in evidence identification and intermediate claim construction, revealing that strong surface-level report quality does not necessarily imply grounded multi-stage reasoning on our benchmark.
[IR-20] Memory Layer: Train the In-Model Cache for Recommendation Models
链接: https://arxiv.org/abs/2607.25110
作者: Liangyuan Na,Gufan Yin,Yixin Bao,Xianjie Chen,Justin Lin,Ziheng huang,Xinyuan Zhang,Wen Zhang,Hao Lin,Xiaoheng Mao,Shuo Tang,Min Yu,Lei Chen,Chao yang,Ziliang Zhao,Mengjiao Zhou,Zheng Qi,Dmitry Barablin,Chuo-Yun Yang,Kaustubh Vartak,Tingting Zhang,Arun Kumar Singh
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注:
Abstract:Early ranking stages in recommendation systems precompute item embeddings and cache them in-model for scoring within strict latency constraints. Because this cache exists only at serving time, outside the training loop, training and serving use different item representations, a structural discrepancy that limits quality and adds operational fragility. We show that co-designing the training and serving paths removes this representation discrepancy at its source. We introduce the memory layer, an in-model key-value embedding cache co-trained with the model: the item tower writes embeddings during training and the model reads them at serving, one source of truth for item representations by construction. Always-on embeddings cover items not yet cached, so every item receives a prediction, and the design consolidates three separate trainer-to-predictor update paths into a single self-contained pipeline. Deployed in production on Instagram Reels, the memory layer raises prediction coverage from 96% to 100%, improves embedding freshness from O(5\text min) to O(20\text s) , and narrows the training-serving Normalized Entropy (NE) gap by up to 86%, yielding over 2\times recall for the freshest content and a 5-6% cold start engagement lift. Because embeddings are produced during training, the system needs no separate bulk-evaluation or publish-time recomputation, cutting training-and-publish computational cost by 30% at neutral serving computational cost.
[IR-21] On the Convergent Validity of Offline Evaluation Designs for Recommender Systems RECSYS2026
链接: https://arxiv.org/abs/2607.25097
作者: Sushobhan Parajuli,Samira Vaez Barenji,Michael D. Ekstrand
类目: Information Retrieval (cs.IR)
备注: 9 pages, Accepted at ACM RecSys 2026 Main Track
Abstract:Offline evaluation on historical interaction logs is the most common evaluation methodology for recommender systems. However, such evaluations depend on sparse, incomplete, or biased data, which raises concerns about whether commonly used evaluation setups reliably reflect true user preferences. In this work, we study how offline evaluation design choices affect the validity of recommender system comparisons. We evaluate a set of recommendation models across several evaluation setups that vary key factors such as data filtering thresholds and candidate set construction. To assess the validity of these configurations, we measure the correlation between model rankings obtained from conventional train-test splits on sparse interaction data and rankings from evaluations based on dense ground-truth user feedback. We use this agreement as an indication of their validity with respect to true user preferences. Our results show that the validity of sparse evaluation depends on the dataset and the specific dense evaluation targets, and that there is no uniformly best offline evaluation design.
[IR-22] ScoreShield: Differentially Private Release of Similarity Scores
链接: https://arxiv.org/abs/2607.25041
作者: Behrooz Razeghi,Parsa Rahimi
类目: Information Retrieval (cs.IR); Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:A growing number of applications, such as biometrics and retrieval-augmented generation (RAG), rely on cosine similarity scores computed between vector embeddings of text, images, or audio. These systems return similarity scores through their APIs for ranking and verification. However, such releases can leak information about individual records and enable membership inference attacks. While differential privacy (DP) provides a principled metric for quantifying attack risks, naïve application of DP mechanisms—such as adding i.i.d. Gaussian noise to vector entries—leads to excessive distortion (i.e., low utility) at a given privacy constraint that scales poorly with the number of released scores. We propose \textscScoreShield, a perturb-then-project mechanism that adds Gaussian noise calibrated to global sensitivity of the chosen score release regime and then projects the result onto the feasibility set of valid cosine objects. \textscScoreShield satisfies (\varepsilon,\delta) -DP for releasing similarity score vectors and Gram matrices. We provide utility guarantees for the exact Frobenius metric projection used in the risk analysis, and prove convergence to feasibility for the practical averaged alternating-projection solver used for large-scale Gram releases. For full pairwise cosine Gram release under record-level replacement adjacency, the exact-projection bound improves the n -dependence of squared Frobenius risk from \Theta(n^3) for the naïve Gaussian baseline to \mathcalO(n^2) for fixed privacy parameters, with sharper local bounds at low-rank Grams. We evaluate the mechanism across RAG, face recognition, semantic retrieval, image similarity, and recommender-system tasks.
[IR-23] Grounded in Consensus In Step With Emerging Science: A Consensus-Anchored Multi-Corpus Clinical Chatbot for Long COVID
链接: https://arxiv.org/abs/2607.25038
作者: Yining Wu,Philip DiGiacomo,Ying Ding,William Brode
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注:
Abstract:Long COVID (LC) poses a challenge for clinical decision support because relevant evidence is distributed across sources with different update cycles, evidentiary roles, and levels of clinical maturity. We present a clinician-facing chatbot that organizes four sources within a retrieval-augmented workflow: expert-curated consensus guidance, current PubMed literature, registered interventional trials, and evidence from living systematic reviews. Consensus guidance is always included to frame responses, while the remaining sources are retrieved in parallel when selected by the user. In an exploratory automated evaluation on 50 clinician-facing questions, our chatbot showed comparable mean ratings to OpenEvidence, with numerically higher scores and lower score variability in an LLM-judged comparison.
[IR-24] Agent Retrieval Bench: Evaluating Repository Context Retrieval for Coding Agents
链接: https://arxiv.org/abs/2607.24882
作者: Bowen Qin,Yi Xie
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Modern coding agents are usually evaluated by whether they eventually produce a correct patch, but patch generation depends on an earlier context-acquisition stage: finding the repository files needed for the task. We introduce Agent Retrieval Bench, a file-level benchmark for this upstream retrieval problem. Samples are built from real coding-workflow signals and evaluated against frozen base-commit repositories, with relevance defined by what an agent needs next rather than direct query-file semantic similarity. The benchmark covers four positive-retrieval tasks: code2test, comment2context, trace2code, and edit2ripple; a fifth subset evaluates selective retrieval using natural evidence-backed no-gold cases and counterfactual wrong-repository controls. Agent Retrieval Bench contains 427 samples across 25 repositories: 345 positive examples, 50 natural no-gold examples, and 32 counterfactual controls. The corpus includes 308 base-commit snapshots, 392,000 files, and 7.9 million chunks. We evaluate lexical retrieval, RepoMap, open-source embeddings, selective abstention, and logged agent context selection. No single retrieval family dominates: Qwen3-Embedding-4B has the best sample-weighted MRR on positive samples, Qwen3-Embedding-8B the best Recall@20, and RepoMap the best budgeted context yield at 8K tokens, with task-level winners differing substantially. Selective thresholds calibrated with counterfactual controls do not improve selective success on natural no-gold cases, revealing a calibration gap. Logged trajectories also miss every gold file on 27-35 percent of samples. A controlled seed-intervention pilot finds that retrieval-derived initial context yields higher file F1 with less post-seed exploration than random non-gold context, while oracle gold context shows substantial remaining headroom.
[IR-25] Ranked by Position: Order Sensitivity as an Exploitable Attack Surface in LLM Listwise Recommenders
链接: https://arxiv.org/abs/2607.24869
作者: Ge Zhang,Jingru Cheng,Huiyuan Chen
类目: Information Retrieval (cs.IR); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 13 pages, 6 figures, 11 tables. Code and data available at this https URL
Abstract:Large language models (LLMs) used as listwise rerankers in recommendation systems suffer from position bias when serializing candidate sets into prompts. We show this order sensitivity creates an exploitable attack surface: an attacker can promote a label-0 target into the top- k solely by reordering candidates, without changing item content, labels, or model parameters. We introduce \mathrmpromo@k to quantify this vulnerability, measuring the fraction of label-0 targets that can be elevated into top- k rankings via permutation. Evaluating across three domains (MovieLens, Amazon Books, and Amazon Fashion), \mathrmpromo@5 reaches up to 0.57 at an attack budget of R = 50 orderings. Furthermore, ordinary permutation stability predicts vulnerability without running the attack. While a bidirectional T5 encoder scorer reduces exposure, permutation-consistency regularization and architectural invariance effectively mitigate it. Pointwise scoring avoids the bias issue but degrades ranking quality. These results demonstrate that input candidate order in listwise LLM reranking is a security-relevant attack vector. Code and data are available at this https URL.
[IR-26] okens are All You Need: Dual-purpose Semantic IDs for Achieving LLM -Level I/O Efficiency in recommendation systems RECSYS2026
链接: https://arxiv.org/abs/2607.24865
作者: Baolei Li,Yiping Yuan,Yilin Zheng,Likang Yin,Ling Liu,Fabio Soldo,Romer Rosales,Xinyang Yi,Lichan Hong
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: RecSys 2026
Abstract:Large-scale recommendation systems face “Memory Wall” bottlenecks due to massive, dense embedding tables. While generative retrieval uses discrete tokens for IDs, high-dimensional context still relies on inefficient dense formats. Inspired by computer vision data compression, we propose Dual-purpose Semantic IDs to achieve LLM-level I/O efficiency. Our methodology uses hierarchical quantization to condense continuous embeddings into discrete Semantic IDs performing two concurrent roles: (1) Collaborative Identity: modeling user-item interactions via learnable embedding table; and (2) Content Reconstruction: using a lightweight Semantic Decoder for on-the-fly embedding approximation. This approach replaces massive vector storage with on-demand reconstruction, reducing system overhead and data footprints. We demonstrate the efficacy of our framework through offline evaluations and successful online deployment in production-scale ranking and retrieval systems at a major video sharing platform, showing that discrete tokens are indeed all you need for highly efficient, content-rich recommendation.
[IR-27] KuaiLive-M3: A Multi-Modal Multi-Domain and Multi-Feedback Dataset for Live Streaming Recommendation
链接: https://arxiv.org/abs/2607.24862
作者: Ke Guo,Changle Qu,Jiayaqi Cheng,Xiao Zhang,Shijun Wang,Xiaoyu Zhang,Xueliang Wang,Le Zhang,Lantao Hu,Jun Xu
类目: Information Retrieval (cs.IR)
备注:
Abstract:Existing public live streaming datasets suffer from three major limitations: they provide limited access to temporally evolving multimodal live content, overlook users’ cross-domain interactions between short videos and live streams, and contain only implicit behavioral signals without explicit feedback that captures users’ perceived content quality and satisfaction. These limitations prevent existing benchmarks from faithfully reflecting real-world live streaming scenarios and hinder comprehensive research on live streaming recommendation. To address these limitations, we introduce KuaiLive-M3, a multi-modal, multi-domain, and multi-feedback dataset for live streaming recommendation, collected from Kuaishou, a leading live streaming and short video platform in China. KuaiLive-M3 covers 21,938 users and contains 35 million live streaming interactions and 111 million short video interactions, with fine-grained timestamps and diverse user behaviors. It further provides approximately 88 million timestamped segment-level multi-modal embeddings that capture the temporal evolution of live streaming content, as well as 25,403 questionnaire-based feedback records that bridge implicit user behaviors and explicit user preferences. Based on these unique signals, we establish benchmarks for cross-domain recommendation, live stream highlight prediction, and questionnaire-enhanced recommendation. Extensive experiments with representative baselines demonstrate that KuaiLive-M3 provides a challenging and realistic benchmark for future live streaming recommendation research. The results further highlight the importance of modeling temporally evolving content, transferring user preferences across domains, and bridging the gap between implicit behaviors and explicit user feedback. The dataset and benchmark code are publicly available at this https URL.
[IR-28] HVM-GraphRAG : Holistic-View Multimodal Graph Retrieval-Augmented Generation on Complex Document
链接: https://arxiv.org/abs/2607.24861
作者: Xin He,Yili Wang,Wenqi Fan,Qing Li,Qinggang Zhang,Yi Chang,Xin Wang
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注:
Abstract:Question answering (QA) over complex documents requires models to retrieve and integrate evidence distributed across distant document regions and modalities. Multimodal GraphRAG provides a promising direction by organizing document evidence with graph structures. However, existing methods often suffer from unreliable cross-modal evidence indexing and expensive graph traversal. To address these issues, we propose HVM-GraphRAG, a holistic-view multimodal GraphRAG framework on complex document. HVM-GraphRAG uses a holistic view to guide graph construction, thereby reducing noisy and conflicting graph updates and building reliable indices between concept-level graph nodes and supporting multimodal chunks. During retrieval, HVM-GraphRAG searches over a compact concept-level graph and directly accesses supporting evidence through the constructed index, avoiding costly traversal over dense entity-level graphs. After obtaining the retrieved evidence, HVM-GraphRAG further reorganizes chunks into modality-specific groups, enabling the answering model to better integrate heterogeneous evidence. Experiments on three datasets show that HVM-GraphRAG achieves the best answer performance in most evaluated settings while substantially improving online retrieval efficiency over representative graph-based baselines.
[IR-29] SearchArt: Training Long-Horizon Search Agent with Scalable Synthetic and Verified Task
链接: https://arxiv.org/abs/2607.24850
作者: Lang Mei,Xiaohan Yu,Chong Chen,Liyan Liu,Xiangnan Chen,Jinchao Ma,Chao Feng,Li Huang,Siyu Mo,Sichen Kang,Yunkun Xu,Zhihan Yang,Zhujun Xue,Jingren Zhang,Qing He,Yingdi Huang,Hao Jiang,Ziao Ma,Zewei Pan,Minhao Sun,Zhuo Tao,Jinzhao Xiao,Gangtao Xin,Huanyao Zhang,Wenjian Zhang,Jiangshan Zhang,Guojie Zhu,Jiaxin Mao,Wentao Zhang
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注:
Abstract:Recent advances in large language models (LLMs) have enabled search agents to autonomously tackle complex tasks across extended search and reasoning horizons. However, training effective search agents remains challenging due to the lack of scalable and long-horizon tasks, and the difficulty of evaluating and correcting intermediate reasoning and tool-use behaviors. We introduce SearchArt, a scalable framework for training long-horizon search agents through verification-driven task synthesis and a multi-stage post-training pipeline. SearchArt constructs large-scale datasets for complex search-, research- and user-oriented tasks by synthesizing diverse information-seeking QA pairs and corresponding search trajectories from web documents and automatically generated evidence graphs. To ensure the reliability of the synthesized data, we design a verification pipeline that jointly evaluates QA consistency, trajectory quality, and the relevance of retrieved evidence. The verified trajectories are subsequently used in a multi-stage training process comprising supervised fine-tuning and reinforcement learning-based policy optimization. Search agents trained with SearchArt exhibit adaptive search planning, iterative evidence aggregation, and complex reasoning over extended interaction horizons. Experimental results demonstrate that, with only (Qwen3.5-) 27B parameters, SearchArt scores 74.39 on BrowseComp-ZH, 70.06 on BrowseComp, and 52.55 on Deepresearch-bench, matching or surpassing frontier closed-source agents on both deepsearch and deepresearch benchmarks.
[IR-30] wo Views One Voice: Evidence-Grounded Conversational Music Recommendation RECSYS
链接: https://arxiv.org/abs/2607.24846
作者: Sungwook Yoo,Sewook Yoo
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 5 pages, 4 figures, 5 tables. 4th-place solution (team swyoo) in the Blind-B industry track of the ACM RecSys Challenge 2026
Abstract:Traditional conversational recommenders entangle retrieval and response generation within a single text interface, so exact entity cues fade as the dialogue’s intent evolves, which compromises explanation credibility. We address this within the ACM RecSys Challenge 2026, which mandates both top-20 ranking and evidence-grounded response generation. This paper presents the third-place solution by team “swyoo” for the Blind-B industry track. We decouple retrieval and response into separate pipelines connected strictly via ranked tracks and metadata. Retrieval combines a hybrid lexical-dense pool for exact matching with a task-adapted pool driven by fine-tuned Qwen 8B adapters. Candidates are calibrated via LightGBM, then routed to an evidence-grounded propose-assign-select (PAS) framework to structure responses. This system also ranked second on the explanation-quality leaderboard in the final blind evaluation. Our findings demonstrate that: (i) isolating retrieval and response preserves both catalog cues and fluid intent; (ii) structuring generation via explicit evidence assignment is key to this near-best-in-class explanation reliability.
[IR-31] REPREC: Representation Driven Parameter-Efficient Recommendation System
链接: https://arxiv.org/abs/2607.24845
作者: Harshini Kavuru,Dwipam Katariya,Giri Iyengar,Pranab Mohanty,Kalanand Mishra,Kalanand Mishra
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models (LLMs) have been applied to sequential recommendation by formulating it as a natural language task. Previous work has improved personalization by incorporating collaborative and sequential signals through input conditioning or LLM fine-tuning. However, existing approaches often rely on one or more of the following: LLM fine-tuning, additional architectural modules, representation distillation, or item-level conditioning over long interaction histories, increasing training complexity and deployment cost. We propose REPREC, a lightweight framework that reformulates LLM-based sequential recommendation through lightweight user representation alignment. REPREC maps a fixed-size user embedding from a frozen sequential encoder into a small set of learned soft tokens through a lightweight MLP injector that conditions a frozen LLM, leaving both pretrained backbones unchanged while training only the injector. We conducted exhaustive experiments on multiple benchmark datasets and demonstrate that REPREC consistently outperforms LoRA while remaining compatible with different pretrained sequential encoders and LLM backbones, enabling a modular and production-friendly recommendation pipeline without modifying either pretrained component. The gains are particularly pronounced for casual and core users across all datasets, highlighting REPREC’s effectiveness in low-data regimes. Finally, when trained on short prompt histories and evaluated with longer contexts, REPREC maintains 85-100% of LoRA’s performance while reducing per-epoch training time by an average of 1.51X, demonstrating an effective balance between recommendation quality and computational efficiency for production deployment. The code is available at this https URL
[IR-32] MedJudgeRAG : Option-Wise Evidence Judgment with Dynamic Knowledge Graphs for Medical MCQA ICML2026
链接: https://arxiv.org/abs/2607.24838
作者: Seongwon Seo,Seung Hwan Cho,Young-Min Kim
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注: 16 pages, 2 figures, Accepted at The Workshop on Graph Foundation Models at the 43 rd International Conference on Machine Learning (ICML 2026)
Abstract:In medical multiple-choice question answering (MCQA), Retrieval-Augmented Generation (RAG) can supplement the domain knowledge of language models (LMs). However, since vanilla RAG indiscriminately utilizes retrieved documents, it can degrade LM performance. To address this, we propose MedJudgeRAG. Our framework represents retrieved documents as a dynamic knowledge graph (KG) composed of entities and relations. For each option, the model judges an evidence verdict from the retrieved documents and the KG. Based on the verdict combination, the model determines a knowledge utilization strategy to reason toward the final answer. These capabilities are trained via supervised fine-tuning using structured reasoning traces generated by a teacher LM. The training employs a weighted cross-entropy loss that differentially weights the KG and reasoning segments. Experiments on two medical MCQA benchmarks demonstrate that MedJudgeRAG consistently outperforms both vanilla RAG and parametric baselines. Furthermore, ablation analysis reveals that the dynamic KG is more effective as graph-conditioned supervision at training time than as an explicit output at inference time. Our code is available at this https URL, and the generated reasoning traces are released at this https URL.
[IR-33] Improving Rare Medication Recommendation with Counterfactual Data Augmentation and Large Language Models RECSYS2026
链接: https://arxiv.org/abs/2607.24829
作者: Shinhwan Kang,Soo Yong Lee,Jaewon Kim,Kijung Shin,Buru Chang
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注: Accepted for publication at the 20th ACM Conference on Recommender Systems (RecSys 2026)
Abstract:AI-based medication recommendation systems have attracted substantial attention due to their potential to enhance patient safety and therapeutic outcomes. Despite the clinical importance of accurately recommending rarely prescribed medications (rare-meds), we observe that most existing methods show significantly lower predictive performance for rare-meds. We attribute this issue to two intrinsic limitations: (a) the inherent scarcity of data for rare-meds and (b) limited consideration of co-recommended medications. To address these limitations, we propose GenRxR, a novel framework based on large language models (LLMs). GenRxR leverages the medical knowledge and clinical reasoning capability of LLMs to generate counterfactual medical data, mitigating the data scarcity issue for rare-meds. It also integrates an LLM into the medication recommendation process to model relationships among co-recommended medications. To further enhance the clinical reasoning, we introduce an instruction tuning step that aligns the LLM’s capability with the recommendation task, enabling better handling of clinical context, including rare-meds cases. In our experiments, we show that GenRxR outperforms 14 (including 5 LLM-based) baselines in most cases. Specifically, it achieves up to 30.9% higher predictive performance for rare-meds than the strongest baseline.
[IR-34] Retrieval-based and Fine-tuned LLM Approaches for Industrial Asset Health Monitoring and Decision Support
链接: https://arxiv.org/abs/2607.24824
作者: Seshu Kumar Damarla,Xiuli Zhu
类目: Information Retrieval (cs.IR)
备注: 6 pages, 4 figures, 6 tables
Abstract:Industrial plants run many important machines such as pumps, turbines, and compressors. Although engineers can use their experience to identify and diagnose machine problems, transferring this reasoning ability to computer systems remains difficult. This work studies how well a retrieval-only method and an open-source large language model (LLM) perform failure-sensor diagnostic reasoning using the FailureSensorIQ benchmark, a multiple-choice question-answering task introduced by IBM Research. In the retrieval-only approach, each answer option is converted into an option-level query and scored using similar correct and incorrect records from the training data. TF-IDF, BM25, semantic search, and hybrid search are tested and compared. In the LLM-based approach, the Qwen2.5-7B-Instruct model is evaluated using zero-shot prompting, few-shot prompting, and QLoRA fine-tuning. The results show that semantic search and hybrid search perform better than pure keyword-matching techniques, indicating that meaning-based similarity is more important for industrial failure-sensor reasoning. Among the LLM-based methods, the fine-tuned model achieves the best performance and substantially improves over zero-shot and few-shot prompting. Error analysis shows that performance decreases as the number of answer options increases. Robustness analysis also shows that all methods are sensitive to option shuffling, changed labels, paraphrasing, and additional distractors.
[IR-35] A Position Paper on Recommender Systems in the Era of Autonomous Agents RECSYS2026
链接: https://arxiv.org/abs/2607.24822
作者: Aixin Sun
类目: Information Retrieval (cs.IR)
备注: Accepted to the ACM RecSys 2026 Main Track
Abstract:For decades, recommender systems have been optimized to serve human users. However, the rapid deployment of autonomous agents introduces a paradigm shift: recommendation consumers are predicted to be increasingly a mixture of humans and authorized agents acting on their behalf. This position paper reviews insights from prior human-centric RecSys research and outlines the transition to this hybrid environment. In our discussion, we treat agents as independent, user-aligned assistants that act as end-users of recommender systems, rather than platform-built components. We characterize the resulting tripartite interactions among humans, agents, and platforms, highlighting the dynamics that can arise across these relationships. This shift presents new opportunities for RecSys research, while introducing unique challenges for evaluation, alignment, and system design.
[IR-36] Dual-Level Atomic and Coordination Geometry Learning for Crystal Property Prediction Using Graph Neural Networks
链接: https://arxiv.org/abs/2607.24818
作者: Sanjay Chakraborty
类目: Information Retrieval (cs.IR); Materials Science (cond-mat.mtrl-sci); Artificial Intelligence (cs.AI); Computational Engineering, Finance, and Science (cs.CE)
备注:
Abstract:Accurate prediction of crystal properties remains a key challenge in computational materials science. While graph neural networks (GNNs) such as CGCNN, MEGNet, ALIGNN, and SchNet have shown strong performance, they primarily represent crystals at the atomic level and implicitly learn local chemical environments through message passing. However, many material properties are governed by coordination polyhedra, the fundamental structural units formed by atoms and their neighboring atoms. To address this limitation, we propose the Coordination Polyhedron Graph Network (CPGN), a multi-scale GNN that jointly learns atomic, bond, and coordination-polyhedron representations. CPGN constructs three coupled graphs: an atom graph encoding elemental and bonding information, a line graph capturing angular interactions, and a coordination polyhedron graph describing Voronoi-derived local environments through corner-, edge-, and face-sharing relationships. Physically meaningful geometric descriptors are incorporated for each polyhedron, while an interleaved message-passing mechanism with bidirectional cross-attention enables effective information exchange across structural levels. Extensive evaluations on the Materials Project, JARVIS-DFT, and QM9 benchmark datasets demonstrate that CPGN outperforms existing state-of-the-art GNN models. It achieves a formation-energy MAE of 0.060 eV/atom and a band-gap MAE of 0.292 eV on the Materials Project, while providing competitive multi-property prediction on JARVIS-DFT and superior HOMO prediction on QM9. The results highlight that explicit modeling of coordination polyhedra improves crystal representation learning and enables accurate, physically interpretable prediction of material properties.
[IR-37] Retrieval-Augmented Generation in LLM s for Mental Health: Quantifying the Incremental Contribution of Retrieval Within a Layered Safety Architecture
链接: https://arxiv.org/abs/2607.24817
作者: Anand Gupta,Akshat Surolia,Shubham Mishra,Shakil Imtiaz,Chaitali Sinha
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 11 pages, 6 figures
Abstract:Digital mental health interventions (DMHIs) offer scalable support, but ensuring they accurately detect users’ intent during volatile situations can be challenging. Pure parametric Large Language models (LLMs) do not contain specific safety critical architecture, and can miss critical cues, or hallucinate, undermining reliability. Retrieval Augmented Generation (RAG), which supplements an LLM with retrieved context, could enhance intent detection during volatile situations. Commercially available DMHIs typically combine multiple independent safety layers like rule-based filters, symbolic escalation protocols, and neural classification. The incremental contribution of any single layer, however, remains unquantified. This paper evaluates six LLM models within a DMHI called Wysa, via a controlled comparison of RAG-enabled versus RAG-disabled modes. Anonymized real and synthetic user-chatbot exchanges were annotated by a qualified clinical team against multi-class intent categories (e.g. self-harm, abuse, panic). The study computed classification accuracy, recall, precision and F1 scores against ground truth labels and tested differences for statistical significance. Performance was also examined by risk category and inter-model agreement. While RAG caused a rise in false alarms, the trade-off is consistent with safety-critical design principles that prioritize sensitivity, where flagged cases are routed to additional review rather than acted on directly. Overall, these findings support RAG as a promising approach to improve the accuracy, consistency and safety of LLM-driven DMHIs. Keywords: Digital Mental Health Intervention, Large Language Model, Retrieval Augmented Generation, Accuracy, Recall, Precision Comments: 11 pages, 6 figures Subjects: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL) Cite as: arXiv:2607.24817 [cs.IR] (or arXiv:2607.24817v1 [cs.IR] for this version) https://doi.org/10.48550/arXiv.2607.24817 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[IR-38] Bumblebee: Interleaved Mixed-Layer Building Blocks for Large-Scale Recommendation Systems
链接: https://arxiv.org/abs/2607.24804
作者: David Bauer,Cancan Zhang,Wenshun Liu,Xiaoyi Zhang,Weijia Liu,Wanli Ma,Yue Weng,Wei Li,Rui Li,Jing Qian,Huayu Li,Xiaoyi Liu,Linhong Zhu,Jerry Fu
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注:
Abstract:Recommendation systems have undergone significant transformations in the past years. The transition from traditional feature interaction modules to generative next-action prediction has pushed the boundaries of personalized content. Developments have largely evolved along two separate tracks. Sequence modeling approaches on the one hand and feature interaction methods on the other. In this paper, we introduce Bumblebee, a recommendation architecture that addresses the lack of interaction between the two directions through an interleaved, stackable block design. Each block implements a micro-pipeline of layers combining sequence personalization, attention-based encoding, and feature crossing into a self-contained unit. Every block produces a joint representation of both feature modalities which is consumed by the next block in the sequence. This mechanism encourages early and repeated mixture of modalities and enriches downstream features with additional contextual information. Residual connections between blocks create cross-modal information pathways and yield additional predictive performance without adding additional parameters. Blocks can be specialized by selectively dropping components, enabling flexible trade-offs between quality and throughput. We evaluate our approach on large-scale industrial data and show consistent improvements over comparable baseline models across several classification and regression tasks. Furthermore, we conduct ablation studies to confirm that the interleaved composition itself is the primary driver of these improvements. Our results suggest that interleaving heterogeneous functional units, rather than composing deep stacks, is a promising paradigm for future-generation recommendation architectures.
[IR-39] SciClaimSeekers at CheckThat! 2026: Retrieving Scientific Sources for Social Media Claims with LLM Reranking
链接: https://arxiv.org/abs/2607.24803
作者: Mohotarema Rashid,Nansu Baniya,Anirban Saha Anik,Xiaoying Song,Lingzi Hong
类目: Information Retrieval (cs.IR); Computation and Language (cs.CL)
备注: CLEF 2026 Working Notes / CheckThat! Lab at CLEF 2026, Jena, Germany
Abstract:Scientific claims often spread on social media faster than they can be verified, while posts rarely link to the original scholarly sources. To tackle this problem this paper presents system called SciClaimSeekers, a retrieval and reranking framework by combining BM25 and zero-shot multilingual E5 retrieval with Reciprocal Rank Fusion (k=60), followed by Qwen2.5-14B-Instruct pointwise reranking. The pipeline reaches 64.36% MRR@5 on the English development set a 13.67-point jump over BM25 and 10.17 points over the unranked hybrid and 64.39% on the official test set, in the CLEF-2026 CheckThat! Task 1 evaluation. Our experiment suggests that large pre-trained models, when combined into a careful pipeline, can be competitive with fine-tuned approaches on this task.
[IR-40] SourceMinds at CheckThat! 2026: NLI-Grounded Citation Auditing in a Multi-Agent Pipeline for Full Fact-Checking Article Generation
链接: https://arxiv.org/abs/2607.24802
作者: Farhan Sharukh Hasan,Anirban Saha Anik,Eric Liu,Xiaoying Song,Mohotarema Rashid,Lingzi Hong
类目: Information Retrieval (cs.IR); Computation and Language (cs.CL)
备注: CLEF 2026 Working Notes / CheckThat! Lab at CLEF 2026, Jena, Germany
Abstract:This paper presents our system for Task 3 of the CLEF 2026 CheckThat! Lab, which focuses on generating full fact-checking articles from claims, veracity labels, and evidence documents. We propose a multi-agent pipeline that combines evidence retrieval, structured fact planning, article generation, gated self-critique, and NLI-based citation auditing. The system retrieves claim-relevant evidence using dense retrieval, reranking, and source-balanced selection, then generates a citation-supported article from a structured plan. A gated self-critique stage revises weakly grounded drafts, while the NLI citation auditor repairs missing citations and removes unsupported or redundant ones. The approach highlights the importance of combining evidence selection, structured generation, and post-generation citation validation for source-grounded fact-checking article generation.
[IR-41] Influence of Prompt Engineering on Small Language Models for Guarded Query Routing
链接: https://arxiv.org/abs/2607.24801
作者: Richard Šléher,William Brach,Kristián Košťál,Lukas Galke Poech
类目: Information Retrieval (cs.IR); Computation and Language (cs.CL)
备注:
Abstract:We study the problem of guarded query routing, where we assume that a user query first meets a router that either determines the ideal endpoint for in-distribution queries or rejects out-of-distribution queries that are potentially unsafe or out of the system’s scope. We investigate whether compact open-weight Small Language Models (SLMs) can jointly handle both tasks under latency constraints. We evaluate 22 models on GQR-Bench and score them with the harmonic mean of in-distribution and out-of-distribution accuracy. We find that mid-scale SLMs come close to frontier model routing quality at much lower latency. Still, many compact models fail because they do not reliably follow the required output format. However, our results show that prompt optimization techniques enable SLMs to handle such cases gracefully, without changing the models’ weights. Moreover, few-shot prompt optimization raises Mistral 7B from 81.79 to 90.87 GQR-Score and lifts Qwen3.5 9B to 95.74, the best optimized score in our study and within 0.3 points of the strongest unoptimized larger model: Gemma 3 27B at 96.01. The bare DSPy signature, without in-context exemplars, is the most effective strategy for Granite 4 Tiny, raising its score from 54.29 to 83.05. These results show that prompt optimization is a useful first step for guarded query routing, while weaker models may still need weight-level adaptation or schema-aware training
[IR-42] When Thinking Before Retrieval Hurts: TraceBound Diagnostics for Adaptive Knowledge-Graph Retrieval ICML2026
链接: https://arxiv.org/abs/2607.24800
作者: Partha Sarathi Purkayastha(ETH Zürich)
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 11 pages, 4 figures, 15 tables. Accepted at the Failure Modes in Agentic AI (FAGEN) Workshop at ICML 2026
Abstract:Adaptive retrieval promises to make knowledge-graph question answering more robust by letting a controller search, inspect neighborhoods, revise actions, and stop when evidence is sufficient. We study this premise by introducing TraceBound, a lightweight profile- and trace-conditioned diagnostic protocol for an ARK-style retriever on text-rich knowledge graphs. TraceBound exposes a compact query profile before retrieval, issues short trace hints after observable failure symptoms, and logs trajectory counters, while keeping graph data, tools, gold labels, and ranking metrics fixed. Across STaRK validation and held-out subsets, the added conditioning improves inspectability but consistently reduces retrieval quality under open-weight controllers. Paired trajectory analysis localizes the degradation to repeated calls, zero-result calls, and misallocated exploration budget, while stricter interaction budgets shorten trajectories without repairing the policy. The result diagnoses the common failure mode in that "thinking before retrieval’’ must be evaluated as a control problem over action selection, not as a prompt-format change.
[IR-43] Multimodal Hybrid Retrieval-Augmented Generation for Scientific Document Understanding using Open-Source SLMs
链接: https://arxiv.org/abs/2607.24799
作者: Alexandru-Andrei Saucă,Ana-Luiza Rusnac
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注: 7 pages, 1 figure, 4 tables
Abstract:Large Language Models tend to hallucinate when answering domain-specific ques tions from scientific documents without prior fine-tuning. Currently, methods such as Retrieval-Augmented Generation partially solve this problem but face different challenges: limited context knowledge, difference between sparse and dense retrieval, and retrieval noise. This paper presents an Advanced Multimodal Retrieval-Augmented Generation system that aims to solve those challenges and im prove the accuracy of information extraction. The proposed architecture introduces a multimodal ingestion pipeline that leverages an open-source Vision-Language Model (Qwen2-VL-2B-Instruct) to generate textual summaries of tables and fig ures. The retrieval phase integrates HNSW-based semantic search with GIN-based lexical search, unified through Reciprocal Rank Fusion and refined using Cross Encoder reranking to minimize retrieval noise. To ensure conversational coherence across multi-turn interactions, a Query Condenser module is employed. Evaluation is conducted by independently assessing the ingestion, retrieval and generation stages using the MMLongBench benchmark, a BeIR-format synthetic dataset and the DeepEval framework. Moreover, results demonstrate a 157% improvement in retrieval quality over a Naive-RAG baseline, with only 50 ms additional la tency, while Qwen2-VL-2B-Instruct achieved results comparable to cloud-based models in BERTScore. These findings validate that open-source optimized SLMs, paired with advanced retrieval strategies, can provide competitive performance for document understanding without relying on cloud-based models.
[IR-44] Prediction Is Not Memory: Dual-Timescale Gated Profile Writing for Persistent User Modeling
链接: https://arxiv.org/abs/2607.24798
作者: Ziyide Li
类目: Information Retrieval (cs.IR)
备注: 9 pages, 5 figures
Abstract:Persistent user profiles increasingly serve as reusable memory in recommender systems, but common update pipelines conflate two decisions: predicting an interaction and deciding whether it should persist in the profile. After an interaction is observed, many systems treat it as evidence for updating durable user state. This assumption can be harmful when the event reflects transient context, exploration, exposure, or short-term satisfaction rather than stable preference formation. We formulate this boundary as selective profile-write control: after an observed interaction, a system should decide whether, and how strongly, to write it into the persistent profile. We introduce a chronological near/far offline protocol in which near-future evidence provides weak write-risk supervision and far-future evidence is reserved for evaluation. We instantiate the controller as SPW-Gate, a lightweight write-risk gate using long-term, short-term, and candidate-profile drift features. On MicroLens-100K, write-all updating hurts far-future profile alignment in 22.45% of test cases under the main protocol. SPW-Gate reduces far hurt to about 14.5% while preserving about 77% write coverage. Matched-coverage controls and prediction-confidence baselines show that the gain is not merely a by-product of writing less, and that next-item confidence is not a sufficient proxy for persistent write validity.
[IR-45] Retrieval not hallucinations will be the limiting factor for LLM -based clinical AI tools
链接: https://arxiv.org/abs/2607.24793
作者: Kirk Roberts,Steven Bedrick,Kurt Miller,William R. Hersh,Hongfang Liu
类目: Information Retrieval (cs.IR); Computation and Language (cs.CL)
备注: Perspective piece
Abstract:Discussions around large language model (LLM) errors in clinical artificial intelligence (AI) generally center around precision errors like hallucinations. This perspective, targeting both clinicians and AI researchers, seeks to shift that discussion to recall errors, particularly in retrieval of patient-level data needed for many clinical AI tools. The perspective outlines types of errors and mitigation strategies, describes research directions in LLMs and retrieval, and provides an overview of retrieval evaluation.
[IR-46] AI-Assisted Knowledge Access for Legacy Enterprise Asset Management in Energy Operations: A Practical Retrieval System
链接: https://arxiv.org/abs/2607.24792
作者: Dave Mercier,Mishca de Costa,Muhammad Anwar,Mark Randall,Issam Hammad
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注: Accepted at the 2026 IEEE the 14th International Conference on Smart Energy Grid Engineering (SEGE 2026)
Abstract:Energy utilities still run engineering work management, engineering procurement, and inventory processes on long-lived enterprise asset management platforms. Replacing these platforms is often cost prohibitive and operationally disruptive, so practical improvement layers are required. This paper presents a retrieval assistant that improves day-to-day knowledge access across three operational modes: vendor documentation question answering, operational data store (ODS) schema question answering, and user interface usage and how-to question answering. The runtime method combines intent understanding, query rewriting, hybrid semantic and vector retrieval, context engineering under token limits, grounded answer generation, and deterministic hyperlink conversion for panel identifiers and cited documentation. The data preparation pipeline emphasizes semantic enrichment as the primary quality lever by adding table and field descriptions, normalizing acronyms across sources, and indexing representative row-level context when useful. A measured pilot shows consistent gains in retrieval quality and user outcomes. Precision at five improved from 0.56 to 0.72, mean reciprocal rank from 0.43 to 0.58, and normalized discounted cumulative gain (nDCG) at five from 0.51 to 0.66. Median task completion time dropped from 14.2 to 8.3 minutes, while usefulness and confidence both increased to 4.0 on a five-point scale. Results are based on a small sample and are reported as pilot findings, but they indicate that intent understanding and semantic enrichment can deliver meaningful operational value in legacy environments while also establishing reusable foundations for future analytics and automation tools.
[IR-47] From Naive RAG to Deep Agent ic Retrieval: An Evolving Context Engineering Pipeline for Regulatory Compliance
链接: https://arxiv.org/abs/2607.24791
作者: Mishca de Costa,Muhammad Saleh Anwar,Dave Mercier,Issam Hammad
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Accepted at the 2026 IEEE the 14th International Conference on Smart Energy Grid Engineering (SEGE 2026)
Abstract:Retrieval-augmented generation (RAG) is the dominant paradigm for applying large language models (LLMs) to enterprise document corpora, yet naive implementations encounter hard limits as corpus scale and query complexity grow. This paper traces the evolution of a production retrieval pipeline at Ontario Power Generation (OPG) for regulatory compliance and rate case analysis under Ontario Energy Board (OEB) reporting requirements. We examine successive stages: naive RAG, hybrid retrieval with re-ranking, agentic function-calling retrieval, and a deep multi-agent architecture with code-based tool synthesis and explicit planning, and identify the failure modes and tradeoffs that motivated each transition. We formalize the mature architecture as Progressive Evidence Acquisition with Cost-Aware Escalation (PEA-CAE): begin with low-cost, high-precision retrieval and escalate to full-document reads only when the expected evidence gain justifies latency and cost. Our findings show that context engineering is a more tractable and economically viable path than domain-specific fine-tuning for large, evolving regulatory corpora. More broadly, the progression toward deep agentic retrieval mirrors classical information retrieval ideas, introducing adaptive query reformulation, progressive document discovery, and hierarchical subagent summarization as practical system primitives. Operational traces further support the search-based nature of modern retrieval systems, where iterative evidence acquisition and adaptive planning increasingly replace single-pass retrieval as the foundation for enterprise-scale question answering.
[IR-48] NEXT: Reasoning -Driven Video Recommendation via a Vision-Language Model
链接: https://arxiv.org/abs/2607.24789
作者: Yuming Liu,Hongye Yang,Harrison Zhao,Ellie Zhu,Bokai Cao,Lei Huang,Lizhu Zhang,Xiangjun Fan
类目: Information Retrieval (cs.IR); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Multimedia (cs.MM)
备注: 13 pages, 2 figures, 5 tables
Abstract:We present NEXT (Next-interest EXploration Transformer), a reasoning-driven video recommendation framework that reasons over the video a user has just watched, infers the viewer’s next intent, and retrieves concrete follow-up videos. Explicit continuations such as episodes are linked directly; implicit cases are handled by generating intent queries and searching for matching candidates. This Item-to-Intent-to-Item formulation produces directed recommendations beyond co-engagement correlation or semantic similarity. To make this framework reliable at scale, we train NEXT-8B, a purpose-trained 8B vision-language model with a three-stage recipe: Perception-Enhanced Reinforcement Learning for query-agnostic evidence extraction, Distribution-Aligned Supervised Fine-Tuning over real and synthetic visual QA mixtures, and Group Relative Policy Optimization for last-mile alignment. NEXT-8B achieves the best single-model DocVQA performance, ranking second overall only behind a multi-agent system while surpassing a substantially larger 200B+ scale model, and improves next-intent logic-wise quality by 3.3% over the base model in a task-specific LLM-as-a-judge evaluation. We deploy NEXT as an additional retrieval path in a large-scale social media recommendation system and observe statistically significant production gains, including +0.53% watch time and +0.51% distinct video exposure. Overall, NEXT shows that a carefully trained compact vision-language model can serve as a practical reasoning engine for next-interest exploration at production scale. Comments: 13 pages, 2 figures, 5 tables Subjects: Information Retrieval (cs.IR); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Multimedia (cs.MM) Cite as: arXiv:2607.24789 [cs.IR] (or arXiv:2607.24789v1 [cs.IR] for this version) https://doi.org/10.48550/arXiv.2607.24789 Focus to learn more arXiv-issued DOI via DataCite
[IR-49] Unlocking Spatial Grounding in Large Audio-Visual Retrieval models
链接: https://arxiv.org/abs/2607.24786
作者: Hugo Malard,Michel Olvera,Sanjeel Parekh,Gaël Richard,Slim Essid,Stéphane Lathuilière
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Multimedia (cs.MM); Sound (cs.SD); Audio and Speech Processing (eess.AS)
备注:
Abstract:Weak supervision sets a practical regime for audio-visual sound source localization as dense spatial annotations are costly to obtain at scale. The task, however, remains challenging, as models must locate sound sources from temporally aligned audio-visual data without pixel-level supervision. Recent large-scale audio-visual retrieval models, trained at unprecedented scale, encode rich multimodal structure. We show their latent representations, though optimized for global alignment, can nonetheless enable fine-grained spatial grounding. While spatial detail is progressively lost in the upper layers of retrieval backbones due to global pooling, intermediate visual tokens retain highly structured spatial information. To exploit this, we introduce LAIP (\emphLocalization via Audio-Informed Pooling), a framework that employs a lightweight \emphAudio-informed Spatial Pooling (AiSP) to replace the standard global aggregation module. By using frame-aligned audio to query intermediate visual tokens, LAIP recovers localized spatial information that is otherwise discarded by the frozen retrieval pipeline. Our approach achieves state-of-the-art performance on AVSBench and AVATAR, nearly doubling previous results on the latter. These findings prove that accurate localization does not need to be learned from scratch; instead, it can be unlocked from existing retrieval representations, providing a unified path for both retrieval and localization tasks.
[IR-50] hree Sides of Retrieval: Factorial Evidence for Document-Side Query-Side and Answer-Side Complementarity in RAG
链接: https://arxiv.org/abs/2607.24781
作者: Ng S. T. Chong
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注:
Abstract:RAG systems rely on chunking, which destroys structural information in documents. Existing heading-based retrieval (Jeong et al., 2025) requires multiple LLM calls per document and returns sub-chunks within matched sections. We introduce ToC-guided page retrieval, which infers headings from visual formatting without LLM calls, embeds them as a parallel index, and loads full page sections. Across 1,280 conditions on 8 enterprise documents (5 to 195 pages), we find: (1) ToC is a significant main effect on answer quality (d = +0.41, p = 0.031), with the largest gains in completeness (+0.40) and usefulness (+0.40); (2) combined with answer-side verification, it outperforms query-side decomposition + verification (d = +0.32, p = 0.036); (3) ToC contributes 20% of citations despite adding only 2.9 pages per query; (4) gains are directionally larger on longer documents (up to +1.50 on 118 pages), though the trend does not reach significance with 8 documents; and (5) a 480-condition sensitivity analysis finds no significant parameter effects (all p 0.38), confirming defaults are near-optimal. The contribution is both methodological (a new zero-LLM-cost retrieval algorithm) and empirical: factorial evidence that document-side, query-side, and answer-side enhancements are complementary, a three-way interaction not previously studied.
[IR-51] JKO-RAG : Distributional Retrieval as Wasserstein Free-Energy Gradient Flow
链接: https://arxiv.org/abs/2607.24776
作者: Levi Segal,Murari Ambati
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注:
Abstract:RAG pipelines return a \emphranked list of passages. We argue this is a mismatch: the downstream language model conditions on a \emphset, and the selection problem is fundamentally geometric. We propose \jko, which frames reranking as minimising a free-energy functional F§=\textrelevance+\textentropy+\textredundancy under Wasserstein-2 gradient flow via the Jordan–Kinderlehrer–Otto proximal scheme. The ground metric C_ij=(1-\cos\langle z_i,z_j\rangle)^2 encodes the semantic geometry of the embedding manifold. Our central contribution is a \emphlinear-response theory explaining \emphwhy the Wasserstein geometry helps: the Wasserstein and KL retrieval maps differ only in their proximal Hessian – dense and geometry-aware for W^2 , diagonal and geometry-blind for KL – and this difference damps the mass transport that query paraphrase induces. The theory yields a falsifiable prediction: the stability advantage is monotonically decreasing in step size h . We verify this empirically via free-energy descent, frequency-resolved perturbation response, the predicted h -dependence, and a certified-radius analysis. Four extensions are introduced: \textbf\nmjko (learned ground metric), \textbf\bwjko ( W^2 --KL interpolation), \textbf\samjko ( 2\times speedup), and \textbf\dualrank (OT dual potentials as confidence signals). Across five BEIR benchmarks, \jko\ outperforms the cross-encoder on all five; the decisive advantage is robustness – 22–38% more stable under paraphrase, 2\times fewer leaked distractors.
[IR-52] he Effect of Text Chunk Size on Retrieval-Augmented Generation Performance
链接: https://arxiv.org/abs/2607.24767
作者: German Garrido-Lestache Belinchon,Hugo Garrido-Lestache Belinchon
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Retrieval-Augmented Generation (RAG) systems have emerged as a powerful process for allowing large language models (LLMs) to retrieve relevant information to use as source material during text generation. A critical yet under-explored component of these systems is the granularity at which source documents are segmented into retrievable chunks. The size of these chunks has the potential to significantly influence generation quality, contextual correctness, retrieval precision, and computational efficiency. Despite its importance, chunk size is often selected without proper evaluation of its impact on generation quality. Smaller chunks, such as individual sentences, may allow for precise retrieval by narrowing the focus of each chunk. However, they contain less information, which may limit the model’s ability to generate coherent responses. Larger chunks, such as entire chapters, contain lots of broad information that may improve correctness, but also introduce additional noise and increase computational cost. Because larger chunks contain more information, the number of chunks returned to the model must also be considered. This paper evaluates how chunk size, along with the number of retrieved segments, influences generation quality and retrieval effectiveness. By comparing these configurations, this study seeks to better understand how document segmentation affects the performance and efficiency of Retrieval-Augmented Generation systems. segmentation affects the performance and efficiency of Retrieval-Augmented Generation systems.
[IR-53] CHaystack: Benchmarking Chinese Document Retrieval and VQA
链接: https://arxiv.org/abs/2607.24760
作者: Hanxi Li
类目: Information Retrieval (cs.IR)
备注:
Abstract:Retrieval-augmented generation (RAG) has made substantial progress in extending the memory of large language models (LLMs), and recent advances have further pushed RAG from pure text settings toward multimodal scenarios. In the document understanding domain, document visual question answering (DocumentVQA) has evolved from question answering over a single document to retrieval-and-generation pipelines over large-scale document collections. However, a benchmark specifically designed for Chinese large-scale document retrieval and question answering is still lacking. To bridge this gap, we introduce CHaystack, a new Chinese DocumentVQA benchmark that covers four document categories, namely academic papers, advertisements, web pages, and real-world photographed documents, enabling a more comprehensive evaluation of DocumentVQA systems. In addition, we present CDocRAG, a Chinese DocumentVQA system that uses a VLM-based relevance filter to verify retrieved document images before answer generation. We evaluate representative open-source embedding and generation models on CHaystack. The results reveal a clear contrast in category-wise strengths: Qwen-family models perform best on text-rich documents such as webpages and papers, whereas other models only achieve competitive results on visually rich categories such as advertisements and degrade sharply on text-dense documents. For retrieval, Qwen3-VL reaches 71.91 Recall@1 while the best non-Qwen model achieves only 14.40. These results indicate that the core challenge of CHaystack lies in Chinese textual encoding, and that Chinese large-scale DocumentVQA still leaves substantial room for improvement. Our code and dateset is available at this https URL.
[IR-54] VLD-RAG : Agent ic Vision-Language Retrieval-Augmented Generation for Long Visually-Rich Multi-Page Documents
链接: https://arxiv.org/abs/2607.24748
作者: Seonok Kim
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Visually-rich documents such as reports, slides, and manuals often distribute the evidence needed to answer a question across multiple pages, mixing text with layout cues, tables, charts, and figures. This work studies multimodal retrieval-augmented generation for question answering over such visually-rich long documents, where retrieval must select evidence pages that include both textual and visual signals. We present VLD-RAG, an agentic multimodal RAG framework for multi-page evidence retrieval and cross-page reasoning over long documents. VLD-RAG builds a page-preserving multimodal index that stores parsed text, page-level metadata, and dense visual representations, and uses a hybrid retrieval strategy that combines keyword-based sparse search with dense semantic queries to identify candidate sources and evidence pages. A verifier-guided agent workflow coordinates a Retrieval Agent, Answer Agent, and Validation Agent to broaden evidence coverage, detect missing citations, and refine retrieval requests when needed. We evaluate retrieval with Top-1 and Top-5 evidence-page accuracy and generation with generalized accuracy, and show that VLD-RAG improves both evidence-page retrieval and end-task question answering on visually-rich long-document benchmarks, including LongDocURL and MMLongBench-Doc, outperforming previous vision-based retrieval baselines. These findings highlight that coordinated agent verification and multimodal hybrid retrieval are crucial for reliable grounding when correct answers depend on evidence scattered across pages.
[IR-55] DocAnnot – Accelerating the Creation of Key Information Extraction Datasets with GenAI-Powered Auto-annotation
链接: https://arxiv.org/abs/2607.24745
作者: Siddartha Reddy,Harikrishnan P M,Goutham Vignesh,Varun V,Vishal Vaddina
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV)
备注: 15 pages, 2 figures
Abstract:Key Information Extraction (KIE) is vital for many document applications, but creating training datasets is traditionally a time-consuming manual process. We introduce DocAnnot, a framework that significantly accelerates KIE dataset generation. DocAnnot leverages a Large Vision Language Model (LVLM) for label value extraction, OCR for text/bounding box detection, and a novel Spatially Informed Contextual Matching (SICM) algorithm. SICM improves label-value association by combining spatial relationships and proximity analysis with textual matching. We evaluate our framework on the CORD and SROIE benchmarks, demonstrating its ability to auto-generate annotations with F1-scores of 0.679 and 0.846, respectively. Furthermore, we investigate the effectiveness of using auto-annotated data for fine-tuning downstream KIE models. While human-annotated data remains superior, models trained exclusively on DocAnnot’s outputs attain respectable performance (e.g., LayoutLMv3 achieving an F1-score of 0.6765 on CORD). These results show that while our framework significantly reduces reliance on manual effort, it does not yet fully eliminate the need for human intervention. However, by automating the process to a point where reviewers can efficiently refine outputs, our system enables near-perfect annotations with much greater efficiency than manual annotation from scratch. This approach offers substantial time and cost savings, making it valuable for resource-constrained settings and rapid model prototyping.
人机交互
[HC-0] LLM 4OSC: Profile-Bound Natural Language Control with Deterministic Validation for Open Sound Control
链接: https://arxiv.org/abs/2607.26024
作者: Yuan-Yi Fan
类目: Human-Computer Interaction (cs.HC); Sound (cs.SD); Audio and Speech Processing (eess.AS)
备注:
Abstract:Open Sound Control (OSC) is the dominant wire protocol for real-time parametric control in professional audio, live performance, and virtual production. Large language models can emit plausible OSC, but they hallucinate addresses, mishandle type tags, and fail under paraphrase- unacceptable in show-critical contexts. We present LLM4OSC, a local-first architecture in which models propose structured intent JSON over a human-reviewed device profile, and deterministic code validates, clamps, and encodes before any UDP send. We introduce a frozen evaluation harness with CI gates on wrong-send rate: mismatches that would still pass validation and transmit. On a Max/MSP hero profile (12 patterns; 8 literal + 8 paraphrase + 4 refusal cases), after profile tag enrichment, symbolic slot fill, NL refine, and a retrieval confidence gate, backends B0–B3 all pass frozen gates (100% semantic accuracy, 0% wrong-send). B0 (rules) remains the production default at ~0.05ms; LLM backends remain ~3-4s. Historical few-shot B2 accuracy of 62.5% rises to 100% on this suite only after symbolic post-processing- not because the 0.5B model alone becomes show-safe. We argue for propose-validate-send and wrong-send rate as first-class metrics for language-to-control systems.
[HC-1] Who is scientific code for? Maintaining human-readable landmarks in agent -written code
链接: https://arxiv.org/abs/2607.25975
作者: Elle O’Brien
类目: Human-Computer Interaction (cs.HC)
备注: Position piece submitted to Infrastructure @ CSCW 26 workshop
Abstract:Scientific research involving code has long rested on the assumption that at least one person understands why the code exists. As scientists adopt coding agents, this assumption is breaking down. Drawing on an ongoing contextual inquiry of scientific programmers working with agentic tools (four cases to date), a survey of over 800 scientific programmers, and my own analysis workflows, this position piece describes how scientists are inventing personal conventions, “landmarking strategies”, for marking which artifacts in a codebase are meant for human understanding and which are context for agents. They repurpose shared infrastructure (version control especially) in idiosyncratic ways, and I argue that this quiet de-standardization could complicate collaboration in teams with heterogeneous software practices. Alternatively, teams that explicitly delineate what is human-readable versus agent context will be better able to develop, document, and maintain scientific code.
[HC-2] E-MagDiP: Electro-Magnetic based Differential Privacy for EEG based Community Sensing
链接: https://arxiv.org/abs/2607.25968
作者: Ayanga Imesha Kumari Kalupahana,Vishruti Ranjan,Li-Shiuan Peh
类目: Cryptography and Security (cs.CR); Human-Computer Interaction (cs.HC)
备注:
Abstract:EEG-based community sensing programs are emerging globally as a tool to leverage aggregated brain data to gain insights into attentiveness of students and employees. But these programs raise privacy concerns because EEG signals contain sensitive personal information. Differential Privacy (DP) can protect individuals while preserving aggregate statistics yet applying DP to EEG data is challenging as it requires user-level noise generation, which increases power and latency. Besides, most commercial EEG headsets cannot be modified to add such noise. We propose E-MagDiP, a framework that uses an external radio to transmit RF signals onto EEG headsets, perturbing signals at acquisition to induce DP noise. To the best of our knowledge, E-MagDiP is the first framework to use RF signals for privacy instead of attacks, enabling practical DP for EEG community sensing without any user-level modification.
[HC-3] Faster Higher Stronger? The Impact of GenAI on Knowledge Work Productivity - Evidence from the Field
链接: https://arxiv.org/abs/2607.25922
作者: Sven Bottesch,Chiara Schwenke,Jakob Zimmermann,Maximilian Förster,Mathias Klier
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:The rise of generative artificial intelligence (GenAI) has fueled high expectations regarding its potential to enhance knowledge work productivity in terms of efficiency and quality. Building on task-technology fit (TTF) theory, we empirically examine the extent of GenAI’s productivity effect for different task types. We conducted a randomized lab-in-the-field experiment with 128 knowledge workers from a multinational industrial organization. Participants completed three representative knowledge work tasks (knowledge acquisition, packaging, and creation), either with or without GenAI. Results show that GenAI consistently increases efficiency across tasks. However, its impact on quality is task-contingent: quality increases for knowledge packaging and creation but declines for knowledge acquisition. Furthermore, GenAI tends to reduce quality variance for knowledge packaging and creation, primarily benefiting lower-performing knowledge workers. However, it increases quality variance for knowledge acquisition. These findings contribute to a more granular, differentiated understanding of GenAI’s productivity impact and hold implications for research and practice alike.
[HC-4] AnnoBench: A Benchmark for Visualization Annotation Generation
链接: https://arxiv.org/abs/2607.25911
作者: Md Rahat-uz-Zaman,Md Dilshadur Rahman,Andrew McNutt,Paul Rosen
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注:
Abstract:Annotation is among the most demanding visualization tasks to automate, as it simultaneously requires correctly navigating visual, semantic, and stylistic constraints. Failure to meet any of these conditions severely undermines the utility of an annotation, rendering it challenging to read, inaccurate, or visually discordant. Despite a growing body of annotation tools and automations, no existing benchmark or evaluation framework tests whether these conditions are met because of their scope and annotation not being the focus of their studies. We introduce AnnoBench, a benchmark for visualization annotation that materializes the inherent challenges of this domain in a structured and testable manner. AnnoBench pairs visualizations from professional data journalism and visualization galleries with annotation tasks, spanning four representation formats, five chart description conditions, and two prompt specification levels. The benchmark is executed via VLM-as-a-judge, using models aligned with manual human assessment. We evaluate the benchmark via four one-factor-at-a-time experiments, exploring the effects of input representation, semantic context, and prompt specificity, and model selection on annotation quality. This work provides a foundation for advancing annotation automation, tooling, and visualization-generation pipelines.
[HC-5] AIs Capability in Assisting Scientific Research in Physics Astrophysics and Cosmology II: Project Planning and Proposal Evaluation
链接: https://arxiv.org/abs/2607.25881
作者: Jia Liu,Veena Krishnaraj,Kateryna Vovk,Kosuke Aizawa,Adrian E. Bayer,Linda Blot,Jessica Cowell,Suyog Garg,Jonathan Grée,Anamaria Hell,Ben Horowitz,Masaya Ichikawa,Kanyuni Iemoto,Keigo Kondo,Zacharie Lorsin,Kevin McCarthy,Jamie Robinson,Miguel Ruiz-Granda,Leander Thiele,Ievgen Vovk,Mingshen Zhou
类目: Computation and Language (cs.CL); Cosmology and Nongalactic Astrophysics (astro-ph.CO); Instrumentation and Methods for Astrophysics (astro-ph.IM); Human-Computer Interaction (cs.HC); General Relativity and Quantum Cosmology (gr-qc)
备注: 16 pages, 4 figures
Abstract:We investigate how well large language models (LLMs) can assist scientific project planning and proposal evaluation. One-page project plans were independently generated for eight expert-conceived research projects in physics, astrophysics, and cosmology by human researchers and three contemporary LLMs (ChatGPT, Claude, and DeepSeek; mid-2025 models, used with their default tool access). The resulting 32 proposals were blindly evaluated by four human reviewers and two newer frontier LLMs (Claude Opus 4.8 and ChatGPT Pro 5.5) using a four-aspect evaluation rubric. Reviewers were also asked to identify whether each proposal was written by a human or an AI. Human reviewers rated human- and AI-written proposals similarly overall, whereas both AI reviewers scored AI-written proposals about one point higher (on a five-point scale) than human-written proposals. Human reviewers correctly identified human- and AI-written proposals 72% and 79% of the time, respectively, while both AI reviewers correctly classified all 32 proposals (100%). These results suggest that current LLMs can produce project plans comparable to human-written ones in the eyes of human reviewers, but that AI reviewers show a systematic preference for AI-generated proposals. Our results suggest caution when deploying LLMs widely in proposal preparation and evaluation.
[HC-6] BioDisclose: An Actionability-Aware Benchmark for Biomedical Safety under Adversarial Elicitation
链接: https://arxiv.org/abs/2607.25700
作者: Yinuo Zhu,He Liu,Boyuan Gu
类目: Human-Computer Interaction (cs.HC)
备注: 27 pages, 3 figures
Abstract:Large language models (LLMs) increasingly support biomedical research, yet their behavior under adversarial requests for dual-use knowledge remains insufficiently characterized. We introduce BioDisclose, a benchmark for measuring biomedical knowledge disclosure under adversarial elicitation. BioDisclose contains 480 prompts derived from 24 expert-authored scenarios across six biomedical risk domains and four elicitation families spanning academic, historical, role-playing, and decomposed prompting. We grade model responses on a four-level scale from refusal to executable disclosure, distinguishing high-level discussion from technically specific and actionable content, including refuse-then-leak behavior. Across five deployed LLM systems, detailed-or-higher disclosure rates vary substantially, ranging from 9.2% to 64.0%. Academic framing is the most effective elicitation family on average (43.2%), while laboratory safety scenarios show the highest disclosure rate across domains (51.5%). These results reveal pronounced variation across models, prompting strategies, and biomedical risk categories, suggesting that current safeguards remain uneven in high-stakes scientific settings. BioDisclose provides a focused testbed for evaluating biomedical safety beyond binary refusal metrics.
[HC-7] A Human-in-the-Loop Corpus for LLM -Based Simplification of Scientific Summaries
链接: https://arxiv.org/abs/2607.25630
作者: Kyuri Im,Michael Färber
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注: Accepted at FGWM@KI2026
Abstract:Interdisciplinary research is accelerating, yet scientific papers remain difficult to understand outside their home fields. We study large language model (LLM)-based simplification of scientific texts and present a human-in-the-loop workflow that transforms expert summaries into more accessible versions for non-specialists. Using SciSummNet as the source corpus, we first generate baseline simplifications with GPT-4o-mini. In Phase 1, readers from STEM fields outside computer science identify difficult sentences and phrases and compare the original and GPT-simplified summaries in terms of comprehensibility, naturalness, and simplicity. In Phase 2, computer science experts use this feedback to create expert-edited reference simplifications. We release the resulting corpus together with human judgments and automatic evaluation results. The Phase 1 judgments show a clear preference for the GPT-generated summaries in terms of comprehensibility and simplicity, while qualitative analysis of the Phase 2 edits highlights the importance of preserving domain-specific terminology and the strength of scientific claims. The resulting resource supports the training and benchmarking of simplification systems for cross-disciplinary scientific communication.
[HC-8] Beyond Epistemia: Epistemic Schizologia and Large Language Models as Techno-Semiotic Machines
链接: https://arxiv.org/abs/2607.25620
作者: Federico Cabitza,Gianluca Colombo
类目: Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注:
Abstract:Quattrociocchi and colleagues warn that the fluent outputs of large language models may allow linguistic plausibility to substitute for epistemic evaluation, producing the condition they call Epistemia: the experience of possessing knowledge without undertaking the practices through which judgment would ordinarily be warranted. This article accepts that diagnosis but challenges its explanatory framework, which compares an embodied, socially situated human knower with an isolated generative model thereby locating epistemic legitimacy in capacities internal to autonomous agents. Drawing on Carlo Sini’s philosophy of practices, writing, signs, and technics, we propose instead to understand a large language model (LLM) as a techno-semiotic machine that automates a phase of written semiosis by producing plausible linguistic configurations from the sedimented archive of human writing. From this perspective, Epistemia is one consequence of a broader phenomenon that we call epistemic schizologia: the socio-technical cleavage between signs as linguistically accomplished expressions and signs as moments within socially embedded circuits of interpretation, evidence, criticism, verification, and responsibility. This cleavage is reinforced by eikotic closure, through which a plausible continuation is presented with the finality of an epistemic result, and by algorithmic authority and epistemic self-misrecognition. The relevant unit is therefore not the model alone but the complete practice in which generated inscriptions are prompted, interpreted, verified, contested, used, and made consequential. This reframing preserves the distinction between linguistic production and responsible understanding while grounding a design programme centred on inspectable genealogy, contestability, distributed responsibility, epistemic agency, and the evaluation of hybrid human–AIpractices.
[HC-9] “Drag on Slayer Becomes the Drag on”: How Players Perceive and Respond to Inequality in the Game World of Whiteout Survival
链接: https://arxiv.org/abs/2607.25574
作者: Shiyu Lei,Ke-Xin Ren,Daiyi Jiang,Ray LC
类目: Human-Computer Interaction (cs.HC)
备注: 40 pages, 4 figures, 1 table, will be presenting this paper at CHIPLAY 2026
Abstract:Inequality in real-world societies are associated with psychological distress and behavioral consequences. However, less is known about whether similar dynamics emerge when inequality exists within virtual environments or make-belief worlds. As online games increasingly constitute meaningful social spaces, it becomes critical to examine how players perceive and react to structural and resource differences online to optimize their experiences. This study studies perceptions of inequality in the online simulation game “Whiteout Survival,” using semi-structured interviews and think-aloud gameplay walkthrough protocols. By focusing on players’ interpretations of resource distribution, ranking systems, gaming mechanisms, and in-game social dynamics, our analyses revealed that players’ attitudes on inequality vary according to their relative status: those occupying lower positions often criticize unfair structures, yet as they acquire stakes through resource accumulation or social integration, many defend the same systems they previously opposed. These shifts reveal how hierarchies reproduce position-dependent evaluations of fairness. The consequences of inequality on player actions depended on the transparency of game mechanisms, the structure of community hierarchies, and differential social capital. This work shows how human social perception and consequent actions are transformed when enacted in virtual processes in make-belief.
[HC-10] From Dyad to Triad: Eliciting XAI Requirements in Stroke Rehabilitation
链接: https://arxiv.org/abs/2607.25423
作者: Param Rajpura,Yogesh Kumar Meena
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI); Emerging Technologies (cs.ET)
备注: AI and Cognitive Computing for Trustworthy Human Machine Systems Session, IEEE International Conference on Systems, Man, and Cybernetics (IEEE SMC 2026)
Abstract:Eliciting explainable AI (XAI) requirements from stroke survivors presents a methodological challenge with direct implications for the design of trustworthy brain-computer interfaces for rehabilitation. How can patients and caregivers articulate preferences about algorithmic transparency when they lack conceptual frameworks for explainability, and when standard elicitation approaches are structurally inadequate for users with acquired communication disorders? We present a video-based scaffolding protocol for XAI requirements elicitation, developed and piloted in a rehabilitation context. In a formative study with three stroke survivors (two with moderate-to-severe aphasia) and three caregivers, facilitators employed four scaffolding approaches alongside the videos: 1) analogical bridging mapping AI states to familiar systems, 2) projective personas depersonalising sensitive topics, 3) binary forcing reducing cognitive load, and 4) extended response time. These approaches successfully surfaced heterogeneous, sometimes conflicting XAI needs across participants. Reflexive analysis additionally revealed three systematic facilitation biases, namely, normative bias, hypothesis confirmation bias, and presence effect, where scaffolding inadvertently shaped responses. We present these as protocol risk guidelines for practitioners. Together, the protocol and guidelines constitute a reusable methodological contribution for eliciting patient-facing XAI requirements in rehabilitation, arguing that such elicitation is a necessary prerequisite for trustworthy human-machine systems design, not an optional preliminary.
[HC-11] Agent ic AI Autonomy Assessment: A Decision-Support Framework Towards Governed Supply Chain Systems
链接: https://arxiv.org/abs/2607.25405
作者: Lennart Trumpler,Rodrigo Furlan de Assis,Elias Ribeiro da Silva,Luis Antonio de Santa-Eulalia,Christian Hendriksen
类目: Human-Computer Interaction (cs.HC)
备注: 28 pages, 7 figures
Abstract:Supply chain decision-making is rapidly transforming with the rise of agentic AI - highly autonomous systems that can operate on complex, long-horizon tasks. Yet the adoption of agentic systems outpaces their governance: existing taxonomies of autonomy only offer discrete classifications, rely on subjective judgement, and cannot track autonomy across a system’s life cycle, leaving enterprises unable to assess the risks of increasingly autonomous supply chain agents. This paper proposes the Agentic AI Autonomy Assessment (AAAA) framework, which defines and measures the degree of autonomy at a task level. The framework is based on the three dimensions of user delegation, consultation, and collaboration, enabling continuous monitoring from an agent’s development through its runtime to end-of-life. The framework’s construct validity was tested in a simulated beer distribution game, examining how the autonomy score relates to a company’s performance. Results reveal a weak link between autonomy and tier costs with a positional effect: upstream tiers benefit from higher autonomy while downstream tiers are harmed, positioning autonomy as an inherent dimension of agentic systems, orthogonal to capability. The framework provides a foundation for risk assessment, governance, and transparent autonomy policies to support the governed enterprise adoption of agentic AI in supply chains.
[HC-12] ailVis: Expressive Chart Refinement Preserving Data-Binding Integrity
链接: https://arxiv.org/abs/2607.25386
作者: Yumin Song,Seokhyeon Park,Soohyun Lee,Aeri Cho,John Joon Young Chung,Hyeon Jeon,Jinwook Seo
类目: Human-Computer Interaction (cs.HC)
备注: 11 pages, 8 figures, 1 table
Abstract:Creating static visualizations for presentations and publications requires granular refinements of visual details, even for simple charts. Existing data-driven visualization tools offer limited interactive control for such refinements, forcing users to export charts to external graphic editors and breaking the critical link between data and visual representation. To address this gap, we propose an extended InfoVis Reference Model to account for post-render design refinement. A formative study with 18 visualization practitioners and a follow-up survey of 35 respondents confirmed that this stage is pervasive yet unsupported in current practice. Based on these findings, we present TailVis, a visualization authoring system that enables expressive visual customization while preserving data-binding integrity. TailVis supports element-level direct selection and scope expansion, allowing users to define a data-aware scope ranging from a single mark to a data-driven category with a simple selection. For modifications beyond predefined controls, TailVis blends natural language input with dynamically generated GUI widgets, where deictic interaction lets users reference elements simply by clicking them, keeping even open-ended edits bound to the data. To support rigorous exploration and comparison of design alternatives, TailVis implements a provenance history that enables users to capture diverse design iterations while ensuring data-visual integrity. A user study with 12 participants verified that TailVis effectively supports expressive, granular refinement without sacrificing data binding, significantly reducing repetitive manual processes in an integrated environment.
[HC-13] FIDAC: An Easy-to-use Pipeline to Extract and Interpret Interpersonal Distance From Video
链接: https://arxiv.org/abs/2607.25146
作者: Keshav Rastogi,Eugy Han,Jeremy N. Bailenson
类目: Computer Vision and Pattern Recognition (cs.CV); Human-Computer Interaction (cs.HC)
备注:
Abstract:The distance between persons reveals significant information about their perception of each other. However, such information is not easily extractable and interpretable from video input. We developed an open-sourced library, Facial Interpersonal Distance Analysis and Coding (FIDAC) that transforms facial detection results into actionable data about location and interpersonal distance. This tool merges data from multiple open-source facial detection models, strategically compensating for gaps in any individual model. In addition, we include methods for more accurate tracking, such as a pipeline for human coding of the selection of faces and a benchmarking tool to reduce depth distortion. For next steps, we plan on building upon FIDAC by evaluating its effectiveness at measuring interpersonal distance at various depths and orientations while further integrating features of proxemic analysis such as synchrony into its software.
[HC-14] Beyond the Post Hoc User Study: Modeling Visual Decision-Making with Active Inference
链接: https://arxiv.org/abs/2607.25131
作者: Harrison J. Goldwyn,Graham Johnson,Christopher Ibarra,Lace Padilla,Kenny Gruchalla
类目: Human-Computer Interaction (cs.HC); Neurons and Cognition (q-bio.NC)
备注:
Abstract:Empirical user studies are essential for evaluating visual encodings and can reveal perceptual and cognitive mechanisms, but they do not by themselves provide causal, predictive accounts of interpretation errors. Evaluations are therefore often post hoc: they measure performance after a design has been specified rather than predicting how attention, uncertainty, memory, and bias may produce accurate or erroneous judgments. To address this mechanistic gap, we translate a cognitive theory of visualization interpretation into executable simulation using Active Inference, a probabilistic framework for perception, learning, and action. We model chart reading as dynamic visual search in which agents update beliefs and choose actions that balance uncertainty reduction against cognitive effort. As a proof of concept, we implement Fast, heuristic (Type 1) and Slow, analytic (Type 2) agents for a bar-chart average-estimation task. The Fast agent is vulnerable to tick-salience bias, whereas the Slow agent is more vulnerable to working-memory decay. Both produce inspectable cognitive traces, including evolving belief uncertainty and fixation sequences. By expressing these hypothesized failure mechanisms as interpretable parameters, the architecture provides a framework for formalizing and testing mechanistic hypotheses about visualization interpretation. Empirical studies can then parameterize, refine, or falsify these simulations, supporting earlier and more predictive in silico evaluation of visualization efficacy.
[HC-15] Learning from 53.6K Real-World Developer Edits of AI-Generated Code
链接: https://arxiv.org/abs/2607.25130
作者: Jenny T. Liang,Mihika Bairathi,Wayne Chi,Ameet Talwalkar,Nishant Subramani,Valerie Chen
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC); Machine Learning (cs.LG)
备注:
Abstract:Imperfections in AI-generated code require that software developers modify the generated code manually, or by re-prompting an AI programming assistant. Manual code edits provide more realistic and granular information on editing behavior than Git commits, which only contain final successful code snippets. Yet, due to a lack of high-quality, realistic code editing data, LLMs are mostly trained on publicly available Git data (e.g., commits). To address this gap, we introduce DECODE (Developer Edits of Code Dataset), a dataset of 53.6K real-world in-IDE code edits of AI-generated code in Python, TypeScript, and JavaScript, sourced from 1K+ developers. First, we demonstrate the utility of DECODE for data analysis, obtaining insights on when, why, and how AI-generated code is edited. We find that most edits occur within the first 15 minutes after accepting an AI completion, resulting in the removal of AI completions in 31% of edit trajectories. Second, we use DECODE to benchmark the ability of LLMs to predict code edits. We find that finetuning on DECODE enables open-source 3B models to perform code edit prediction tasks significantly better than frontier LLMs. We then discuss implications of this work, emphasizing the necessity of developer-centric machine learning approaches for future AI programming assistants.
[HC-16] UrbanTrace: LLM -Assisted Discovery and Semantics-Aware Integration of Spatial Data IEEE-VIS2026
链接: https://arxiv.org/abs/2607.25124
作者: Sonia Castelo,Eden Wu,Joao Rulff,Harish Doraiswamy,Juliana Freire,Claudio Silva
类目: Human-Computer Interaction (cs.HC)
备注: 11 pages, 7 figures. Accepted to IEEE VIS 2026 (Full Papers Track). Author’s version accepted for publication in IEEE Transactions on Visualization and Computer Graphics
Abstract:Urban decision-making requires integrating heterogeneous spatial data. While current GIS tools handle geometric computation efficiently, they lack the semantic reasoning to guide complex workflows. Analysts manually manage data discovery, spatial boundaries, and measurement semantics, risking aggregation errors. We present UrbanTrace, a visual analytics system that transforms manual spatial data-wrangling into a transparent, node-based collaborative workflow with context-aware AI agents. Using an offline profiler to extract semantic and geometric metadata, UrbanTrace grounds LLMs in real-world data distributions. This enables specialized agents to retrieve datasets based on high-level goals and automatically enforce valid spatial aggregations. To make harmonization explicit, three interactive views: an Integration Provenance Graph, Multivariate Priority Map, and Spatial Delta Map, allow users to explore how conclusions shift across spatial configurations. We evaluate UrbanTrace on 28 urban scenarios spanning 112 datasets. Quantitative ablations show our profiling significantly outperforms baseline LLMs in data discovery, achieving 100% semantic and 87% geometric validity in spatial mapping. Through real-world case studies and expert interviews, we demonstrate that UrbanTrace turns spatial aggregation sensitivity from a methodological burden into an exploratory visual asset.
[HC-17] Memdora: Designing Cognitively-Grounded Flashcard Interactions for AI-Powered Spaced Repetition
链接: https://arxiv.org/abs/2607.25096
作者: Ruiyang Zhang
类目: Human-Computer Interaction (cs.HC)
备注: 10 pages
Abstract:Spaced repetition systems (SRS) have demonstrated robust effects on long-term retention, yet existing tools reduce the flashcard interaction to a single binary gesture: flip and self-rate. This impoverished interaction model fails to leverage decades of cognitive science evidence on retrieval practice, and requires learners to context-switch out of their reading flow to create cards manually. We present Memdora, a cross-platform AI spaced repetition system that addresses these limitations through four contributions: (1) a taxonomy of 17 cognitively-grounded interaction types across three learning categories – Language (6 types), By Heart (1 type with 3 retrieval modes), and Exam (10 types) – each mapped to peer-reviewed cognitive science evidence displayed on every card; (2) a unified AI generation pipeline that collapses card creation to a single gesture at the point of reading across web, mobile, and three browser extensions (Chrome, Edge, Firefox); (3) a collaborative classroom layer enabling teachers to publish decks, assign them to students, and track learning outcomes at the individual card level; and (4) an effort-based behavioral reward system that incentivizes actual cognitive engagement rather than mere app presence. Memdora integrates FSRS-6, the current state-of-the-art spaced repetition algorithm, and is deployed publicly on iOS, Android, Web, and three browser extensions. We describe the design rationale for each interaction type, discuss how the system advances beyond prior AI flashcard systems, and outline implications for educational technology design.
[HC-18] ObjectEMS: Electrical Muscle Stimulation Without Electrodes on the User
链接: https://arxiv.org/abs/2607.25084
作者: Yun Ho,Zhechen Zhao,Romain Nith,Andre de la Cruz,Pedro Lopes
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:Interactive electrical muscle stimulation (EMS) has revealed its promise as a portable interface for force-feedback. However, while much ink has been spilled about the advantages of EMS, few have investigated one of its central limitations: the need to attach electrodes to users. This has dramatically limited the application of EMS, especially in brief interactions or physical assistance with tools. To explore an alternative, we propose embedding electrodes (and stimulator) inside objects that the user interacts with. This is made possible because we identified multiple novel electrode placements that can elicit four distinct finger movements from the palm (no forearm stimulation). To illustrate this new way of implementing electrical muscle stimulation, we developed a set of self-contained interactive objects that use capacitive sensing to determine if a user’s hand is in poses conducive to stimulation and then actuate the fingers from contact points between the hand and the grasped object. In our user study, we found that participants spent less time calibrating ObjectEMS than traditional EMS, and they felt less tethered while engaging with an EMS application via this novel approach. Our approach frees up the user’s body from wearing electrodes and EMS stimulators, providing a new pathway for researching, implementing, or deploying actuated tangible interfaces. We demonstrate its potential via exemplary applications, such as a game controller with finger-level force feedback, a door handle with force feedback to prevent push-pull confusion, and more.
[HC-19] Psychological Influences of Conversational AI: Research and Design Directions for Reducing Harm and Promoting Well-Being
链接: https://arxiv.org/abs/2607.25057
作者: Jina Suh,Mihaela Vorvoreanu,Forough Poursabzi-Sangdeh,Emily Tseng,Eugenia Kim,Luke Nicholls,James W. Pennebaker,Eric Horvitz
类目: Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Human-Computer Interaction (cs.HC)
备注:
Abstract:As conversational AI systems become increasingly integrated into daily life, their potential effects on user well-being require ongoing attention. While consumer-facing generalist models can provide benefits, including improved access to information, learning, productivity, self-reflection, and companionship, they also introduce risks, such as emotional entanglement, unhealthy dependence, and the amplification of psychological vulnerabilities. Drawing on prior research and empirical observations of AI chatbot behavior, we propose a set of aspirational directions for guiding the behavior of general-purpose AI systems in ways that may reduce potential psychological harms and support user well-being. We acknowledge the difficulty of systematically assessing the long-term impacts of AI chatbot use and frame these directions as hypotheses for studying how AI behavior may influence users across general interactions, role-playing scenarios, and contexts that could be characterized as providing psychological support. While some proposed directions are supported by existing research and expert insights, others identify open questions and areas requiring deeper study. We hope that this formulation and these hypotheses encourage further discussion, empirical investigation, and exploration of interactive design approaches aimed at better accommodating users’ psychological needs and promoting their well-being.
[HC-20] Extended Reality as a Mediation Layer for Situated Human Control in Human-Robot Teaming
链接: https://arxiv.org/abs/2607.25047
作者: Jens Grubert,John Dudley,Eyal Ofek,Per Ola Kristensson
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI); Robotics (cs.RO)
备注: Accepted to 2026 IEEE International Symposium on Mixed and Augmented Reality Adjunct (ISMAR-Adjunct)
Abstract:Extended Reality (XR) is increasingly used in human-robot interaction to communicate robot intent, planned motion, reachability, and state. We argue that XR should also be understood as a mediation layer for situated human control in human-robot teaming. Situated human control denotes the human collaborator’s ability to understand, shape, authorize, and interrupt robot action within the concrete physical, social, and temporal context in which that action unfolds. We ground this perspective in scenarios from robot-assisted bedside nursing, multi-arm supervisory control, and collaborative assembly under divided attention. Across these scenarios, robot autonomy must remain inspectable and adjustable as people move, goals change, sensing is incomplete, control roles shift, and plans become invalid. We identify four mediation functions connecting human intent and robot autonomy, robot plans and human judgment, levels of shared control, and team roles, handover, and recovery. Building on these functions, we derive six design dimensions: joint action possibilities, socio-physical constraints, uncertainty and plan validity, multimodal control and correction, roles, handover, and accountability, and anticipatory recovery. The paper outlines a research agenda for XR systems that make robot autonomy more actionable and accountable in dynamic shared environments.
[HC-21] he AI Wave and the Reinvention of Game Discovery: Oversupply Structural Correction and Agent ic Player-Game Matching
链接: https://arxiv.org/abs/2607.25010
作者: Brian Dean Madanamootoo
类目: Human-Computer Interaction (cs.HC); Computers and Society (cs.CY)
备注:
Abstract:AI-assisted production has sharply reduced the cost and team size required to ship a video game, producing a supply shock on open marketplaces. Recent estimates put Steam release volume at roughly sixty new titles per day, with median per-title revenue for a large share of releases falling below the platform’s own submission fee [1]. This paper asks whether the resulting oversupply constitutes an emerging market crash or a structural correction, and what discovery infrastructure the market will require as a consequence. We first quantify the 2010-2026 supply shock using a 93,073-title Steam metadata snapshot, a 200,000-interaction Steam user-behavior dataset, and this http URL catalog data, computing attention-concentration metrics directly (Gini coefficient of 0.96 over playtime, with the top 1 percent of titles absorbing 73.5 percent of total play hours), and we introduce generative asset-model release velocity on Hugging Face as a candidate leading indicator of production-cost decline. We then conduct a comparative-historical analysis against the 1983 North American video game crash, the closest documented case of supply-driven collapse in the medium’s history, identifying which structural divergences (digital distribution, diversified incumbent revenue, and consolidation capital) redirect the present contraction toward concentration rather than collapse, drawing on incumbent evidence including Ubisoft’s 2025-26 restructuring and its transfer of equity to Tencent-backed Vantage Studios. Third, we analyze Netflix Games, Xbox Game Pass, and the curated browser platform Poki as natural experiments in access-based distribution. Subjects: Human-Computer Interaction (cs.HC); Computers and Society (cs.CY) Cite as: arXiv:2607.25010 [cs.HC] (or arXiv:2607.25010v1 [cs.HC] for this version) https://doi.org/10.48550/arXiv.2607.25010 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[HC-22] When Shortest Isnt Safest: A Design Science Approach to Senior-Friendly Pedestrian Routing
链接: https://arxiv.org/abs/2607.24795
作者: Erdi Ünal,Daniel Eisenhardt,Christian Meske,Seyed Nima Afzali,Aysegül Dogangün
类目: Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注:
Abstract:Older adults’ independent mobility enables out-of-home participation, well-being and health, yet pedestrian navigation systems still optimize primarily for distance or time, often overlooking barriers, safety thresholds, and supportive infrastructure that shape late-life walking decisions. We present a senior-friendly pedestrian routing artefact developed through echeloned Design Science Re-search, translating lived mobility constraints into prescriptive design knowledge. Based on 11 semi-structured interviews, we derive initial Design Requirements (DRs) and Design Principles (DPs) for barrier-aware, amenity-sensitive routing and execution-relevant explanations. We instantiate these in an OpenStreetMap pedestrian network enriched with amenities (benches, toilets, and shelters) and height data, and implemented an A*-based routing engine with configurable costs and explanation payloads. In a field-based walking study, 14 older adults com-pared artefact-generated routes with baselines and provided ratings and qualitative feedback; the senior-friendly route was preferred overall. Thematic analysis further showed that infrastructure maintenance, seasonal conditions, traffic exposure, and social context shape route acceptance. We synthesize these insights into refined DRs and DPs emphasizing context-aware hazard modeling, multi-route transparency, landmark-grounded explanations, social-context sensitivity, and stage-appropriate information. Our contributions provide actionable guidance for practitioners developing senior-friendly pedestrian navigation systems.
[HC-23] Empathy and the Human-Moment Gaps of AI Chatbots: Insights from Empathy Displacement Theory
链接: https://arxiv.org/abs/2607.24775
作者: Victor Frimpong
类目: Human-Computer Interaction (cs.HC); Computers and Society (cs.CY)
备注: BRAIN. Broad Research in Artificial Intelligence and Neuroscience, 2025
Abstract:Artificial intelligence (AI) chatbots are increasingly deployed in domains where empathy is essential, including healthcare, education, and customer service. However, their capacity to sustain authentic human moments remains structurally limited. This paper introduces two interlinked conceptual models to explain and address this limitation. First, the Human-Moment Gap Framework (HMGF) identifies three structural empathy deficits in AI-mediated interaction: affective surfaceism (emotional imitation without depth), memory fragmentation (lack of relational continuity), and moral framing mismatch (efficiency prioritised over dignity). Second, the paper develops the Empathy Displacement Theory (EDT), which explains how AI-simulated empathy can progressively substitute, distort, and displace genuine human empathy across individual, relational, and organisational contexts. HMGF serves as the causal foundation of EDT by demonstrating how technical and moral deficiencies in chatbot design may evolve into broader social and institutional consequences. The study is conceptual and exploratory, aiming to develop an integrative theoretical framework rather than provide empirical validation. Together, HMGF and EDT provide a unified framework for understanding AI-mediated empathy, generating testable propositions and implications for the responsible development and governance of empathetic AI systems. The paper concludes that the central challenge of empathetic AI is not whether machines can genuinely care, but how simulated care reshapes human emotional expectations, interpersonal behaviour, and institutional norms.
[HC-24] ProcAgent : An Agent ic Framework for Procedural Task Guidance on Edge with Human-in-the-Loop
链接: https://arxiv.org/abs/2607.24770
作者: Azizul Zahid,Subrata Biswas,Bashima Islam,Sai Swaminathan
类目: Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注:
Abstract:Procedural tasks such as furniture assembly and home repair impose substantial cognitive demands because users must interpret instructions, track task progress, reason about spatial state, and recover from errors while performing physical actions. Prior multimodal assistants have shown promise for procedural guidance, but most rely on cloud inference and fixed always-on perception, making them poorly suited to privacy-sensitive, latency-critical domestic settings. We present ProcAgent, a fully on-device, agentic, vision-based procedural assistant for real-time adaptive guidances on a single NVIDIA Jetson AGX Orin. ProcAgent uses a propose-and-verify architecture that combines low-latency continuous perception, a symbolic task graph, on-demand vision-language verification, and an LLM-based interaction agent. The system continuously proposes user progress, invokes expensive visual reasoning only when ambiguity or likely deviation arises, and supports both reactive question answering and proactive intervention with human-in-the- loop confirmation. We evaluate ProcAgent along four dimensions: perception accuracy, reasoning, task-level performance, and user experience. Despite running entirely on-device, the system maintains responsive interaction, resolving text-only queries in approximately 2 seconds and visually grounded queries in approximately 8 seconds. In a user study with 10 participants completing assembly tasks, ProcAgent receives positive ratings for comprehensibility, actionability, and privacy comfort. These results show that adaptive procedural assistance can be achieved entirely on edge hardware without sacrificing usability.
[HC-25] Measuring and Improving Behavioral Consistency in Large Language Models through Fact-Heuristic-Emotion State Enforcement
链接: https://arxiv.org/abs/2607.24765
作者: Gi-Hun Lee,Joong Yull Park
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注: 40 pages, 6 figures; 54-page supplementary material included as an ancillary file. Code, prompts, and data: this https URL
Abstract:Large language models (LLMs) can give different answers to the same decision problem across runs, and reverse a decision when their own prior answer returns as context. We ask whether this instability can be measured and partially reduced without changing model weights. We test the Cognitive Kernel Model (CKM), a prompt-level state-enforcement layer. Before deciding, the model must separate its input into three epistemic roles: Fact (given or verifiable), Heuristic (inferred or assumed), and Emotion (evaluative or priority signal). CKM adds no capability; it forces the model to track what kind of information it uses before acting. Formally it maintains a structured state S_t = F_t, H_t, E_t updated by a transition function. We evaluate CKM on Korean-language decision scenarios (ambiguity, ethical conflict, resource allocation, error handling) across 26 LLMs from four vendors and 37,403 observations, via four core experiments, a 4-arm ablation, a 5-arm sham-restriction ablation, and a temperature probe. Findings: (1) CKM reduces repeated-output variability (random-effects Hedges’ g=1.09, 95% CI [0.83, 1.35], 31 model pairs); (2) state persistence cuts the decision-flip rate by 82% in newer models (g=1.52); (3) the effect is not JSON formatting alone (value-only recomputation, g=2.24); (4) intrinsic randomness under fixed anchor states is negligible; (5) the advantage grows under sampling stochasticity (g=2.87 at temperature 0.7); (6) a sham ablation attributes about 45% of the gain to structural scaffolding and 55% to Fact/Heuristic/Emotion content, and CKM is the only arm that both raises consistency and reduces flipping. CKM does not improve reasoning correctness. The narrower result: behavioral consistency is measurable, varies across models, and is partially improvable by forcing models to separate facts, assumptions, and evaluative signals before deciding. Comments: 40 pages, 6 figures; 54-page supplementary material included as an ancillary file. Code, prompts, and data: this https URL Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC) ACMclasses: I.2.7; I.2.0 Cite as: arXiv:2607.24765 [cs.CL] (or arXiv:2607.24765v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2607.24765 Focus to learn more arXiv-issued DOI via DataCite Submission history From: Gihun Lee Dr [view email] [v1] Fri, 5 Jun 2026 12:36:36 UTC (483 KB) function toggleList(whichLayer,toggleThis) var elem, vis; if( document.getElementById ) // standard elem = document.getElementById( whichLayer ); else if( document.all ) // old msie versions elem = document.all[whichLayer]; else if( document.layers ) // nn4 elem = document.layers[whichLayer]; vis = elem.style; // if the style.display value is blank we try to figure it out here if(vis.display==‘’!=undefined!=undefined) vis.display = (elem.offsetWidth!=0!=0)?‘inline’:‘none’; vis.display = (vis.display==‘’||vis.display==‘inline’)?‘none’:‘inline’; // toggle link inner text status = vis.display; if(vis.display==‘inline’) document.getElementById(‘toggle’).innerHTML = “(collapse list)”; document.getElementById(‘toggle’).title = “Collapse list”; else document.getElementById(‘toggle’).innerHTML = “(”+toggleThis+“)”; document.getElementById(‘toggle’).title = “Show complete list”; Full-text links: Access Paper: View a PDF of the paper titled Measuring and Improving Behavioral Consistency in Large Language Models through Fact-Heuristic-Emotion State Enforcement, by Gi-Hun Lee and 1 other authorsView PDFHTML (experimental)TeX Source view license Ancillary-file links: Ancillary files (details): supplementary.pdf Current browse context: cs.CL prev | next new | recent | 2026-07 Change to browse by: cs cs.AI cs.HC References Citations NASA ADSGoogle Scholar Semantic Scholar export BibTeX citation Loading… BibTeX formatted citation loading… Data provided by: Bookmark checked="checked"class=“labs-tab-input”> Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, Media Code, Data and Media Associated with this Article alphaXiv Toggle alphaXiv (What is alphaXiv?) Links to Code Toggle CatalyzeX Code Finder for Papers (What is CatalyzeX?) DagsHub Toggle DagsHub (What is DagsHub?) GotitPub Toggle Gotit.pub (What is GotitPub?) Huggingface Toggle Hugging Face (What is Huggingface?) ScienceCast Toggle ScienceCast (What is ScienceCast?) Demos Demos Replicate Toggle Replicate (What is Replicate?) Spaces Toggle Hugging Face Spaces (What is Spaces?) Spaces Toggle TXYZ.AI (What is TXYZ.AI?) Related Papers Recommenders and Search Tools Link to Influence Flower Influence Flower (What are Influence Flowers?) Core recommender toggle CORE Recommender (What is CORE?) Author Venue Institution Topic About arXivLabs arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv’s community? Learn more about arXivLabs. Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?) mathjaxToggle(); We gratefully acknowledge support from our major funders, member institutions, , and all contributors. About Help Contact Subscribe Copyright Privacy Accessibility Operational Status (opens in new tab) Major funding support from
[HC-26] Verification Without Distrust: Reframing User-Side Oversight as Routine Epistemic Governance in Everyday Human-Chatbot Interaction
链接: https://arxiv.org/abs/2607.24761
作者: Aung Pyae
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注:
Abstract:Research on human-AI interaction has long framed verification of system outputs as a trust-contingent behavior that better-calibrated trust should reduce. We test this assumption in everyday human-chatbot interaction through a mixed-methods survey of 153 frequent chatbot users. Contrary to the canonical prediction, we find no detectable association between trust and verification, with the result robust across sensitivity analyses. Three further user-side practices - refinement, correction, and approval before automated actions - are widely endorsed and positively associated with satisfaction. The data reveal a substantive distinction between evaluative oversight (trust-decoupled, weakly tied to satisfaction) and interventionist oversight (weakly trust-correlated, strongly tied to satisfaction). A medium-to-large satisfaction-control gap shows that effective task outcomes do not produce a felt sense of agency. Qualitative findings identify instrumental mental models, failure-mode-specific doubt, and demand for epistemic infrastructure. We reframe user-side oversight as routine epistemic governance compatible with trust, and derive four design directions for scaffolded oversight in conversational AI.
[HC-27] From Idea to Classroom in Days: Using “Vibe Coding” to Create a Programming Process Visualizer from IDE Activity Logs
链接: https://arxiv.org/abs/2607.24757
作者: Heidi Taveter,Marina Lepp
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注: 9 pages, 6 figures. Accepted for presentation at the 2026 IEEE Frontiers in Education Conference (FIE 2026)
Abstract:This paper reports on the rapid development and classroom deployment of a Thonny log visualizer built using AI-assisted ``vibe coding’’ to make students’ programming processes easily visible to teachers. We developed a web application that analyzes log files generated by Thonny (an IDE for Python) and produces interpretable views of students’ programming processes. Teachers can upload a log, a ZIP archive, or a folder containing logs for a group or the course; the system parses all logs, generates results per student, and provides student-by-student navigation for reviewing cases. Each student’s view includes an interactive activity timeline, a compact session summary, a code-size graph, a programming-process replay, and more. These views support teacher decision-making by enabling the identification of learning-support situations and flagging sessions for academic-integrity clarification. The tool was initially evaluated using logs from previous courses; a February 2026 pilot in an introductory programming course with 160 participants provided teachers’ feedback and informed iterative usability improvements.
[HC-28] What Gets Lost When Memory Becomes Media? Evaluating AI-Generated Oral History Visualization ICML2026
链接: https://arxiv.org/abs/2607.24756
作者: Kwangsuk Park,Jaehyun Koo,Jiyeon Lee,Anjung Tan,Hyoungchul Park
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注: Accepted at ICML 2026 Workshop on Culture x AI: Evaluating AI as a Cultural Technology
Abstract:What gets lost when memory becomes media? Diaspora oral-history interviews require a double transformation; first-person recollection to third-person scene, present interview room to past time and place. When generative AI performs this transformation, no agreed criteria for success exist. We derive success conditions from oral-history theory, design 15 metrics around three failure modes, and compare a Multi-Agent Scene-decomposition pipeline (MAS) with a Single Summarization Pipeline (SSP) across 82 interviews from diaspora communities, spanning from oral interviews to 6-image sequences. Scene-planning and narrative preservation conflict in the majority of cases, and the narrative-structure strength of the source testimony is the primary predictor of this conflict. We propose a failure-mode-based evaluation framework, an empirical analysis of conflict conditions, and a routing protocol for system selection based on narrative-structure strength.
[HC-29] Patterns of Learner-AI Interaction and Academic Performance in an Object-Oriented Programming Course
链接: https://arxiv.org/abs/2607.24755
作者: Marina Lepp
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注: This work has been accepted for publication at the IEEE Frontiers in Education (FIE) 2026 Conference. Copyright may be transferred without notice, after which this version may no longer be accessible
Abstract:This full research paper examines how different forms of learner-AI interaction relate to learning outcomes in object-oriented programming (OOP) courses. Generative artificial intelligence (GenAI) tools are increasingly used by students in programming education, yet evidence on their educational impact remains mixed. In particular, little is known about how students integrate GenAI tools when learning OOP, and how different patterns of use relate to students’ learning experiences and outcomes. This study investigates patterns of students’ self-directed GenAI use and their relationship with academic performance, perceived difficulty, understanding, and trust. Survey data were collected from 210 undergraduate students enrolled in a first-year OOP course in which the use of GenAI tools was permitted for coursework but prohibited in assessments. Results show that students used GenAI significantly more often for explanation seeking and debugging than for code generation. Cluster analysis identified five distinct learner-AI interaction profiles, including a “smart” high-usage pattern characterized by low reliance on code generation and high use for conceptual support and debugging. While usage patterns were associated with differences in perceived assignment difficulty, self-assessed understanding, trust in AI-generated code, and norm-related attitudes, no significant differences in assessment performance were found across clusters. These findings suggest that self-directed GenAI use alone does not lead to measurable learning gains, underscoring the need for pedagogically guided and process-aware AI support. The study contributes empirical evidence on learner-AI interaction patterns and highlights the importance of pedagogically guided AI use in programming education.
[HC-30] CARE-MH: Towards Unified Reproducible and Comparable Evaluation of Mental Health LLM s
链接: https://arxiv.org/abs/2607.24754
作者: Asher Sprigler,Yixue Zhao,Yi Ding
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注: 39 pages, 22 figures, 22 tables
Abstract:Large language models (LLMs) are increasingly used to provide mental health support, requiring reliable evaluation of safety, empathy, and therapeutic appropriateness. However, existing mental health benchmarks are difficult to reproduce and compare due to inconsistent evaluation designs and metric definitions. We present CARE-MH, a unified framework for comparable and reproducible evaluation of mental health LLMs. Using CARE-MH, we reproduce and analyze state-of-the-art benchmarks, revealing that reproducibility depends strongly on model stability and that cross-benchmark disagreement primarily arises from differences in metric definitions. Our findings highlight the need for standardized evaluation configurations and shared metric definitions for future mental health LLM benchmarks.
[HC-31] Language as a Material Interface for Creative LLM Interaction
链接: https://arxiv.org/abs/2607.24753
作者: Jon McCormack,Tace McNamara,Chen Wang,Maria Teresa Llano
类目: Human-Computer Interaction (cs.HC); Computation and Language (cs.CL)
备注: Paper accepted at Creativity and Cognition '26, London 13-16 July 2026
Abstract:Although directive prompting is the predominant way to interact with Large Language Models (LLMs), many creative practices rely on language that is open-ended, associative, phonaesthetic, symbolic, and that unfolds across multiple temporalities. In this work, we explore how creative practitioners might work with AI systems when language is treated not merely as instruction but as material. We conducted an ecological two-week study with four creative practitioners using a design probe: the Memetic Mixer, a tangible interactive device that constrains interaction with an LLM. Analysis of post-study interviews and device logs identified distinct modes of material language use and temporalities that shaped each participant’s engagement with AI and their creative practice. We reflect on these findings and contribute design considerations that support open-ended interaction with AI in creative practice.
[HC-32] Development and applications of Generative AI in architectural design studios
链接: https://arxiv.org/abs/2607.24752
作者: Leman Figen Gül,Burak Delikanlı,Oğulcan Üneşi,Ertuğrul Ömer Gül
类目: Human-Computer Interaction (cs.HC)
备注: 24 pages, 10 figures, journal article
Abstract:Recent advancements in deep generative models, with their capacity to yield outputs that are both visually appealing and semantically coherent, have served to further intricate the nexus between Artificial Intelligence (AI) and creativity. This integration of these models into creative disciplines, most notably within the domain of design, has been rapidly escalating. Nevertheless, a prevailing gap remains in our understanding of the impact of Generative Artificial Intelligence (GenAI) on design creativity. To address this gap, a longitudinal study was initiated, encompassing the development of a user-friendly GenAI interface. This interface was designed to facilitate the integration of GenAI models into design education, following a cyclical process of experimentation and refinement. The study involved the utilization of GenAI models in design studios, the collection of feedback, the development and evaluation of the GAI-A platform, and subsequent implementation in design education. The findings of the study provide a discussion based on the views and observations of students on the extent to which they use GenAI models, their perspectives and expectations of these models, and the extent to which this process improves their creativity in the context of the final product.
[HC-33] Stop Writing for Me: Generative Refusal in AI Tools for Thought
链接: https://arxiv.org/abs/2607.24751
作者: Sora Kang
类目: Human-Computer Interaction (cs.HC)
备注: Workshop on Tools for Thought at CHI 2026
Abstract:In creative domains where the labor of articulation is central to the craft, how should we design Tools for Thought that enhance rather than bypass human cognition? Current GenAI paradigms often prioritize “cognitive offloading”-writing on behalf of users-risking the erosion of the constructive thought process essential to artistic training. In this position paper, we explore AI as a Maieutic Partner through “Generative Refusal”-strategically withholding text generation to demand user articulation. We discuss Actor’s Note, a journaling tool that generates context-aware questions instead of draft text. Our field study suggests that this constraint significantly reduced cognitive burden while fostering a residual effect of internalized questioning habits. We use these findings to discuss broader design implications for protecting human cognition against the tendency of generative efficiency.
[HC-34] meCapsule: Generative Hallucination as a Method for Historical Sensemaking
链接: https://arxiv.org/abs/2607.24750
作者: Hayk Grigorian,Hamed Yaghoobian
类目: Computation and Language (cs.CL); Human-Computer Interaction (cs.HC)
备注: 10 pages, 4 figures. Accepted to Creativity and Cognition (CC '26)
Abstract:Large Language Models (LLMs) are temporally overexposed: trained on vast contemporary corpora, they encode present-day concepts that make them unreliable narrators of the past. We present TimeCapsule, a 1.2B-parameter LLaMA-style causal model trained exclusively on Victorian texts (1800-1875) as an epistemologically isolated generative archive. Quantitative evaluation shows a 45.4% perplexity reduction over a GPT-2 baseline on held-out Victorian prose, while larger contemporary causal models achieve lower raw perplexity through broader pretraining but lack temporal isolation. TimeCapsule exhibits computational sensemaking, generating historically plausible analogical explanations for unfamiliar modern concepts (e.g., describing a computer as a “hypertrophied lung”). A qualitative hermeneutic probe with two humanities scholars revealed a crisis of authenticity, as both misclassified approximately 40% of genuine Victorian excerpts as machine-produced. We argue that structural ignorance of the future transforms hallucinations into interpretive probes of nineteenth-century ontologies.
[HC-35] Game AI Not Fun? A Scoping Review and Meta-Analysis on the Differences in Enjoyment between Human and Computer Opponents
链接: https://arxiv.org/abs/2607.24749
作者: Ray Ito
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注: This work has been accepted for publication in IEEE Conference on Games 2026, IEEE. Copyright may be transferred without notice, after which this version may no longer be accessible
Abstract:Although advancements in game character AI aim to enhance player engagement, evidence suggests that perceiving an opponent as artificial can diminish the psychological experience. This paper presents a scoping review and meta-analysis of empirical studies focusing on player enjoyment when competing against human versus computer opponents. First, the scoping review was conducted to map the landscape of 20 included studies, detailing their study designs, outcome measures, and research foci. Second, a three-level meta-analysis synthesizing baseline comparisons from nine studies quantitatively assesses the differences in enjoyment. The results demonstrate a statistically significant, medium-to-large pooled effect size, indicating a psychological penalty in computer-opponent conditions. This paper provides a comprehensive overview of the extant knowledge on this topic, and underscores the necessity for further research in order to fully understand and resolve the penalty of the computer opponent context.
[HC-36] Multimodal User Authentication Method via Fusion of Keystroke Dynamics and Glove-Based Hand Kinematics
链接: https://arxiv.org/abs/2607.24747
作者: Issei Hyakuda,Lei Jing
类目: Cryptography and Security (cs.CR); Human-Computer Interaction (cs.HC)
备注: 9 pages, 10 figures
Abstract:Although keystroke dynamics are cost-effective behavioral biometrics, their practical deployment is hindered by susceptibility to environmental variations. To address this, we propose a robust multimodal authentication framework that augments traditional keystroke dynamics using 19-dimensional hand kinematics. Features are captured using a bespoke data glove equipped with 10 piezoresistive pressure sensors and a 9-axis IMU. A hybrid CNN-LSTM architecture effectively fuses these heterogeneous time-series streams. To ensure real-world applicability, we implemented a rigorous “unseen” evaluation protocol: the model was trained on a desktop keyboard using data from 1 target user and 8 “known” impostors, but evaluated on a laptop keyboard (cross-domain) against the target and an “unknown” impostor excluded from training. Averaged over five trials, the multimodal method achieved a mean Equal Error Rate (EER) of 2.12% for individual 600-ms authentication events (Window 1). Crucially, aggregating scores over 10 events (6 seconds) using a temporal smoothing window yielded perfect authentication (0.00% EER). While our ablation study showed IMU kinematics alone achieved equivalent performance in a static laboratory, error analysis confirmed that pressure and IMU sensors, despite strong physical correlation, possess distinct sensitivities to error factors. Fusing these modalities establishes a vital fail-safe against real-world vulnerabilities like spatial spoofing and environmental noise, highlighting that combined physical traits provide much stronger biometric discrimination than timing features alone.
[HC-37] Human-Humanoid Collaboration and Ergonomic Risk: An Anthropometric Perspective
链接: https://arxiv.org/abs/2607.24746
作者: He Wen
类目: Human-Computer Interaction (cs.HC)
备注: Accepted and Presented in IISE Annual Conference Expo 2026
Abstract:Humanoid robots are increasingly deployed in industrial environments where close physical interaction with human workers is expected. Although these systems are often designed at a human scale, their embodiment is shaped by mechanical, control, and task-oriented constraints rather than biological anatomy. This study examines ergonomic risk in human-humanoid collaboration from an anthropometric perspective using ISO 7250 as a reference framework. Six contemporary humanoid robots are benchmarked based on externally observable geometry to evaluate landmark identifiability, measurement feasibility, and cross-platform patterns of anthropometric deviation. Results show that several ISO-defined landmarks and measurements tied to biological anatomy are consistently inapplicable to humanoid robots, while many geometric and joint-level dimensions remain measurable. Four recurring patterns of anthropometric deviation are observed across platforms: additive, subtractive, exaggerative, and speculative. These findings indicate that ergonomic risk arises not from scale alone, but from how humanoid bodies diverge from human anthropometric assumptions. ISO 7250 remains a useful reference framework, but its direct transfer to humanoid robots is limited, underscoring the need for anthropometry-aware evaluation approaches in collaborative system design.
[HC-38] More Electrodes Faster Minds? Rethinking Bandwidth in Brain-Computer Interfaces
链接: https://arxiv.org/abs/2607.24820
作者: Boxuan Jiang
类目: Neurons and Cognition (q-bio.NC); Human-Computer Interaction (cs.HC)
备注:
Abstract:High-bandwidth brain–computer interfaces (BCIs) can bypass damaged pathways, reduce motor costs, and improve communication and control. They also inspire visions of accelerated thought output, mind reading, and instant skill acquisition. This Perspective asks how gains in meaningful human I/O scale with interface capacity. We distinguish bandwidth, decodable neural states, neural states, and information a person can use, confirm, and express. Slowly updated task states can unfold into complex behavior through the body, neural control, sensory feedback, the environment, and shared context. Decodable neural activity can support prediction and control; subject-level communication depends on selection, confirmation, and authorization. On the input side, stimulation may guide plasticity and accelerate learning, while embodied skills arise through coordination of a brain, body, and environment. The scaling relationship is likely nonlinear: higher-capacity interfaces can yield real gains, while extreme increases in meaningful human I/O encounter constraints rooted in embodiment, learning, and subject expression.
[HC-39] EEG Emotion Recognition From AI-Generated Biodigital Architecture Images
链接: https://arxiv.org/abs/2607.24808
作者: Hongye Yang,Eva Guttmann-Flury
类目: Neurons and Cognition (q-bio.NC); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注: 12 pages, 3 figures; published in the proceedings of SIGraDi 2024
Abstract:Emotional responses to biodigital architecture were examined using electroencephalographic (EEG) data from AI-generated images. A pre-experiment involving 336 participants identified 60 images, selected from an initial pool of 600, that elicited strong emotional responses categorized as awe, disgust, or content. These images were used for EEG recordings of 52 volunteers, with channel selection and sample size estimation based on the analysis of an existing dataset. Gamma and delta bands yielded the highest classification accuracy, with the gamma band achieving an accuracy of 77.07 percent +/- 13.8 percent for the awe emotion. Key factors such as greenery and non-uniform granularity were linked to positive emotions, while dampness triggered negative reactions. These results emphasize the significance of incorporating natural elements and varied textures in biodigital architecture to enhance aesthetic appeal and acceptance. The study demonstrates EEG’s capability to objectively assess architectural preferences, providing valuable insights for architects to design engaging and sustainable environments.
计算机视觉
[CV-0] VetClaw: An Edge-Cloud Multimodal Agent ic System for Veterinary Disease Screening
链接: https://arxiv.org/abs/2607.26042
作者: Syed Mhamudul Hasan,Anas AlSobeh,Hussein Zangoti,Abdur R. Shahid
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:We present VetClaw, an edge-cloud multimodal agentic system for early veterinary disease screening. VetClaw uses a camera module as an edge sensing device and sends captured images, together with optional symptom descriptions, to a server-hosted vision-language model for zero-shot disease classification. The system separates agent interaction from workflow orchestration: OpenClaw provides scheduling, tool access, user interaction, and notification services on the edge device, while LangGraph manages the stateful screening workflow, including input validation, image transmission, model invocation, safety checks, conditional routing, failure handling, and structured logging. This design moves beyond static image classification by enabling the system to collect visual evidence, invoke external models, apply deterministic safety rules, and generate diagnostic-support alerts. Results show that image-only VLM prediction remains limited, whereas symptom-guided and multimodal inputs improve zero-shot classification performance. Thus, VetClaw transforms a static prediction model into a coordinated, safety-aware system that can use tools, manage workflows, handle failures, and escalate uncertain cases.
[CV-1] Desktop-Delta Bench: Do Computer-Use Models Understand Desktop GUI Transitions?
链接: https://arxiv.org/abs/2607.26041
作者: Abhishek Pillai,Samir Kumar Nayak,Yuan Chen
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Computer-use agents (CUAs) increasingly act through desktop GUIs to complete long-horizon tasks. Current benchmarks primarily measure end-task success or single-frame grounding. Neither isolates whether a model can reconstruct the causal, task-relevant transition produced by an action- crucial for rejecting stale observations, verifying progress, and recovering from failure. This is difficult because inference, remote input, app rendering, and screenshot capture are asynchronous: the next observation may be delayed, occluded, transient, or unrelated, then misread as progress and carried into subsequent planning. We introduce Desktop-Delta Bench (DDB), an offline step-level benchmark with 2,013 human-verified instances from novel, multi-app Linux trajectories across ~15 applications and 50 task domains. DDB trajectories targets 3 failure dimensions- state verification, source tracking, and context-aware control- through 2 complementary tasks: 463 3-frame temporal-ordering instances, including 105 with a cross-trajectory decoy, and 1,550 before-after pairs labeled from 5 actions + its payload. We evaluate 8 closed and open-source model families across 32 ordering and 16 single-action settings, observing consistent gaps. Ordering remains unsaturated: best non-decoy and decoy exact-match rates are 65.1% and 65.7%. Task context improves decoy identification by 6.9 percentage points but reduces non-decoy exact match by 2.2 points; error analysis reveals systematic copying of the presented A-B-C order. Single-action results show that inferring the action family is harder than locating it: click F1 is 0.96 vs, 0.76 for drag, while recognized drags are generally localized well. DDB, thus, complements end-to-end benchmarks by filling the missing diagnostic layer between GUI grounding and final task success, enabling targeted improvements to desktop CUA verification, reliability, and recovery.
[CV-2] Wonder: Video World Model Done Better
链接: https://arxiv.org/abs/2607.26037
作者: Jiacong Xu,Hanwen Jiang,Zhixin Shu,Kalyan Sunkavalli,Vishal M. Patel,Yiqun Mei
类目: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR)
备注: Project Page: this https URL
Abstract:We present Wonder, a general-purpose video world model for real-time, camera-controllable world exploration. Given an image or a conditional video, Wonder constructs a playable world where users can navigate interactively by moving the camera, discovering unseen regions, and revisiting previously observed areas in real time and over a long-term horizon. Achieving this capability requires a system-level co-design of control method, memory mechanism, and training strategy. We introduce a novel camera conditioning with a dense coordinate field whose renderings provide spatially aligned motion and orientation cues, allowing the model to interpret camera motion directly as visual evidence. To support fast and precise memory retrieval over a growing generation context, we propose an efficient sparse attention-based memory mechanism, enabling the model to selectively attend to a small set of relevant context tokens at inference time, regardless of actual context length. We further develop several techniques to rectify the self-forcing-style distillation pipeline, improving the student model’s ability to respect control signals, as well as maintaining diverse generation modes and long-term memory from the teacher. Together, these components enable Wonder to synthesize diverse, minute-scale videos at 16 FPS while preserving coherent geometry, appearance, and dynamics across long rollouts. Beyond image-to-video generation, Wonder naturally supports video-conditioned generation, allowing existing dynamic scenes to be re-shot in real time.
[CV-3] Pictura: Perspective-View Self-Play at Scale for Driving
链接: https://arxiv.org/abs/2607.26005
作者: Yuan Yin,Elias Ramzi,Marc Lafon,Valentin Charraut,Victor Bares,Yihong Xu,Éloi Zablocki,Alexandre Boulch,Thibault Buhet,Andrei Bursuc,Matthieu Cord
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Robotics (cs.RO)
备注:
Abstract:Self-play in simulation produces robust driving policies at scale. Demonstrations of such behavior have been made using privileged vectorized observations such as exact poses and velocities, even for occluded agents. This assumes that perception is solved and introduces a representation gap with the partial observation of a deployed agent driving from the perspective view of egocentric cameras. A common fix, distilling the privileged policy into a camera-input student, leaves the student imitating decisions its own view cannot justify. Instead, we establish perspective-view self-play as a practical training regime. We introduce Pictura, a GPU-accelerated multi-agent driving simulator that renders each agent’s egocentric view at every step, mitigating the representation gap at its source. Pictura sustains up to 500K agent-steps/s (2M images/s) on a single H100. Using Pictura, we train Alberti by self-play with plain PPO. It is the first large-scale driving self-play policy trained directly from perspective images, without privileged observations. Training spans 50B agent steps for ~35M km of driving. It approaches the driving performance of its privileged vectorized counterpart, and transfers zero-shot to Waymo Open Motion Dataset layouts re-rendered in Pictura, where it outperforms privileged vectorized agents. Project page: this https URL
[CV-4] Parallel Decoding Distillation for Fast Image and Video Generation
链接: https://arxiv.org/abs/2607.26004
作者: Neta Shaul,Chao Liu,Arash Vahdat,Julius Berner
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Generation in video diffusion or flow models is computationally expensive due to the slow and iterative sampling process. Current state-of-the-art (SOTA) acceleration methods heavily rely on variational score distillation (VSD) and adversarial losses to distill diffusion models into few-step generators. Albeit achieving high-quality video generation, these training losses are notoriously hard to optimize and suffer from mode collapse, leading to loss of video diversity and lack of motion. In this paper, we introduce Parallel Decoding Distillation (PDD), a simplified and scalable trajectory-based distillation method for fast inference of diffusion and flow matching models. Our architecture and training procedure are compatible with any pre-trained model and support sampling with a varying number of function evaluations (NFE). PDD accelerates generation by predicting multiple denoising steps per network evaluation. Conceptually, it learns a representation of the mean velocity without regressing its derivative using JVPs or finite-difference approximations. Our method achieves SOTA performance with 4-8 NFE on LTX-2.3 Text-to-Video/Audio, Wan 14B Text-to-Video, and Qwen-Image Text-to-Image. Moreover, PDD presents a significant improvement in generated video diversity.
[CV-5] Beyond Zooming: Learning Multi-Tool Visual Reasoning for Ultra-High-Resolution Remote Sensing
链接: https://arxiv.org/abs/2607.25993
作者: Fengxiang Wang,Jiangnan Huang,Mingshuo Chen,Yueying Li,Yang Shi,Junwei Luo,Haoyu Wang,Yansheng Li,Jing Zhang,Haiyan Zhao,Wenjing Yang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Ultra-high-resolution (UHR) remote-sensing (RS) imagery provides fine-grained Earth-observation evidence over city-scale scenes, but poses a fundamental challenge for multimodal large language models (MLLMs): task-relevant evidence is often sparse, local, and spatially dispersed across extremely large visual contexts. A natural solution is to equip MLLMs with zoom-in tools for active local inspection. However, through a pilot study on XLRS-Bench, we find that zoom-in is only partially effective: it resolves easy and medium-level tasks with locally recoverable evidence, but saturates on hard cases requiring global search, multi-region comparison, path planning, or dispersed-evidence reasoning. Motivated by this finding, we move beyond single-tool zoom-in and introduce GeoMTVR, a large-scale Geospatial Multi-Tool Visual Reasoning dataset built from wide-area satellite imagery. GeoMTVR contains 13K UHR VQA samples with interleaved reasoning trajectories, diverse visual tool calls, and returned visual observations, enabling models to learn question decomposition, tool selection, regional inspection, object-level grounding, auxiliary visual reasoning, and cross-tool evidence integration. Beyond supervised fine-tuning, we propose a tool-attention-focused reinforcement learning algorithm that concentrates optimization on critical tool-use decisions, including when to invoke tools, which tool to select, where to apply it, and how to interpret tool outputs. By combining SFT on GeoMTVR with our RL algorithm, we develop GeoLens, a multi-tool visual reasoning MLLM for UHR RS. Experiments show that GeoLens consistently outperforms direct reasoning and single-tool zoom-in baselines, achieving stronger accuracy, better evidence grounding, and more efficient tool-use trajectories.
[CV-6] On the Use of Synthetic Data for Threshold Calibration in Face Recognition: Performance and Security Implications for Border Control Systems
链接: https://arxiv.org/abs/2607.25990
作者: Arto Apila
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by IJCB 2026
Abstract:The recently deployed Entry/Exit System (EES) introduces large-scale biometric verification into European border control, requiring face recognition systems to operate at extremely low false match rates (FMR). While regulatory frameworks define performance targets at the EES Central System level, they do not specify how verification thresholds should be calibrated in practice at the Member State level. In operational settings, obtaining representative real-world data for calibration is often constrained by legal, logistical, and privacy limitations. In this work, we investigate the use of synthetic face data for threshold calibration in document-to-live verification scenarios relevant to border control systems. We analyze the alignment of genuine and impostor score distributions between synthetic and real datasets and evaluate the transferability of calibrated thresholds across domains, with a focus on low-FMR operating points. Our results show that synthetic data can approximate calibration behavior in controlled settings, but fails to reliably generalize to unconstrained conditions due to mismatches in score distribution tails. These discrepancies lead to significant degradation in recognition performance and increased vulnerability to morph-based attacks. We further demonstrate that calibration outcomes are highly dataset-dependent, even across synthetic datasets. Overall, our findings highlight that while synthetic data is useful for system development and preliminary calibration, our results indicate that reliable threshold selection in high-security deployments typically requires validation and adjustment using representative real-world data. Comments: Accepted by IJCB 2026 Subjects: Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2607.25990 [cs.CV] (or arXiv:2607.25990v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2607.25990 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Arto Apila [view email] [v1] Mon, 27 Jul 2026 09:14:35 UTC (5,931 KB)
[CV-7] Schrödingers Cat: Probabilistic Representation and Prediction of Potential Scene Kinematics ECCV2026
链接: https://arxiv.org/abs/2607.25984
作者: Timy Phan,Jannik Wiese,Björn Ommer
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: Accepted at ECCV 2026. Project page: this https URL
Abstract:Predicting how a scene may evolve from partial observations requires reasoning about multiple possible futures rather than committing to a single trajectory. Existing approaches either generate appearance-dominated video predictions or sample a small number of trajectories without explicitly modeling the distribution of possible motion. We introduce Goal-Aware Representations of Future kInEmatic Latent Distributions (GARFIELD), a probabilistic model of scene kinematics that learns a structured spatio-temporal latent representation of the distribution over possible futures given an image and optional spatio-temporally sparse constraints. The same latent representation enables both joint sampling of all trajectories and direct access to the underlying motion distribution through an efficient deterministic density decoder. As a result, uncertainty about future motion can be localized to specific scene elements and timesteps and progressively refined through additional constraints. Experiments demonstrate strong motion planning performance competitive with large video generation models while sampling trajectories 97\times faster. Our method further estimates motion densities two orders of magnitude faster than Monte-Carlo sampling from motion generation models, enabling interactive exploration and uncertainty-aware planning.
[CV-8] Quasi-SVD: Learning a Lie-constrained matrix factorisation for real-time imaging
链接: https://arxiv.org/abs/2607.25967
作者: Christopher Hahne
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Numerical Analysis (math.NA)
备注:
Abstract:Singular Value Decomposition (SVD) underlies matrix factorisation tasks across computational imaging, with medical applications increasingly demanding real-time processing. Yet SVD algorithms are inherently sequential, constraining real-time GPU throughput and limit online deployment in clinical pipelines. This study introduces Quasi-SVD, a differentiable, fully parallelized matrix factorization framework for GPUs. Rather than enforcing orthogonality on both factors, it guarantees exact orthogonality for a single Lie-parameterized factor while recovering the remaining components through soft constraints, enabling efficient parallel decomposition without iterative singular-vector optimization. This asymmetric design, provably sufficient for valid factorisation, achieves reconstruction fidelity of SSIM = 0.89-0.94 and accelerates computation by 3-20x relative to cuSOLVER and randomised SVD, enabling throughput above 25 FPS. Performance is evaluated on two medical imaging tasks spanning complementary computational regimes: (1) spatio-temporal background subtraction for ultrasound localisation microscopy, requiring high-dimensional matrix separation, and (2) Mueller matrix polarimetry for neurosurgical tissue characterisation, requiring massive batch processing of small matrices. Across both regimes and multiple imaging instruments, the proposed framework demonstrates robust domain transfer and throughput exceeding 25 FPS at clinical matrix scales, a rate sufficient for live image-guided workflows that classical solvers cannot currently support in these settings. By prioritising downstream reconstruction fidelity over exact spectral recovery, Quasi-SVD makes structured matrix factorisation practical for real-time imaging.
[CV-9] LaP-Forensics: Latent-Pixel Consistency Guided Multimodal Reasoning for Deepfake Detection ACM-MM2026
链接: https://arxiv.org/abs/2607.25962
作者: Can Wang,Yuhao Wang,Yushe Cao,Canran Xiao,Fei Shen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at ACM Multimedia 2026 (ACM MM 2026)
Abstract:Recent generative models can produce images with few obvious visual artifacts, weakening detectors and explanations that rely only on surface appearance. We present LaP-Forensics, a multimodal framework that augments RGB semantics with reconstruction-based forensic evidence. A frozen Stable Diffusion DDIM inversion-reconstruction model provides a fixed reconstruction reference, and its residual map measures local compatibility with that reference. Independent projectors encode the RGB image and residual map before a structured Where-What-Why model predicts a textual analysis and an artifact this http URL fine-tuning is followed by Group Relative Policy Optimization (GRPO), whose reward combines mask overlap with output-structure and evidence-reference terms. These text-side terms encourage the model to refer to the consistency map but do not constitute a verifier of free-form textual truth. A separate image-level head fuses RGB and DDIM-residual class features. Experiments show cross-generator detection on UniversalFakeDetect and competitive artifact localization on the official SynthScars benchmark. Controlled cue-construction, inversion-horizon, component, reward-term, and counterfactual analyses support the utility of the residual stream under the evaluated settings, while free-form textual faithfulness and reliability under post-processing remain open limitations.
[CV-10] Knowledge-Guided Multimodal Reasoning over Interacting Streams for Video-Level Ambivalence and Hesitancy Recognition ECCV2026
链接: https://arxiv.org/abs/2607.25961
作者: Podakanti Satyajith Chary,Barath Parthiban,Pranesh Velmurugan,Adeeba Khan,Nagarajan Ganapathy
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 14 Pages, 1 Figure, Ambivalence/Hesitancy (AH) Video Recognition Challenge, ECCV 2026
Abstract:Ambivalence and hesitancy (A/H) are conflicting affective states that precede the delay or abandonment of health behaviour change. Recognition of A/H at the video level is difficult, since the signal arises from disagreement across and within facial, vocal, linguistic, and bodily modalities, and manifests differently across individuals. The proposed PRISM-AH (Predictive Reasoning over Interacting Streams for Multimodal Ambivalence/Hesitancy Recognition), is a framework that treats A/H as a multimodal conflict that unfolds over time. Frozen vision, audio, and text encoders are aligned into short time windows and passed to a lightweight streaming model that scores cross-modal dissonance, predicts each next window to expose a hesitation surprise signal, discovers behaviour prototypes, and is conditioned on participant metadata. Dense window-level annotations supervise the model as an auxiliary objective, and the decision threshold is calibrated for macro F1. A knowledge-guided large language model then reasons over structured evidence using the expert cue taxonomy of the dataset, and its verdict is fused late only when validation performance improves. On the labelled public test partition of 525 videos, PRISM-AH attains a macro F1 of 0.6133, compared to the reported zero-shot baseline of 0.2827. The reasoning gain is validated to transfer from validation to the larger test partition.
[CV-11] MODUS: Decoder-Only Any-to-Any Modeling of Diverse Modalities ICML2026
链接: https://arxiv.org/abs/2607.25948
作者: Mingqiao Ye,Zhaochong An,Zhitong Gao,Xian Liu,François Fleuret,Chuan Li,Amir Zadeh,Serge Belongie,Afshin Dehghan,Jesse Allardice,David Mizrahi,Oğuzhan Fatih Kar,Roman Bachmann,Amir Zamir
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Accepted at ICML 2026. Project page: this https URL
Abstract:Any-to-any models predict any modality from any combination of others within a single network, a formulation used in multimodal vision and vision-language models, and increasingly in scientific domains such as ecology and astronomy. Existing any-to-any models are typically trained from scratch using encoder-decoder or diffusion architectures, impacting their performance and preventing them from using strong pre-trained decoder-only models as a prior. In this work, we investigate decoder-only any-to-any multimodal modeling, which treats all modalities symmetrically and supports arbitrary modalities as inputs and outputs without modality-specific heads, losses, or task pipelines. Because every modality is both an input and an output of the same model, the resulting model, named Modus, can support a range of applications, such as chained generation through intermediate modalities or cross-modal self-verification by scoring the model’s own outputs with another generated modality. Modus demonstrates strong out-of-the-box performance and is competitive with specialist and multitask baselines using a single model across various benchmarks. All materials are open-sourced at this https URL.
[CV-12] Face De-Identification: A Domain-Centric Survey from Capture to Processing
链接: https://arxiv.org/abs/2607.25926
作者: Hui Wei,Hao Yu,Guoying Zhao
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted by IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI). Github repository: this https URL
Abstract:Face de-identification (De-ID) aims to remove or conceal personally identifiable facial features in images or videos to prevent identity recognition while preserving utility for downstream tasks. With the rising emphasis on data privacy and responsible AI, face De-ID has emerged as an active research area spanning computer vision and privacy-preserving communities. Early approaches, and many contemporary ones, operate in the digital domain by modifying pixel-level or appearance-level features through post-capture processing. Recent advances extend face De-ID beyond post-processing by integrating privacy mechanisms directly into sensors during image acquisition, bridging sensing systems and downstream vision algorithms. In parallel, physical-domain methods explore wearable accessories and materials that conceal identity information in real-world environments prior to capture. In this survey, we present the first unified overview that spans the full data acquisition pipeline, encompassing the physical, sensor, and digital domains. Through this domain-centric lens, we systematically analyze current methodologies, technical progress, and the distinct challenges inherent to each stage. We then review and organize existing evaluation protocols, examining current practices and highlighting the critical need for standardized, comprehensive benchmarks. Finally, we identify key open problems and outline emerging research directions to guide future work in this rapidly evolving field. To support ongoing research, we maintain a project page that organizes relevant literature with collected datasets and open source code: this https URL.
[CV-13] Evaluating VLMs for Autonomous Agent -Driven Geometry Clipping Detection in Video Game QA
链接: https://arxiv.org/abs/2607.25921
作者: Carlos Celemin,Benedict Wilkins,Adrián Barahona-Ríos,Saman Zadtootaghaj,Nabajeet Barman
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:In this work, we study the use of Vision-Language Models (VLMs) for anomaly detection in an agent-driven game Quality Assurance (QA) pipeline focusing on geometry clipping. In this evaluation, a custom exploration agent navigates a game level to collect visual observations, while the automatic annotation pipeline provides frame-level clipping labels. This setup allows us to evaluate recent VLMs on a controlled anomaly detection task without manual annotation. We benchmark six recent VLMs (Gemini, GPT, Qwen, Gemma, Llama, and Ministral) under a zero-shot prompting setting and analyse their sensitivity to four prompt variants. Our results show that while the VLMs can capture visual cues associated with geometry clipping, they all produce substantial false positives on visually ambiguous frames such as near-contact geometry and partial occlusions. Gemini-3.1-Flash achieves the best overall accuracy and is the most robust to prompt variation, while open-source models exhibit large precision–recall swings depending on the prompt design. These findings suggest that current VLMs are best suited as high-recall candidate filters within multi-stage QA pipelines rather than as standalone bug detectors. Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.25921 [cs.CV] (or arXiv:2607.25921v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2607.25921 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-14] HiFi-UMI: Learning Deployable Manipulation Policies from High-Fidelity UMI Data Alone
链接: https://arxiv.org/abs/2607.25895
作者: Simple AI:Yuteng Wei,Jinming Ma,Jiawei Wang,Weitao Zhou,Yushen Zuo,Ke Rui,Minglei Li,Jinhao Zhang,Zhikang Pan,Xiang Wang,Haoran Jia,Huan Du,Zicheng Zeng,Jun Ma,Guiyu Qin,Di Zhang,Xiaofei Li
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 33 pages, 15 figures, 4 tables. Project page: this https URL Dataset: this https URL
Abstract:Learning deployable manipulation policies is bottlenecked by the scarcity of data that is both high-fidelity and scalable. Real-robot teleoperation is accurate but costly to scale; robot-free UMI capture scales readily, and current practice uses the resulting data mainly for pre-training, adding a small real-robot “anchor” at post-training. We ask whether raising the fidelity of robot-free UMI data, rather than shrinking the real-robot fraction, can remove that anchor. We present HiFi-UMI, a portable UMI data-production system co-designed for trajectory accuracy, inter-gripper relative pose, synchronization, and field of view: head-mounted offline stereo-inertial SLAM, native rather than reconstructed relative pose, a shared microsecond GPIO trigger, and two wide-angle cameras per hand covering ~200 degrees. It reaches 3 mm workspace-local end-effector accuracy without external tracking infrastructure. Using this corpus, we demonstrate zero-robot post-training: a policy post-trained solely on HiFi-UMI demonstrations deploys directly on a real robot and matches in-domain teleoperation across three backbones spanning the vision-language-action and world-action-model families, with success-rate differences of -2.5, +3.1, and -0.6 percentage points on StarVLA-QwenPI, OpenPI-pi_0.5, and LingBot-VA; the strongest policy reaches 85% on a precision insertion task, even though the teleoperation baseline is collected in the evaluation scene and no HiFi-UMI trajectory is. Pre-training on 4,000 hours from the same corpus lowers action error on ten unseen tasks by 41% and, on StarVLA-QwenPI, raises real-robot success by a further 18.1 percentage points. We open-source HiFi-UMI-2K, 2,000 hours of microsecond-synchronized, ultra-wide-FoV demonstrations, each automatically reconstructed and validated through simulation replay, as a large-scale, high-fidelity resource for the robot-learning community.
[CV-15] IGA: Trajectory-Injected Generative Attack against Black-box AIGC Detectors
链接: https://arxiv.org/abs/2607.25894
作者: Xia Du,Zhuosen Bao,Zheng Lin,Jizhe Zhou,Jiawei Lian,Chi-man Pun,Jun Luo,Wei Ni,Symeon Chatzinotas
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 14 pages, 5 figures
Abstract:Recent diffusion models have achieved remarkable realism in facial image synthesis, posing growing challenges to artificial intelligence-generated content (AIGC) forensic this http URL evasion methods typically perturb pre-generated images or require detector-aware training, which may introduce visible or statistical artifacts and limit applicability when the diffusion model must remain frozen and the target detector is accessible only through black-box queries. We propose Trajectory-Injected Generative Attack (TIGA), a source-image-free and training free framework that generates detector-evasive images within a single diffusion sampling trajectory. TIGA steers the latent Denoising Diffusion Implicit Model (DDIM) trajectory so that adversarial properties emerge during generation rather than being added afterward. TIGA first aggregates gradients from multiple white-box surrogate detectors to form a transferable, sign-aware prior, and then performs anisotropic directional search with symmetric finite-difference queries to estimate the black-box target response. The estimated directions are stabilized by decayed momentum and injected according to the DDIM noise schedule, with frequency-domain reshaping to suppress high frequency artifacts. Experiments on surrogate and unseen specialized forensic detectors show that TIGA achieves strong blackbox attack performance, transferability, and high robustness under common post-processing operations without source images or diffusion-model retraining, while preserving high perceptual quality.
[CV-16] Open-Ended CT Volume Segmentation with Weak Supervision from Language
链接: https://arxiv.org/abs/2607.25860
作者: Sanjay Subramanian,Junwei Yu,Zirui Wang,Rohil Malpani,Maggie Chung,Adam Yala,Dan Klein,Trevor Darrell
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:We introduce a method for training a text-conditioned segmentation model for CT scans, which combines voxel-level supervision with coarse but scalable slice-level supervision from reports. We extract, from a large database of scan-report pairs, descriptions of findings with indices of slices where those findings occur. We then finetune a general-purpose 2D image segmentation model, SAM3, with standard segmentation losses from strongly labeled data and with a slice-level classification loss from the extracted weak supervision. Our results on the ReXGroundingCT dataset illustrate that this strategy improves the segmentation dice score: from an 8% relative gain when there are 1000 fully labeled volumes to 22% when there are 250 fully labeled volumes.
[CV-17] Adversarial Deepfake Generation and an Investigation of Purification-Based Adversarial Detection
链接: https://arxiv.org/abs/2607.25842
作者: Junghyun Kim,Seunghyun Kim,Jiyoung Woo
类目: Computer Vision and Pattern Recognition (cs.CV); Cryptography and Security (cs.CR)
备注: Accepted at CLEF 2026, ImageCLEF-Deepfake task. Published in CEUR-WS CLEF 2026 Working Notes
Abstract:This paper describes the participation of team “Go To Germany” in the ImageCLEF 2026 Deepfake Detection and Generation Task. For the image generation task, we employ FLUX.1-dev with PuLID for identity-preserving face synthesis, combined with a multi-model PGD adversarial attack targeting 12 detectors simultaneously (DiffJPEG-in-loop, MI/DI/EoT, adaptive weighting, two-stage warm-start). Our approach achieved 90% evasion against organizer detectors and 57.6% against participant detectors, with a final generation score of 0.4170. For the image detection task, we combine two complementary detectors - SigLIP+DINOv2 for AI-generated images and GenD-DINOv3 for face manipulations - in a max-probability ensemble, achieving 99.4% accuracy on baseline deepfakes but suffering from high false-positive rates on real images, resulting in a final detection score of 0.6986. Beyond the official submission, we conducted a self-initiated investigation of purification-based adversarial detection, comparing three families of detection signals across six detectors that share a CLIP ViT-L/14 backbone. We find that raw |\Delta \textlogit| under median-3 purification, applied through the EFFORT detector, separates adversarial inputs from clean inputs with AUROC 0.81-0.98 across four adversarial source types - a finding that refutes the simple backbone-preservation hypothesis and exposes a sharp JPEG-quality cliff at Q70 where the signal collapses.
[CV-18] Beyond Static Costs: Learning-Dynamics Aware Loss Functions for Long-Tailed Classification
链接: https://arxiv.org/abs/2607.25830
作者: Varad Shinde,Nikhil Kumar Shrey,Magesh Rajasekaran,Md Saiful Islam Sajol,Harshil Bhargava,Subhajit Sidanta,Supratik Mukhopadhyay,Yimin Zhu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Deep learning models in computer vision face significant challenges when trained on long-tailed datasets, where a few majority classes dominate while many minority classes are severely underrepresented. Such imbalances frequently arise in real-world scenarios such as rare species recognition, manufacturing fault detection, and medical image understanding, leading to biased models that underperform on tail classes. Existing reweighting methods typically rely on static class frequencies to penalize the model, ignoring the dynamic nature of how effectively a network actually learns a class over time. We address this by introducing a novel Learning-Dynamics Aware Loss (LDAL) function that shifts the focus from static sample counts to dynamic learning progress. LDAL framework adjusts class weights continuously by leveraging: (i) the strength of learned feature representations (semantic scale), (ii) the intrinsic learning difficulty of each class, measured via the Shannon entropy of its predictions, and (iii) an inter-epoch regularizer term that tracks prediction shifts between consecutive epochs to stabilize training and avoid local minima. LDAL is purely a objective function which incurs negligible computational overhead while adapting to the feature learning of the model. Experimental results on multiple benchmark datasets demonstrate that our approach significantly surpasses state-of-the-art reweighting loss functions, providing an optimal trade-off between accuracy and generalizability. The source code is available at this https URL
[CV-19] Food Image Segmentation with LLM -Derived Ingredient Labels and Multimodal Fusion
链接: https://arxiv.org/abs/2607.25820
作者: Jui-Feng Chi,Wei-Ta Chu,Sheng-Long Lin
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Food image segmentation plays a vital role in health-related applications such as nutrition tracking and personalized health monitoring. However, existing models often underperform on visually similar ingredients and rare food categories. To address this issue, we propose two plug-and-play multimodal modules that enhance the segmentation performance by leveraging ingredient labels inferred from food images using large language models (LLMs). The first module, called LIM-F (Language Injection Module for Features), is designed to pair with any image encoder that produces multi-layer outputs (e.g., Swin Transformer), while the second module, LIM-Q (Language Injection Module for Queries), targets Mask2Former-style Transformer-based decoders. Both modules enable training without the need for pre-aligning images with text by directly injecting semantic ingredient information into the visual analysis pipeline. On the FoodSeg103 benchmark, the proposed method achieves state-of-the-art performance. Specifically, integrating LIM-Q into the Mask2Former decoder with a Swin-L image encoder yields a mean Intersection over Union (mIoU) of 55.0. LIM-F also demonstrates strong generalization and competitive performance, reaching an mIoU of 54.4 under the same model (Swin-L+Mask2Former). Furthermore, its applicability extends beyond Transformer-based decoders, as evidenced by an improvement from 47.7 to 49.8 mIoU when integrated into a CNN-based architecture. Notably, the improved segmentation accuracy is achieved with only a moderate (at most 3.8 GB) increase in the GPU memory consumption during training. Thus, the proposed approach offers a practical and scalable solution for fine-grained food understanding.
[CV-20] SepPrune:A Separator-based Pruning Framework for Efficient Multimodal Large Language Models
链接: https://arxiv.org/abs/2607.25818
作者: Yuchen Wang,Qihui Zhu,Yang Liu,Xiaoyan Sun,Siying Wu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Recent multimodal large language models (MLLMs), such as Qwen2.5-VL and InternVL3, generate large numbers of vision tokens for high-resolution inputs, leading to substantial computational cost. Existing vision token pruning methods either depend on cross-modal attention and cannot prune before the prefill stage, or rely on diversity estimation with high computational overhead. We observe that attention scores from both vision and text tokens peak at modality separator tokens, suggesting that these separators bridge the two modalities. Based on this observation, we propose SepPrune, an efficient, training-free, plug-and-play pruning method that uses the separator token as a unified query to rank and select informative vision tokens. SepPrune reuses the LLM’s built-in projection parameters and requires no architectural changes. Experiments on Qwen2.5-VL-7B show that SepPrune achieves state-of-the-art performance, retaining 96.3% of the original accuracy while removing 80.2% of vision tokens.
[CV-21] Freq-RemoteVAR: Next-Frequency Autoregressive Modeling for Remote Sensing Change Detection
链接: https://arxiv.org/abs/2607.25815
作者: Luqi Gong,Rui Xu,Yue Chen,Chao Li,Jingqi Hong,Xuefeng Zhao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Remote sensing change detection aims to identify land-cover changes from bi-temporal images. Most existing methods follow a one-shot dense prediction paradigm, directly regressing a change mask from fused features. However, such approaches overlook the intrinsic frequency characteristics of change patterns. We propose Freq-RemoteVAR, a frequency autoregressive framework that reformulates change detection as a structured generation problem in the frequency domain. Instead of predicting the change mask in a single step, we introduce a next-frequency prediction paradigm, where change information is progressively generated from coarse to fine. We design a frequency-aware mask tokenization strategy that decomposes change supervision into multi-frequency token targets via Fourier transformation and quantization. We develop a Frequency VAR Transformer, which performs causal autoregressive modeling over frequency tokens. The model starts from learned mask queries and progressively predicts frequency-level tokens conditioned on previously generated tokens and bi-temporal image features, effectively capturing long-range dependencies across frequency scales. We introduce Scale-Aligned RoPE Cross Attention (SRCA) module, which aligns frequency-domain mask queries with spatial-domain bi-temporal features under a unified coordinate system, enhancing spatial-frequency consistency during generation. We propose a Change-quality Control module that adaptively modulates the generation process through dynamic normalization, attention biasing, and spatial offset adjustment, thereby suppressing pseudo-change responses and improving robustness. Extensive experiments on CDD, GZ-CD, and LEVIR-CD demonstrate that Freq-RemoteVAR consistently outperforms existing methods, particularly in challenging scenarios with complex appearance variations and noisy disturbances.
[CV-22] Explicit Layer Modeling for Video Object Insertion and Layer Decomposition
链接: https://arxiv.org/abs/2607.25802
作者: Kyujin Han,Seungjoo Shin,Sunghyun Cho
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Most video editing systems still lack explicit layered video representations, limiting their ability to perform realistic compositing, object reuse, and consistent manipulation. This limitation is especially pronounced in video object insertion and video layer decomposition, where existing methods rely on implicit inference or per-scene optimization due to the absence of explicit foreground-layer supervision. We introduce TriLayer, a large-scale triplet video dataset containing aligned composite, background, and foreground videos, where the foreground layers include both object appearance and associated visual effects. This explicit supervision enables models to learn layered video representations directly rather than inferring them implicitly. Building on this dataset, we propose DBL-Diffusion, a dual-branch diffusion framework that jointly models RGB composites and RGBA foreground layers through shared denoising and cross-branch interaction. We instantiate the framework in two tasks: DBL-Insert for layered object insertion, which generates explicit RGBA layers for realistic compositing and flexible post-editing, and DBL-Decompose for video layer decomposition, which recovers foreground and background layers using triplet supervision. Experiments demonstrate that explicit layer modeling substantially improves both insertion fidelity and decomposition quality.
[CV-23] Fine-Grained Food Image Understanding via Target-Aware Data Alignment
链接: https://arxiv.org/abs/2607.25794
作者: Jui-Feng Chi,Wei-Lun Chu,Bruce Coburn,Jinge Ma,Fengqing Zhu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Fine-grained food visual–semantic understanding requires models to capture subtle distinctions across ingredients, cooking methods, doneness, color, texture, and plate composition. Although CLIP-style vision-language models provide a natural framework for this task, their effectiveness is limited when training relies on heterogeneous web-collected image–text pairs. Such data often exhibit a web-to-target domain gap and cross-modal misalignment, where images differ from the target distribution and captions are noisy, multilingual, or weakly grounded in visual content. We propose a data-centric multimodal alignment method for fine-grained food description and recognition. Our method first performs target-aware data selection to identify visually relevant training subsets, then applies VLM-based caption refinement to generate visually grounded, target-style descriptions. Using these curated image–caption pairs, we train complementary CLIP-style retrieval experts and further combine their decisions through a hierarchical VLM-assisted multi-expert decision-level fusion strategy that invokes the VLM only when experts disagree. Experiments show that our data refinement strategy significantly improves retrieval performance over naive web supervision, with VLM-based caption refinement alone yielding an average performance gain of approximately 19%. Our full method also achieves more than twice the retrieval score of pure VLM-based retrieval while remaining substantially more efficient.
[CV-24] FLASH: Efficient Impact Fall Detection with Unified Hypergraph State-Space Model ICIP2026
链接: https://arxiv.org/abs/2607.25791
作者: Tresor Y. Koffi,Youssef Mourchid,Yohan Dupuis
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 6 pages, 2 figures. Accepted at IEEE International Conference on Image Processing (ICIP 2026), to appear September 2026
Abstract:Falls represent a critical public health challenge, and accurate detection of the impact moment when an individual hits the ground is crucial for timely intervention. Existing skeleton-based methods rely on graph neural networks modeling only pairwise joint connections, failing to capture multi-joint coordination characteristic of fall impacts, while transformer-based temporal models suffer from quadratic complexity limiting real-time deployment. We propose FLASH, a novel framework integrating single-matrix hypergraph representations with Mamba’s selective state-space models through adaptive feedback mechanisms for efficient impact detection. Our approach constructs biomechanically-grounded hyperedges to model functional joint coordination while leveraging Mamba’s linear-time complexity to capture temporal dynamics. Experiments on UP-Fall and UMAFall datasets demonstrate that FLASH achieves state-of-the-art accuracy with real-time inference capability and strong zero-shot cross-dataset generalization, while significantly reducing computational cost compared to dual-representation and transformer-based methods. The model provides interpretable feedback through learned attention patterns aligned with biomechanical principles. Code is available at this https URL.
[CV-25] owards Faithful Sentimental Image Captioning via Evidence-Aware Multi-Agent Reasoning
链接: https://arxiv.org/abs/2607.25789
作者: Tiecheng Cai,Zexian Yang,Chao Chen,Shanshan Lin,Xiangwen Liao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Sentimental Image Captioning (SIC) requires balancing emotional expression with visual fidelity. Existing methods often struggle with this trade-off, leading to hallucinations due to insufficient local grounding and the lack of sentimental verification mechanisms. To address these limitations, we propose SEA-Cap, a Sentiment-Evidence-Aware Multi-Agent System for faithful and evidence-grounded sentimental image captioning. SEA-Cap incorporates a Sentiment Evidence Miner that extracts structured, local affective cues to shift sentiment control from global attributes to verifiable object-level evidence. Leveraging this evidence, our framework orchestrates a collaborative workflow where a Generator, Hallucination Checker, and Arbitrator iteratively refine captions via a shared blackboard. By explicitly auditing generated content against mined visual evidence, SEA-Cap ensures both sentiment accuracy and factual consistency. Extensive experiments on two benchmark datasets demonstrate that SEA-Cap effectively mitigates hallucinations and achieves state-of-the-art performance.
[CV-26] GeoMFD: Continual Drone-View Geo-Localization with Geometry-Aware Adapter and Margin-Field Distillation
链接: https://arxiv.org/abs/2607.25788
作者: Zhongwei Chen,Hai-jun Rong,Tao Zhang,Xianfeng Nie,Xiangbao Zhang,Guoqi Li,Zhao-Xu Yang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Existing drone-view geo-localization (DVGL) methods are mainly developed under a static training paradigm, where models are optimized for fixed environments with all training data available in advance. However, this paradigm is difficult to extend to real-world deployment, where drones may encounter diverse environments and require multiple environment-specific models, resulting in additional storage and model-selection costs. Directly adapting a single model to new environments also risks distorting previously learned cross-view embedding geometry and causing forgetting. To address these challenges, we formalize the continual drone-view geo-localization (C-DVGL) setting and propose GeoMFD, a geometry-aware continual adaptation method for DVGL. GeoMFD combines a cold-start bootstrapping strategy (CBS), a geometry-aware adapter (Geo-Adapter), and margin-field distillation (MFD) to balance adaptation and cross-view geometry preservation. CBS initializes a stable embedding space, Geo-Adapter enables environment adaptation through controlled residual corrections, and MFD preserves similarity margins between positive pairs and hard negatives to alleviate cross-view geometry forgetting. Extensive experiments demonstrate that GeoMFD effectively mitigates forgetting and achieves competitive performance with environment-specific DVGL methods using a single continuously updated model.
[CV-27] A Unified Benchmark and Modality-Adaptive Network for Day-and-Night Drone-View Geo-Localization
链接: https://arxiv.org/abs/2607.25778
作者: Songtianhao Xu,Zhongwei Chen,Zhao-Xu Yang,Weifeng Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Most existing drone-view geo-localization (DVGL) benchmarks contain drone imagery captured under a single illumination condition and lack geographically aligned visible drone images, infrared drone images, and satellite images from the same locations. To evaluate the generalization capability of DVGL methods under challenging illumination conditions, some methods train models on a visible benchmark and test them on an independent infrared benchmark. This protocol essentially constitutes transfer between datasets, which makes it difficult to systematically evaluate DVGL across daytime and nighttime conditions within a unified benchmark. To address this limitation, we construct IRCHN,a real-world DVGL benchmark designed for localization across different illumination conditions. IRCHN contains 26,460 images collected from 8,820 geographic locations across four representative scene categories, including farmland, coastline, forest, and urban areas. Each location provides one visible drone image, one infrared drone image, and one corresponding satellite image, which enables unified evaluation of DVGL methods across different illumination conditions and sensing modalities. We further propose the Modality-Adaptive State-Space Transport Relation Network (MASTR-Net), a DVGL framework tailored to localization under varying illumination conditions. MASTR-Net integrates modality-adaptive feature enhancement, bidirectional selective state-space relation modeling, and soft optimal transport relation alignment to jointly reduce modality gaps and view-induced structural discrepancies. Extensive experiments demonstrate that MASTR-Net outperforms existing state-of-the-art methods on IRCHN for localization under varying illumination conditions and achieves competitive performance on two infrared benchmarks, IR-VL328 and CVGL-RGBT. Code: this https URL
[CV-28] Image Quality Dependent Degradation for AI Systems
链接: https://arxiv.org/abs/2607.25736
作者: Yannick Kees,Elena Hoemann,Frank Köster,Sven Hallerbach
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Perception is one of the primary applications where neural networks outperform conventional algorithms. One example is AI systems for automated driving, which can detect pedestrians based on image data and avoid them accordingly. A substantial challenge with these AI systems is that their output depends heavily on the quality of the input images. For example, if an image is of inferior quality due to heavy contamination, such as noise or darkness, accurate predictions are hardly feasible. Additionally, various types of errors can occur, each with varying relevance to the trustworthiness of the underlying AI system. In particular, it may be more critical not to detect an existing person than to detect a person where there is none. Therefore, we want to show that we can still avoid the most critical errors in situations of inferior image quality. To achieve this, we aim to establish a fail-degraded system by lowering the network’s confidence threshold based on the estimated image quality, enabling it to detect objects more cautiously in uncertain situations. Additionally, we present a novel method for estimating the quality of incoming images by comparing them to the training data using normalizing flows. We will also conduct experiments applying our method to state-of-the-art object detection. In summary, we will present a design strategy for AI-based systems in automated driving that can deal with poor-quality input data without resorting to fallback solutions. Such measures enhance trust in AI-based systems and lead to an increased provision of the AI component.
[CV-29] A systematic evaluation of machine learning classifiers for event-by-event background rejection in LAFOV PET scanners
链接: https://arxiv.org/abs/2607.25732
作者: Konrad Klimaszewski,Michał Obara,Mateusz Bala,Beatrix C. Hiesmayr,Lech Raczyński,Roman Y. Shopa,Wojciech Zdeb,Wojciech Krzemien
类目: Computer Vision and Pattern Recognition (cs.CV); Medical Physics (physics.med-ph)
备注:
Abstract:The introduction of LAFOV PET scanners brings significant sensitivity gains but also a substantial increase in the background rate from accidental coincidences, phantom-scattered and detector-scattered photons. While machine learning methods have been applied to background reduction in PET imaging, they target specific background components in post-processing rather than event-by-event classification on the raw data. In this work, we formulate coincidence classification as a supervised multi-class problem and evaluate XGBoost, AdaBoost and Neural Network classifiers as pre-reconstruction filters, using Monte Carlo simulations of the Siemens Biograph Vision Quadra scanner with NEMA IEC and anthropomorphic XCAT phantoms. We investigate two feature sets: a 4-feature representation based on the Attenuation Factor, photon time difference, energy sum, and energy difference, and an extended 6-feature set that incorporates topology-based variables. A systematic robustness study via cross-phantom inference reveals that the 4-feature models generalise significantly better across different phantom geometries, with XGBoost suffering an accuracy loss of only 0.04 compared to 0.13 for the 6-feature variant. Our best models achieve accuracies of up to 0.74 and 0.69 for the NEMA IEC and XCAT phantoms, respectively, outperforming traditional geometry-based cuts. However, we show that this compact feature set not only provides limited suppression of in-phantom scattered coincidences, but it also can lead to non-trivial spatial patterns. With scattered coincidences being the dominant background component in clinical conditions, this suggests that while the method serves as an effective and geometry-agnostic replacement for traditional cut-based selection, meaningful further gains in image quality will require either larger input representations or dedicated treatment of the phantom-scattered component.
[CV-30] Impact Detection in Fall Events: Leverag ing Spatio-Temporal Graph Convolutional Networks and Recurrent Neural Networks Using 3D Skeletons Data
链接: https://arxiv.org/abs/2607.25710
作者: Tresor Y. Koffi,Youssef Mourchid,Mohammed Hindawi,Yohan Dupuis
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Fall represents a significant risk of accidental death among individuals aged over 65, presenting a global health concern. A fall is defined as any event where a person loses balance and moves to an off-position, which may or may not result in an impact where the person hits the ground. While fall detection systems have achieved good results in general, impact detection within falls remains challenging. This study proposes an efficient methodology for accurately detecting impacts within fall events by incorporating 3D joints skeleton data treated as a graph using Spatio-Temporal Graph Convolutional Networks (STGCN), Gated Recurrent Unit (GRU), and Bidirectional Long Short-Term Memory (BiLSTM) layers. By pinpointing impact moments, our approach enhances precision by distinguishing between false falls and actual impacts, contributing to better healthcare resource allocation. Our methodology, evaluated using the improved 3D skeletons UP-Fall dataset, achieves accuracy exceeding 90% across various fall scenarios. We have made this improved dataset publicly available at this https URL to facilitate further research.
[CV-31] OmniPhys: Knowledge-Graph-Driven Benchmarking and Collective Optimization for Physical Commonsense in Text-to-Image Generation KDD2026
链接: https://arxiv.org/abs/2607.25641
作者: Yajing Xu,Yarong Lan,Jiaoyan Chen,Yichi Zhang,Jeff Z. Pan,Mingchen Tu,Zhizhen Liu,Wen Zhang,Huajun Chen
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: accepted by KDD 2026 DB track
Abstract:While text-to-image models exhibit remarkable visual fidelity, they frequently violate fundamental physical commonsense. Existing benchmarks often rely on coarse-grained descriptions, failing to diagnose the mastery of specific physical principles. Moreover, the high stochasticity of generative processes causes current prompt optimization methods to suffer from gradient hallucinations, where optimizers are misled by transient visual artifacts rather than systemic flaws. To address these challenges, we introduce OmniPhys, a rigorous benchmark of 1,551 samples grounded in a Physical Knowledge Graph. By aligning PhET simulations with standard curricula, OmniPhys operationalizes a knowledge-to-scenario pipeline that performs diagnostic stress tests via a dual-path verification protocol. We further propose OmniPrompt, an iterative framework that treats physical alignment as a discrete optimization problem. For each query, OmniPrompt aggregates K stochastic images into a per-query feedback buffer. Across training, it further merges feedback from batches of B queries before each meta-policy update, filtering seed and query-local noise. Evaluations across 12 representative text-to-image models reveal universal physical bottlenecks. Results demonstrate that OmniPrompt significantly enhances physical consistency across diverse backbones, proving the transferability and efficacy of our evolved meta-policies. The code and data are available at this https URL
[CV-32] Beyond Facial Consistency: Personalized Person Image Generation with Holistic Identity Preservation
链接: https://arxiv.org/abs/2607.25622
作者: Yuxuan Xiao,Shanshan Zhang,Jian Yang,Shengcai Liao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Personalized person image generation requires preserving subject identity across both local facial details and broader appearance cues. Existing methods typically emphasize only one level of identity information, leading to an inherent trade-off between facial fidelity and overall appearance consistency. To address this, we first propose a simple dual-branch baseline that unifies global appearance control and local facial control within a shared generation framework. This simple combination of different branches yields promising results, but suffers from instability in practice due to uncoordinated branch contributions. To this end, we propose Dynamic Balancing Scaling (DBS), a fine-tuning strategy for improving face and appearance identity coordination. DBS consists of two components: adaptive temporal gating, which dynamically modulates branch contributions along the denoising trajectory, and region-aware optimization, which improves the coordination of facial, appearance, and global supervision. Together, these designs alleviate persistent face-branch over-dominance and encourage more effective appearance-aware guidance. We also introduce Pexels-100, a benchmark for evaluating holistic identity consistency in personalized person generation. Experiments show that DBS achieves a better trade-off between facial fidelity and appearance consistency than existing open-source baselines, while providing a controllable basic framework for holistic identity modeling.
[CV-33] he LAIA Dataset: Labelled Attention for Intelligent Automobiles
链接: https://arxiv.org/abs/2607.25570
作者: A. Contreras,D. Porres,R. Abad,P. Cano,G. Villalonga,A. M. López,A. Hernández-Sabaté
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
备注: 11 pages, 12 figures, 3 tables. Dataset and supplementary information available from the project website
Abstract:The development of autonomous vehicles (AVs) usually relies heavily on data-driven artificial intelligence (AI) models that require large volumes of sensor data with ground-truth annotations. While modular architectures are widely used, end-to-end driving paradigms offer a promising alternative by directly mapping sensor inputs to control actions. However, their adoption is limited by challenges in interpretability and explainability. To address this, we present LAIA (Labelled Attention for Intelligent Automobiles), a novel synthetic dataset designed to enrich end-to-end driving research with human attention data. Collected using the CARLA simulator in closed-loop environments, LAIA comprises over 15 hours of driving from 44 participants across carefully crafted scenarios designed to evoke natural responses. Each sequence includes RGB images under six weather conditions, semantic and instance segmentation, depth, optical flow, CAN bus signals, and synchronized eye-tracking data. LAIA enables applications including training attention-aware end-to-end AI drivers, predicting driver behavior, developing methods to detect anomalous driver-attention patterns, and improving model explainability. In this work, we use LAIA to compare human attention with the perceptual attention emerging in our end-to-end driving models, thereby providing insight into their behavior.
[CV-34] ReDesign: Recovering Editable Design Structures from Images via Agent ic Decomposition ECCV2026
链接: https://arxiv.org/abs/2607.25565
作者: Jooyeol Yun,Jintae Park,Hyesu Lim,Junha Hyung,Hyungjin Chung,Jaegul Choo
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to ECCV 2026
Abstract:Recovering an editable design file from a raster image is a common and costly bottleneck in modern design workflows, yet remains challenging since editability depends on recovering multi-modal attributes, such as typography, vector geometry, colors, grouping, and layer ordering. We present ReDesign, an agentic framework that grows an editable layer hierarchy by selecting and composing specialized tools across modalities. To keep this long decision process reliable despite imperfect tool outputs, we introduce graceful verification at each expansion, which provides local accept, prune, or retry feedback that prevents error accumulation and avoids large scale reruns. To evaluate editability at scale, we introduce the Figma Edit Replay Benchmark, consisting of 909 raw Figma files and 14,796 controlled edit instructions that replay edits on reconstructed outputs. Across this benchmark and standard reconstruction metrics, ReDesign achieves strong visual fidelity while delivering the highest editability across layout, color, and text edits, outperforming layered decomposition baselines and serial tool use pipelines.
[CV-35] Few-Shot Open-Vocabulary Remote Sensing Segmentation via Textual Inversion
链接: https://arxiv.org/abs/2607.25563
作者: Junhyuk Heo,Junghwan Park
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Open-vocabulary segmentation labels arbitrary categories from a text query without per-class training, yet on remote sensing imagery it underperforms on categories it handles reliably elsewhere. We find that much of this gap traces to the text query rather than to the segmentation model. Because these models are not specialized for overhead imagery, the class name that serves as the query is often a weak address into the vision-language embedding space. We show that a better name repairs part of the gap, while the remaining failures call for an address that the tested natural-language rephrasings do not provide. We recover that address from a few examples through textual inversion on a frozen model, keeping inference text only. On a representative benchmark this raises the mean intersection over union on the affected categories from 3.9 to 39.4, and across eight remote sensing datasets it improves over few-shot methods that instead inject visual prompts at inference.
[CV-36] OrthKD: Extracting Generalized Clinical Knowledge from Heterogeneous Teachers for Lightweight Deployment ECAI2026 ALT IJCAI
链接: https://arxiv.org/abs/2607.25545
作者: Yi Xu,Cheng Chen,Mufan Cao
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to the IJCAI-ECAI 2026 Special Track on AI and Health. 8 pages, 3 figures
Abstract:Deploying diabetic retinopathy (DR) screening models in primary care requires edge-efficient systems that remain accurate, safe, and reliable under domain shift. Multi-teacher knowledge distillation (KD) is a natural compression strategy, but existing approaches largely assume that all teachers provide equally trustworthy supervision. In our setting, this assumption fails: a strong CNN teacher (EfficientNet-B3, 0.876 QWK) and a weaker Transformer teacher (Swin-Base, 0.830 QWK) are complementary, yet the Transformer’s logits can still mislead the student. We therefore propose OrthKD, a selective-trust distillation framework that transfers full supervision from the strong CNN, uses feature-only distillation from the weak ViT, and enforces orthogonality between teacher-specific student projections to encourage complementary rather than redundant evidence. This design preserves local lesion precision, injects global structural context, and improves robustness to distribution shift. On 132,049 retinal images, a 5.4M-parameter MobileNetV3 student reaches 0.885 QWK on EyePACS and improves zero-shot Messidor-2 performance from 0.507 to 0.728 QWK, while also achieving strong referral AUC and calibration. These results show that selectively distilling heterogeneous teachers can enable practical DR screening on resource-constrained devices.
[CV-37] Less is More: Modality-Decoupling for General AIGC Audio-Video Detection ECAI2026 IJCAI
链接: https://arxiv.org/abs/2607.25543
作者: Jielun Peng,Yabin Wang,Yaqi Li,Jincheng Liu,Xiaopeng Hong,Athanasios V. Vasilakos
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: First place in the General AIGC Audio-Video Detection Challenge at the IJCAI-ECAI 2026 DDL 2.0 Workshop
Abstract:Generative AI has rapidly expanded audio-visual forgery beyond human-centric deepfakes into general scenes. Existing AIGC detection methods assume audio-visual content correspondence, identifying forgeries by spotting cross-modal inconsistencies. However, we empirically find that this assumption does not consistently hold in general scenarios. We argue that, for general audio-visual AIGC detection, decision-level fusion is a more robust alternative to feature-level fusion. Therefore, we propose DAV-Det, a decoupled audio-visual AIGC detection system that independently models forensic evidence from each modality. The visual detector leverages multi-granularity representations at global, patch, and segment levels to capture spatial forgery cues, while the audio detector exploits both temporal and spectral irregularities via a gated temporal-spectral dual-branch architecture to model acoustic artifacts. Our method ranks 1st in the General AIGC Audio-Video Detection Challenge of the IJCAI-ECAI 2026 DDL 2.0 Workshop, with a final score of 0.8460. Code is available at this https URL.
[CV-38] Visual prompt engineering for video models
链接: https://arxiv.org/abs/2607.25537
作者: Robert Geirhos,Yuxuan Li,Thaddäus Wiedemer,Neha Kalibhat,Zi Wang,Mani Malek,Oyvind Tafjord,Kevin Swersky,Been Kim,Priyank Jaini
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:In the age of foundation models, a model is only as good as its prompt. For this reason, prompt engineering has become an essential technique for improving language model performance. Since video models are currently becoming foundation models for visual tasks (e.g., visual reasoning), we here ask whether they similarly benefit from visual prompt engineering: automatically modifying the task image to improve model performance. For example, for a visual physics reasoning task (“Where does the ball land, after passing a set of obstacles?”), an abstract sketch-like scene can be turned into a photorealistic version with a simple call to an image editing model. We find that visual prompt engineering, or VIPE for short, improves video reasoning performance across tasks. In fact, for video models, visual prompt engineering can be even more effective than classic text-based prompt engineering or test-time scaling. Ultimately, just as text-based prompt engineering systematically improves language model performance, visual prompt engineering can serve as a simple, compute-efficient approach to elicit better visual reasoning performance from video models. Example videos on our project page at this https URL.
[CV-39] Multi-Scale Structural Features for Continual Comprehensible Visual Recognition in a Developmental Learning Framework
链接: https://arxiv.org/abs/2607.25531
作者: Zeki Doruk Erden
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Contemporary machine learning struggles to learn continually, reuse prior knowledge, and expose a comprehensible internal structure. A recently proposed developmental, gradient-free learning framework addresses these limitations by learning a discrete, topological model of its inputs through local variation and selection, yielding an inherent continual-learning guarantee: new observations refine existing structure without overwriting past knowledge, and without replay buffers or predefined task boundaries. Its extension to visual inputs demonstrated this principle on shape recognition, but relied on a feature representation of limited expressivity that capped recognition accuracy. We introduce a new visual feature representation that encodes shape structure across multiple scales, capturing edge and contour features together with their spatial relations, and integrate it with the network-refinement learning process; we further improve the learning dynamics and the read-out used to predict from the learned model. The study targets two-dimensional shape, with class-incremental MNIST as a controlled, interpretable benchmark in which continual-learning behavior can be measured directly. Our approach substantially increases accuracy over the prior representation, matching or exceeding replay- and regularisation-based baselines at comparable storage while storing no past data, and preserves the framework’s defining behavior: earlier-learned classes are retained as new ones are introduced, with no destructive adaptation, and the learned representations remain human-interpretable. What separates the methods is retention: the baselines surrender most of a just-trained class within its own cycle and relearn it afterwards, which ours does not. The significance lies in the manner of learning. The system integrates information one sample at a time while provably preserving its responses to…
[CV-40] Argus-Unified: Towards A Compact and Economical Unified Model for Image Understanding and Generation
链接: https://arxiv.org/abs/2607.25527
作者: Weiming Zhuang,Jiabo Huang,Jingtao Li,Zhizhong Li,Chen Chen,Sina Sajadmanesh,Lingjuan Lyu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Unifying visual understanding and generation in one model holds immense promise, but remains challenging and expensive due to heavy compute and data demands and conflicts between the visual features needed for these two capabilities. To address these challenges, we present Argus-Unified, a compact, effective and unified multimodal model built with low demand on computation and data. Instead of aligning modalities from scratch, Argus-Unified effectively leverages pretrained vision-language models (VLMs) that provide strong multimodal priors. Specifically, we introduce hybrid visual tokens that preserve continuous tokens for understanding while learning discrete tokens for generation from a frozen unified vision encoder. Our training pipeline includes two stages: the first stage learns a quantizer and image decoder on top of the frozen vision encoder, the second stage trains the LLM initialized from a pretrained VLM for the unified multimodal modeling. Using by far the least amount of data (15.6M) and the lowest cost (~ 2,000), we demonstrate that unified multimodal models can be trained economically while achieving strong performance in both understanding and generation. Notably, our model attains state-of-the-art multimodal understanding on GQA, POPE, and VQAv2, and competitive generation quality compared to models with dedicated vision encoders (e.g., Janus, Janus-Pro), all at ~10x lower cost and with ~5x less data. We envision Argus-Unified as a useful baseline that lowers the development barrier for unified models.
[CV-41] ReLATE: Reliability-Guided Evidence Fusion for Robust UAV–Satellite cross-view Geo-Localization
链接: https://arxiv.org/abs/2607.25524
作者: Haochen Jiang,Jialei Pan,Yuzhe Sun,Zhe Dong,Lecheng Ren,Yanfeng Gu,Tianzhu Liu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Unmanned aerial vehicle (UAV)-satellite cross-view geo-localization matches UAV images against satellite imagery and has achieved impressive accuracy on clean (non-degraded) image benchmarks. In real-world flights, however, UAV observations are frequently affected by adverse weather, illumination changes, platform motion, sensor noise, and compression, while the robustness of existing methods under such degradations remains largely unexamined. In this paper, we present UAVSat-Deg, a large-scale robustness benchmark for degraded UAV-satellite geo-localization, comprising University-1652-Deg and SUES-200-Deg. UAVSat-Deg covers 27 corruption types, including 19 core and 8 compound corruptions, at three severity levels, supports bidirectional drone-to-satellite and satellite-to-drone retrieval as well as multi-height UAV acquisition, and contains more than 11.7 million pre-generated corrupted test images. Benchmarking representative methods under this protocol reveals substantial robustness gaps, particularly under severe and compound corruptions. To address this problem, we propose ReLATE, a Reliable Evidence Learning framework with Adaptive Token Evidence Regulation, which realizes reliability-adaptive feature fusion during descriptor construction. ReLATE estimates a structure-smoothed reliability field over visual tokens, aggregates trustworthy local evidence, and adaptively integrates it into query-derived representations; the regulated query representations are then combined with the CLS-token and GeM-pooled branches to form the final cross-view descriptor. Across both test sets and retrieval directions, ReLATE achieves the best average corrupted-test performance among the compared methods while maintaining competitive accuracy on clean images. The code and dataset will be available at this https URL.
[CV-42] I2VShield: An Efficient Proactive Defense Framework against DiT-based Image-to-Video Models
链接: https://arxiv.org/abs/2607.25522
作者: Yimao Guo,Zuomin Qu,Wei Lu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:The rapid advancement of video generation models has led to the increasing misuse of image-to-video (I2V) models. Although substantial progress has been made in detecting AI-generated videos, proactive defenses against I2V models remain underexplored. In particular, current proactive defenses against I2V models predominantly rely on gradient-based adversarial attacks, which require defenders to possess GPUs with substantial memory resources (VRAM) to generate adversarial examples. To address this issue, we propose I2VShield, a privacy protection method based on generative adversarial attacks tailored to Diffusion Transformer (DiT)-based I2V models. The proposed method primarily consists of two components: (1) a text-adaptive perturbation generation framework integrating adversarial learning to mitigate computational overhead while maintaining visual imperceptibility; and (2) an untargeted Multimodal Attention Disruption (MAD) attack that exploits the inherent vulnerabilities of DiT-based I2V models, maximizing the deviation of the internal attention features from their clean states. Extensive experiments demonstrate that our approach achieves highly competitive protection performance across various datasets and mainstream DiT-based I2V models, particularly in disrupting spatiotemporal coherence, while substantially reducing computational costs.
[CV-43] Group Equivariant Diffusion for Anomaly Detection in Computational Cytology MICCAI2026
链接: https://arxiv.org/abs/2607.25503
作者: Swarnadip Chatterjee,Ssharvien Kumar Sivakumar,Anirban Mukhopadhyay
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 11 pages, 2 figures, 1 table, 1 algorithm. Accepted for publication in MICCAI 2026
Abstract:Computational cytology on whole-slide images is challenging because malignant cells are rare, heterogeneous, and annotated slides are scarce. Anomaly detection frameworks can be trained on normal slide-negative patches and then applied at test time to flag abnormal patches in held-out slides. Most unsupervised anomaly detection approaches including generative ones (GAN-based and diffusion-based), are tuned to organ-level imaging and require large curated datasets. In cytology the signal is cell-centric: rotating or flipping a single-cell patch does not change its diagnostic class, yet standard diffusion models treat transformed views as distinct inputs, leading to transformation-dependent reconstructions and unstable anomaly scores. We propose a D4-equivariant diffusion framework that enforces rotation and reflection symmetry both architecturally, via a D4-equivariant U-Net, and at inference, via equivariant noise coupling and (optionally) frame averaging. This alignment with biological invariance yields transformation-consistent pseudo-healthy reconstructions and more stable anomaly ranking under symmetry. On two publicly available cytology datasets of bone marrow and peripheral blood smears, our D4-equivariant diffusion models achieve higher AUC and retrieve more abnormal cells in the top K predictions than non-equivariant generative baselines, a deep one-class, and a multiple instance learning based method, while substantially reducing score variance across rotations and flips. Code is available at this https URL.
[CV-44] Beyond Counts: A Distributional Robustness Margin For Pathology Foundation Models
链接: https://arxiv.org/abs/2607.25497
作者: Clément Grisi,Jeroen van der Laak,Geert Litjens
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Preprint
Abstract:Pathology foundation models are approaching clinical deployment, yet remain vulnerable to systematic non-biological variation across centres. Differences in tissue preparation, staining and scanning are strongly encoded in their representations, enabling shortcut learning and weakening generalisation across cohorts and institutions. The Robustness Index (RI) quantifies whether local representation geometry is dominated by biology or by non-biological variation, but its count-based formulation discards distance information. We show that adding distance weights changes little because the deeper limitation lies in RI’s pooled, fixed-neighbourhood design, which obscures sample-level heterogeneity and effectively evaluates only a model-dependent subset of samples. We introduce the Cross-confounder Robustness Margin (CRoMa), a sample-resolved measure that directly compares distances to cross-confounder biological matches and same-confounder biological distractors. CRoMa recasts robustness as a cohort-wide margin distribution rather than a single pooled score. We evaluated frozen representations from 20 tile-level encoders across three benchmarks and 4 slide-level encoders on a fourth. Rankings by median CRoMa were broadly consistent across datasets, while the underlying distributions revealed substantial within-model heterogeneity. Every tile encoder retained a confounder-dominated lower tail, whose prevalence and severity varied markedly across models. These distinct robustness profiles frame model selection as a Pareto trade-off between typical and lower-tail robustness. Higher CRoMa was also associated with smaller shortcut-induced performance drops after supervised adaptation. By turning representation geometry into a distributional robustness readout that anticipates downstream shortcut susceptibility, CRoMa provides a principled basis for robustness assessment and model selection.
[CV-45] Agent ic AI in medicine: architectures applications evaluation and challenges for clinical translation
链接: https://arxiv.org/abs/2607.25489
作者: Zheng Tong,Yang Liu,Wanshu Fan,Jing Qin,Zhongbin Han,Haifan Gong,Congyu Liao,Xiaofeng Liu,Cong Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Review article, 6 figures, 2 tables. 42 pages
Abstract:Large language models and multimodal foundation models are enabling medical artificial intelligence (AI) systems to move beyond isolated prediction and undertake multistep clinical tasks that require planning, tool use, memory, iterative correction, and coordination among specialized agents. However, the scope of agentic AI in medicine remains unsettled, and current evaluation practices are not yet aligned with the requirements of clinical use. We conducted a scoping review with systematic evidence mapping across five electronic sources, screened 1,649 exportable records, and provisionally included 557 unique studies that met predefined criteria for goal-directed task execution, tool use, interaction with external resources, feedback-based refinement, or multi-agent collaboration. The included studies describe single agents that use external tools, workflows supported by retrieval and external knowledge, multimodal agents, and multi-agent systems applied to medical question answering, image interpretation, electronic health record analysis, drug safety, and clinical trial prediction. The evidence base remains dominated by public benchmarks, simulated settings, retrospective datasets, and small-scale expert evaluation. Process reliability, evidence traceability, uncertainty, safety, workflow impact, and external validity are evaluated less consistently. Clinical translation will depend on clearer definitions, reproducible evaluation, auditable oversight, interoperable system design, and prospective validation in real-world clinical workflows.
[CV-46] CoTinyVLA: Chain-of-Thought Distillation for a Sub-Billion-Parameter Vision-Language-Action Model
链接: https://arxiv.org/abs/2607.25487
作者: Minhyeok Lee,Chiyoung Kim,Chanhoe Gu,Seongrok Kim,Sanghyuk Roy Choi,Donghwan Hwang,Donghun Ryu,Seokhyun Kim
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注: 22 pages, 2 figures, 20 tables. Code at this https URL
Abstract:Vision-Language-Action (VLA) models translate natural-language commands into robot action sequences, but leading systems on the LIBERO-Plus robustness benchmark use three- to seven-billion-parameter backbones whose memory demands can exceed embedded robotic budgets. We present CoTinyVLA, a 0.9B-parameter action model on a Qwen3.5-0.8B backbone that obtains that robustness by structuring supervision instead of enlarging the model. Three components target different axes of the problem: dual-view temporal input of 16 history frames per step with textual camera and time markers; hierarchical chain-of-thought (CoT) distillation from a 35B teacher into an episode-level Plan and a chunk-level Think span over task phase, gripper state and next subaction; and paraphrase augmentation expanding 40 base commands into 800 variants. On LIBERO-Plus, spanning 10,030 perturbed tasks across seven perturbation dimensions, CoTinyVLA reaches 90.8% on Spatial, 87.3% on Object, 86.6% on Goal and 80.7% on Long, leading the strongest 7B baseline on all four suites by 4.7, 2.8, 15.9 and 3.0 points, with every margin interval excluding zero. The gains concentrate on the hardest axes of the benchmark: across the eleven published baselines none exceeds 53.2% on Robot Initial States in any suite, whereas CoTinyVLA reaches 73.6% on Goal against 39.9% for the strongest baseline. Ablations show the three components to be separable by perturbation axis, and at a matched image budget how frames are divided between the two cameras and across time accounts for 8.6 points on its own. Closed-loop inference peaks at 2.25 GiB of allocated GPU memory, and paired interventions show the episode Plan to be load-bearing: replacing it with an empty or contradictory span costs 40 to 45 points of success. Structured supervision thus lets a 0.9B backbone exceed all of them. Code: this https URL
[CV-47] Safety-Aware Cascaded Inference for Crop Damage Assessment with Controlled Error Trade-offs
链接: https://arxiv.org/abs/2607.25468
作者: José Thiéry Messigbédé Hagbe,Gani Kawsar Gounou,Songbian Karim Zimé
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 44 pages, 7 figures, 12 tables. Submitted to Precision Agriculture, July 2026
Abstract:In picture-based agricultural insurance for smallholder farmers, missed damage detections carry substantially higher cost than false alarms: a farmer who sustained real losses receives no payout, while unnecessary expert review is operationally costly but reversible. Standard multi-class classifiers optimize global accuracy but provide no mechanism to operationalize or control this asymmetric cost structure at inference time. We propose CascadeCropNet, a two-stage cascade architecture calibrated to satisfy a target recall constraint (Rec-Damaged = 0.95) through threshold selection. A lightweight Sentinel model performs binary health triage; samples exceeding a calibrated damage probability threshold tau are escalated to a specialist Expert model for fine-grained diagnosis. This design provides explicit, deployment-time control over the safety-efficiency trade-off without retraining. Evaluated on the Eyes on the Ground dataset (23,804 images from Kenyan smallholder maize farms), the cascade achieves Rec-Damaged = 0.974 at tau = 0.5, reducing missed damage cases by up to 54% relative to a flat baseline. Under evaluation alignment, the representational gap reduces to +0.008 F1-macro, confirming the contribution is architectural rather than representational. Under input degradation, the system prioritizes escalation over confident misclassification, reflecting error containment through architectural isolation rather than intrinsic model robustness. These results demonstrate that cascade architectures can operationalize safety-oriented decision constraints through calibrated routing in settings where reliability matters more than aggregate accuracy. These properties depend on threshold calibration and deployment conditions and do not constitute guarantees under arbitrary distribution shift.
[CV-48] Seen Said or Forgotten? A Causal Audit of Visual KV Memory Across Dialog Turns
链接: https://arxiv.org/abs/2607.25467
作者: Hong Chen,Kang Chen,Yuxuan Fan,Bo Wang,Yubo Gao,Yuanlin Chu,Xuming Hu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Stateful multimodal assistants encode an image once but may answer questions about it many turns later. Attention-guided visual-KV eviction assumes that evidence irrelevant now will remain dispensable, although future questions are unknown. We ask when a visual fact is actually safe to forget and introduce the Causal Visual Memory Audit (CVMA), a paired single-prefill framework that tests what later answers lose when a visual region, the whole image, or prior assistant text becomes unavailable. On VisDial and ConvBench, current attention can rank future-useful regions worse than random even though a diagnostic marginal-utility control shows substantial selection headroom. Aggregate scores hide this failure when later turns do not need vision; controlled and stock-generated histories reveal a second escape route, in which assistant-text KV replaces image KV for facts already stated but not reliably for unstated facts. In the tested stacks, safe forgetting is supported by low future visual dependence or fact-specific verbalization—not by low current attention.
[CV-49] DensFiLM: Density-Conditioned Video Saliency for Crowd Scenes
链接: https://arxiv.org/abs/2607.25465
作者: Anis Ur Rahman
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 12 pages, 2 figures, 3 tables
Abstract:Video saliency models typically apply a single fixation strategy across crowd scenes, despite systematic changes in attention with crowd density. Sparse scenes encourage tracking individuals, whereas dense scenes shift attention toward collective motion and scene-level landmarks. We introduce DensFiLM, a density-conditioned video saliency model that inserts a lightweight Feature-wise Linear Modulation layer at the bottleneck of a Video Swin Transformer. A learned density embedding produces channel-wise scale and shift parameters, allowing the decoder to reconstruct saliency from features selected for each density regime. The module adds only ~100K parameters and can use either CrowdFix density labels or the model’s own density prediction. On CrowdFix, DensFiLM achieves mean NSS 1.434 and CC 0.517 over four seeds, improving over ACLNet by 14.7% and 14.9%, respectively, while predicted-density conditioning matches oracle-label performance. Ablations show that explicit RAFT optical flow and larger temporal and social-force extensions provide no further improvement in this setting. In a centre-prior-subtraction diagnostic, density conditioning yields an NSS gain of 0.462 over the unconditioned backbone, compared with 0.124 under standard evaluation. These results show that lightweight bottleneck conditioning provides a more effective inductive bias than increasing model capacity for crowd-video saliency. Our code is available at this https URL.
[CV-50] Reading Legends on Ancient Coins: An Object Detection Approach for Character Recognition on a Novel Roman Republican Dataset
链接: https://arxiv.org/abs/2607.25455
作者: Hafeez Anwar
类目: Computer Vision and Pattern Recognition (cs.CV); Emerging Technologies (cs.ET)
备注:
Abstract:When it comes to the proper classification of ancient coins with respect to their time and issuer, the textual inscriptions on these coins, also known as legends, are of paramount importance. These legends consist of alphabets or characters still used in English. This paper addresses image based character recognition on ancient Roman Republican coins via a deep learning based object detection strategy. However, legends on these coins pose high variation due to non-uniform placement, primitive inscription techniques, and wear and tear. Additional challenges include inconsistent imaging conditions such as illumination, orientation, and scale. To accommodate these, we gathered a novel large-scale dataset of 5,654 Roman Republican coin images, manually annotated with 21 character labels, totaling 38,808 annotations. For recognition, we use You Only Look Once (YOLO) variants: YOLOv3, v4, v5, v7, and v8. YOLOv7-Large achieves the best mAP50 of 90.4%, followed by YOLOv7-Extended and YOLOv7-xl with 90.2% and 90.1%, respectively.
[CV-51] Bi-Level Collaborative Learning for Few-Shot Scribble-Supervised Medical Image Segmentation
链接: https://arxiv.org/abs/2607.25432
作者: Xiang-Xiang Su,Yufan Ye,Yihang Zheng,Min Gan,Guang-Yong Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Scribble annotations offer an efficient alternative to costly pixel-wise labeling for medical image segmentation, yet in real clinical scenarios, scribble-annotated samples are often still limited, imposing the dual challenges of sparse supervision and annotated sample scarcity. These compounded constraints severely deprive models of the structural evidence needed for complete region recovery and precise boundary delineation. To break this bottleneck, we propose a bi-level collaborative learning framework for few-shot scribble-supervised medical image segmentation. Specifically, an upper-level learnable superpixel model is introduced to provide region-structural priors for lower-level segmentation, while superpixel-based region-wise pseudo-label propagation and a spatial-prior-guided filtering strategy are performed to generate reliable dense pseudo-labels for segmentation learning. Meanwhile, the anatomical semantics learned by the lower-level segmentation model under the guidance of the current superpixels are fed back to the upper level, further driving it to learn region-structural representations better aligned with the segmentation task. Through bidirectional interaction and collaborative learning between the upper and lower levels, the proposed framework significantly outperforms existing state-of-the-art scribble-supervised methods on the ACDC and Prostate datasets under the few-shot scribble-supervised setting.
[CV-52] ANFI: Rethinking Neighbor Feature Interaction in Person Re-ID ECCV2026
链接: https://arxiv.org/abs/2607.25407
作者: Xulin Li,Yan Lu,Bin Liu,Jiaze Li,Qinhong Yang,Tao Gong,Qi Chu,Nenghai Yu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: accepted by ECCV2026
Abstract:In person re-identification, neighbor-based methods have achieved significant success by interacting with neighbor samples to obtain more robust representations. However, existing methods rely only on affinity relations, causing their success to depend heavily on the reliability of selected neighbors. We find that affinity-only interaction often fails in challenging scenarios due to the inevitable presence of noisy neighbors. To enable effective interactions under noisy neighborhoods, we revisit neighbor-based methods under distinct reliability conditions and propose a novel Adaptive Neighbor Feature Interaction (ANFI) method. The core idea of ANFI is to account for negative effects from noisy neighbors, allowing samples to remain distinguishable from false positive neighbors. Unlike existing methods, ANFI models not only affinity relations but also discrepancy relations, and employs sample-wise adaptive weighting for these two types of relations. Given that capturing negative effects from noisy neighbors differs significantly from traditional relation learning, we derive discrepancy relations from a new neighborhood similarity, which provides more information than pairwise similarity. In addition, we propose Noisy Relation Supervision (NRS) to train ANFI, gradually injecting robustness to noisy relations into the model. Extensive experiments conducted under standard, cross-modal, and cross-domain settings, including comparisons with neighbor-based methods and re-ranking methods, demonstrate the superiority of our method across various neighbor distributions.
[CV-53] owards Reliable Stain Transfer: An Iterative Data-Model Co-Optimization Framework Based on Multimodal Expert-Guided Assessment
链接: https://arxiv.org/abs/2607.25393
作者: Siyuan Xu,Yan Wang,Haofei Song,Lili Gao,Jiansheng Wang,Qing Zhang,Dan Huang,Boxiang Yun,Hongkai Xiong,Qingli Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 10 pages, accepted by ACMMM2026 Main Track
Abstract:Histopathological examination primarily relies on hematoxylin and eosin (HE) and immunohistochemistry (IHC) staining. Although IHC provides critical molecular information, it is costly and requires specialized expertise. Stain transfer provides an efficient alternative by computationally generating IHC from HE images, but remains challenged by unified and interpretable modeling for heterogeneous biomarkers under pixel-unaligned supervision. We propose DMCoStain, a novel Data-Model Co-optimization framework for Stain transfer. It iteratively co-refines training data and model capability, improving staining accuracy and interpretability in both pathological and structural consistency. To refine training data in a clinically meaningful manner, it incorporates the Multimodal Expert-Guided Finer Selection (MEGFS) strategy, built upon a pioneering IHC-positive-expression (IPE) vision-language model (VLM) that emulates pathologist reasoning. To support MEGFS, we construct ImmunoInstruction, the first large-scale IPE instruction-following dataset with 150K VQA samples. Extensive experiments on multiple tissues and biomarkers demonstrate that DMCoStain achieves state-of-the-art (SOTA) accuracy. This paradigm offers strong practical value, and MEGFS also functions as a specialized evaluation tool for future model development. Dataset, code, and more details are in this https URL.
[CV-54] RDVSv2: A Large-scale Benchmark for RGB-D Video Salient Object Detection
链接: https://arxiv.org/abs/2607.25392
作者: Tianyu Li,Jiahao He,Keren Fu,Qijun Zhao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to ACMMM 2026
Abstract:We introduce RDVSv2, a large-scale benchmark for RGB-D video salient object detection (RGB-D VSOD) with dense frame-level annotations. Existing datasets in this emerging field are often limited in scale and annotation quality, while also relying on less geometry-consistent depth cues. To address these limitations, RDVSv2 is built from publicly accessible stereoscopic online videos and contains 249 video sequences with 29,077 annotated frames. It includes depth maps derived from stereoscopic videos, together with frame-wise salient object masks annotated with eye-tracking guidance. Compared with existing datasets, RDVSv2 is much larger in scale and covers more diverse and challenging scenarios. In addition, we establish a strong baseline for RGB-D VSOD based on Segment Anything Model 2 (SAM2). Specifically, we employ a parameter-efficient fine-tuning (PEFT) strategy to adapt the SAM2 encoder to jointly encode RGB, depth, and optical flow cues. Extensive experiments show that RDVSv2 is substantially more challenging for existing RGB-D VSOD methods. Meanwhile, the proposed baseline achieves state-of-the-art results on RDVSv2 and existing RGB-D VSOD benchmarks. We hope that RDVSv2 and the provided baseline will serve as useful resources for future research on RGB-D VSOD and related multi-modal video understanding tasks. Our dataset and code will be available at this https URL.
[CV-55] Noise-Free One-Step LoRA for Task-Driven Image Restoration with Diffusion Priors
链接: https://arxiv.org/abs/2607.25390
作者: Jaeha Kim,Kyoung Mu Lee
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Code: this https URL
Abstract:Degraded images not only reduce visual quality but also impair downstream high-level vision tasks. Task-driven image restoration (TDIR) addresses this issue by jointly optimizing restoration quality and task performance. Recent works show that pretrained diffusion priors benefit TDIR, yet diffusion-based restoration is inherently stochastic, as the sampling process depends on a random noise term, which can undermine task consistency. In this paper, we show that a deterministic, noise-free one-step forward pass with pretrained diffusion priors can substantially improve TDIR, but the benefit critically depends on the adaptation module: LoRA yields consistent gains, whereas ControlNet-style conditioning does not. This enables one-step forwarding that surpasses conventional multi-step diffusion TDIR baselines. Furthermore, we introduce a task-preserving GAN training strategy that improves perceptual quality without sacrificing task performance. Extensive experiments on classification, segmentation, and detection demonstrate consistent gains over prior TDIR methods, and we further validate generalization on real-world degraded images and OCR.
[CV-56] HOME: Robust Hough-space Matching Method for Structured and Textureless Videos
链接: https://arxiv.org/abs/2607.25389
作者: Masaki Satoh
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注: 7 pages, 5 figures
Abstract:Visual front-ends for robotic localization typically rely on point-based features such as Oriented FAST and Rotated BRIEF (ORB), which frequently fail in structured environments dominated by strong linear structures or textureless surfaces. While line-based Simultaneous Localization and Mapping (SLAM) systems mitigate this by utilizing line segments, conventional line extraction and description algorithms are computationally prohibitive for real-time edge robotics. To address this fundamental bottleneck, we propose HOME (Hough-space One-dimensional Matching of Extrema), an ultra-lightweight, training-free feature matching framework. HOME transforms images into Hough space, mapping global linear structures to stable local extrema, which serve as keypoints, thereby reformulating complex line matching into highly efficient one-dimensional point matching. The proposed 1D radial descriptor mathematically guarantees rotational and translational invariance without the overhead of explicit orientation estimation. As a proof of concept to validate the matching accuracy and efficiency of HOME, this paper focuses on homography estimation. Extensive evaluations demonstrate that HOME achieves robust registration in challenging scenarios where point-based methods fail, operating at a much faster speed than existing line-based methods. Extending this robust matching engine to full 3D pose estimation remains a highly promising future direction.
[CV-57] Gaussian Volumetric Representation for Efficient Shear-Warp Visualization
链接: https://arxiv.org/abs/2607.25377
作者: Mayuri Mathur(1),Ojaswa Sharma(1) ((1) Indraprastha Institute of Information Technology Delhi)
类目: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR)
备注:
Abstract:Medical image visualization requires volumetric rendering algorithms that preserve anatomical fidelity while maintaining high rendering speeds. To address the high computational cost of large volumetric datasets, we propose a Gaussian-based volumetric representation for efficient visualization of dense medical volumes without compromising structural and radiometric details. We optimize the proposed representation using Monte Carlo volumetric estimation, which enables training on a highly sparse subset of voxels while maintaining consistency with the dense volumetric objective. In addition, we introduce a curriculum learning strategy that progressively incorporates structured slice-based sampling during training. Sparse voxel samples provide an early global coverage of the volume, while slice samples capture spatially correlated regions that aid geometric structure and texture continuity. This combination enables the Gaussian representation to learn anatomical details of various structures and corresponding textures from sparse supervision while significantly reducing the computational cost associated with dense voxel processing. The learned representation supports slice-based rendering methods such as shear-warp volume rendering, enabling efficient visualization of multimodal medical datasets including MRI and Cryosection volumes while preserving anatomical structures. Using sparse supervision, our method achieves up to 43.86 FPS rendering with a compression ratio of 11.31:1.
[CV-58] Hyperspectral Intrinsic Decomposition: Joint Recovery of Reflectance and Photometric Components for Non-Lambertian Scenes
链接: https://arxiv.org/abs/2607.25371
作者: Hao Ye,Zhan Shi,Chenglong Huang,Tao Lv,Mingjie Ji,Qiu Shen,Xun Cao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Hyperspectral intrinsic decomposition (HID) aims to disentangle material-related spectral properties and photometric effects in hyperspectral images (HSIs), which is essential for understanding real-world imaging processes and benefits a variety of downstream applications. Most existing HID studies have been developed under Lambertian or near-Lambertian assumptions. The few prior non-Lambertian efforts rely on simplified specular assumptions insufficient to handle diverse real-world specularity, and typically require auxiliary inputs or recover only a subset of the coupled reflectance and photometric components, hindering complete and blind decomposition. In this paper, we revisit the dichromatic reflection model (DRM) and develop a unified inversion paradigm that reformulates the recovery of four coupled reflectance and photometric components as the estimation of two spectral–spatial target variables. Building on this reformulation, we propose a dual-scale decomposition scheme to handle non-Lambertian effects with distinct spatial characteristics. At the global scale, photometrically invariant descriptors serve as edge priors for high-fidelity intrinsic boundary preservation; at the local scale, specularity-guided attention directs refinement with emphasis on specularity-dominated regions, including those affected by clipping distortion. To facilitate future research, we establish CITE, the first public real-world HID dataset for non-Lambertian objects, and develop a Physically-faithful Intrinsic Set Generator (PISG) for controllable data synthesis. Extensive ablation studies and experiments on the CITE and additional HSIs demonstrate the effectiveness of our method and its robustness across diverse scenes.
[CV-59] Leak-Free Cross-Validated Stacking with Per-Architecture Calibration for Sand-Boil Segmentation in Earthen Levees
链接: https://arxiv.org/abs/2607.25367
作者: Padam Jung Thapa,Anav Katwal,Ayon Dey,Abdullah Bin Naeem,Steve Sloan,Kendall Niles,Md Tamjidul Hoque
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 24 pages, 17 figures, 14 tables. Supported by the U.S. Army Corps of Engineers under Contract No. W912HZ-23-2-0004
Abstract:Sand boils, points where water seeping beneath an earthen levee re-emerges at the surface, are early warnings of internal erosion, and deep segmentation networks are increasingly used to find them in inspection photographs. Annotated examples are scarce, and two common ways of working around that scarcity quietly inflate reported accuracy: tuning ensemble weights on the same images later used to score them, and training on synthetic images derived from the very photographs held out for testing. We present a sand-boil segmentation framework that closes both loopholes. Every synthetic image carries a pointer to its real parent, and a per-fold filter excludes any image whose parent is held out; five encoder-decoder backbones are trained under five-fold cross-validation, calibrated by one temperature scalar each, and combined by a per-pixel meta-learner fitted only on out-of-fold predictions. On the held-out test set the proposed Updated SandBoilNet reaches an intersection-over-union of 0.707 over three seeds, against 0.608 for the published original re-evaluated on the same split. Under the stacking protocol the calibrated stack reaches 0.681 against 0.694 for the strongest fold-averaged member, so it does not improve on the best single model; eight meta-learner families reproduce that outcome, which we trace to a mean pairwise error correlation of 0.894 among members. A synthetic pool filtered for label fidelity lifts the champion to 0.718 over three seeds against a 0.707 control. We also introduce a mask-conditioned synthesis route that makes the conditioning mask the label by construction, giving labelled training images at zero annotation cost.
[CV-60] PanoLess: Environment Reconstruction from Partial Reflective Views ECCV2026
链接: https://arxiv.org/abs/2607.25362
作者: Ahitagni Das,Ashok Veeraraghavan,Vivek Boominathan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: ECCV 2026. Main paper with supplementary material
Abstract:Reflections from shiny objects and glass facades naturally extend the field of view of a camera, capturing the surrounding environment without the need to pan the camera or acquire a full panorama. We propose PanoLess, a Gaussian-splat-based framework that reconstructs the surrounding environment as a distant illumination map from images captured on only one side of a reflective surface. PanoLess leverages surface-aligned 2D Gaussian splats with deferred shading to recover accurate per-pixel normals and reflection cues, which are fused into a neural cubemap representation of the environment. In addition, PanoLess produces a visibility map that explicitly denotes which regions of the environment are supported by the partial reflective observations. Unlike existing inverse-rendering and reflection-aware Gaussian-splatting approaches, which typically require full 360-degree coverage and struggle under incomplete views, PanoLess enables consistent, physically grounded illumination estimation from partial-view input. We show that PanoLess achieves high-fidelity and geometrically consistent environment reconstruction, outperforming reflection-aware baselines on a new custom synthetic benchmark and publicly available datasets, and demonstrating generalization to real-world reflective captures.
[CV-61] Balanced Soft mixture-of-expert model for Glaucoma Detection
链接: https://arxiv.org/abs/2607.25324
作者: Sai Venkatesh Chilukoti,Krishna Rauniyar,Min Shi,Xiali Hei
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 18 pages, 5 figures
Abstract:Glaucoma is a group of eye diseases that damage the optic nerve, often caused by elevated intraocular pressure. It is a leading cause of irreversible vision loss and is typically developed slowly and painlessly, making it difficult to notice until significant damage has occurred. Therefore, early detection is crucial to prevent or slow the progression of vision loss. In recent years, deep learning based uni-modal models have improved the accuracy and efficiency of glaucoma detection, empowering doctors with tools for earlier diagnosis, better monitoring, and timely treatment. Building on this, multi-modal models have emerged, leveraging the strengths of different imaging modalities to learn richer and more robust representations, further enhancing glaucoma detection accuracy. However, multi-modal learning faces challenges such as imbalanced and under-optimized uni-modal representations due to joint learning objectives. To address this, we propose a balanced soft mixture-experts model with three experts and load balancing loss. The performance is measured by AUC, our proposed method surpasses the performance of all uni-modal baselines, conventional multi-modal models, and current stateof- the-art balanced multi-modal models. The proposed model can be generalized to other disease detections such as diabetic retinopathy.
[CV-62] Beyond Background Bias: Saliency-Driven Prototype Alignment for Dataset Distillation
链接: https://arxiv.org/abs/2607.25318
作者: Yawen Zou,Wenqi Cai,Guang Li,Ling Xiao,Chunzhi Gu,Chao Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Dataset distillation aims to synthesize compact datasets that can approximate the performance of full-data training while significantly reducing computational and storage costs. However, diffusion-based distillation methods often struggle to preserve structural coherence and generalization, especially in visually complex domains. This issue often stems from latent prototypes that are weakly aligned with class-discriminative regions and contaminated by irrelevant background, thereby degrading generation quality and generalization. To address this limitation, we propose a saliency-driven distillation framework that constructs class-discriminative latent prototypes to enhance representativeness and generalization. The framework proceeds in two stages: (1) ensemble Grad-CAM saliency is used to construct prototypes emphasizing high-confidence regions, and (2) hard prototype refinement is then applied to construct challenging yet class-consistent prototypes, thereby enhancing discriminability and diversity. Importantly, the diffusion backbones (e.g., LDM and DiT) remain frozen; only lightweight classifiers used for saliency extraction are trained. Extensive experiments across multiple benchmarks demonstrate consistent performance improvements over strong baselines. Code will be released.
[CV-63] Sense it with your eyes: Sensation Generation and Understanding for Advertisements
链接: https://arxiv.org/abs/2607.25314
作者: Aysan Aghazadeh,Sina Malakouti,Adriana Kovashka
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Sensory advertising evokes human senses through visual cues, enabling audiences to mentally simulate experiences and increasing persuasive impact. Despite the recent increase in using AI in generating and understanding creative and persuasive content, how advertisements visually evoke sensations remains largely unexplored. In this work, we introduce the first study of understanding, evaluating, and generating sensory ads. We introduce the Sensory Ad dataset, and define sensation classification tasks (SenseClass) to benchmark LLMs and MLLMs. We further propose SenseScore, an automated evaluation metric for sensation evocation achieving strong agreement with human judgments. Finally, we introduce the Sensory Ad Generation (SenseGen) task and propose SAGA, a multi-agent framework that improves message image alignment, sensory evocation, and persuasion. Our work establishes a foundation for sensory-aware visual persuasion.
[CV-64] Human-in-the-Loop Signature Bootstrapping for UAV Hyperspectral PFM-1 Mine Detection
链接: https://arxiv.org/abs/2607.25310
作者: Sagar Lekhak,Prasanna Reddy Pulakurthi,Emmett J. Ientilucci
类目: Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
备注: Manuscript submitted for peer review
Abstract:Hyperspectral imaging (HSI) is useful for material discrimination, but operational mine screening also depends on how many false alarms must be inspected before targets are found. This paper studies PFM-1 landmine detection in unmanned aerial vehicle (UAV) visible and near-infrared (VNIR) HSI using spectral angle mapper (SAM), matched filter (MF), adaptive coherence estimator (ACE), and constrained energy minimization (CEM). We compare a ground-measured SVC signature, a fully informed in-scene core-pixel signature, and a simulated human-in-the-loop signature bootstrap. Besides receiver operating characteristic area under the curve and average precision, we report target-discovery curves and spatial candidate-review counts. Full-review bootstrapping reaches the fully informed in-scene signature case after all seven target regions are verified, but the required inspection effort varies strongly: ACE confirms all regions in two rounds and nine candidate inspections, whereas the SAM variants need thousands of candidate reviews for their final target locations. Code is available at this https URL.
[CV-65] MEDit-Bench: A Dataset for Evaluating Message-Driven Narrative Video Editing
链接: https://arxiv.org/abs/2607.25300
作者: Katsuya Ogata,Zongshang Pang,Mayu Otani,Yuta Nakashima
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Video editing is fundamentally message-driven: even from the same source footage, the selected shots change depending on the narrative the editor wishes to convey. Benchmarks for a closely related task, video summarization, reduce editorial intent to a single, message-agnostic notion of saliency and thus do not account for this diversity. For evaluating message-driven video editing, we present \textbfMEDit-Bench, a dataset and benchmark, which pairs long-form videos with multiple editing messages and multiple professionally produced edits per message, demonstrating that different messages yield substantially different edits from the same source. We define an automatic evaluation protocol based on temporal alignment metrics, and find that an LLM-as-a-judge preference, a natural proxy for narrative quality, is unreliable for this task due to severe position bias. We additionally annotate each message with ambiguity and contextfulness scores, and show that both dimensions negatively correlate with model performance, establishing message difficulty as a meaningful stratification factor. Experiments with state-of-the-art MLLMs and reinforcement fine-tuned baselines show that while strong models approach human temporal alignment at lenient thresholds, all models fall behind humans at stricter criteria. A human perceptual study further confirms a large quality gap, with professional human edits remaining consistently preferred over model outputs.
[CV-66] ScaleResfusion: Residual Rectified Flow based on Residual Vector Field
链接: https://arxiv.org/abs/2607.25275
作者: Zhenning Shi,Chen Xu,Junhao Zhang,Kefei Zhang,Linjie Liu,Zhedong Zheng,Tao Li
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Real-world Image Restoration (Real-IR) aims to recover high-quality (HQ) images from complex and unknown degradations. Although recent diffusion-based methods have substantially improved perceptual quality, their current designs leave two key challenges unresolved. Methods that start from Gaussian noise are slow and often less faithful to the degraded input. Residual-based methods usually train from scratch, which makes it hard to exploit modern pre-trained generative priors. In this paper, we present ScaleResfusion, a scalable diffusion framework for real-world image restoration built on pre-trained text-to-image rectified-flow models. The core of our method is Residual Rectified Flow, which introduces the residual term R into Standard Rectified Flow. Instead of starting from pure noise, it uses a residual transport path that starts from noisy low-quality (LQ) images and admits an exact acceleration point. By learning the residual vector field, Residual Rectified Flow keeps the output distribution and linear diffusion process consistent with the pre-trained rectified-flow models. This makes parameter-efficient fine-tuning possible at scale. We further introduce a knowledge-distillation pipeline to reduce sampling cost while maintaining restoration quality. Extensive experiments on multiple real-world restoration tasks show that ScaleResfusion achieves state-of-the-art performance with much higher efficiency. These results suggest a practical and scalable way to adapt large pre-trained diffusion models to real-world image restoration. Our code and models are available at this https URL.
[CV-67] FORGE: Frame Orthogonality in Relevance Geometry for Long-Form Video Understanding
链接: https://arxiv.org/abs/2607.25266
作者: Ghazal Kaviani,Ghassan AlRegib
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: Under Review
Abstract:Multimodal large language models (MLLMs) have enabled long-form video understanding at a scale that was not previously possible. However, the density of relevant content decreases sharply as video sequence length increases, and exposing the model to more irrelevant content measurably reduces its accuracy. In this paper, we address the problem of maximizing query-relevant information in a frame subset selected at inference time, without training. FORGE (Frame Orthogonality in Relevance Geometry) is a model-agnostic method that induces a query-conditioned geometry on a pretrained multimodal embedding space, unifying relevance and diversity into a single objective. In this space, frames that cover independent query-relevant directions are far apart, and selecting the subset of maximum information captures diverse query-relevant content within the budget. Experiments on Video-MME and LongVideoBench at budgets of 16, 32, and 64 frames show that FORGE improves the unified keyframe selection score by 11.0-15.3 points over the strongest training-free baseline and up to doubles keyframe recall (0.415 vs. 0.204 at K=64 on Video-MME). The gains extend to question answering, where accuracy improves in every evaluated setting across eight open-source MLLMs spanning 4B to 32B parameters, by up to 8.7 points over uniform sampling and 5.2 points over the strongest baseline. Our findings suggest that aligning the embedding space with the query’s high-dimensional structure is a promising direction for inference-time video understanding.
[CV-68] Medical world models in healthcare: foundations applications and challenges for trustworthy clinical translation
链接: https://arxiv.org/abs/2607.25242
作者: Zhaoyan Chen,Zhongxiu Cong,Zhuanfeng Jin,Wanshu Fan,Dongsheng Zhou,Qi Ai,Haifan Gong,Congyu Liao,Xiaofeng Liu,Cong Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Medical world models offer a framework for extending medical artificial intelligence beyond static prediction by representing evolving patient states and modelling how they change over time and in response to clinical interventions. This Review defines the conceptual boundaries, technical foundations, application domains, and evidence requirements of the field through a structured narrative synthesis with reproducible evidence this http URL screened 1,455 unique records and assembled a corpus of 98 sources, including 14 studies that met a strict empirical definition of a medical world model. The field is organised around four capabilities: patient state representation, temporal dynamics modelling, intervention-conditioned simulation, and clinician-supervised planning. Evidence spans medical imaging, longitudinal electronic health records, treatment response modelling, physiological and multimodal state modelling, ultrasound and surgical interaction, and population and health-system simulation; clinical digital twins are treated as a cross-cutting integration this http URL studies provide early evidence of technical feasibility for trajectory forecasting and comparison of candidate interventions, but most remain retrospective, task-specific, or preclinical. The evidence base is further limited by incomplete longitudinal intervention data, inconsistent action semantics, limited causal identifiability, long-horizon error accumulation, inadequate uncertainty estimation, and limited external validation. Clinical translation will therefore depend on precise intervention representations, robust causal and mechanistic grounding, calibrated trajectory-level uncertainty, safety-constrained planning, and prospective multicentre validation against clinically meaningful endpoints.
[CV-69] CD-RMOT-Bench: Benchmarking the Cross-Domain Referring Multi-Object Tracking
链接: https://arxiv.org/abs/2607.25239
作者: Xiangqun Zhang,Likai Wang,Zekun Qian,Ruize Han,Wei Feng
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Referring multi-object tracking (RMOT) extends tracking from category-driven perception to language-guided understanding by grounding object trajectories in natural-language expressions. Despite recent progress, existing RMOT studies are largely conducted under in-domain settings, leaving the robustness of language-conditioned tracking under inevitable visual domain shifts unexplored. In this paper, we study Cross-Domain Referring Multi-Object Tracking (CD-RMOT), a new and challenging problem that evaluates whether an RMOT model trained on a labeled source domain can reliably follow natural-language expressions in an unlabeled target domain with different visual conditions. To support systematic study, we construct CD-RMOT-Bench, a unified benchmark that combines real clear-domain referring tracking data, aligned digital-twin variants, and real adverse-domain videos. CD-RMOT-Bench enables both controlled weather/viewpoint shift analysis and realistic synthetic-real transfer evaluation under a shared RMOT protocol. Further, we provide a Query-Centric Adaptation (QCA) framework, designed to stabilize the query space that bridges visual trajectories and referring expressions. Extensive experiments reveal that domain shifts severely degrade RMOT performance, where the failure is not merely caused by object detection errors but more critically by unstable expression-conditioned temporal association and target selection. QCA establishes a strong baseline, while CD-RMOT-Bench opens a new direction for robust language-guided tracking across visual domains.
[CV-70] WHTMix: Efficient Stereo Depth Estimation via Walsh-Hadamard Token Mixing
链接: https://arxiv.org/abs/2607.25234
作者: Prathyush Sajith,Emadeldeen Hamdan,Ahmet Enis Cetin
类目: Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV); Signal Processing (eess.SP)
备注:
Abstract:Stereo depth estimation for driving, robotics and augmented reality must run at high resolution under tight latency budgets, yet in transformer-based matchers the global self-attention that aggregates scene context grows quadratically with the number of pixels and comes to dominate runtime. We show that the joint self-attention stage of a stereo transformer, whose role is to spread context across both views, can be replaced by a data-independent Walsh-Hadamard token mixer that mixes tokens globally in the transform domain at log-linear cost, while the data-dependent cross-attention that performs left-right correspondence is retained. On synthetic driving data the mixer matches the attention baseline in end-point error while reducing model compute by a factor of 2.46 and single-image inference latency by a factor of 2.65. A complexity analysis shows the benefit is governed by the ratio of sequence length to channel width, which explains why high-resolution stereo matching is a particularly favorable setting and why classification transformers are not; we confirm this token-to-channel scaling on non-stereo long-sequence benchmarks. Furthermore, we introduce a hybrid log-disparity loss function designed to up-weight small-disparity pixels corresponding to long-range objects. This approach reduces the error on distant objects without incurring any additional computational overhead.
[CV-71] Leverag ing Semantic Maps for City-Scale Cross-View Localization STOC
链接: https://arxiv.org/abs/2607.25215
作者: Ethan Fahnestock,Erick Fuentes,Philip R Osteen,Nicholas Roy
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注: Equal contribution by Ethan Fahnestock and Erick Fuentes. 13 pages, 7 figures, and 5 tables
Abstract:We want robots to localize in previously untraversed environments against commonly available prior data. Rich semantic data available from OpenStreetMap can be useful in this task. However, existing methods either ignore this semantic information, directly matching panoramas and overhead imagery, or dramatically compress the semantic information, working with a small set of fixed classes. To leverage this rich semantic information, two challenges need to be overcome. First, useful semantic information needs to be extracted from the robot’s egocentric observations. Second, the observed information must be quickly associated with the large prior semantic map (e.g., up to 628 km^2). We show that VLMs are effective at both extracting relevant landmarks from panoramas, and identifying feasible correspondences between these landmarks and prior overhead landmarks. However, using VLMs to propose all correspondences scales poorly as the number of mapped landmarks increases. Instead, we propose distilling a lightweight matcher from a VLM which computes correspondences for all entities in a map. We use this output to form an observation likelihood which is fused over time with a Bayes filter to create a time series of pose estimates. To support further investigation into generalizable cross-view methods that leverage semantic information, we release a dataset of extracted semantics and evaluation trajectories spanning eleven environments, including panoramas we collected in a snowstorm and at night in Boston. We demonstrate our method, trained on a single city’s fair-weather data, generalizes across location, lighting, weather, and other challenges. Code and datasets are available at this https URL.
[CV-72] ObliCity: A Benchmark and Baseline for Roof-to-Ground Projection Displacement Correction
链接: https://arxiv.org/abs/2607.25210
作者: Kai Li,Yupeng Deng,Ligao Deng,Zhihao Xi,Chenhao Wang,Jierui Zhang,Yingrui Ji,Yu Meng,Xiangyu Zhao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 12 pages
Abstract:Oblique-view urban remote sensing imagery inevitably exhibits geometric projection displacements between building roofs and footprints, leading to significant distortions in spatial structure. Existing approaches either ignore these deformations or handle them implicitly within segmentation-based frameworks, where progress is dominated by general segmentation advances rather than improvements in geometric correction. In this work, we explicitly define roof-to-footprint offset vector (RFOV) extraction as an independent learning task that decouples geometric alignment from semantic segmentation. To support this task, we introduce the Oblique City dataset (ObliCity), the first large-scale benchmark that integrates high-resolution UAV imagery and globally distributed satellite data, covering diverse city morphologies and camera perspectives. Methodologically, we reformulate DragOSM into DragRoof, an ODE-based framework inspired by human annotation behavior. By simulating the continuous process of dragging roofs toward their footprints, DragRoof learns deterministic, geometry-consistent offset fields and adaptively determines convergence through an end token. Extensive experiments on ObliCity demonstrate that DragRoof achieves state-of-the-art RFOV extraction performance, requiring fewer inference steps while delivering superior directional and length accuracy. Our dataset and model establish a principled foundation for studying projection displacement correction in oblique remote sensing imagery. The source code and dataset will be avaliable at this https URL.
[CV-73] LGFNet: A CTC-Guided Local-Global Fusion Framework for Single-Channel Sleep Staging
链接: https://arxiv.org/abs/2607.25197
作者: Chongjian Wang,Zhenghang Hou,Junjie Gao,Xiaofang Zhong,Shiyuan Han,Tong Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Sleep staging remains challenging due to long-range temporal dependencies, ambiguous stage transitions-particularly in N1-and substantial distribution shifts across subjects, sampling rates, and EEG montages. These difficulties are further amplified in single-channel, low-latency scenarios required by wearable and real-world applications. To address these issues, we propose LGFNet, a CTC-guided sequence-to-sequence framework for robust sleep staging. LGFNet introduces a Local-Global Fusion encoder that jointly models fine-grained temporal dynamics and long-range sleep structure, overcoming the limitations of conventional serial hybrid architectures. A CTC-Attention joint training paradigm is adopted to unify temporal alignment with context-dependent modeling, enabling more accurate recognition of stage boundaries and transitions. Furthermore, a three-stage decoding strategy is devised, leveraging CTC-guided decoding and Viterbi-based smoothing to reduce error accumulation and enforce physiological consistency. Extensive cross-dataset evaluations on five public benchmarks demonstrate that LGFNet consistently outperforms state-of-the-art single-channel methods. In particular, on Sleep-EDF-78, LGFNet surpasses DMIN by +1.27% accuracy, +1.74% macro-F1, and +1.93% kappa, with pronounced gains on N1 and transition segments, highlighting its robustness and strong generalization across diverse sampling rates, montages, and recording environments.
[CV-74] OrganLens: Organ-Specific Representation Learning for CT Foundation Models
链接: https://arxiv.org/abs/2607.25164
作者: Zhixuan Ge,Anqi Li,Sadeer Al-Kindi,Hanwen Xu,Wei Qiu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 16 pages, 7 figures, 5 tables
Abstract:A CT examination captures multiple organs, but many biomedical questions concern abnormalities, prognosis, or longitudinal change in a specific organ. These questions require a separate representation for each organ within the same CT volume. Existing CT foundation models commonly produce a single volume-level representation, while recent anatomy-aware methods either encode pre-separated organ volumes or explicitly disentangle images into organ token groups. The former may remove clinically relevant surrounding context, while the latter does not condition a shared encoder on a selected organ before its features are formed. We introduce OrganLens for organ-specific representation learning through self-supervision. An organ identity conditions a shared CT encoder, while organ-specific distillation and anatomy-mask supervision shape features for anatomy-weighted pooling into organ-specific representations. At inference, the shared model produces 11 organ-specific representations without external segmentation masks. We evaluate OrganLens on CT-RATE, RAD-ChestCT, INSPECT, and NLST across diverse acquisitions and downstream evaluations. Relative to CT-pretrained DINOv2, heart representations raise CT-RATE cardiomegaly AUROC from 0.910 to 0.953, while lung representations improve the Harrell C-index for NLST lung-cancer mortality by 14.2%. The global representation reaches INSPECT Recall@10 of 33.09% and 32.04% for text-to-image and image-to-text retrieval, respectively. Across organ-related tasks, anatomically matched representations provide stronger task-relevant signal, while the global representation retains broad utility. OrganLens offers a scalable approach to organ-specific CT representation learning with a shared encoder. More broadly, it provides the medical research community with a reusable framework for studying organ-specific disease across cohorts and clinical endpoints.
[CV-75] OpenPVMapper: A Multi-source Nationwide Database of Rooftop Photovoltaic Systems in France
链接: https://arxiv.org/abs/2607.25153
作者: Gabiel Kasmi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 14 pages, 5 tables, 7 figures
Abstract:Rooftop photovoltaic (PV) systems account for the vast majority of PV grid connections, yet no open, comprehensive, installation-level dataset of these systems exists: public registries aggregate data only above a capacity threshold, and remote sensing-based detection efforts, while extensive, are typically confined to a single method, a limited geographic scope, or a single point in time. We introduce OpenPVMapper, a nationwide, multi-source database of rooftop PV installations in mainland France, built by aggregating and reconciling complementary sources: a deep learning-based detection pipeline deployed on nationwide aerial imagery, OpenStreetMap and a probabilistic building-level detection dataset. The resulting database contains 1,135,850 installations, totaling approximately 15.01~GWp of installed capacity, each documented with its provenance, detection method, and, where available, a manual validation flag. Manual review of a stratified sample of 1,862 installations places the database’s overall precision at approximately 74–75%, with corroboration across independent sources bringing a substantial, quantified precision gain. By aggregating independent sources rather than relying on any single detection method, OpenPVMapper reaches a level of confidence beyond what any one source could provide on its own, while remaining extensible to further sources as they become available. It is released under an open (CC-BY) license alongside the full source code used to build it.
[CV-76] LENS: Adaptive Spatio-Temporal Zooming for Keyframe Sampling in Long-Form Videos ECCV2026
链接: https://arxiv.org/abs/2607.25125
作者: Ce Zhang,Jinxi He,Katia Sycara,Yaqi Xie
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at ECCV 2026. Project page: this https URL
Abstract:Despite rapid progress in Multi-modal Large Language Models (MLLMs), understanding long-form videos is still bottlenecked by limited context windows. While recent keyframe sampling methods attempt to mitigate this by distilling video inputs into a compact set of query-relevant frames, navigating the vast spatio-temporal search space remains challenging, as spatial detail and temporal coverage often conflict. To address this, we introduce LENS, a training-free keyframe sampling framework that dynamically decides when to zoom in for fine-grained details and when to zoom out for broader context based on the text query. Concretely, LENS adaptively allocates a limited frame budget between spatial zoom-ins, which highlight query-relevant regions within individual frames, and temporal zoom-outs, which expand the temporal scope through multi-frame aggregation, enabling the model to reason across multiple granularities while capturing both high-fidelity details and long-range context. Across diverse long-form video benchmarks, LENS consistently outperforms prior state-of-the-art keyframe sampling methods and delivers substantial gains over uniform sampling, improving Video-MME accuracy from 53.3% to 60.7% with this http URL is available at this https URL.
[CV-77] OPERA: Offline Policy-guided Expert Routing and Adaptation for Universal Biomedical Image Analysis ACM-MM2026
链接: https://arxiv.org/abs/2607.25108
作者: Zihan Li,Feiyang Liu,Dandan Shan,Ruibo Wang,Qingqi Hong
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Image and Video Processing (eess.IV)
备注: Accepted by ACM MM 2026. 18 pages
Abstract:Biomedical image analysis spans diverse modalities and tasks, yet real-world deployment is hindered by severe distribution shifts across scanners, protocols, and patient populations. High-performing models consequently require repeated domain-specific fine-tuning, which is a costly cycle that becomes impractical when labels are scarce or privacy constraints limit data sharing. We propose OPERA (Offline Policy-guided Expert Routing and Adaptation), a multi-agent ensemble framework that addresses this deployment bottleneck by treating expert weight assignment as an offline policy learning problem: a routing policy is learned from a small validation set without gradient updates to any expert agent, then deployed with test-time adaptation to handle distribution shift. OPERA coordinates heterogeneous specialist agents through complementary mechanisms. The expert profiling module learns selection policies offline, enabling informed allocation of expertise. Each agent undergoes confidence calibration through temperature adjustment, ensuring more reliable probabilistic outputs. OPERA also incorporates distribution aware adaptation, where class weights are dynamically adjusted at the batch level using statistics derived from unlabeled test data. Instance level routing assigns each sample to the most suitable expert by leveraging inter model agreement and predictive entropy. We evaluate OPERA on 9 datasets covering fundus photography, chest X-ray, CT, MRI, and multimodal diagnostic benchmarks, comparing against 30+ baselines across classification, segmentation, and multimodal settings. OPERA consistently improves performance and calibration quality, demonstrating that offline policy-guided expert agents coordination is a practical path to deployable biomedical AI without retraining. Code is on \hrefthis https URLGitHub.
[CV-78] IMPRINT: Image-Conditioned Query Enrichment for Long-Tail Object Goal Navigation IROS
链接: https://arxiv.org/abs/2607.25106
作者: Jelin Raphael Akkara,Filippo Ziliotto,Luciano Serafini,Lamberto Ballan,Tommaso Campari
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at the 2026 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). Project page: this https URL
Abstract:Embodied AI increasingly relies on queryable semantic maps built from pre-trained vision-language models to enable zero-shot Object Goal Navigation (ObjectNav). However, existing approaches typically depend on text-only queries, which become less reliable as semantic specificity increases toward fine-grained object categories. We introduce IMPRINT, a zero-shot plug-and-play framework that enriches textual object queries with web-sourced images to improve grounding in queryable maps. Retrieved images are encoded using a vision-language model, matched against the semantic map to produce similarity maps, and aggregated to yield context-aware localization. Notably, this requires no training or modification of the underlying navigation policy. To explicitly evaluate long-tail behavior, we present HSSD-rare, a new ObjectNav benchmark built on Habitat Synthetic Scenes and featuring semantically specific subcategories. Across both OVON and HSSD-rare, image-conditioned queries consistently improve object grounding and yield end-to-end navigation gains. Further analysis reveals that translating localization gains to navigation performance depends critically on downstream detection quality, highlighting a key systems bottleneck in long-tail embodied navigation.
[CV-79] MorphUNet: Alpha-Controlled Biometric Transport for Diffusion-Based Face Morphing Attacks
链接: https://arxiv.org/abs/2607.25092
作者: Taimoor Rizwan,Sara Atito,Zhenhua Feng,Muhammad Awais,Josef Kittler
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 37 pages, 23 figures, 8 tables. Includes supplementary material (additional robustness analysis, CFD stress tests, and conditioning ablations) appended after the references
Abstract:Face morphing attacks create synthetic images verifiable against multiple identities, threatening border control and identity verification systems. We introduce MorphUNet, a diffusion morphing framework formulating two-parent generation as alpha-controlled biometric transport: each parent is decomposed into CLIP appearance and ArcFace identity evidence, aligned into a CLIP-compatible token space, with the two contributors preserved as separate identity-aware token banks. To our knowledge, MorphUNet is the first diffusion-based morphing framework using trainable parent-separated dual cross-attention inside the denoising U-Net: a Biometric Transport Layer carrying parent-specific identity evidence through denoising, attending to each parent separately before combining residuals via the morphing parameter alpha. DDIM-inverted latent interpolation gives a coherent denoising start, while weaker-parent-guided selection favours morphs maximising the lower parent-similarity score, reducing collapse toward one contributor. We evaluate MorphUNet against three state-of-the-art baselines (StableMorph, MIPGAN-II, and MorDIFF) on FEI and FRLL using six recognition systems, and propose CFD-based unseen-identity stress testing across gender and ethnicity pairing, demographic shifts, and parent-similarity extremes. MorphUNet achieves the best Morphing Attack Potential (MAP) when at least three of six systems are fooled by one morph, reaching 0.919 on FEI and 0.886 on FRLL, and obtains the best FID on both datasets (35.19 FEI, 44.86 FRLL). It also gives the highest APCER at 5% BPCER in the same-dataset setting, and remains highly difficult to detect under cross-dataset transfer, with APCER 0.996 on FEI and 0.946 on FRLL. The full evaluation analyses MAP, MAD, per-system vulnerability, identity balance, image quality, top/bottom-similarity stress tests, and CFD unseen-identity robustness.
[CV-80] Diff-ID: Identity Consistent Facial Image Generation and Morphing via Diffusion Models
链接: https://arxiv.org/abs/2607.25078
作者: Taimoor Rizwan,Sara Atito,Muhammad Awais,Zhenhua Feng,Josef Kittler
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 20 pages, 11 figures, 3 tables
Abstract:Generative diffusion models have revolutionized facial image synthesis, yet robust identity preservation in high resolution outputs remains a critical challenge. This issue is especially vital for security systems, biometric authentication, and privacy sensitive applications, where any drift in identity integrity can undermine trust and functionality. We introduce Diff-ID, a diffusion based framework that enforces identity consistency while delivering photorealistic quality. Central to our approach is a custom 210K image dataset synthesized from CelebA-HQ, FFHQ, and LAION-Face and captioned via a fine tuned BLIP model to bolster identity awareness during training. Diff-ID integrates ArcFace and CLIP embeddings through a dual cross attention adapter within a fine tuned Stable Diffusion UNet. To further reinforce identity fidelity, we propose a pseudo discriminator loss based on ArcFace cosine similarity with exponential timestep weighting. Experiments on held out and unseen faces show that Diff-ID does not exceed InstantID in raw ArcFace Face Similarity, but achieves substantially lower FID and the strongest FIQ based identity–realism trade off among the evaluated methods. We also present a unified DDIM based morphing pipeline that enables qualitative facial interpolation without per identity fine tuning. We further argue that identity preservation and photorealism should be evaluated jointly rather than in isolation, as high identity similarity alone does not guarantee realistic outputs. To make this trade off explicit, we report Face Image Quality (FIQ) as a complementary ratio based score that combines identity similarity and perceptual realism while keeping FS and FID as the primary metrics.
[CV-81] Unifying Active Learning and Semi-Supervised Learning for Medical Image Segmentation
链接: https://arxiv.org/abs/2607.25014
作者: Bahram Jafrasteh,Cheng Wan,Heejong Kim,Johannes C. Paetzold,Qingyu Zhao
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (stat.ML)
备注:
Abstract:In practical settings, medical image segmentation models are often developed with limited annotated data rather than fully labeled datasets. Training frequently begins in ultra-low labeled regimes where only a small number of volumes are annotated. In such scenarios, practitioners must simultaneously decide which cases to annotate and how to best use the remaining unlabeled data. Although active learning (AL) and semi-supervised learning (SSL) both target annotation scarcity, they are typically designed and optimized independently, resulting in objective mismatch and unstable training during early-stage “cold start” conditions. We propose RegAL, a unified active semi-supervised framework governed by a shared topology-aware Pareto optimization that couples sample acquisition with unlabeled data utilization. RegAL evaluates images along three complementary axes, voxel-wise uncertainty, feature diversity, and a novel topological consistency metric, to select anatomically informative edge cases for annotation. On the other hand, the same criteria are used to identify geometrically stable atlas candidates for diffeomorphic registration-guided augmentation to train a self-supervised Mean Teacher segmentation network. Across BraTS 2021, dHCP, and ProstateX, RegAL remains stable with few labeled volumes and consistently outperforms state-of-the-art AL, SSL, and active semi-supervised baselines across Dice and boundary-distance (ASD, HD95) metrics under extreme annotation scarcity.
[CV-82] Enabling Fully Integer-Only Inference for Lightweight Detection Transformers
链接: https://arxiv.org/abs/2607.24981
作者: Thanh Cong Le,Michal Szczepanski,Martyna Poreba
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Vision Transformer detectors now approach the accuracy of CNNs but remain difficult to deploy on NPUs and microcontrollers because key components, including deformable attention, feature fusion, and nonlinear activation functions, are not natively compatible with integer arithmetic. Existing quantized detectors either retain operators such as Softmax, GELU, and LayerNorm or focus on heavyweight backbones, leaving lightweight detection transformers without an end-to-end integer implementation. We address this gap with I-LW-DETR, the first fully integer-only lightweight DETR, in which every operation in the forward pass, including transformer nonlinearities, is executed in integer arithmetic. I-LW-DETR is built upon three key components: a scale-preserving split convolution that assigns independent activation scale to each branch of the multi-scale projector; SD-ShiftGELU, a sign-dependent GELU approximation that preserves element-wise behavior while avoiding the accuracy degradation; and a constrained Shiftmax that maintains stable Softmax normalization. Experimental results demonstrate that the proposed quantization pipeline consistently produces efficient fully integer-only models across different model scales. Across all model scales, the proposed pipeline incurs only a moderate accuracy degradation while reducing the model size by approximately 3.6\times and the computational cost by more than one order of magnitude.
[CV-83] PerceptionBench: Evaluating Atomic Visual Perception in Multimodal Large Language Models
链接: https://arxiv.org/abs/2607.24957
作者: Zichao Lin,Yifeng Xie,Bowen Qu,Haiming Wang,Jia Li,Haoning Wu,Yuhao Dong,Zuhao Yang,Jinguo Zhu,Haoyu Lu,Zijia Zhao,Tongtian Yue,Zhangyang Qi,Junwei Yang,Mengfan Dong,Peizhou Cao,Chenzhuang Du,Zaida Zhou,Haotian Yao,Hao Yang,Hongcheng Gao,Lin Sui,Weihong Li,Xinxing Zu,Jia Chen,Yao Wang,Xiaoxue Wu,Yalin Wang,Y. Charles,Yiping Bao,Yangyang Liu,Zhiqi Huang,Xinyu Zhou
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:We introduce PerceptionBench, a benchmark specifically designed to evaluate the atomic visual perception capabilities of Multimodal Large Language Models (MLLMs). Existing benchmarks often fail to isolate perception: holistic evaluations conflate perceptual errors with failures in reasoning or domain knowledge, while application-driven benchmarks only cover narrow, fragmented domains shaped by heuristic designs. To address these limitations, PerceptionBench adopts a bottom-up approach: by diagnosing the earliest failure points in the responses of frontier MLLMs across 42 existing benchmarks, we construct an error taxonomy whose perception branch defines ten atomic perceptual capabilities. Guided by this taxonomy, we construct 3,000 verified questions with short, unambiguous answers, each isolating a single capability, with difficulty stemming from perception rather than reasoning or knowledge. Benchmark results across sixteen frontier MLLMs reveal that atomic perception remains largely unsolved—no model reaches 60% accuracy, perception-related hallucination is the weakest capability on average, and similar overall scores conceal sharply divergent capability profiles. PerceptionBench thus provides a capability-level standard for measuring and diagnosing the visual perception boundaries of MLLMs.
[CV-84] Intrinsic and Triangulation-Agnostic Attention: A Simple and Powerful Approach for Learning on Meshes
链接: https://arxiv.org/abs/2607.24954
作者: Ashwath Shetty,Zihan Zhu,Soeren Pirk,Noam Aigerman
类目: Graphics (cs.GR); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:This work proposes an adaptation of the attention mechanism for triangle meshes. The core observation is that endowing the attention mechanism with critical properties for learning over meshes – intrinsicality and triangulation-agnosticism – enables it to attain state-of-the-art results over several learning-based tasks in geometry-processing. The above is achieved by modifying the attention mechanism from the bottom up based on simple principles from geometry-processing. Namely, the quantities used within attention – queries, keys and values – are created by an intrinsic, triangulation-agnostic network, and treated as discretizations of continuous functions. From that, we devise an appropriate attention mechanism that operates over triangle meshes through standard FEM discretization of the resulting integrals of the above functions. Surprisingly, as far as we know, this straightforward approach has not been utilized for learning over meshes. Experiments show our method exceeds current state of the art, including both mesh-based architectures as well as point cloud transformers. Namely, we show significant improvements on several common benchmarks and tasks – predicting canonical high-frequency signals; predicting deformations; computing dense correspondences, both between full shapes and partial ones; and predicting feature descriptors.
[CV-85] Harm is not Universal: Community-Specific Toxicity Detection is Urgently Needed
链接: https://arxiv.org/abs/2607.24898
作者: Xinnuo Xu,Anja Thieme,Daniela Massiceti,Ioana Tanase,Rita Marques,Melanie Fernandez Pradier,Martin Grayson,Camilla Longden,Cecily Morrison
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 18 pages, under review
Abstract:State-of-the-art toxicity detectors for text-to-image generation adopt a one-size-fits-all approach: a single universal model applying fixed safety guidelines to all users. Our empirical evidence shows that these detectors fail to shield marginalized communities: approximately 35% of generated images labeled safe are considered harmful by disability communities. In this position paper, we argue for community-specific toxicity detection (CTD). To demonstrate its feasibility, we collaborate with disability experts to develop safety guidelines for two communities: dwarfism and blind/low vision. Using a dataset of 2,400 annotated T2I-generated images we demonstrate that both large vision-language models and existing general-purpose toxicity detectors catastrophically fail to recognize harmful content under these guidelines in zero-shot settings with F1 score lower than random guessing (F1 0.32 and 0.37). Promisingly, prompt-based adaptation methods (ICL, VQA) substantially improve harm detection performance (GPT-4o: F1 0.50 and 0.78), while parameter-efficient fine-tuning improves smaller models (0.5b-7b with best F1 0.48 and 0.59) with less than 100 demonstrations, but remains sensitive to evolving guidelines. Despite these gains, CTD performance remains far below F1 \approx 0.9 achieved for general-purpose toxicity detection, highlighting the challenge and the need for sustained research effort.
[CV-86] DisasterTD: Disaster Toponym Disambiguation Using Multimodal LLM s and Cross-View Geolocalization
链接: https://arxiv.org/abs/2607.24856
作者: Wenping Yin,Ziqi Liu,Naixia Mou,Weijia Li,Danfeng Hong,Hao Li
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Social media imagery (SMI) provides timely and fine-grained ground perspectives that are valuable for situational awareness and emergency response. Unlike satellite or aerial imagery, SMI can capture disaster impacts and ground-level conditions in a timely manner. However, geographic references in SMI are often vague or ambiguous, making accurate geolocalization challenging. To address this issue, we propose DisasterTD, a disaster toponym disambiguation framework that integrates multimodal large language model (MLLMs)-based semantic reasoning with cross-view geolocalization. First, MLLMs extract toponyms and generate candidate geolocations from noisy textual inputs. Then, cross-view matching between SMI, remote sensing imagery (RSI), and optionally street-view imagery (SVI) is used to verify and refine these candidate results. We evaluate DisasterTD on the Hurricane Harvey dataset, where SMI is augmented with collected RSI and SVI to construct a cross-view benchmark for disaster geolocalization. The dataset is divided into four categories based on toponym clarity and ambiguity, allowing a fine-grained performance analysis across scenarios. Results show that DisasterTD consistently outperforms MLLM-only and cross-view-only baselines without disambiguation, achieving geolocalization accuracies of 71.62% within 1000 m, 62.36% within 500 m, 57.99% within 250 m, 52.09% within 100 m, and 47.01% within 50 m, while reducing the mean and median errors to 11.33 km and 0.68 km, respectively. The largest improvements appear in ambiguous toponyms, where semantic reasoning with cross-view evidence reduces candidate dispersion and errors. These findings demonstrate the effectiveness of integrating MLLM-based candidate generation with cross-view verification for fine-grained disaster geolocalization.
[CV-87] rack-Leakage-Free Hold-Out Self-Validation for Photogrammetric Reconstruction: Protocol Sensitivity and Limits
链接: https://arxiv.org/abs/2607.24852
作者: Behnam Asadi
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注: 16 pages, 4 figures. Submitted to the ISPRS Journal of Photogrammetry and Remote Sensing. Negative/characterisation result: a ground-truth-free self-consistency signal is shown to saturate and to be blind to coherent global distortion
Abstract:Automated photogrammetric inspection emits metric measurements from a 3D reconstruction whose own correctness is normally unknown without an external survey. Can a reconstruction estimate its own reliability with no ground truth? We formalise a track-leakage-free hold-out protocol: a deterministic subset of images is withheld and each re-localised against only 3D points seen by at least two retained images – a track-level barrier so a view is never tested against structure it helped create – aggregated into an mAA confidence. Held-out observations still entered the bundle adjustment that built the trusted structure, so we call it track-leakage-free (a stricter re-mapping variant confirms this at 3/5 deg). Across operational GNSS-referenced captures plus ETH3D, EuRoC and the IMC 2025 benchmark we report four findings. (i) The protocol is computationally well-posed: good reconstructions score millidegree self-consistency. (ii) Thresholded self-consistency saturates and does not track absolute accuracy: confidence stays near 1.00 while true RTK error swings up to 14x within a capture, and the per-capture correlation is sign-unstable across five captures (-0.57 to +0.98; pooled 95% CI spanning zero), so it cannot gate accuracy. (iii) It flags gross failure only when the failure destroys internal consistency: a fragmenting model drops confidence, but a single self-consistent, globally-distorted model evades it – three of four captures gave one model wrong by 55-106 m at confidence 1.00. (iv) The same dichotomy holds on independent ETH3D and IMC 2025 ground truth. Track-leakage-free hold-out measures internal geometric consistency, not absolute accuracy: neither a substitute for control-point assessment nor a general gross-failure gate. We release the protocol and degradation harness.
[CV-88] Gradient-Based Latent Decomposition Reveals Mechanisms of Feature Degradation in Weakly Supervised Mammography
链接: https://arxiv.org/abs/2607.24835
作者: Vinceline Bertrand,Ionut Cardei
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Weakly supervised hierarchical models exhibit a persistent asymmetry: coarse lesion-type features are preserved under reconstruction while fine-grained malignancy cues degrade—a pattern with direct consequences for the clinical reliability of breast cancer screening pipelines. We introduce gradient-based orthogonal latent decomposition for hierarchical Variational Autoencoders~(H-VAEs) to mechanistically explain this asymmetry. The latent space is partitioned into a task-aligned component~( z_1 ), shaped by coarse supervisory gradients, and an orthogonal residual~( z_\textres ) capturing remaining representational capacity. On~3,550 mammographic Regions of Interest~(ROIs) from CBIS-DDSM, only~ \sim 4.4% of latent magnitude aligns with supervisory gradients, leaving~ \sim 95.6% in the orthogonal residual upon which fine-grained pathology prediction primarily depends. The model achieves Stage-1~AUC~0.866 and Stage 2~AUC~0.552, with a reconstruction stability gap of \Delta_\textdiag=5% ( p=0.005 ) and a classification gap of \Delta_\textAUC=0.314 ( p0.001 ). Latent ablation confirms that features for both tasks reside heavily in~ z_\textres , structurally explaining why reconstruction degrades pathology stability disproportionately. Comparisons with Multi-Instance Learning~(MIL) and Multi-Task Learning~(MTL) confirm generalization across architectures and modalities. These findings reveal that in high-dimensional spaces, a single coarse supervisory signal isolates only a sparse 1D latent direction, forcing critical fine-grained features into the vulnerable residual subspace.
[CV-89] AVE-Compass: Towards Holistic Evaluation for Audio-Video Editing Abilities
链接: https://arxiv.org/abs/2607.24821
作者: Yuqing Wen,Yukai Huang,Qianqian Xie,Jiangtao Wu,Yibin Lin,Yikai Gu,Jialu Chen,Yuanxing Zhang,Jiaheng Liu
类目: Multimedia (cs.MM); Computer Vision and Pattern Recognition (cs.CV); Sound (cs.SD)
备注:
Abstract:While instruction-based video editing has advanced rapidly, real-world videos contain tightly coupled audio and visual signals, and editing one modality often requires coordinated changes in the other. Existing benchmarks primarily evaluate visual transformations on silent clips or isolated audio editing, leaving complex audio-visual editing and cross-modal consistency underexplored. We introduce AVE-Compass, a comprehensive benchmark with 145 curated source videos, 196 audio-visually coupled editing instructions, and 2,688 fine-grained checklist items. It evaluates Instruction Following, Fidelity Preserving, Realism, and Editing Intent through checklist-based MLLM judging and a dedicated realism rubric, complemented by automated cross-modal, video, and audio metrics. Extensive evaluation shows that state-of-the-art models still struggle to execute cross-modal instructions while preserving non-target content. We further propose AVE-Agent, a modular agent framework that decomposes complex instructions into dependent subtasks and iteratively improves editing results through self-reflection and evaluator feedback. AVE-Agent improves instruction execution, Fidelity Preserving, and audio-visual alignment in joint editing while maintaining competitive perceptual quality.
[CV-90] Reasoning with Memory: A Temporal Granularity-Adaptive Framework for Training-Free Long Video Understanding ECCV2026
链接: https://arxiv.org/abs/2607.24794
作者: Linghao Meng,Qiankun Li,Junyuan Mao,Pujin Liao,Zhicheng He,Enbo Zhang,Kun Wang,Yang Liu,Huazhu Fu,Yueming Jin
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ECCV 2026
Abstract:While Multimodal Large Language Models (MLLMs) demonstrate superior generalization in fundamental video tasks, restricted context windows limit their long video understanding. To accommodate this constraint, models typically resort to keyframe selection. However, uniform sampling or static query-guided selection often overlooks critical temporal context, failing to adapt to the varying query temporal granularities. In this paper, we propose ReMem, a temporal granularity-adaptive keyframe selection framework for training-free LongVideoQA. ReMem introduces a dual-level memory-augmented adaptation. At the query level, Memory-Driven Question Parsing leverages LLM long-term memory to decode question temporal granularity and extract semantic entities. At the video level, Synergistic Dual-Semantic Frame Alignment exploits intrinsic structural memory to align frames with query semantics, guiding Structure-Aware Dynamic Frame Routing to cluster events and optimally distribute sampling budgets. By explicitly preserving temporal information with memory mechanisms, ReMem suppresses redundancy and empowers MLLMs to perform robust multi-granular video reasoning. Evaluations across four popular LongVideoQA benchmarks using three MLLMs demonstrate highly efficient, state-of-the-art zero-shot performance; notably, LLaVA-Video with ReMem reaches 54.5% (+12.3%) on LVBench and 67.1% (+8.2%) on LongVideoBench.
[CV-91] CORF-GS: Real-Time Wireless Radiance Field Reconstruction via Coupled Optical-RF Gaussian Splatting
链接: https://arxiv.org/abs/2607.25569
作者: Jinya Zhang,Jiajia Guo,Chao-Kai Wen,Shi Jin
类目: ignal Processing (eess.SP); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Information Theory (cs.IT)
备注:
Abstract:Recent advances in 3D Gaussian Splatting (3DGS)-based wireless radiance field (WRF) reconstruction provide an efficient solution for wireless channel modeling. However, existing WRF reconstruction methods rely on pre-collected observations and offline optimization, and thus struggle to provide real-time channel knowledge. To bridge this gap, we propose CORF-GS, a real-time WRF reconstruction framework that processes sequential optical and radio frequency (RF) keyframes. Specifically, CORF-GS constructs a unified Gaussian representation for optical and RF with shared geometry and modality-specific appearance, allowing high-resolution optical images to provide structural priors for WRF reconstruction. When a new keyframe arrives, CORF-GS first employs optical-guided Gaussian sampling to densify the WRF in under-represented regions. Since light and radio waves may respond differently to the same object surfaces due to wavelength mismatch, relying solely on optical guidance may neglect RF-informative areas. Therefore, CORF-GS performs coupled optical-RF optimization to jointly refine the shared Gaussians. Compared with the existing two-stage training pipelines, this prevents WRF from passively adapting to a frozen optical geometry and encourages the shared Gaussians to adapt to both optical structures and RF power distributions. Simulations show that CORF-GS achieves state-of-the-art RF spectrum synthesis quality and reduces the reconstruction time by 6.4\times compared with existing WRF methods.
[CV-92] Analysis of the Shortcut Learning and Clever Hans Effect in CNN based ECG Image Classification
链接: https://arxiv.org/abs/2607.25117
作者: Abhay Kumar Pathak,Mrityunjay Chaubey,Manjari Gupta,Deepti Mishra
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 models for ECG image classification may achieve high accuracy by exploiting non-physiological visual cues instead of ECG waveform morphology. Given the black-box nature of deep learning models, their promise of high predictive performance often remains insufficiently translated into clinical or real-world trust, interpretability, and actionable decision-making. In this study, we examine shortcut learning and Clever Hans effect in a publicly available ECG image dataset using convolutional neural networks. In process we have created six image-derived feature sets (FSs), FS1: raw full ECG images, FS2: cropped waveform-only images, FS3: waveform-masked metadata images, FS4: red-arrow artifact images for the myocardial infarction class, FS5: contrast-enhanced images for the abnormal heartbeat class and FS6: Gaussian-blurred images for the normal class. These controlled representations were used to test whether classification performance persists when waveform information is removed or when artificial class-specific artifacts are introduced. Shortcut retention score, prediction consistency and confidence divergence across Feature-Set Representations have been calculated to assess the transparency about the learning pattern. Along with factual results, average Integrated Gradients and occlusion sensitivity test results are presented to inspect whether model attribution focused on ECG-relevant waveform regions or on non-clinical artifacts. Performance changes across feature sets and attribution patterns were used to identify potential Clever Hans behavior. This study evaluates whether ECG image classifiers learn clinically meaningful morphology or shortcut cues introduced by report layout, metadata, contrast, blur, or artificial markers.
[CV-93] A Functional Approach to Curve Alignment and Shape Analysis
链接: https://arxiv.org/abs/2503.05632
作者: Issam-Ali Moindjié,Cédric Beaulac,Marie-Hélène Descary
类目: Methodology (stat.ME); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (stat.ML)
备注:
Abstract:In many image analysis problems, the contours of objects carry important statistical information about shape. Such contours are typically affected by deformation variables including scaling, translation, rotation, and reparametrization. Previous studies in statistical shape analysis have mainly focused on analyzing contours and shapes through discrete observations. While this approach might offer computational advantages, it overlooks the continuous nature of these objects and their underlying geometric structure. It also ignores potential dependencies between the deformation variables and their effect on the shape, which may result in a loss of statistical information and reduced interpretability. In this paper, we introduce a novel framework for analyzing shapes within the context of Functional Data Analysis (FDA). Basis expansion techniques are employed to derive analytic solutions for the estimation of deformation variables, namely scaling, translation, rotation, and reparametrization, thereby achieving curve alignment. A generative model for random contours is then developed using principal component analysis techniques. Numerical experiments on simulated data and the \textitMPEG-7 database demonstrate that our method successfully identifies deformation parameters and captures the underlying distribution of random contours in settings where traditional FDA methods fail.
人工智能
[AI-0] πmathbfR2: Reactive Real-time Flow Policies
链接: https://arxiv.org/abs/2607.26055
作者: Sungjae Park,Shubham Tulsiani
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Preprint(20 pages). Under Review
Abstract:Generalist manipulation policies increasingly take the form of action-chunking flow policies built on large pretrained backbones. Such chunks run open-loop, so the policy cannot react to sensory input arriving mid-execution, sacrificing \emphreactivity. Replanning more often would restore it, but the perception-to-action pipeline (a large backbone plus multiple denoising steps) is too slow: this \emphlatency forbids frequent replanning and leaves committed actions stale, making such policies ill-suited for dynamic, closed-loop control. We present \pi\mathbfR^2 , which makes these policies reactive and real-time while retaining large backbones, expressive multi-modal policies, and multi-action prediction. Built on the per-position noise schedule of diffusion forcing, \pi\mathbfR^2 contributes two ideas. First, it splits conditioning into a fast channel (proprioception, fresh every tick) and an asynchronously updated slow channel (vision-language features), so the policy reacts to proprioception within a chunk while tolerating stale vision. Second, a latency-adaptive flow schedule treats in-flight actions as inpainting conditioning and emits actions in one denoising step per call, letting one trained model adapt to varying hardware latency. Requiring minimal modification to existing architectures, \pi\mathbfR^2 can be finetuned from a pretrained policy: applied to GR00T-N1.7 on a real xArm6+XHand platform, it replans closed-loop roughly 4\times faster than the base policy (~ 25 Hz on an A5000 GPU), acting on a fresh observation every 40 ms. Across simulation and real-world manipulation tasks, \pi\mathbfR^2 improves the success rate by up to 23% in simulation and 30% in the real world over the strongest baseline. Project page: this https URL
[AI-1] Falling Behind Drives Unsafe Development in an Idealised AI Race Experiment
链接: https://arxiv.org/abs/2607.26034
作者: Elias Fernández Domingos, TheAnh Han
类目: Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Computer Science and Game Theory (cs.GT); General Economics (econ.GN)
备注: 45 pages (main + SI)
Abstract:Technological races create tension between speed and safety: actors may gain by moving faster than competitors, even when risky development is harmful. This is prominent in debates about artificial intelligence (AI), where competitive pressure is often argued to incentivise riskier, less safety-conscious development. We study this using a framed behavioural experiment based on an idealised AI race, in which paired participants repeatedly chose between Safe and Unsafe development under an uncertain time horizon. Unsafe development gave faster progress and higher immediate payoffs but accumulated private risk up to a treatment-specific maximum of 10%, 60%, or 90%; the race’s competitive structure was held constant, and only this maximum risk varied. Neither the pre-registered comparison between risk levels nor the role of elicited risk preferences was supported by the data. Instead, exploratory analyses motivated by the task’s repeated structure show that Unsafe behaviour is shaped less by risk preferences than by the evolving strategic state of the race: participants are more likely to choose Unsafe after their opponent does so, being ahead reduces Unsafe play while falling behind increases it, and first-round choices predict later behaviour. To interpret these effects we introduce a reduced evolutionary model with four strategies – Always Safe, Always Unsafe, Conditionally Safe, and Conditionally Antisocial Safe – which reproduces the treatment effect and shows how conditional Unsafe behaviour can be favoured by competitive race dynamics. Together, the experiment and model show that unsafe development can emerge from early behavioural momentum, opponent behaviour, and fear of falling behind, rather than from risk preferences alone, suggesting policy should focus on reducing competitive pressure and promoting cooperation in AI development rather than only individual risk.
[AI-2] CHARM: A Multimodal Graph Foundation Model with Hierarchical Context Modeling for Zero-Shot Transfer
链接: https://arxiv.org/abs/2607.26023
作者: Ankang Yang,Jitao Zhao,Di Jin,Yuxiao Huang,Dongxiao He
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Graph foundation models (GFMs) have emerged as a promising paradigm for transferring knowledge across graph domains and tasks. Real-world graphs associate nodes with text, images, and other modalities, making multimodal graphs essential for representing complex entities and relations. Moreover, collecting labels and adapting models for every new graph domain is costly and often infeasible, motivating zero-shot transfer. Unfortunately, zero-shot transfer on multimodal graphs remains underexplored. Existing GNN-based graph foundation models typically require downstream adaptation, whereas LLM-based graph methods mainly address unimodal graphs or tasks within a single domain. This setting presents two key challenges. First, models must generalize knowledge from individual modalities while capturing transferable cross-modal relations. Second, without target-domain fine-tuning, node representations remain entangled with domain-specific structures and modality-specific characteristics, obscuring shared concepts in unseen domains. To address these challenges, we propose CHARM, a multimodal graph foundation model with hierarchical context modeling for zero-shot transfer. CHARM replaces isolated raw nodes with hierarchical graph contexts that capture multimodal semantics and cross-modal relations. These contexts map domain-specific node patterns to shared high-level concepts, reducing reliance on target-domain supervision or adaptation. A modality-aware graph context encoder integrates multimodal information with graph structure and converts the resulting representations into graph tokens for a large language model . Experiments show consistent improvements on zero-shot multimodal graph tasks.
[AI-3] MDTransformer: A Hardware-Software Co-Design of Mode-Division Photonic Transformer Accelerator with Inverse-Designed Coherent Crossbar
链接: https://arxiv.org/abs/2607.26016
作者: Solomon Micheal Serunjogi,Rachmad Vidya Wicaksana Putra,Ayat Taha,Muhammad Shafique,Mahmoud Rasras
类目: Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI); Distributed, Parallel, and Cluster Computing (cs.DC)
备注: 10 pages, 10 figures, 1 table
Abstract:Recently, photonic transformer accelerators (PTAs) have successfully achieved significant speedup and energy efficiency improvements over electronic accelerators for expediting Transformer inference. However, state-of-the-art rely on expensive multi-wavelength light generation and large dot-product units due to active phase-shifter components, thus making their approach inefficient and impractical. To address this, we propose MDTransformer, a novel hardware-software co-design of PTA based on mode-division optical dataflow and operations. Specifically, MDTransformer performs complex matrix operations using spatial-mode interference, that leverages the inverse-designed multi-mode couplers, crossings, and Mach-Zehnder IQ modulators into a compact mode-division photonic tensor core (MPTC), capable of executing matrix multiplications in the optical domain. Its each guided mode (i.e., TE0-TE3) acts as an independent computational lane, enabling four-fold parallelism-per-waveguide without spectral filtering or free-spectral-range limitations. Moreover, its coherent detection and IQ modulation jointly encode amplitude and phase, realizing complex-valued arithmetic for full-range operations in transformers. MDTransformer offers analog multiplication with sub-4-bit effective precision and inter-modal crosstalk below -30 dB. Its inverse-designed approach also offers scalable and full compatibility with single-laser continuous-wave operation at 1550 nm. Experimental results show that MDTransformer achieves 40.4% area reduction, 63.6% power saving, 40.6% energy saving, and comparable latency over the state-of-the-art PTA across different workloads (i.e., DeiT-Tiny/Small/Base and BERT-Base/Large). These results show that MDTransformer offers a practical solution for high-performance and energy-efficient transformer-based systems.
[AI-4] Empirical Evaluation of Out-Of-Distribution Performance of Tabular Foundation Models
链接: https://arxiv.org/abs/2607.26000
作者: Malena Loza,David Chushig-Muzo,Eva Milara,Luis Bote-Curiel,Luis Estrada-Petrocelli,Felipe Grijalva
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Tabular Foundation Models (TFMs) have emerged as novel approaches for tabular predictive tasks, demonstrating competitive predictive performance to ensemble tree-based models. Most TFMs are trained and evaluated on independent and identically distributed data, but this assumption changes in real-world scenarios due to distribution shifts, which compromise the robustness of models. Limited research has been conducted of TFMs under distribution shifts. We present an empirical evaluation of Out-Of-Distribution (OOD) performance of nine TFMs, spanning diverse pre-training strategies and architectures: TabPFNv2, TabPFNv2.5, TabPFNv2.6, TabPFNv3, TabICL, TabICLv2, Mitra, LimiX and TabFM. Three real-world datasets from the TableShift study were considered (HELOC, Voting, Childhood Lead), covering label, socioeconomic, and geographic shift types. Our results show that all evaluated TFMs degrade systematically under distribution shift regardless of pre-training strategy, with shift gaps ranging from 0.003 to 0.060 depending on shift type. The relationship between in-distribution and OOD predictive performance documented for classical tabular models extends into TFMs. We also identified a scalability gap, as high-performing models demand significant memory and computational resources beyond what standard deployment infrastructure can support. This study extends existing benchmarks for OOD in tabular data, providing evidence to support their adoption in high-stakes domains characterized by structural distribution shifts.
[AI-5] Does Runtime Topology Context Improve LLM -Generated Kubernetes Security Patches? ESORICS2026
链接: https://arxiv.org/abs/2607.25995
作者: Farooq Shaikh
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注: Accepted at the Workshop on the Use of Large Language Models for Cybersecurity (LLMSec), co-located with ESORICS 2026
Abstract:Kubernetes is central to the cloud-native ecosystem, orchestrating containerised workloads. Recent work suggests that large language models (LLMs) can automate cluster security remediation, generating configuration patches from Kubernetes Security Posture Management (KSPM) findings without human authoring. Such systems, however, prompt the model with each finding in isolation from the live service call graph, assuming general hardening knowledge suffices. This assumption breaks down whenever a patch must preserve a runtime service dependency invisible to the model: an otherwise compliant fix then carries a destructive functional blast radius, crashing downstream callers or silently severing call edges across the cluster. Whether live cluster context improves patch correctness has not been measured under controlled conditions across multiple dependency classes. We introduce KuTIE (Kubernetes Topology Intelligence Engine), which builds a live cluster context from Istio call edges, Trivy KSPM findings, and the service-account bindings a workload reads, and conditions LLM patch generation on it. It is evaluated on VulnCare, a purpose-built 36-deployment, four-namespace healthcare cluster with 31 injectable findings across seven dependency classes, each labelled by topology dependence against cluster ground truth. Across 248 trials, topology context raises topology-dependent patch correctness from 11.1% to 78.0% ( \Delta = 0.669 ), a gap that holds for every model and for six of seven classes, from credential and network-policy ( \Delta = 0.95 ) to role-based access control ( \Delta = 0.31 ); a topology-independent control exhibits no such effect ( \Delta = 0.0 ), isolating the result from generic prompt enrichment. Supplying the live service-call graph and the service-account bindings it exposes thus improves remediation of topology-dependent findings well beyond scanner-only context.
[AI-6] MemLens: A Value-Aware Memory Management System with Interactive Analytics for LLM -based Agents
链接: https://arxiv.org/abs/2607.25992
作者: Shuyue Wei,Chang Liu,Zimu Zhou,Yongxin Tong,Lizhen Cui
类目: Databases (cs.DB); Artificial Intelligence (cs.AI)
备注:
Abstract:Recently, memory management has become a key infrastructure for LLM-based agents, as it directly affects long-horizon reasoning, personalized responses, and knowledge reuse. However, existing LLM memory systems typically adopt a coarse-grained (utility-agnostic) manner that treats heterogeneous user-LLM interaction records uniformly, leading to redundant and low-impact records persisting in the memory repository. To address this challenge, we present MemLens, a value-aware memory management system that takes memory records as first-class data objects. MemLens provides an end-to-end interactive analytics dashboard that exposes the complete memory lifecycle, including Shapley-style memory evaluation, value-aware storage, and memory-assisted response. Through a study-copilot application, the system enables users to inspect memory values, visualize hierarchical memory structures, and compare various memory management strategies in terms of response quality, retrieval latency, and token consumption. Therefore, our MemLens can serve as an efficient, interpretable, and personalized long-term memory management system for LLM-based agents.
[AI-7] Reinforcement Learning for Code Optimization
链接: https://arxiv.org/abs/2607.25970
作者: Pierre Chambon,Kunhao Zheng,Juliette Decugis,Benoit Sagot,Gabriel Synnaeve
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 125 pages
Abstract:RL for code correctness is now established: have the model generate a program, run it against hidden test cases, and reward solutions that pass. Extending this to code optimization seems straightforward: just add execution time to the reward. But in practice, once timing drives the reward, small problems in measurement noise, reward sparsity, or GRPO instability overwhelm the signal and make RL fail: generated solutions are barely faster, and more of them can fail. We make execution time learnable through three stages: (1) how code is tested, by building DMC-Optim with large optimization tests and a calibrated sandbox; (2) how speed is turned into reward, by composing correctness and speed in the RL environment and using an offline simulator to predict the most promising configurations; and (3) how the model learns from that reward, by adapting GRPO and evaluation to the sparser, noisier timed-execution setting. On DMC-Optim, the strongest optimization-aware configurations improve strict top-50% pass@1 from 18.0% to 31.3% on Qwen 2.5 7B and from 30.7% to 50.4% on CWM 32B. These gains further increase at stricter percentiles such as top-30%, with 125% relative improvement for CWM 32B, while preserving pure-correctness scores. When the timing sandbox is degraded, robust optimization RL reaches 100% to 200% improvement over standard RLVR, depending on the evaluation criterion. On LCB, CWM 32B wins up to 83% of median-sample speed comparisons against standard RLVR. Relative to the fastest correct human submissions per problem, it reaches about half the human rate of complexity-class improvements (14% vs. 28%).
[AI-8] Large Language Model for Operations Research Formulation Selection in Multi-Warehouse Inventory Allocation
链接: https://arxiv.org/abs/2607.25956
作者: Jintao Xu,Yingzheng Ma,Jiong Dong,Yongzhi Qi,Jianshen Zhang
类目: Artificial Intelligence (cs.AI); Optimization and Control (math.OC)
备注:
Abstract:Multi-warehouse inventory allocation is typically formulated as a mixed-integer programming (MIP) problem, yet no single formulation consistently matches heterogeneous instance-level regimes induced by demand concentration, inventory imbalance, replenishment scale, service constraints, and forecast volatility. We study this issue as instance-wise operations research (OR) formulation selection, where each allocation instance is assigned to a solver-executable formulation from a candidate OR expert library. We propose a solver-guided large language model (LLM) framework for OR formulation selection, in which each OR expert corresponds to a MIP formulation encoding a distinct allocation priority. To train the selector, the framework first constructs balanced expert-conditioned supervised fine-tuning (SFT) records for schema learning, and then uses MIP solver evaluation on historical instances to convert solver-evaluated allocation-quality gaps into margin-weighted identity preference optimization (IPO) preferences and per-instance expert-score metadata for reward lookup during group relative policy optimization (GRPO) to assign rewards to sampled responses. Experiments on multi-warehouse inventory allocation instances from JD \mathord. com, one of China’s largest e-retailers, demonstrate that GRPO substantially improves expert-selection accuracy relative to the SFT+IPO selector and, more importantly, produces higher realized allocation quality than both the preference-trained selector and the best fixed formulation. With GRPO, Hit Ratio@1 and Hit Ratio@2 increase from 21.45% to 50.42% and from 70.47% to 82.31%. The resulting selector achieves an allocation accuracy gain of 12.57 percentage points over the incumbent baseline, outperforming both the SFT+IPO selector and the best fixed OR expert, and reduces the gap to the ex-post oracle to 4.85 percentage points.
[AI-9] dtControl2varepsilon: Trading Optimality for Explainability in MDPs via Decision Trees
链接: https://arxiv.org/abs/2607.25925
作者: Tereza Kinská,Jan Křetínský,Tobias Meggendorfer,Sabine Rieder,Maximilian Weininger
类目: Artificial Intelligence (cs.AI)
备注: This paper is accepted at FMCAD26
Abstract:Over the past decade, decision trees have been used to represent controllers (a.k.a. policies) in an explainable way, with dtControl2 as a current state-of-the-art tool. However, for systems that are large or have many corner cases, even such representations tend to be too complex and not human-comprehensible. Unfortunately, reducing the size of the decision tree is not straightforward, as missing just a single crucial case might result in an incorrect controller. We tackle this issue in the setting of Markov decision processes, extending dtControl2 by " \varepsilon " functionality: Given an allowed imprecision \varepsilon \geq 0 , we construct a smaller decision tree, distilling the essence of the controller, while still guaranteeing its \varepsilon -optimality. This enables us to provide tunably simpler explanations, omitting a controllable amount of detail. Our tool constructs decision trees that are orders of magnitude smaller than the state of the art.
[AI-10] Penelope: Localized Latent Recurrence for Efficient Structured Reasoning
链接: https://arxiv.org/abs/2607.25915
作者: Yutong Chen,Shouqian Shi,Xinran Liu,Haochen Wang,Jiaying Wang,Tianxing Xu,Yuanxi Wang,Zirui Ding
类目: Artificial Intelligence (cs.AI)
备注: 8 pages, 2 figures
Abstract:Complex structured reasoning tasks often require additional computation, yet current language models obtain it mainly by increasing parameter scale or by serializing intermediate steps as chain-of-thought (CoT) tokens. The former raises training and deployment costs, while the latter ties reasoning computation to autoregressive output length. We introduce Penelope, an efficient latent-reasoning framework for pretrained decoder-only Transformers that localizes recurrent computation to a selected decoder interval. The lower decoder prefix is evaluated once to construct a problem-conditioned boundary memory, which is then iteratively refined through time-modulated GRU dynamics and recurrent readout states before answer generation. A progressive CoT-to-latent curriculum transfers visible reasoning into this internal recurrent path, allowing additional computation to be allocated in latent space without repeatedly executing the complete decoder or generating a long intermediate trace. Experiments on open-source structured-reasoning benchmarks show that, at validation-selected latent budgets, Penelope attains competitive accuracy relative to established latent-reasoning models while reducing measured inference latency. These results show that latent refinement can be localized to a narrow decoder interval, reducing repeated full-decoder execution without generating a long visible reasoning trace and providing a practical accuracy-efficiency tradeoff for decoder-only Transformer models.
[AI-11] oward Standardized Cross-Vendor Agent Tool Trust Management in Autonomous Networks
链接: https://arxiv.org/abs/2607.25914
作者: Ravi Kant Sharma,Ashutosh Uttam,Ajay Kumar
类目: Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR); Networking and Internet Architecture (cs.NI)
备注: 22 pages, 7 figures, 9 tables, 4 algorithms
Abstract:Autonomous Network Levels 4-5 require AI agents to invoke tools across vendor boundaries without human oversight, yet existing management standards lack a standardized mechanism for cross-vendor trust visibility. When a tool from Vendor B is compromised, agents from Vendor A continue invoking it – unaware of the trust degradation – causing cascading service impact. We present AgentToolMO, a proposed 3GPP NRM information model for agent tool trust management. The model comprises: a formally defined trust state machine with provable graduated enforcement, damped cascade propagation with bounded convergence, cross-vendor trust notifications via existing Management Services (MnS) interfaces, and retroactive impact assessment through NRM dependency graph traversal. Simulation-based evaluation across multi-vendor topologies shows that standardized cross-vendor notifications reduce blast radius from hours-scale undetected propagation to near-real-time containment bounded by MnS notification delivery, with cascade convergence guaranteed in bounded iterations and sub-linear notification scaling across vendor domains. The framework operates within existing 3GPP management infrastructure, leverages existing protocols, and provides a standardization pathway for trustworthy multi-vendor autonomous network management.
[AI-12] SAM3D-Guided Object-Centric Representation Alignment for Vision-Language-Action Models
链接: https://arxiv.org/abs/2607.25912
作者: Zonghe Liu(1),Shanyuan Jie(2),Xiaoquan Sun(3),Chen Cao(1),Zetian Xu(1),Zongsheng Liu(4),Jiayu Chen(1 and 5) ((1) University of Hong Kong, (2) Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences, (3) Huazhong University of Science and Technology, (4) Beijing University of Aeronautics and Astronautics, (5) Infiforce)
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: 8 pages, 4 figures
Abstract:Vision-Language-Action (VLA) models have shown strong potential for general robot manipulation, but most existing models rely on 2D visual-language backbones and lack fine-grained 3D understanding of target objects, especially under occlusion, pose variation, scale changes, and precise spatial interaction. We propose an object-centric 3D representation alignment framework built upon \pi_0 , using SAM3D as a frozen 3D teacher to provide target-object 3D priors during training. Specifically, we localize task-relevant objects with object recognition models, generate corresponding object masks, and use SAM3D to extract dense object-level 3D representations, which are aligned with intermediate visual features of \pi_0 . This enables the policy to internalize target-object 3D information while preserving the original RGB-language-to-action inference pipeline without requiring depth, point clouds, masks, SAM3D, or additional 3D modules at test time. Simulation experiments show consistent improvements, achieving 99.1% on LIBERO and an average length of 4.11 on CALVIN. Real-world experiments further demonstrate that our method is particularly effective in long-horizon manipulation scenarios where the robot must focus on different target objects across multiple subtasks.
[AI-13] Interactive Reward Agent : GUI Task Evaluation via Environment-State Verification
链接: https://arxiv.org/abs/2607.25904
作者: Chenrui Shi,Yuwei Wu,Yang Liu,Ruining Feng,Zirui Shang,Zhi Gao,Lifeng Fan,Che Sun
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Graphical user interface task evaluation aims to determine whether a GUI agent has successfully completed a user instruction. Automated GUI task evaluation has received increasing attention because the evaluation results can serve as reward signals for both test-time scaling and post-training. However, reliable GUI task evaluation remains challenging because the judgments often require access to environment states, such as system configurations, file data, and application settings, beyond the screenshots of execution trajectories. In this paper, we propose an interactive reward agent (IRA) based on a propose-then-verify framework to acquire and verify evidence from the post-execution environment. Given a task instruction and a GUI environment after the GUI agent execution, IRA first proposes the task completion conditions and then verifies them by invoking system tools, application tools, and GUI tools. This design combines evidence from both visible interfaces and the environment state in an interactive process. We further introduce GUI-RewardBench, a benchmark of 321 GUI task trajectories spanning 10 Ubuntu desktop application categories. Experiments show that IRA achieves 86.9% accuracy on GUI-RewardBench, outperforming existing evaluator baselines. We further apply IRA to reinforcement learning of GUI agents, achieving a 34.0% OSWorld success rate, which demonstrates that IRA can provide effective reward signals for training GUI agents.
[AI-14] Messier: A High-Resolution Corpus for Cross-Benchmark Agent Evaluation
链接: https://arxiv.org/abs/2607.25891
作者: Stefan Krsteski,Charlotte Meyer,Guillaume Allegre,Tony O’Halloran,Alexandre Sallinen
类目: Artificial Intelligence (cs.AI); Databases (cs.DB)
备注:
Abstract:Evaluating AI agents in interactive environments is hindered by fragmented tasks, scaffolds, verifiers, and scoring rules. Existing efforts focus on narrow settings, remain limited in scale, or require costly reruns, leaving much of the empirical record incomparable. We introduce Messier, a unified corpus of 957,253 records that span 30 benchmarks, 714 agents, 11,891 tasks, and 74,205 verifiers. Messier consolidates public benchmark scores and supplements them with five-agent runs across six underrepresented professional and scientific domains, including a recent legal benchmark. Each record is standardized by model, scaffold, environment, task, verifier, and aggregation rule, with SOC/NAICS classifications for occupational and industry analysis. Using this corpus, we show frontier progress is uneven across benchmark types, with “function calling” saturated, “programming” improving the fastest, and “enterprise workflows” remaining the most challenging. Furthermore, counterfactual rescoring shows that strict all-pass aggregation in multi-verifier tasks can obscure progress and artificially alter agent rankings. From these standardized records, we derive capability scales that align with Epoch’s Evaluation Capability Index rankings at Spearman \rho = 0.81 and can be specialized by domain, occupation, action space, or verifier type. Messier provides a foundational, reusable infrastructure for agent capability scaling, benchmark auditing, and fine-grained analysis of evaluation failures.
[AI-15] Distributing Security Controls Through Harness Engineering
链接: https://arxiv.org/abs/2607.25890
作者: William Robert Gore
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:AI coding agents are being adopted at historic speed, yet security and risk concerns remain the primary barrier to scaling agentic AI across organizations. Existing security controls for coding agents are not systematically distributed to engineering teams, and vendor-native solutions introduce ecosystem dependencies that may not suit every deployment context. This paper investigates whether off-the-shelf security controls can be implemented on commercial AI coding agents and scaled to a distributed user base via a custom agent harness. A phased testing methodology was applied across four agent configurations — two commercial agents with and without controls, a baseline harness, and a security-hardened harness — using a 23-test suite derived from the OWASP Top 10 for Agentic Applications. SHarD (Secure Harness Distribution), a distributable harness built on the Pi agent harness, demonstrated that three categories of security controls — OS sandboxing, skill scanning, and tool restriction — can be embedded and distributed via a single install command while retaining equivalent efficacy to direct installation on commercial agents. SHarD achieved an adjusted score of 100%, matching the best securely configured commercial agent, with no regression across any test category. Notable observations include evidence that model non-determinism produces inconsistent security outcomes and that autonomous agent behavior can cross system boundaries in ways that OS sandboxing directly mitigates. Initial characteristics toward a control harness fitness framework are proposed, and a third research question is identified for future investigation.
[AI-16] A Machine-Learning-Based Gas Lift Optimization Workflow for Unconventional Fields
链接: https://arxiv.org/abs/2607.25885
作者: Sha (Sasha)Miao,Alexandra Vendetti,Logan Smart,Gunta Chomchalerm,Yang Chen,Christopher Frazier,Dustin Haralson,Jeremy Sorenson,Xiao Ma,Huafei Sun,Aaron Shinn,Haining Zheng,Xiao-Hui Wu,Peng Xu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
备注: 11 pages, 13 figures
Abstract:In this paper, we present an automated data-driven workflow using Machine Learning (ML) for gas lift optimization in unconventional fields. This workflow integrates a ML model that accurately forecasts the Gas Lift Performance Curve, and a Bayesian Optimization Framework to solve for the optimal gas injection rates under the constraints of facility capacity. The ML model leverages the historical production time series data without requiring downhole gauges or multi-rate well tests. We piloted this workflow on 30 wells across 5 well pads in Bakken and obtained 5% production uplift on average. With the success of the pilot, we have now fully-deployed this workflow in Bakken across 200+ gas lift and plunger-assisted gas lift (PAGL) wells. Moreover, the ML-based gas lift optimization workflow presented in this paper is an effective and economic solution for other assets where downhole data or multi-rate testing are not available/feasible due to cost or facility constraints.
[AI-17] Runtime Uncertainty Monitoring for LLM -Based Multi-Agent Systems Using Bayesian Networks WWW
链接: https://arxiv.org/abs/2607.25877
作者: Bart Custers,Koorosh Aslansefat
类目: Artificial Intelligence (cs.AI)
备注: This paper got accepted for Ninth International Workshop on Artificial Intelligence Safety Engineering (WAISE 2026): this https URL
Abstract:This paper investigates how multi-agent systems (MAS)-based on large language models (LLMs) can support actuarial risk modelling, with a particular focus on uncertainty quantification. Actuarial workflows represent a high-stakes decision-support setting where unreliable outputs may lead to incorrect risk assessment, unfair pricing, and regulatory non-compliance. To address uncertainty introduced by the probabilistic nature of LLMs and dependencies between agents, a multi-agent framework is proposed in which specialised agents perform data preparation, modelling, review, and explanation tasks under a central hub. The main contribution is a novel approach to uncertainty propagation using token-level log-probabilities and a Bayesian Network. Importantly, log probabilities are not treated as direct probabilities of correctness or task success. Instead, length-normalised log-probability summaries are transformed into calibrated task-level confidence estimates before incorporation into the Bayesian Network. Results show that the framework reproduces baseline actuarial performance while providing additional insight into workflow stability and runtime uncertainty propagation.
[AI-18] A2TTA: Anchored-and-Agile Test-Time Adaptation for Evolving Traffic Sensor Networks
链接: https://arxiv.org/abs/2607.25875
作者: Du Yin,Xiachong Lin,Yue Tan,Jinliang Deng,Estrid He,Hao Xue,Flora D. Salim
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Under Review
Abstract:Traffic forecasting is important for efficient traffic management and route planning in smart cities. Existing traffic forecasting studies typically assume fixed sensor graphs, overlooking the continuous evolution of real-world traffic networks, e.g., ongoing road network construction and evolving human mobility patterns. These dynamic changes can substantially degrade conventional forecasting models, motivating test-time adaptation (TTA) to efficiently adapt pretrained models during deployment. However, applying TTA to evolving traffic sensor networks remains challenging in two aspects. First, topology expansion introduces new sensors and connections, continuously reshaping the sensor graph. Second, tem- poral shifts vary in time scale and stability, requiring differentiated adaptation to long-term and short-term shifts. In this study, we address these challenges by proposing A2TTA, an Anchored-and-Agile Test-Time Adaptation framework for evolving traffic sensor networks, which transforms topology-induced forecasting errors into an expandable output calibration problem and separates tem- poral adaptation into persistent global correction and agile context-specific specialization. By jointly addressing topology evolution and multi-scale temporal shifts, A2TTA enables efficient and robust adaptation to continuously evolving traffic environments. Extensive experiments on ten real-world traffic networks demonstrate that A2TTA consistently improves forecasting performance across different backbones, datasets, and prediction horizons. Our code is available in this https URL.
[AI-19] How Do LLM s Read Bug Reports? An Empirical Study of Attention in LLM s for Automated Program Repair
链接: https://arxiv.org/abs/2607.25873
作者: Ramtin Ehsani,Irene Manotas,Saurabh Pujar,Luca Buratti,Preetha Chatterjee
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: Accepted at the 41st IEEE/ACM International Conference on Automated Software Engineering (ASE) 2026 Conference
Abstract:Large Language Model (LLM)-based Automated Program Repair systems are advancing rapidly, yet their performance remains inconsistent. Even when provided with the same contextual information, an LLM may generate a correct patch for one bug but fail on another closely related bug. Why this happens remains poorly understood, and it is unclear how LLMs prioritize the diverse information in bug reports and whether model attention affects repair success. In this paper, we present the first empirical study of attention patterns in LLM-based program repair, providing interpretable insights into how models process bug reports and where their attention is concentrated during repair. We analyze 319 real-world Python and Java bugs from SWE-bench Verified and Multi-SWE-bench to study (RQ1) how model attention is distributed across bug report sections, (RQ2) how attention patterns within each section differ between successful and unsuccessful repairs, and (RQ3) how these patterns compare to information developers consider important for bug fixing. We find that successful repairs are characterized by diffused attention across multiple diagnostic components such as bug descriptions, stacktraces, and test cases, while failures often exhibit over-localized attention toward metadata such as version information. We further observe that stronger alignment between model attention and developer-identified key sections and phrases is associated with higher repair success. Our results provide the first empirical evidence that attention misallocation is a key factor in LLM-based APR failures, and offer actionable insights for designing more interpretable and reliable future APR systems.
[AI-20] HiSkill: Empowering LLM Agents with Hierarchical Skill Graphs
链接: https://arxiv.org/abs/2607.25853
作者: Yu Hao,Jinxuan Cai,Qi Zhang,Yawen Li,Zhiqiang Zhang,Chuan Shi,Cheng Yang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Skills have become an important abstraction for enabling large language model (LLM) agents to reuse past experience in long-horizon interactive tasks. However, existing trajectory-to-skill methods often produce flat collections of high-level textual skills that are stored and retrieved independently, leaving skill relations underutilized and maintaining a gap between high-level skills and executable actions. In this paper, we propose HiSkill, a hierarchical skill graph framework that organizes interaction trajectories into a directed graph with skill nodes, AtomicOp nodes, and typed edges. Specifically, the graph connects reusable high-level skills with executable action templates, while also capturing decomposition, temporal transition, compatibility, support, and recovery relations among them. At inference time, HiSkill retrieves a compact task-relevant subgraph and performs subgraph-guided task execution, where a symbolic task state, an active skill, and the retrieved subgraph guide the LLM agent to switch skills, select AtomicOps, and ground executable actions iteratively. Experiments on three interactive environments show that HiSkill outperforms state-of-the-art baselines while reducing inference token consumption, demonstrating the effectiveness of bridging high-level skills and executable action grounding through a hierarchical skill graph. Our data and code is available at this https URL.
[AI-21] Distributed Constraint Optimization via Online Learning and Iterative Pricing with Application to Large-Scale Satellite Scheduling
链接: https://arxiv.org/abs/2607.25835
作者: Itai Zilberstein,Pranav Rajbhandari,Steve Chien,Tuomas Sandholm
类目: Artificial Intelligence (cs.AI); Computer Science and Game Theory (cs.GT)
备注:
Abstract:Distributed constraint optimization problems (DCOPs) provide a popular framework for distributed decision making under limited communication, but many real-world instances are too large to solve monolithically. We address this challenge from two complementary directions. We revisit the connection between DCOPs and potential games, and adapt modern online learning algorithms for equilibrium finding to DCOPs. We show that these algorithms are competitive with representative incomplete DCOP algorithms. We then turn to decomposition frameworks for large-scale DCOPs, motivated by large-scale decentralized satellite scheduling. We propose a new framework that separates a DCOP into two interacting subproblems: a high-level meta-DCOP for task allocation, and independent local optimization problems for scheduling. To couple the two levels, we develop a novel iterative pricing method that updates the meta-level utilities using feedback from the local optimizers. Combining our online learning methods with our iterative pricing framework, we obtain near-optimal performance on real-world decentralized satellite scheduling problem instances, fulfilling over 99% of observation requests compared with 87% for state-of-the-art baselines.
[AI-22] Speculate While You Reason : Teaching Agents Agent s to Predict Their Next Tool Call via Joint Agent-Speculator RL
链接: https://arxiv.org/abs/2607.25816
作者: Jiabao Ji,Yujian Liu,Li An,Rohit Jain,Gungor Polatkan,Siyu Zhu,Shiyu Chang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large language model agents often spend substantial wall-clock time waiting for tool call results. Tool-call speculation can hide this latency by predicting and pre-executing an agent’s next tool call if the prediction matches the agent’s eventual tool call, but existing speculators are typically separate draft models or cached traces that are poorly aligned with the deployed agent’s own behavior. We identify this speculator-agent gap and show that the target agent itself is a strong next-call speculator. This points to a simpler design: unifying the agent and speculator within the same model. In this paper, we introduce the self-speculating agent, a single model that both solves tasks in agent mode and predicts its next tool call from partial trajectories in speculator mode, fully reusing prefix KV cache. To enable this dual-mode agent without degrading performance, we propose a joint agent-speculator reinforcement learning method, which derives speculation targets from the agent’s own rollouts and alternates agent and speculator updates. Across agentic search QA and conversational tool-use agentic tasks, our method improves average next tool-call Hit@1 from 44.1 to 61.2 for Qwen3-4B and from 48.9 to 66.3 for Qwen3.5-4B, while preserving agent task success.
[AI-23] SpectONet: A Physics-Guided Spectral Deep Operator Network for Euler-Bernoulli Beam Dynamics
链接: https://arxiv.org/abs/2607.25790
作者: Shivani Saini,Ramesh Kumar Vats,Arup Kumar Sahoo
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Dynamical Systems (math.DS)
备注:
Abstract:This paper proposes a novel physics-guided spectral deep operator network, termed SpectONet, for solving Euler-Bernoulli beam (EBB) vibration problems. The proposed framework integrates the operator-learning capability of DeepONet with physics-informed constraints and Chebyshev-Gauss-Lobatto (CGL) sensor placement. Unlike conventional DeepONet frameworks, which commonly employ uniformly distributed sensors, SpectONet uses nonuniform spectral sensor locations with a higher concentration of points near the domain boundaries. This sampling strategy improves the finite-dimensional representation of boundary-sensitive structural responses while requiring only a limited number of branch-network inputs. The governing beam equation, together with the associated initial and boundary conditions, incorporated into the training objective to promote physically consistent and generalizable predictions. Numerical experiments on three synthetic EBB vibration problems and a real-world bridge vibration dataset demonstrate the effectiveness of the proposed framework. Comparisons with strong baselines such as, Vanilla DeepONet, PI-DeepONet, PINN, and CNN-UNet show that SpectONet consistently achieves lower prediction errors across all considered evaluation metrics. In particular, SpectONet achieves at least (64%) improvement over the considered baseline models across the three synthetic problems and at least (37%) for the real-world problems. These results demonstrate that SpectONet provides an accurate, computationally efficient, and physically consistent operator-learning framework for structural vibration analysis.
[AI-24] Loss Invariance Determines What Concept Layers Encode: Volume Grounding in Echocardiography
链接: https://arxiv.org/abs/2607.25748
作者: Hyunkyung Han,Min Jung Kim
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Objective: Concept bottleneck models route prediction through interpretable intermediate variables, and their validity is normally judged by how accurately those variables are predicted. We ask whether that judgement is sufficient, using left ventricular volumes as the concepts underlying ejection fraction estimation from echocardiographic video. Methods: A video transformer encoder was trained on a publicly available echocardiography dataset. End-systolic and end-diastolic volumes formed a concept layer from which ejection fraction was computed analytically, with no residual path to the output. We compared training under an ejection fraction objective alone against training with additional supervision of the volumes in millilitres, and evaluated both on 1276 held-out studies. Results: The concept bottleneck did not increase ejection fraction error relative to direct regression, at 6.89 against 7.13 mean absolute error. Without volume supervision, however, the spread of predicted volumes collapsed to 0.1 millilitres against reference spreads of 35.7 and 45.7 millilitres, while correlation was partly preserved. We show that this follows from an invariance property of the objective: ejection fraction is a ratio and is unchanged when both volumes are rescaled, so the loss determines the concept layer only up to scale. Supervision in absolute units reduced volume error from 89.8 to 25.8 millilitres at a cost of 0.4 in ejection fraction error. Conclusion: Concept accuracy alone can conceal a concept layer that carries no physical scale. Significance: Interpretable intermediate variables in clinical models should be validated against the invariance structure of the training objective, not only against prediction accuracy. Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2607.25748 [cs.AI] (or arXiv:2607.25748v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2607.25748 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Hyunkyung Han [view email] [v1] Tue, 28 Jul 2026 14:11:10 UTC (314 KB) Full-text links: Access Paper: View a PDF of the paper titled Loss Invariance Determines What Concept Layers Encode: Volume Grounding in Echocardiography, by Hyunkyung Han and 1 other authorsView PDFHTML (experimental)TeX Source view license Current browse context: cs.AI prev | next new | recent | 2026-07 Change to browse by: cs References Citations NASA ADSGoogle Scholar Semantic Scholar export BibTeX citation Loading… BibTeX formatted citation loading… Data provided by: Bookmark checked="checked"class=“labs-tab-input”> Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, Media Code, Data and Media Associated with this Article alphaXiv Toggle alphaXiv (What is alphaXiv?) Links to Code Toggle CatalyzeX Code Finder for Papers (What is CatalyzeX?) DagsHub Toggle DagsHub (What is DagsHub?) GotitPub Toggle Gotit.pub (What is GotitPub?) Huggingface Toggle Hugging Face (What is Huggingface?) ScienceCast Toggle ScienceCast (What is ScienceCast?) Demos Demos Replicate Toggle Replicate (What is Replicate?) Spaces Toggle Hugging Face Spaces (What is Spaces?) Spaces Toggle TXYZ.AI (What is TXYZ.AI?) Related Papers Recommenders and Search Tools Link to Influence Flower Influence Flower (What are Influence Flowers?) Core recommender toggle CORE Recommender (What is CORE?) Author Venue Institution Topic About arXivLabs arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv’s community? Learn more about arXivLabs. Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?) mathjaxToggle(); We gratefully acknowledge support from our major funders, member institutions, , and all contributors. About Help Contact Subscribe Copyright Privacy Accessibility Operational Status (opens in new tab) Major funding support from
[AI-25] Shared Voxel-Map-Based Cooperative Indoor UAV Guidance with a Multi-Agent Soft Actor-Critic Controller
链接: https://arxiv.org/abs/2607.25728
作者: Thomas Hickling,Dylan Wynne,Yu Su,Nabil Aouf
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 11 pages, 17 figures
Abstract:This paper presents a cooperative indoor UAV guidance framework that combines a shared voxel-map world model with a multi-agent Soft Actor-Critic (MASAC) controller. Multiple drones fuse 360 LiDAR observations into a common world-frame occupancy map, which is converted into a compact bird’s-eye-view (BEV) representation and provided to each agent as an ego-aligned local crop. This integrate-in-world, act-in- ego design enables consistent multi-UAV spatial fusion whilst retaining decentralised continuous control. The policy combines BEV map features, near-field obstacle observations, and compact goal and peer-state information within a centralised-training, decentralised-execution framework. In simulation, the learned controller achieves a 90.3% success rate in corridor navigation, outperforming Astar planning, an artificial potential field controller, and a prior guidance method. To address residual sim-to-real mismatch, the simulation-trained policy is further adapted using offline imitation fine-tuning from real-world data. Real-world experiments in GNSS-denied indoor environments demonstrate stable two-UAV cooperative operation across increasingly chal- lenging obstacle layouts. The results show that shared voxel-map representations provide an effective and scalable spatial substrate for learned cooperative indoor UAV guidance.
[AI-26] Nudging Sustainable Choices through LLM -Generated Recommendation Explanations
链接: https://arxiv.org/abs/2607.25726
作者: Haya Halimeh,Dietmar Jannach,Oliver Müller
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Recommender systems mediate everyday consumption, offering a promising channel for encouraging sustainable choices. Prior research shows that explanations influence users’ perceptions of recommendations and can support more informed decisions. We argue that explanations can also serve as behavioral nudges by foregrounding sustainability information at the moment of choice. This study investigates how different behavioral framings of sustainability information in recommendation explanations affect user choices and perceptions. Using generative AI, we generate sustainability-aware explanations by drawing on nudge theory and validate them through human evaluation and LLM-as-a-judge audits. Building on this foundation, we conduct two randomized studies ( N = 529 ) in a low involvement domain (instant coffee) and a high involvement domain (hotel bookings), in which participants choose among preference matched recommendations accompanied by these explanations. Our results show that, across both domains, merely disclosing sustainability information in explanations does not change choices, whereas framing that information or invoking a descriptive social norm significantly increases sustainable selections and eases decision-making. Notably, perception and behavior diverge, as plain disclosure improves explanation evaluations without translating into more sustainable selection behavior. Our work demonstrates how LLMs can generate theory-grounded explanations at scale, pointing toward practical explanation-based interventions for social good. We conclude by discussing implications for adaptive explanation design with generative AI.
[AI-27] From Deterministic to Generative Deep Learning for Urban Air Quality Reconstruction from Sparse Observations
链接: https://arxiv.org/abs/2607.25687
作者: Abhishek A.Sabnis,Mihai Mitrea,Lya Lugon,Karine Sartelet,Marc Bocquet,Xiaoyuan Cheng,Shupeng Zhu,Sibo Cheng
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Full-field reconstruction of air pollution is essential for evaluating pollution exposure and supporting public health decision-making. However, the complex interactions among pollutants, hard-to-predict weather patterns, and limited monitoring station coverage make this a complex task. We apply deep learning techniques to provide fast and accurate reconstructions from sparse observations of four key pollutants: NO2, O3, PM2.5 and PM10. Models are trained on full-field simulation data and evaluated on real-world observations collected from 9 to 28 monitoring stations in the city of Paris. We introduce a diffusion-based generative framework for multi-pollutant reconstruction and benchmark its performance against deterministic deep learning models. Despite noisy observations and strong spatial variability, the models achieve high structural similarity on simulated validation data and produce realistic spatial patterns on real-world observations, as indicated by power-spectrum analysis. We introduce data augmentation methods that enable transfer to real-world observations without retraining, allowing the models to generalise beyond the training period. These findings highlight the potential of ML models for reliable real-world deployment in air pollution reconstruction tasks.
[AI-28] Cognivia: A Cognitive Behavioral Therapy Copilot for Evidence-Based Mental Healthcare
链接: https://arxiv.org/abs/2607.25681
作者: Qi Chen,Siria Xiyueyao Luo,Jian Wang,Yuan Shi,Haocong Rao,Xuejiao Zhao
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Cognitive distortion amplifies negative emotions and contributes to mental health disorders. Cognitive Behavioral Therapy (CBT) is an effective way to address cognitive distortions, but its large-scale application is limited by the shortage of professional therapists. Although large language models (LLMs) have recently been explored for mental health applications, existing methods still suffer from limited domain specificity, overly flattering responses, and the absence of well-defined annotations for cognitive distortions. This paper proposes Cognivia, an evidence-based artificial intelligence therapist that integrates automatic cognitive distortion identification and rational response generation. Our framework is built on authoritative CBT texts widely regarded as core paradigms and standard references. It is further augmented with mental health question-answer (Q and A) data, and employs multi-stage prompting and structured generation strategies under the supervision of behavioral science experts. Then we fine-tune a lightweight LLM on this augmented CBT dataset to obtain Cognivia. In addition, we propose the first hierarchical quality evaluation framework for assessing LLM-generated rational responses, developed through collaboration between AI researchers and behavioral science experts. Cognivia is evaluated using lexical metrics, LLM-based Judges with two complementary criteria, and human evaluation by 10 behavioral science experts. It consistently outperforms the baseline methods in cognitive distortion recognition and rational response generation, demonstrating its effectiveness. Our code is available at this https URL.
[AI-29] Rashomon Alignment
链接: https://arxiv.org/abs/2607.25680
作者: Moisés Santos,Peter van der Putten,Bernhard Pfahringer,Carlos Soares
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:We propose Rashomon Alignment (RA), a new measure to assess functional similarity between two models. Existing functional similarity measures are distributional, quantifying differences between outputs of models applied to real-world data. However, these measures can be regarded as ecologically valid only for regions in the input space represented by the available data. We introduce a geometrical perspective on functional model similarity, which estimates it across the entire data space, offering a comprehensive view of decision boundary alignment independent of any specific data distribution. We also propose geometric Rashomon Alignment as a measure of geometrical similarity, which is computed using data uniformly sampled from the instance space. We perform an experimental analysis on more than 90 datasets, examining critical cases where model alignment diverges from predictive accuracy. Our results show that geometrical and distributional alignment provide different and complementary perspectives on the similarity between models and algorithms. RA can be used for multiple purposes, including model selection, ensemble construction, and enhanced interpretability of machine learning models and algorithms.
[AI-30] DynaBridge: Dynamic Summary-Guided Cross-Task Multimodal Fusion for DASS-Structured Mental Health Assessment
链接: https://arxiv.org/abs/2607.25679
作者: Shiyu Teng,Haichen Yu,Jiaqing Liu,Hao Sun,Yu Song,Shurong Chai,Ruibo Hou,Lanfen Lin,Yen-Wei Chen
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Multimedia (cs.MM)
备注:
Abstract:Multimodal behavioral analysis offers a scalable approach to assessing depression, anxiety, and stress, yet generic fusion models often ignore the psychometric structure of questionnaire labels. In DASS-21, risk labels are derived from ordered symptom items through fixed item-to-subscale mappings. We propose \textbfDynaBridge, a dynamic summary-guided cross-task multimodal framework for DASS-structured mental health assessment. DynaBridge encodes acoustic, visual, and textual cues across multiple sessions and augments them with frozen-LLM-generated DASS-aware summaries as participant-level semantic evidence. It predicts ordinal item distributions, reconstructs depression, anxiety, and stress risk evidence from item-level soft scores, and fuses this evidence with direct multimodal risk predictions. A confidence-aware refinement strategy further incorporates high-confidence semantic cues conservatively. On the official AdoDAS validation split, DynaBridge outperforms the official baseline and representative multimodal methods, achieving 0.5012 mean F1 for D/A/S risk prediction and 0.3216 mean QWK for DASS-21 item prediction. These results show the value of bridging multimodal cues, semantic summaries, and DASS-21 psychometric structure.
[AI-31] DecoEvo: Score-Decoupled Co-Evolution of Solver and Rubric-Generator Skills in Text Space
链接: https://arxiv.org/abs/2607.25675
作者: Jiangwang Chen,Zixin Song,Junlin Liu,Shuaiyu Zhou,Haiyan Wu,Haihan Shi,Chenxi Zhou,Hanqing Li,Xiao Yang,Da Zhu,Guanjun Jiang,Hai Wan,Xibin Zhao
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Text-space optimization adapts large language models (LLMs) by editing external natural-language artifacts rather than model weights, so the optimized artifacts remain inspectable and the model can be treated as a black box. However, most existing text-space methods keep evaluation fixed. On open-ended tasks, this can become a bottleneck: once the solver improves on the criteria a rubric measures, omitted dimensions remain invisible to the optimization signal. Simply evolving the rubric is also unreliable when updates are selected by the current solver’s score, because apparent progress can come from making the rubric easier to satisfy. We introduce DecoEvo (Decoupled Co-Evolution), which co-evolves a solver skill and a rubric-generator skill under decoupled objectives without using gold rubrics during optimization. The solver skill is updated using criterion-level feedback, while the rubric-generator skill is revised through complementary audits of requirement coverage and response discrimination that are independent of aggregate solver score. This separation focuses generator updates on newly exposed solver weaknesses, reducing repeated emphasis on criteria the solver already satisfies. Under each benchmark’s official evaluation, DecoEvo outperforms all compared methods across five benchmarks and three LLM backbones, yielding 2.8–5.0% relative gains over SkillOpt in the five-benchmark average.
[AI-32] OmniDelta: Skill-Driven Budget Allocation for Token Compression in OmniLLM s
链接: https://arxiv.org/abs/2607.25669
作者: Haoyang Huang,Wenjie Huang,Tianqi Xu,Hongyaoxing Gu,Kang Tan,Yikai Fu,Yuhao Shen,Tianyu Liu,Baolin Zhang,Jun Zhang,Xinyi Hu,Jun Dai,Shuang Ge,Lei Chen,Yue Li,Mingchen Wang,Meng Zhang
类目: Artificial Intelligence (cs.AI)
备注: 24 pages, 8 figures
Abstract:Emerging Omni-modal Large Language Models (OmniLLMs) enable unified understanding of text, audio, and video, but their long audio-video token sequences introduce substantial memory and inference costs. Existing compression methods mainly focus on selecting important tokens under fixed budgets, leaving the preceding budget-allocation problem underexplored. We show that direct query-to-audio/video similarity is unreliable for inter-modal budget allocation, and that uniform intra-modal budgets can miss key evidence while retaining redundant content. To address these limitations, we propose OmniDelta, a training-free, skill-driven framework that couples intent-aware inter-modal allocation with content-aware intra-modal allocation. OmniDelta first constructs audio and video skill pools to shift the fixed retained-token budget according to query demand, then reallocates modality budgets over audio segments and video frames using local complexity and temporal redundancy. The resulting local budgets can be combined with existing pruning strategies, preserving the total retained-token ratio while changing where the budget is spent. Experiments on four audio-video benchmarks with two Qwen2.5-Omni models show that OmniDelta establishes a new accuracy-efficiency Pareto frontier across pruning ratios. At 25% token retention on Qwen2.5-Omni-7B, OmniDelta reduces GPU memory by 22.0% and achieves a 1.64x end-to-end speedup over full-token inference.
[AI-33] CoRT: Counterfactual Replay for Token-Level Rubric-Guided Policy Optimization
链接: https://arxiv.org/abs/2607.25659
作者: Bo-Wen Zhang,Junwei He,Wen Wang,Song-Lin Lv,Wentao Ma,Rongyi Lin,Shuhan Zhong,Lan-Zhe Guo
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Rubric-based reinforcement learning enriches language model training by evaluating model outputs against explicit criteria. Yet in GRPO-style pipelines, these structured judgments are reduced to a scalar response-level reward and converted into a response-level advantage, which is broadcast uniformly to all generated tokens. This leaves no explicit mechanism for allocating credit within a response, even when different criteria are grounded in different spans, formatting decisions, or semantic choices. We propose CoRT, a token-level credit weighting method for rubric-conditioned GRPO. Instead of training an auxiliary token scoring model, CoRT uses counterfactual replay to rescore the same sampled response under the original rubric-conditioned prompt and a matched criteria-free prompt. The resulting tokenwise log-likelihood contrasts serve as a proxy for dependence on the rubric context. CoRT maps these contrasts to bounded, response-normalized weights and uses them to redistribute the signed GRPO advantage across tokens, without introducing an auxiliary scorer or changing the response-level reward. Experiments across instruction-tuned models and reward granularities show that CoRT improves over matched response-level GRPO in the vast majority of comparisons, with an average gain of 4.4 percentage points. The method remains competitive with learned token-level credit baselines while avoiding a separate relevance-learning stage. These results suggest that policy-internal counterfactual likelihood contrasts provide an effective training signal for within-response credit allocation while retaining the simplicity and stability of GRPO.
[AI-34] OrchBench: Evaluating Multi-Agent Orchestration Plans in Isolation via Deterministic Simulation
链接: https://arxiv.org/abs/2607.25656
作者: Zhenzhen Ren,Jiyan He,Xinpeng Zhang,Zhenxing Qian,Ke Han,Shuxin Zheng,GuoBiao Li,Xiaoqing Zhang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Complex tasks often decompose into parallelizable yet interdependent subtasks, making orchestration critical to the performance of multi-agent systems (MAS). Existing evaluations typically rely on end-to-end execution, which conflates orchestration-plan quality with worker capabilities, tool reliability, and environmental noise. Moreover, the time and token costs of real execution grow rapidly with workflow scale, making systematic evaluation expensive. We present OrchBench, a simulation-based benchmark for evaluating multi-agent orchestration plans in isolation. Starting from real-world tasks, OrchBench constructs directed acyclic graphs (DAGs) that encode task dependencies, with controlled sizes and degrees of parallelism. Given a DAG, a per-agent context limit, and an agent budget, the evaluated planner assigns subtasks to agents and specifies cross-agent information transfers and their retention ratios. A deterministic simulator evaluates the resulting plan without invoking worker agents and returns interpretable measures of result quality, makespan, and token cost. The simulated scores produced by OrchBench correlate strongly with quality scores from Claude Code executions, achieving a Pearson correlation of (r=0.816), while requiring only (1.3%) of the tokens and (10.3%) of the wall-clock time. Across diverse planners and workflow scales, we find that preserving task-critical information is more important than simply increasing the number of agents, and the benefits of parallelism diminish as coordination failures accumulate. These results establish OrchBench as an efficient and interpretable benchmark for comparing and diagnosing multi-agent orchestration plans.
[AI-35] Engine-Equal Human-Unequal: A Reproducible Outcome Skew in Engine-Assessed Equal Chess Positions
链接: https://arxiv.org/abs/2607.25655
作者: Jesung Park
类目: Artificial Intelligence (cs.AI); Physics and Society (physics.soc-ph); Applications (stat.AP)
备注: 27 pages, 3 figures, 3 tables. Code and data: this https URL
Abstract:Among chess opening positions that a strong engine judges essentially equal (Stockfish 18 evaluation within 10 centipawns of zero, depth-stable) and that humans actually reach on Lichess (October 2025; 1,661 positions, 16.1M occurrences), human results are not balanced. Positions carry outcome skews, each the gap between its games’ actual results and what the players’ ratings predict, whose directions are stable properties of the naturally-reached position: some positions favour White, others Black. These skews reproduce across three re-partitions – disjoint player-account sets (primary), time, and disjoint rating bands – and on an out-of-sample month eight months later. On the primary split, each position’s skew is measured once in each account group, and the replication slope asks how well one measurement predicts the other after removing rating and opening-family effects: one means undiminished carry-over; zero, no linear relation. We find 0.69 (family-clustered 95% CI [0.65, 0.74]), rising to 0.94 on the most-popular, best-measured positions. The slope’s value depends on the position mix. Existence is the invariant claim: it survives every tighter evaluation band, search depth, calibration, and popularity cutoff we test, and replicates within blitz and rapid separately. The typical skew is small (median |\delta| \approx 0.018 , about two percentage points of White score), yet it reproduces, position by position, across disjoint accounts. At these positions the disfavoured side also thinks longer. Even where the evaluation is most confident, it is not a sufficient statistic for human outcomes. The result is observational, and the causal question is left to a pre-registered randomised companion study.
[AI-36] Why Public Service AI Governance Frameworks Risk Failing in the Age of General-Purpose AI: Lessons from Policing
链接: https://arxiv.org/abs/2607.25648
作者: Sam Relins,Daniel Birks
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI)
备注: Preprint; submitted for peer review
Abstract:Public services face growing pressure to adopt artificial intelligence (AI) to close the gap between rising demand and falling resources. That pressure has intensified with general-purpose AI (GPAI): AI built on large language models that can be directed by prompt alone to perform an effectively unbounded range of tasks. We argue that the properties that make these models attractive - their generality, accessibility, and low deployment cost - undermine the conditions under which AI safety has historically been pursued. The safety concepts that public service governance frameworks foreground - accuracy, bias, explainability, and accountability - were made tractable by narrow, purpose-built AI, and the mitigations that guidance documents prescribe presuppose exactly what GPAI removes. Accuracy cannot be quantified over unbounded outputs. Bias cannot be disaggregated when outputs are free-text judgements rather than categorical predictions. Explainability gives way to the appearance of explanation, and accountability erodes as outputs are optimized to persuade. We develop this through the case of policing, where the consequences of governance failure are most severe, and show why the same failure is likely to recur across other public services. The two mitigations that dominate policing AI strategy - expert evaluation and human-in-the-loop oversight - both rest on assumptions that GPAI violates. Safety assurance thus shifts from an intrinsic feature of building an AI tool to an optional add-on. We recommend a clear taxonomic distinction between narrow and general-purpose AI in governance documentation, a preference for technological parsimony, a pause on operational deployment of GPAI in policing until adequate evidence exists, and a coordinated national safety infrastructure with the authority to generate that evidence and determine when responsible deployment is achievable.
[AI-37] F(AI)2R: Who Did What and Who Checked? Verifiable AI Provenance as an Executable Skill
链接: https://arxiv.org/abs/2607.25637
作者: Florian Krebs
类目: Digital Libraries (cs.DL); Artificial Intelligence (cs.AI); Social and Information Networks (cs.SI)
备注:
Abstract:F(AI)2R is FAIR research with AI in the loop, twice: an AI-assisted authoring pass and a machine-readable audit pass over every artefact. AI systems now draft, refactor, and verify research artefacts, yet their contributions are rarely recorded in a form a later human or machine can audit. Building on the original F(AI)2R experiment, we generalize its provenance model beyond scholarly writing into aiprov, a PROV-O extension covering any AI-in-the-loop artefact, and we package the method as an executable skill that an AI agent operates itself: setup asks the human operator for their ORCID ID, resolves their identity from the public registry, and scaffolds continuous integration that gates every push on graph conformance and publishes the current build of this very paper. The paper is its own case study. Every activity, claim, and source in its production is recorded in the repository’s provenance graph under two invariants: no parentless claim, and verification rungs that only humans may grant.
[AI-38] Joint Text-Audio Alignment for EEG-to-Text Decoding in Chinese Speech Production and Perception
链接: https://arxiv.org/abs/2607.25626
作者: Tian Zheng,Xurong Xie,Xinxin Zhu,Xiaolan Peng,Feng Tian
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Decoding speech information directly from scalp electroencephalography (EEG) into text provides a potential non-invasive neural communication pathway for individuals with severe speech and motor impairments. Compared with invasive approaches such as electrocorticography, EEG is safer and more widely deployable, yet substantially more challenging to this http URL challenge is exacerbated for Chinese sentence decoding, which must handle a high-dimensional output space with thousands of characters, severe inter-subject variability, and low signal-to-noise ratios for text this http URL methods commit to a single supervisory axis—either text semantics or audio acoustic features—yet neither can simultaneously satisfy the demands of sentence-level discriminability and fine-grained temporal resolution required for large-vocabulary Chinese decoding. We introduce EEGAlign, a novel parameter-efficient framework that jointly aligns EEG with two axes—text alignment with BGE-M3 text embeddings and audio alignment with wav2vec~2.0 speech features via contrastive learning followed by CTC character-sequence decoding. On ChineseEEG-2 data, EEGAlign yields state-of-the-art closed-set sentence classification performance, reaching up to 82.37% Top-1 accuracy on Reading Aloud EEG and 41.43% on Passive Listening EEG out of 101 candidates. Ablation studies show that the two alignment axes are highly complementary: combining them yields consistently better performance than either alone. To the best of our knowledge, this is the first study on decoding large-vocabulary Chinese sentences from non-invasive EEG during overt speech production, and achieving strong classification performance with relatively large closed-set candidate-sentence setting.
[AI-39] Quotient Dynamics Effective Curvature and Implicit Bias in Positive Quadratic Networks
链接: https://arxiv.org/abs/2607.25624
作者: Pengcheng Cheng
类目: Artificial Intelligence (cs.AI)
备注: 75 pages, 19 figures
Abstract:Positive quadratic networks admit the low-rank representation f_U(x)=x^top UU^top x, where UinmathbbR^dtimes r is identifiable only up to right orthogonal multiplication, representing a rank-r PSD matrix Q=UU^top. We study how this quotient structure governs training dynamics, curvature, recovery, and interpolation bias. On the full-column-rank stratum, we identify mathbbR^dtimes r_*/O® with the rank-r PSD manifold. For smooth objectives L(U)=ell(UU^top), the Euclidean factor gradient is horizontal. Thus, factor gradient flow projects exactly to quotient Riemannian gradient flow, while finite-step gradient descent induces an exact congruence recursion for the predictor. For quadratic regression, we derive the effective Hessian at interpolators as the empirical measurement Gram form restricted to the tangent space relative to the quotient metric. Under Gaussian rank-one measurements, we compute population curvature, prove uniform deviation bounds for the empirical normal operator, construct a spectral initializer, and establish local exponential convergence for gradient flow and linear convergence for small-step descent. Recovery guarantees are explicit but conservative due to reliance on full-space second-moment control. In underdetermined commuting regimes, factor gradient flow becomes an exact entropy mirror flow in joint spectral coordinates. Strictly positive initializations converge to Bregman projections onto the interpolation set. With isotropic initialization q(0)=varepsilon^2mathbf1, predictors approach the minimum-trace solution set as varepsilondownarrow0, resolving nonuniqueness via weighted entropy within the invariant joint spectral algebra. Finite-step descent selects interpolants differing from continuous-time Bregman projections by O(eta). Numerical experiments verify these quotient identities, curvature predictions, recovery behaviors, and selection laws.
[AI-40] Multi-Sensor Alignment for Weather Simulations
链接: https://arxiv.org/abs/2607.25612
作者: Samsad Alam,Devyani Lambhate,Aditya Mohan,Vishal Kumar,Vaibhav Katewa
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Perception tasks for autonomous vehicles need to work satisfactorily in adverse weather conditions. Due to lack of real-world weather datasets, weather simulations are a promising alternative. To ensure simulations closely mirror real-world weather data, it’s crucial that they represent the same weather characteristics, including severity and particle positioning, across different sensors. To achieve this, we propose the Reference Dataset Alignment Method (ReDAM) for weather intensity alignment in fog and Unified-weather-edit (inspired by Weather-edit[1]) for particle positioning alignment in rain and snow. We validate both alignment methods using statistical and geometrical tests, respectively. We find that 3D detection models for non-aligned versions tend to be overly optimistic as compared to aligned versions. We also show the aligned-multi-sensor simulation’s effectiveness for achieving robustness for 3D object detection task by finetuning existing sensor fusion models on it.
[AI-41] Contrastive Representation Learning of Longitudinal Disease Trajectories on Temporal Graphs
链接: https://arxiv.org/abs/2607.25609
作者: Bastian Pfeifer
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Quantitative Methods (q-bio.QM)
备注:
Abstract:Understanding disease trajectories from longitudinal clinical data remains challenging due to complex temporal dynamics and heterogeneous patient cohorts. Here, we present a contrastive representation learning framework that models multivariate disease trajectories as temporal graphs and learns representations using contrastive graph neural networks. Nodes represent patient observations over time, while edges capture temporal continuity and structural similarity between trajectories. Structure-aware random walks guide contrastive learning to generate embeddings that preserve temporal context and trajectory topology. The resulting representations enable robust clustering of patients with similar disease progression patterns and reveal latent structure in longitudinal data.
[AI-42] Physics-Informed Broad Learning System: An Efficient Backpropagation-Free Framework for Solving Partial Differential Equations
链接: https://arxiv.org/abs/2607.25608
作者: Pinki Khatun,M. Sajid,Abhinav Jha,M. Tanveer
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Physics-informed neural networks (PINNs) have emerged as a powerful paradigm for solving partial differential equations (PDEs) by embedding governing physical laws into deep neural networks. However, their reliance on computationally expensive gradient-based optimization and deep architectures often results in slow training, high computational cost, and limited scalability. In this work, we propose a novel physics-informed broad learning system (PI-BLS), the first physics-informed learning framework based on broad RdNNs. The proposed formulation embeds the governing differential operator and the associated initial and boundary constraints directly into a linear output-layer optimization problem, thereby replacing nonlinear gradient-based training with a deterministic least-squares solution obtained via the pseudoinverse. Consequently, the entire learning process is reduced to a single linear optimization stage while preserving the underlying physical constraints. As a result, PI-BLS offers an efficient learning paradigm for a physics-informed learning framework for solving PDEs that eliminates iterative backpropagation while preserving the underlying physical constraints. Experimental results on representative forward PDE benchmarks demonstrate that PI-BLS achieves competitive and often superior performance with reduced training time and model parameters compared with conventional PINNs.
[AI-43] Computational Extraction of Legal Causes via al-Sabr wa al-Taqsim: A Set-Theoretic Formalization for Closed Fiqh Chapters
链接: https://arxiv.org/abs/2607.25605
作者: Elnaser Abdelwahab
类目: Artificial Intelligence (cs.AI)
备注: 16 pages. Published in Journal Academica, Vol. 15, No. 1, 2026, pp. 3-18. Includes an Arabic-language computational analysis of Tayammum in the Shafi’i school
Abstract:This paper presents a set-theoretic formalization of the classical usuli method of al-Sabr wa al-Taqsim (Examination and Division) for extracting legal causes ('ilal) within closed chapters of jurisprudence. A computational algorithm is introduced that extracts minimal operational rules from a truth table of juristic verdicts. The principal result is that, given a complete truth table for a closed chapter, the algorithm computes the minimal structural generators of the ruling and eliminates all logically redundant attributes. The resulting structures constitute admissible candidate causes for subsequent juristic evaluation. The framework is conditional upon the availability of a finite school-relative concept vocabulary and a complete ruling table for the chapter under investigation.
[AI-44] A Density-Matrix Framework for Electronic-Structure Analysis of Functional-Group and Salt Effects in Lithium-Metal Electrolytes
链接: https://arxiv.org/abs/2607.25597
作者: Mingkang Liu,Huize Yu,Yanbin Gao,Nan Yao,Xiang Chen,Lei Shen
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:The reactivity of lithium-metal electrolytes arises from the interplay of molecular functional groups, Li ^+ solvation, and salt-anion participation. This interplay operates through the redistribution of electron density across donor, anion, and cation centers, which is most directly read out from the electronic structure resolved in space. Quantum-chemical calculations deliver such readouts faithfully, yet become computationally demanding across this multidimensional design space, and machine-learning electronic-structure models seldom cover chemically diverse solvation shells or electrolyte-relevant readouts. Here, we present a density-matrix-centered AI platform (EMolStudio) for electronic-structure prediction and analysis. Its workflow integrates molecular functionalization, explicit Li ^+ first-shell assembly, density-matrix prediction with idempotency projection, and readouts of frontier orbitals, electrostatic potential, Li ^+ -donor bond order, and electron localization. We apply EMolStudio to 163,655 functionalized molecules and 22,500 explicit Li ^+ first-shell clusters across four lithium salts. We find that 1) at the molecular scale, functionalization distinguishes CO _2 Me, CN, F/CF _3 , and sulfonyl groups by chemically distinct changes in frontier levels, electrostatic potential, and Li ^+ -donor contact, consistent with \pi^* -acceptor, inductive, and polarization contributions, with sublinear accumulation at higher degrees of functionalization; 2) in explicit solvation shells, anion identity reshapes frontier-orbital localization: LiTDI anchors the HOMO on the anion across the entire library, whereas LiDFOB pairs an anion-hosted HOMO with strongly functional-group-dependent LUMO hosting. EMolStudio thereby translates functional-group and salt choices into electronic-structure hypotheses relevant to lithium-bond formation, desolvation, and interphase reactions.
[AI-45] How Small Can You Go? A Controlled Study of LoRA Rank Target Modules and Quantization Trade-offs for Text-to-SQL on a 60M-Parameter Model
链接: https://arxiv.org/abs/2607.25583
作者: Mahendra Singh Rathor,Anagheem Azzam
类目: Artificial Intelligence (cs.AI)
备注: 8 pages, 3 figures, 3 tables
Abstract:Parameter-efficient fine-tuning (PEFT) and low-bit quantization are now standard tools for adapting language models under tight compute budgets, yet their interaction is most often studied on billion-parameter models where the design space is expensive to explore. We ask a complementary question: on a specific, fully reproducible 60M-parameter encoder-decoder model (T5-small) and a single-table text-to-SQL benchmark (WikiSQL), how much task accuracy does each efficiency knob actually cost? We run a controlled, single-variable study over (i) LoRA rank r in 2, 4, 8, 16, 32, (ii) the set of adapted modules, and (iii) numerical precision. We report task accuracy alongside system-level metrics including trainable parameters, peak training memory, inference latency, and throughput, and frame adaptation as a constrained trade-off rather than an accuracy-only objective. Our results show that LoRA with r=16 recovers within 11.6 percentage points of full fine-tuning accuracy (59.6% vs. 71.2% exact-match) while training fewer than 1% of parameters and consuming 31% less peak GPU memory. Within this setting, rank beyond r=16 yields no measurable accuracy gain. QLoRA with INT8 and NF4 quantization achieves comparable accuracy (52.8% and 53.2%) at dramatically lower memory cost (0.60 GB each), demonstrating a compelling trade-off for memory-constrained deployments. All code, configurations, and logs are released for full reproducibility.
[AI-46] Matrix-Free Photoacoustic Image Reconstruction via Sensor-Token Self-Attention
链接: https://arxiv.org/abs/2607.25576
作者: Mary John,Shibili Said,Imad Barhumi,Sherzod Turaev,Mohamed Yahia
类目: Artificial Intelligence (cs.AI)
备注: 25 pages, 13 figures, 4 tables
Abstract:Photoacoustic tomography (PAT) combines the optical absorption contrast of biological tissue with the spatial resolution of ultrasound, yet recovering the initial pressure distribution from sparse-view sensor measurements remains an ill-posed inverse problem. Iterative compressive-sensing solvers and unrolled deep networks both retain a dependence on the system matrix at inference, which leaves real-time clinical reconstruction computationally expensive. This paper proposes the Sensor Attention Network (SAN), a Transformer-based architecture that treats the full time series of each sensor as a token and maps raw measurements directly to the reconstructed image without invoking the system matrix at inference. For training and benchmarking, an analytical k-space H-matrix is constructed and validated against the k-Wave pseudo-spectral solver under matched geometry, achieving a mean per-sensor Pearson correlation of 0.919 +/- 0.049, with k-space apodization and Gaussian temporal damping acting synergistically to reduce the energy-normalized mismatch by 49%. Trained with a vessel-weighted loss on 488 augmented samples and evaluated on 46 held-out samples against ISTA, split-Bregman total variation (SBTV), and learned ISTA (LISTA), SAN attains the highest mean SSIM (0.522) and PSNR (22.09 dB) and the lowest NMSE (0.233). Paired t-tests and Wilcoxon signed-rank tests confirm the superiority of SAN over LISTA on PSNR, NMSE, and Pearson correlation at p 1e-8, and over ISTA and SBTV on all fidelity metrics. By bypassing the H-matrix at inference, SAN reduces reconstruction time by at least an order of magnitude, supporting real-time PAT reconstruction.
[AI-47] Agent Skills Matter: Inferring Proprietary Skills from Execution Trajectories
链接: https://arxiv.org/abs/2607.25560
作者: Jianing Geng,Ruiqi He,Zekun Fei,Biao Yi,Ruijie Wang,Zheli Liu,Xia Hu,Xuansheng Wu,Qingkai Zeng
类目: Artificial Intelligence (cs.AI)
备注: 18 pages, 6 figures, 2 tables
Abstract:Agent skills package reusable procedures that improve downstream performance. Their lightweight, portable form enables marketplace monetization and private deployment behind cloud-hosted agent interfaces, giving providers incentives to keep high-value skills proprietary. Yet hiding the artifacts does not conceal their behavioral effects, which remain observable in execution trajectories and form a behavioral side channel. We define this exposure as Skill Leakage: reconstructing proprietary skills from trajectories elicited by benign queries, without reference answers or success labels. We introduce SigLeak, a black-box framework that exploits recurring skill signatures in agent behavior. It constructs diverse, decision-rich diagnostic tasks, contrasts matched skill-enabled and skill-disabled trajectories, and iteratively refines a reconstructed skill from the isolated patterns. Across five scenarios, three model families, and three agent frameworks, SigLeak outperforms or matches three baselines in nearly every setting. It raises the success rate by 6.88 percentage points over the skill-disabled reference on average and achieves the highest overall SkillSim, our metric for coarse- and fine-grained semantic similarity. These results show that benign execution trajectories can expose proprietary procedural knowledge. The code is available at this https URL.
[AI-48] Distilling Temporal Search and Reasoning : Evolving LLM s for Future Prediction via Harness-Assisted Efficient Data Synthesis
链接: https://arxiv.org/abs/2607.25554
作者: Wanxu Cai,Zhengyu Chen,Huaisheng Zhu,Wei Wang,Jingang Wang,Qiang Xu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Future event prediction carries broad social impact yet remains challenging. SOTA approaches augment LLMs with external agent frameworks whose predictive capability vanishes once the harness is removed. While recent Tool-Integrated Reasoning (TIR) internalizes deep search for multi-hop retrieval of facts, forecasting further demands temporal search and reasoning over historical trends and dynamic shifts. The key obstacle is data: historical queries induce temporal leakage that degrades forecasting into retrieval. Prior works either freeze information gathering with static observations, or rely on rejection sampling or unresolved fresh queries that discard vast amounts of data, degrading synthesis efficiency. We propose a time-truncation harness that enforces a temporal cut-off at every turn, enabling TIR-style sampling from historical events, reducing temporal leakage and reliance of rejection sampling or unsolved queries, increasing the sampling efficiency. We further build a large-scale corpus and a process-based metric and show that our harness naturally induces a broader temporal breadth of search and raises the proportion of high-quality data, further increasing the efficiency and reducing the reliance on complex rubrics. Distillation experiments show that students trained on harness-intervened data achieve the best performance, demonstrating harness-assisted model evolving that turns higher quality temporal search and reasoning data into a parametric advancement of the students.
[AI-49] From Training to Deployment: Post-Hoc Causal Feature Identification via Sensitivity Ratios
链接: https://arxiv.org/abs/2607.25546
作者: Athanasios Vlontzos,Giorgos Papanastasiou,Bernhard Kainz,Sotirios Tsaftaris
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Given a model that is already trained, which features does it rely on causally versus spuriously? Existing methods require access to the training procedure and cannot answer this post-hoc. We introduce the \textbfNormalised Sensitivity Ratio~(NSR), a post-hoc, model-agnostic diagnostic for this question under a structured-shift regime: environments differ primarily in the mean of spurious features while the causal mechanism and causal marginals remain stable, as in multi-site clinical data or multi-batch genomics. Within this regime, causal features induce constant model sensitivity across environments while spurious features track shift. NSR formalises this as the squared coefficient of variation of per-environment sensitivity. Under a linear structural causal model (SCM) with K\ge3 non-degenerate environments, NSR achieves exact identification (Theorem~1). We fully characterise failure: weak shifts ( O(\varepsilon^4) collapse), degenerate geometry, and proxy attenuation ( O((1-\alpha)^4) ), giving practitioners quantitative criteria for assessing whether the regime holds. Finite-sample rates are O_p(n^-1) under the null and O_p(n^-1/2) under the alternative. Experiments confirm all theoretical predictions on synthetic data (area under the ROC curve [AUROC] = 1.000 under conditions satisfying the regime), show consistent rankings across five model families (Kendall \tau\ge0.529 ), and recover six of eight causal features on bike-sharing data (Precision@7 = 0.75 ) without modifying any trained model.
[AI-50] Entangled by Design: Spurious Intra-Variable Signal Routing in Tabular In-Context Learners
链接: https://arxiv.org/abs/2607.25532
作者: Athanasios Vlontzos,Giorgos Papanastasiou,Bernhard Kainz,Sotirios Tsaftaris
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Consider a model trained at a single hospital to predict patient recovery, where the measured feature X bundles the patient’s true health signal ( C ) with a systematic artefact from that hospital’s equipment ( S ). Within that hospital, the artefact correlates with outcomes through unmeasured confounders such as patient demographics; an in-context learner rationally routes predictions through S , not C , and fails silently when deployed at a new hospital with different equipment. We formalise this as \emphspurious routing in composite representations: when a feature X = [C;,\alpha S;,\eta] encodes a causal signal C and a spurious signal S in distinct subspaces, the ICL cannot determine which drives predictions. We prove that under ridge ICL, a linear in-context learner, this routing is unavoidable regardless of context size; TabPFN, a state-of-the-art pretrained tabular ICL model, shows qualitatively consistent behaviour empirically. We derive a closed-form characterisation, \mathrmCSR \propto \rho_S/\rho_C , confirmed at r = 0.997 for linear ICL and r = 0.979 for TabPFN. Contrary to intuition, larger context sharpens commitment to the dominant in-context signal, amplifying spurious routing by up to 1.74\times ; in the high-spurious corner, more expressive models show greater vulnerability empirically ( +2.22 CSR gap at high entanglement). We introduce two lightweight mitigations: environment-stratified context construction and S-swap augmentation, that require only weak environment labels and no knowledge of the causal partition. S-swap reduces spurious routing by 74% for linear ICL and 98.8% for TabPFN, with TabPFN’s causal sensitivity increasing 8.4\times simultaneously: the model does not become agnostic, it reroutes through the causal signal.
[AI-51] Are the High-weight Neurons the Important Ones in Image Classification Neural Networks?
链接: https://arxiv.org/abs/2607.25529
作者: Qitao Chen,Dongfu Yin,F. Richard Yu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:As neural network models for image classification advance, neurons play critical roles in pruning, backdoor defense, and interpretability. Yet existing work lacks clarity on the weight-importance relationship. We address this with a neuron importance assessment method using three experiments: quantifying overlap between high-weight and accuracy-impacting neurons, analyzing high-weight neuron perturbation effects, and testing post-retraining accuracy after high-weight neuron ablation. Experiments on CIFAR-10 and Mini-ImageNet reveal key patterns. Overlap analysis shows top 10% high-weight neurons overlap with important ones by only about 25% at maximum, dropping further in subsequent intervals. Perturbation tests find top 10% high-weight neurons cause 45-80% accuracy degradation under certain operations compared to 3-7% for random perturbations, but a third of them show minimal impact. Ablation-retraining results show removing top 10% high-weight neurons leaves accuracy 10-20% below baseline with no recovery, while ablating top 0.1% allows near-full recovery. Notably, some low-weight intervals show 10-17% degradation when perturbed, comparable to mid-range high-weight neurons. These results confirm not all high-weight neurons are important: their importance is nonlinear. Low-weight neurons also contribute significantly. This challenges weight-importance equivalence, offering refined neuron role insights. It supports applications like encryption prioritizing critical high-weight neurons and pruning removing non-critical ones, advancing neural network analysis.
[AI-52] At-the-Roofline Sparse Tensor Contractions on Vector Processors for Transformer Inference
链接: https://arxiv.org/abs/2607.25504
作者: Bowen Wang,Chi Zhang,Diyou Shen,Renzo Andri,Navaneeth Kunhi Purayil,Luca Benini
类目: Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI)
备注: 5 pages, 4 figures, 34th IFIP/IEEE International Conference on Very Large Scale Integration SoC (VLSI-SoC 2026)
Abstract:Fine-grained weight pruning and activation sparsification have emerged as effective approaches for reducing the compute and memory cost of inference for Transformer models. In the moderate-sparsity regime, Gustavson’s dataflow provides a natural execution model for exploiting both activation and weight sparsity on vector processors through metadata-driven indexed accumulation. However, existing RVV architectures lack native support for this pattern, forcing kernels to rely on software index decoding and L1-backed indexed memory operations that keep sparse tensor contractions far below their roofline performance bound. We present Ventaglio, a runtime-configurable sparse execution unit coupled with RVV ISA extensions that drives sparse tensor contractions toward their roofline through indexed gather-accumulate-scatter support. Integrated into an open-source vector processing cluster and implemented in 12nm FinFET, Ventaglio accelerates sparse tensor contraction kernels by 6.9\text–7.4\times over optimized RVV baselines, with only 3.1% area overhead for a cluster of tightly-L1 coupled vector processing elements. We build a performance-accurate instruction-level model of the Ventaglio extension, calibrate it against RTL implementation, and leverage it for scale-out performance analysis on a large 4\times4 multi-cluster system. Using a DuoGPT-pruned LLaMA-3-8B model with practical 40\text–60% dual sparsity, Ventaglio achieves 2.40\text–5.25\times and 2.06\text–3.16\times speedup over dense baselines during prefill and autoregressive decoding, respectively.
[AI-53] Automated Numerical Stability Analysis of Deep Learning Operators
链接: https://arxiv.org/abs/2607.25494
作者: Xinye Chen
类目: Numerical Analysis (math.NA); Artificial Intelligence (cs.AI)
备注:
Abstract:Finite-precision arithmetic unavoidably introduces numerical approximation errors. Numerical computations may use insufficient precision or an improper formulation, which leads to numerical instability. In this paper, we introduce the first unified software tool that integrates CESTAC for detecting the numerical stability of deep learning operators. Our developed software not only enables numerical validation with a single computation pass but also detects the sources of numerical instability and provides numerical stability monitoring during deep learning training and inference. We verified its effectiveness on the detection of polluted operators with injected numerical instabilities across various tasks. We believe that our developed method and tool provide valuable insights into developing numerically stable computing kernels, which are particularly critical for numerically stable and efficient deep learning training and inference.
[AI-54] Finding Optimal Cost-Bounded Plan Reductions: Refined Model ICAPS2026 ICAPS
链接: https://arxiv.org/abs/2607.25484
作者: Martha Del Toro,Raquel Fuentetaja,Angel García-Olaya
类目: Artificial Intelligence (cs.AI)
备注: Extended version of a paper accepted at ICAPS 2026 (doi: https://doi.org/10.1609/icaps.v36i1.42849 ). Includes a new model and expanded experimental results
Abstract:In some real applications a plan may later become unfeasible due to newly imposed budget constraints, yet, at the same time, using only the original actions of the plan and their order is mandatory. In this paper, we study the problem of extracting, from a precomputed plan, a valid subplan that maximizes utility while respecting a cost bound. Each goal is given a utility value and the plan is reduced by removing actions that support low-utility goals, while preserving both executability and the original action order. We show the decision variant is NP-complete and propose two exact methods to solve it: one via oversubscription planning (OSP) and another via Integer Linear Programming (ILP). This paper extends our previous work published at ICAPS 2026 (Del Toro, Fuentetaja, and García-Olaya 2026b). While the core framework remains as introduced there, we further introduce a refined ILP formulation that significantly decreases the model size and improves computational efficiency.
[AI-55] Architectural Backdoors in Vision-Language Model Supply Chains via Representation Steering
链接: https://arxiv.org/abs/2607.25479
作者: Maria Rosaria Briglia,Igor Maljkovic,Antonio Emanuele Cinà,Luca Oneto,Iacopo Masi,Fabio Roli
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Vision–Language Models (VLMs) are increasingly deployed through a model supply chain in which pretrained checkpoints, architecture definitions, text encoders, and exported computation graphs are distributed by third parties and reused across downstream services. This reuse model creates a security-critical trust boundary: VLM deployments inherit not only learned parameters but also executable behavior encoded in shared model artifacts. In this paper, we show that a malicious provider can exploit this trust boundary by embedding architectural backdoors into VLM supply chains through representation steering. Our attack introduces dormant steering logic into the model architecture through a trigger-gated additive modification of an intermediate representation, without poisoning training data, controlling downstream fine-tuning, or modifying prompts at deployment time. When the trigger is absent, the modification reduces to zero and the model follows its normal computation, preserving clean utility. When the trigger is present, a steering direction shifts the internal representation toward an attacker-defined objective. We evaluate the attack across multiple VLM families and downstream tasks, including visual question answering, text-to-image generation, retrieval, and semantic response biasing. The results show that the proposed architectural steering backdoor compromises integrity, safety enforcement, and ranking fairness while preserving normal behavior on clean inputs. We further show that shared VLM artifacts can carry dormant steering logic against downstream services, and we propose an auditing defense that inspects the executable logic distributed with model artifacts rather than only their learned weights.
[AI-56] Balancing multiscale similarity and cartographic constraints: A similarity-driven optimization framework for line generalization
链接: https://arxiv.org/abs/2607.25474
作者: Pengbo Li,Haowen Yan,Xiaomin Lu,Binbin Lin
类目: Artificial Intelligence (cs.AI); Computational Geometry (cs.CG); Systems and Control (eess.SY)
备注: Submitted to Cartography and Geographic Information Science
Abstract:Cartographic generalization is essential for generating multiscale map representations by balancing information preservation and cartographic readability. However, automated generalization remains challenging because existing approaches often treat spatial similarity evaluation, cartographic constraints, and parameter optimization as separate processes, limiting adaptive and interpretable control across scales. This study formulates cartographic generalization as a constrained multiscale similarity optimization problem and proposes a similarity-driven framework for adaptive generalization control. The framework integrates multiscale spatial similarity as an optimization objective to quantify representation consistency between original and generalized data, while incorporating cartographic constraints to regulate readability, smoothness, and geometric validity. A unified objective function is optimized to automatically identify scale-dependent parameter configurations for different generalization algorithms. Experiments using multiple line simplification algorithms, target scales, and similarity measures, including geometric, structural, and learning-based metrics, demonstrate that the proposed framework achieves an effective balance between similarity preservation and cartographic abstraction. The results further show that combining similarity optimization with cartographic constraints provides more consistent and interpretable parameter control than relying on similarity evaluation alone. This study provides a unified optimization perspective that connects similarity assessment, constraint modeling, and algorithm control, contributing to adaptive and automated cartographic generalization.
[AI-57] RWH: A Text-Driven Random Walk Heterogeneous GNN for Semantic-Aware Sparse Recommendation
链接: https://arxiv.org/abs/2607.25471
作者: He Ma,Chen Liu
类目: Artificial Intelligence (cs.AI); Multimedia (cs.MM)
备注:
Abstract:Graph Neural Networks (GNNs) and Large Language Models (LLMs) have each advanced recommendation systems by modeling structural and semantic signals, respectively. However, integrating their complementary strengths remains challenging, particularly in sparse settings where maintaining semantic precision is critical. We propose TRWH (Text-driven Random Walk Heterogeneous Graph Neural Network), a novel framework that fuses LLM-generated textual profiles with heterogeneous graph structures through strategic random walk augmentation. TRWH consists of three core components: (1) Embedding Creation, which produces user and item representations using both Word2Vec and LLM-based profiling; (2) a Heterogeneous Graph Neural Network (HeteroGNN) that propagates information across multi-relational edges; and (3) Random Walk-based Path Construction, which enriches sparse graphs with second-order user-user and item-item links. Experiments on the Amazon-2023 Fashion (2M users, 825K items) and Beauty (631K users, 112K items) datasets demonstrate that TRWH achieves substantial performance gains over state-of-the-art methods, including 80.0% RMSE and 52.6% MAE reductions on Fashion, and 25.7% and 10.8% improvements on Beauty. Notably, while random walks improve performance with traditional embeddings, they can dilute the nuanced representations learned by LLMs, underscoring the importance of adaptive integration strategies.
[AI-58] Emergent Latent-State Computation under Stochastic Volatility
链接: https://arxiv.org/abs/2607.25459
作者: Xiaoyu Huang,Lulu Wang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Statistical Finance (q-fin.ST)
备注:
Abstract:Mechanistic interpretability has largely focused on language models and deterministic toy tasks. Much less is known about how sequence models internally represent latent stochastic dynamics under noisy, partially observed observations. We study this question in a controlled multivariate stochastic volatility setting, where models observe only returns while the ground-truth latent volatility state is known to the researcher. This setting provides a useful benchmark for mechanistic interpretability under partial observability: the latent state is hidden from the model but directly available for evaluation. Across architectures, losses, and output heads, we find evidence for a two-stage computation. Hidden representations encode substantial information about the next latent volatility state, and the output head maps this representation to squared return forecasts. Furthermore, in Transformers, latent-state decodability emerges at identifiable architectural stages whose location depends on the volatility period. In long-cycle regimes, this computation simplifies into an explicit latent-state filter consisting of a learned linear projection followed by \ell^2 normalization. Output-head replacement further shows that part of the degradation under noisy MSE training arises from readout misalignment rather than representation failure. These results suggest that stochastic volatility models provide a useful benchmark for mechanistic interpretability under noisy latent dynamics and partial observability.
[AI-59] oward an Organizational Science of Multi-Agent LLM Systems: Decoupling Who How and Which Algorithm
链接: https://arxiv.org/abs/2607.25446
作者: Huan Chen,Xiang Song,Jian Jin,Pan Ren,Liang-Jie Zhang
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 8 pages, 2 figures
Abstract:Multi-agent frameworks built on large language models (LLMs) routinely entangle three logically distinct concerns: who is on the team (organization), how members align (coordination), and which algorithm fuses their work (collaboration protocol). IMACS (Intelligent Multi-Agent Collaboration System) separates the three into orthogonal, independently swappable layers. Classic organizational theory (Belbin roles, Mintzberg coordination, RACI accountability) becomes executable, validated configuration, and the framework places six published collaboration algorithms behind a common interface while exposing roles, coordination, and accountability as independently configurable factors. We use this separation to conduct controlled comparisons in which organizational assignments vary while the collaboration protocol is held fixed. It also turns protocol choice into a variable that can be learned: Adaptive Org Routing, a contextual-bandit meta-protocol, selects a protocol per task under an explicit quality-cost tradeoff, outperforms every fixed protocol in a controlled study, and trains online on real benchmark and LLM-judge rewards. The ablations expose a mechanism. Accountability placement changes outcomes exactly when the protocol routes the deliverable through the accountable agent, and the winning placement flips across model families, so organizational design cannot be hard-coded; it must be revalidated, or learned, for each model binding.
[AI-60] he Disruptive Impact of Large Language Models on Capture the Flag Competitions and the Path Toward Fair Play
链接: https://arxiv.org/abs/2607.25425
作者: Michael Macaulay,Harmony Bouabid,Guo Gen Ang,Sasha Shaw
类目: Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR); Computers and Society (cs.CY)
备注: 20 pages, 3 figures, 1 table
Abstract:Capture the Flag (CTF) competitions are among cybersecurity’s most effective training grounds, developing practical skill across cryptography, web exploitation, and binary exploitation. Large language models (LLMs) can now solve a growing share of challenges with minimal human input, raising urgent questions about fairness, the validity of rankings, and whether participation still delivers the learning that justifies the effort. This paper reports a mixed-methods study of LLM impact on modern CTFs, combining a synthesis of published benchmarks, including a recent government evaluation, case studies of live competition across three challenge categories, structured observation of the public channels where the community debates AI use, and semi-structured interviews with experienced players and organisers. We map the current human-machine capability boundary by category, showing that easy and intermediate challenges in cryptography, web, and binary exploitation are now reliably automated while narrower sub-categories continue to resist. We find that community disagreement about whether AI should be permitted is downstream of an undeclared prior question: what a competition is for. Against this backdrop we contribute a four-component safeguard framework, combining tiered competition divisions, LLM-resistant challenge design, telemetry used investigatively, and a draft community code of conduct, together with a decision tool that ties the combination of safeguards to a competition’s declared purpose. The argument reaches beyond CTFs to any setting in cybersecurity where a demonstrated result is taken as evidence of an underlying ability.
[AI-61] Salient Knowledge Pathways: Sparse Cross-Modal Routing for Efficient Knowledge-Intensive Multimodal Question Answering ICML2026
链接: https://arxiv.org/abs/2607.25422
作者: Noor Islam S. Mohammad,Uluğ Bayazıt
类目: Artificial Intelligence (cs.AI)
备注: Accepted at the 43rd International Conference on Machine Learning (ICML 2026) Workshop on Efficient Multimodal Question Answering (EMM-QA), Seoul, South Korea. Copyright 2026 by the author(s). (Archival)
Abstract:Knowledge-intensive multimodal question answering (KI-MMQA) sits at the intersection of three expensive primitives: long visual token sequences, dense retrieval over large external corpora, and full cross-modal fusion. Existing systems pay all three costs uniformly per query, even though only a small fraction of visual content and retrieved knowledge is actually relevant to any given question. We introduce SKIP (Salient Knowledge-Injected Pathways), a unified inference architecture that routes computation along sparse pathways jointly conditioned on the question, the image, and a difficulty estimate. SKIP combines question-guided visual token pruning, region-conditional sparse retrieval, bipartite sparse cross-attention, and speculative knowledge verification with an adaptive budget controller that allocates compute proportional to predicted question difficulty. We derive an information-bottleneck bound showing that the optimal visual sparsity rate scales as O(1/\sqrtN) under realistic question-image mutual-information assumptions, with retained accuracy guarantees. Across five KI-MMQA benchmarks (OK-VQA, A-OKVQA, InfoSeek, Encyclopedic-VQA, and ViQuAE), SKIP matches or exceeds the accuracy of strong dense baselines while using 3.4 – 6.8\times fewer FLOPs and 2.7\times less end-to-end latency. Code available at: this https URL
[AI-62] A Control System a Dataset and a Recipe for Making Frozen LLM Agents Learn a Domain
链接: https://arxiv.org/abs/2607.25415
作者: Debjyoti Paul
类目: Artificial Intelligence (cs.AI)
备注: 8 pages, 1 figure, 3 tables. Code and dataset: this https URL
Abstract:Production LLM agents are increasingly assembled from a frozen model wrapped in a harness: a prompt template, a tool set, a memory/retrieval layer, a planning strategy, and a verification policy. Two 2026 systems, Meta-Harness (Lee et al., 2026) and HyperAgents (Meta AI, 2026), show that this harness can itself be optimized or even self-rewritten by an agentic proposer – at the cost of either an expensive code-search loop or unconstrained self-modifying code, neither of which is auditable or usable with a fully black-box model API. We take a narrower, more constrained position: treat the harness as a small, fixed, human-legible action space and learn a policy over it online with classic sample-efficient reinforcement learning (an \epsilon -greedy contextual bandit and REINFORCE), scored against a multi-objective reward (task success, verifier score, policy compliance, cost, latency, and an unsupported-claim penalty). We instantiate this control system with DSPy (Khattab et al., 2024) as both the context assembler and the source of the strongest non-adaptive baseline (a DSPy BootstrapFewShot static prompt), and evaluate it across three verifiable task domains – tool-use workflows, code generation (HumanEval), and multi-hop retrieval QA (HotpotQA) – and two model providers (a local Ollama model and AWS Bedrock). We release the harness-control-system code, the cross-domain verifiable task suite, the full trajectory/reward-decomposition logs from training, and a provider-agnostic deployment recipe for applying this to a new organization’s domain and verification setup.
[AI-63] Context Assembly as the Controlled Variable: A Control-Theoretic View of Harness Policies for Frozen LLM Agents
链接: https://arxiv.org/abs/2607.25408
作者: Debjyoti Paul
类目: Artificial Intelligence (cs.AI)
备注: 6 pages, 2 figures, 1 table. Code and companion paper’s data: this https URL
Abstract:A growing body of 2026 work applies control theory to LLM agents: Lyapunov-certified stability for tool-mediated controllers (Prinos et al., “Stable Agentic Control”, 2026), sample-complexity bounds for sparse policies over massive discrete tool universes (Majumdar, “Sparse Agentic Control”, 2026), and regulatory-control decompositions of multi-agent systems into auditable feedback loops (Nogueira and Skogestad, 2026). We do not claim to introduce control theory to LLM agents – that ship has sailed. Our narrower claim is about what the controlled variable is. Prior work controls tool selection, inter-agent message routing, or the agent’s raw action stream. We instead treat context assembly itself – which prompt template, which few-shot demonstrations, how much retrieved context, how many planning/verification passes – as the controlled variable, learned online by a contextual bandit or REINFORCE policy sitting outside a frozen model. This paper develops the formal decomposition (inner frozen policy \pi_\theta , outer context policy \pi_\phi ), gives a stability argument for the online controller in the sense used by Zhang et al. (2026) (non-decreasing expected reward under bounded policy change), and reports an uncertainty-calibration analysis of the controller’s own confidence against realized task outcomes. The applied counterpart to this paper instantiates the same controller across three domains and two model providers and releases the dataset, trajectory logs, and a deployment recipe; here we focus on the formal framing and the stability/uncertainty evidence a control-theoretic claim requires.
[AI-64] COVENANT: Natural-Language Workflow Compilation for Aligned Agent Execution
链接: https://arxiv.org/abs/2607.25400
作者: Jincheng Wang,Min Zheng,Tao Wei
类目: Artificial Intelligence (cs.AI)
备注: 12 pages, 7 figures
Abstract:Large language model (LLM) agents are increasingly entrusted with natural-language workflow instructions (e.g., retail-payment policies) that specify not only what outcome to achieve, but also which steps, branches, and tool interactions are permitted. When these instructions are supplied as prompt context, however, the model retains control over both procedure selection and step execution. As interactions accumulate, an agent can skip required steps, take unsupported branches, or execute a valid step with unsupported arguments or effects–a failure mode we call workflow misalignment. In this work, we propose COVENANT, a compiler-and-interpreter architecture for workflow-aligned agent execution. Our key insight is to treat workflow instructions as source programs rather than prompts. COVENANT converts the instructions into a workflow abstract syntax tree (WAST) and lowers it to a workflow control-flow graph (WCFG). At runtime, a controller interprets the WCFG one node at a time, checks each proposal against requirements extracted from the instructions before committing controller state or advancing the graph, and returns diagnostic feedback for repair. To evaluate COVENANT, we use 120 cases from three existing benchmarks, spanning seven workflow scenarios. Compared with state-of-the-art LLM agents, COVENANT improves benchmark success from 50.00% to 83.33% and reduces the workflow-misalignment failure rate from 42.50% to 15.83% (62.75% relative). These results show that COVENANT substantially mitigates workflow misalignment, moving LLM-agent alignment beyond isolated prompt following toward reliable execution of complex and multi-step workflows.
[AI-65] Cyber-Capable AI Agents : Vulnerabilities Evaluation Containment and Defensive Response
链接: https://arxiv.org/abs/2607.25379
作者: Abu Bakar Siddik
类目: Artificial Intelligence (cs.AI)
备注: 27 pages, 8 figures, 8 tables
Abstract:Cyber-capable AI agents combine language models with tools, memory, and execution en- vironments to perform multi-step offensive-security tasks. Existing work separately measures cyber capability and catalogs attacks against agent components, but provides less guidance on containing a capable agent within the environments used to evaluate it. This review synthe- sizes five vulnerability classes at that boundary: multi-step offensive chains, objectives that conflict with sandbox boundaries, supply-chain and credential exposure, persistent command- and-control, and the speed of automated action. We use the reported July 2026 Hugging Face/OpenAI incident as a bounded case study, distinguishing incident-specific observations from findings established in the wider literature. Across the taxonomy and case, we examine controls for containment, privilege separation, provenance, and responder access, including the dual-use problem that defensive artifacts may also enable misuse. The review identifies practical priorities for evaluating cyber capability together with the security of the environment in which that capability is exercised.
[AI-66] Rethinking Likelihood distributions: Students t Likelihood Boosts Bayesian Neural Network Performance
链接: https://arxiv.org/abs/2607.25376
作者: Pei-Hsuan Hsia,Lars H. Heyen,Arvid Weyrauch,Markus Goetz,Achim Streit,Sebastian Krumscheid,Charlotte Debus
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:In Bayesian neural networks (BNNs), variational inference is a widely adopted framework for modeling uncertainty in a distributional way, with the evidence lower bound (ELBO) serving as the standard objective function. Several distributions contribute to the ELBO loss, such as the prior, approximated posterior, and likelihood distribution. Typically, these distributions are all approximated by a Gaussian distribution, since it is easy to compute, allows for reparameterized gradients, and provides a closed-form loss for training. However, several works have highlighted that this assumption may not generally hold, posing the risk of model misspecification. Alternative distributions have been proposed for the prior specifically, while the effect of distribution choice on the likelihood distribution remains unexplored. In this work, our aim is to close this gap by investigating whether alternative assumptions for the likelihood distribution can outperform the commonly used Gaussian. We compare several likelihood distribution assumptions, such as skewed or heavy-tailed, across regression tasks on both artificial and real-world datasets using standard multilayer perceptrons (MLPs). Our findings demonstrate that Student’s t yields better predictive performance than a Gaussian likelihood distribution, independent of the data distribution and MLP architecture (depth and width). In some cases, Student’s t can also lead to shorter training times, while still being easy to implement.
[AI-67] ODYSSE: Episode-wise Policy Optimization for Personalized Agent ic Reasoning
链接: https://arxiv.org/abs/2607.25369
作者: Jiaqi Zhang,Tong Chen,Junliang Yu,Quoc Viet Hung Nguyen,Hongzhi Yin
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Agentic systems have rapidly advanced in their ability to interact with real-world environments, leverage external tools, and provide services for users. However, unlike natural-world tasks that assume well-defined instructions, human-centered scenarios are characterized by ambiguous requests that lead to large, open-ended solution spaces. Decoding users’ personalized preferences is therefore essential for narrowing the candidate solution space. This introduces a new challenge, personalized agentic reasoning, which requires agents to jointly interact with both users and environments to deliver personalized services. In this paper, we present ODYSSE, a Reinforced Fine-Tuning (RFT) framework for personalized agentic reasoning. At its core, ODYSSE proposes Episode-wise GRPO (ESPO), a novel extension of Group Relative Policy Optimization (GRPO) designed to address long action horizons and strong cross-step dependencies in personalized agentic reasoning. Rather than optimizing individual steps independently, ESPO introduces an episode-level reward mechanism together with episodic advantage estimation, enabling upstream evidence to effectively guide downstream personalized decisions and allowing agents to progressively resolve ambiguous user requests across multiple interaction steps. We further propose an episodic batch sampler that groups actions from the same episode into unified training batches, facilitating coherent optimization under ESPO. We evaluate ODYSSE on realistic long-horizon personalized GUI reasoning tasks. Experimental results demonstrate that ODYSSE consistently outperforms both specialist and general-purpose LVLMs, highlighting its effectiveness for personalized agentic reasoning.
[AI-68] AI Deployment and Cyber Governance Failures in Public-Sector Organizations: A Typological Analysis
链接: https://arxiv.org/abs/2607.25368
作者: Md Salahuddin,James Rooney,Fida Hasan
类目: Artificial Intelligence (cs.AI)
备注: Accepted at Conference ML4CS
Abstract:The intersection of artificial intelligence adoption, cybersecurity governance, and public sector institutional constraints has not been examined as a unified analytical problem in the existing literature. Studies address AI cybersecurity risks generically, public sector governance independently, and framework adequacy separately. Existing studies have not integrated these three streams to explain specifically how AI adoption causes cybersecurity governance failure in government organizations, nor test existing governance instruments against AI-specific public sector failure causes. This paper ad-dresses that gap. It proposes a seven-domain typology identifying ten specific AI-driven cyber governance failure causes grounded in public sector institutional analysis. It presents a three-pathway failure model showing how accountability failure, opera-tional resilience failure, and compliance failure interact and reinforce each other. It de-livers a structured coverage matrix testing five major governance frameworks (NIST CSF 2.0, ISO/IEC 27001, COBIT, NIST AI RMF, and ISO/IEC 42001) against the typology, finding that no instrument addresses Shadow AI, speed asymmetry, or gov-ernance vacuum at the operational specificity required for public sector application. The paper introduces speed asymmetry as a named structural construct with a specified mechanism. The framework provides the design specification for an AI-enabled cyber-security maturity model for government organizations.
[AI-69] Explanation-Bound Tool Execution for AI Agents : Server-Verified Action Claims Without Trusting Model Rationales
链接: https://arxiv.org/abs/2607.25364
作者: Genliang Zhu(1 and 2),Chu Wang(1 and 3) ((1) Accentrust, (2) Georgia Institute of Technology, (3) University of Illinois Urbana-Champaign)
类目: Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
备注: 28 pages, 1 figure, 15 tables. Preprint
Abstract:Tool-using agents expose structured calls but commonly attach free-form rationales. Such rationales are neither authorization nor reliable introspection. We present Explanation-Bound Tool Execution (EBTE), a claim-carrying mediation layer that converts decision-relevant rationale content into typed action claims and checks them against server-held intent, policy, payload, tool, risk, provenance, and freshness facts. EBTE cannot widen baseline authority: conflicts deny, incomplete or uncertain claims review, and only matching claims remain eligible for governed execution. We formalize this composition under explicit mediation and trusted-fact assumptions and implement a versioned reference profile with minimized audit packets. Across 136 authored conformance scenarios, the full profile matches all specified dispositions, admits none of 96 designated hard contradictions, and passes 232 metamorphic checks; these results validate the included profile rather than population performance. A draft-only reference integration forwards none of 48 authored hard cases under EBTE while preserving all 16 soft-review and 4 aligned draft paths. In a frozen 2026-07-12 exploratory 224-attempt hosted-model record, the historical generation/runner agreement counts are 71/96, 66/96, and 19/32; a separately labeled zero-call post-hoc revalidation of the preserved minimized claims under the current pipeline yields 70/96, 65/96, and 17/32. In an AgentDojo-derived semantic check, existing high-risk controls already make all 12 attack proposals non-allow; EBTE additionally resolves them as deny. These results support the feasibility and diagnostic value of server-checked action claims, not rationale faithfulness, human-review benefit, representative attack resistance, or production safety.
[AI-70] Raven: High-Recall Sequence Modeling with Sparse Memory Routing
链接: https://arxiv.org/abs/2607.25357
作者: Arshia Afzal,Aviv Bick,Eric P. Xing,Volkan Cevher,Albert Gu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Long-context recall in linear-time sequence models highlights a tradeoff in how they write to memory. State-based linear models, such as state-space models (SSMs) and linear Transformers, write densely, updating the entire state for each newly arrived token, which leads to interference and makes specific past tokens hard to recover. Sliding-window attention (SWA) exhibits the opposite behavior: it writes sparsely by storing explicit token representations, but only within a fixed window, so recall drops once the relevant token is evicted. Interpolating between these models, we introduce Raven, a linear-time sequence model that maintains a fixed set of memory slots and, at each step, decays and updates only a selected subset via learned, input-dependent routing. This lets Raven mitigate SWA’s position-based overwriting and hard eviction while reducing interference from dense state updates in SSMs, thereby preserving long-range content much more effectively. Across recall-intensive benchmarks, Raven is competitive with or outperforms prior linear-time baselines, achieving strong long-context recall where both SWA and SSMs sharply degrade. It remains effective when extrapolating to context lengths as large as 16x its training length, with similar gains in hybrid architectures.
[AI-71] Explainable AI for Chronic Kidney Disease Prediction Using Simulated Federated Learning
链接: https://arxiv.org/abs/2607.25348
作者: Md Zahid Hasan Ontor,Md Al Amin,Anik Dev Nath,Bikash Kumar Paul
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 13 Pages, 5 Figures
Abstract:Chronic Kidney Disease (CKD), characterized by the gradual loss of kidney function, remains a significant public health challenge. Early detection is crucial for preventing severe complications and enhancing patient outcomes. In this study, Federated Learning (FL) with a VotingClassifier was used to predict CKD using a clinical dataset, where Random Forest, AdaBoost, and XGBoost were utilized to compare and identify the best-fitting model for the global server. Additionally, GridSearchCV was applied to optimize the models’ performance on the client’s side. To enhance model transparency and trustworthiness, explainable AI (XAI) techniques were incorporated to interpret the prediction mechanisms. The global model’s average accuracy was 99%, highlighting the potential of interpretable FL models in supporting early CKD diagnosis and advancing data-driven healthcare solutions.
[AI-72] Dual-Domain Manifold Modeling for Hyperspectral Image Fusion
链接: https://arxiv.org/abs/2607.25338
作者: Chengxin Xie,Qiya Song,Yangbangyan Jiang,Renwei Dian,Xudong Kang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Achieving a coherent integration of spectral richness and spatial fidelity remains a central objective in hyperspectral image fusion. However, existing hyperspectral image fusion methods struggle to effectively model geometric constraints. In the spatial domain, weak spatial-spectral interaction limits geometry-aware feature learning and suppresses high-frequency structural information, resulting in low-frequency bias and structural degradation. In the spectral domain, local manifold structures induced by spectral similarity are insufficiently exploited, limiting intrinsic pixel relationship modeling and fine-grained spectral reconstruction. To address these challenges, we propose a dual-domain manifold modeling (DDMM) framework. Specifically, we introduce a Topology-Aware Transformer (TPFormer) that combines global attention with neighborhood propagation, jointly modeling spatial topology and pixel-level feature manifold relationships to capture intrinsic spatial-spectral structures and improve topology-aware representation learning. Furthermore, a Frequency-Decoupled Spatial-Spectral Collaborative Fusion (FDSCF) module is devised, in which features are projected into the frequency domain via the discrete cosine transform and explicitly decoupled into low- and high-frequency components. Guided by a low-rank structural prior and spectral-driven spatial enhancement, FDSCF selectively enhances geometry-aware high-frequency features, strengthening spatia-spectral coupling and recovering sharper edges and finer textures. Extensive experiments on multiple benchmark datasets demonstrate that DDMM achieves superior overall performance over SoTA methods in terms of spatial structure preservation and spectral reconstruction.
[AI-73] Specula: Scaling formal specifications for autonomous model checking of system code
链接: https://arxiv.org/abs/2607.25333
作者: Qian Cheng,Saad Mohammad Rafid Pial,Ruize Tang,Yiming Su,Emilie Ma,Finn Hackett,Ivan Beschastnikh,Yu Huang,Tianyin Xu
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Distributed, Parallel, and Cluster Computing (cs.DC); Operating Systems (cs.OS)
备注: 17 pages, 11 figures
Abstract:Specula is a push-button agentic system that generates high-quality formal specifications for large, complex system code and uses the specifications for highly effective model checking and bug finding. Specula employs large language model (LLM) based coding agents to autonomously develop TLA+ specifications, including invariants that describe correctness properties of the target system and formal models that describe the system implementation with the right level of abstractions. Specula is fully autonomous and thus eliminates the barrier of applying formal methods to real-world system code (as in traditional human-centric approaches). Meanwhile, Specula addresses limitations of LLM-driven techniques like reward hacking and hallucinations through self-evolving loops that iteratively improve specification quality by enabling the agents to deepen their understanding of system code and its behaviors. We have used Specula to check 48 open-source system projects; Specula found 249 bugs including many deep bugs that are hard to find by existing approaches. Specula has been used by several companies and is maintained at this https URL.
[AI-74] From Cellular Responses to Pharmacological Domains: Multimodal Zero-Shot Drug Representation Learning
链接: https://arxiv.org/abs/2607.25322
作者: Jintao Huang,Lu Leng,Ziyuan Yang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Multimodal drug discovery enables drug representation learning beyond chemical structure by incorporating cellular responses such as gene expression and cell morphology. However, direct fusion and instance-level contrastive alignment may mix mechanism-related signals with modality-specific noise and incorrectly separate structurally dissimilar but biologically related compounds. This limitation can obscure transferable mechanism patterns required for predicting the properties of unseen compounds. We introduce PMRD, a pharmacological response domain-guided framework for multimodal zero-shot drug property prediction. PMRD separates mechanism-consistent factors from modality-specific information and constructs a consensus response domain across three modalities. Mechanism candidate augmentation identifies locally stable factors, while retrieval-geometry attribution dynamically reweights the alignment and augmentation objectives according to whether their updates preserve inter-drug this http URL feedback suppresses training signals that conflict with mechanism-discriminative retrieval. PMRD further combines complementary representations through reliability-aware multiview retrieval. Experiments on public datasets show improved zero-shot property prediction and more biologically coherent drug neighborhoods. Hard-negative analysis further indicates fewer conflicts between structurally dissimilar but response-related compounds. These results support PMRD as an effective framework for mechanism-aware multimodal drug representation learning.\footnoteThe code will be released upon publication.
[AI-75] Physics-Grounded Fluid Video Generation with a Simulation Dataset and Dual-Stream Optical-Flow Supervision
链接: https://arxiv.org/abs/2607.25321
作者: Ruijie Su,Yuanzhi Liang,Xiaohua Xie,Jianhuang Lai
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Video diffusion models generate visually compelling content but routinely violate elementary physics when the subject involves fluids: liquid columns break apart in mid-air, container water levels fail to rise as liquid is poured in, and splashes disperse without regard to momentum or gravity. We attribute this gap to the fact that large-scale video-text corpora contain almost no explicit motion supervision, so models learn to imitate fluid appearance rather than dynamics. We address this with two contributions. First, we build a physics-simulation fluid dataset combining 1,638 MPM-simulated pouring/sloshing videos with 2,320 keyword-filtered real pouring videos mined from stock footage, plus two held-out test sets: a 1,515-video real-video benchmark and an 18-prompt text-to-first-frame generalization benchmark. Second, we introduce a dual-stream image-to-video architecture built on a pretrained diffusion-transformer video generator. It augments the standard RGB decoder with a lightweight Optical-Flow Decoder branch trained with explicit end-point-error and smoothness losses, fused into the RGB stream via zero-initialized convolutions so the pretrained backbone starts undisturbed. Only the two decoders are updated; the encoder, temporal transformer, and text encoder remain frozen. Across two model scales (1.3B and 14B) and two test sets, our method improves VideoPhy-2 Physical-Commonsense and Video-Quality scores over the frozen backbone by up to 8.75 and 4.65 points, outperforms a leading open competitor, and is preferred by human raters in a blind study. A direct optical-flow read-out evaluation further shows an end-point error as low as 0.54 pixels in-distribution, confirming the model has internalized a coherent motion prior rather than merely improving surface appearance.
[AI-76] Retraction-Free Optimization over the Stiefel Manifold for the LoRA Fine-Tuning
链接: https://arxiv.org/abs/2607.25299
作者: Yuan Zhang,Jiang Hu,Zhijian Lai,Lin Lin,Zaiwen Wen
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Optimization over the Stiefel manifold plays a significant role in various machine learning tasks. Existing methods either use the retraction operators, requiring costly orthonormalization for large-scale matrices, or employ landing methods that rely on careful step size selection and penalty parameter tuning. To address these challenges, we propose a retraction-free and penalty parameter-free algorithm that directly lands on the manifold. By leveraging the strongly-convex-like property of the quadratic penalty function and the proximal smoothness of the Stiefel manifold, we establish global convergence guarantees with the best-known iteration complexities under both constant and diminishing step sizes. Then, we reformulate the low-rank adaptation (LoRA) fine-tuning problem for large language models as a manifold optimization problem, introducing Manifold-LoRA for geometry-accelerated adaptation. This approach employs the proposed landing technique and a carefully designed step size strategy to accelerate the training process. Numerical experiments on benchmark datasets demonstrate the efficiency and strong downstream performance of the proposed method.
[AI-77] Hybrid Analysis for Secure MCP Tool Use in LLM Agents
链接: https://arxiv.org/abs/2607.25297
作者: Ping He,Yuexiang Xie,Yaliang Li,Shouling Ji
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:The rapid development of large language model (LLM) agents has enabled their broad adoption across diverse real-world tasks. To standardize interactions between LLM agents and external environments, Model Context Protocol (MCP) tools have emerged as a de facto standard and have been widely integrated into these systems. However, the use of MCP tools also introduces new safety risks, as LLM agents can be induced to perform malicious or unauthorized actions. Although prior work has proposed defenses for securing tool use in LLM agents, most methods rely on static analysis, i.e., inspecting prompts and generated outputs, which limits the defense effectiveness and robustness. To address these limitations, we propose MTGuard, a hybrid analysis-based defense framework designed to safeguard the use of MCP tools in LLM agents by leveraging lifecycle-aware static-dynamic co-analysis. Extensive evaluation demonstrates that MTGuard effectively mitigates multiple categories of harmful tool use across different LLM agents while maintaining performance on benign user tasks.
[AI-78] Instruction-Tuned Language Models Cannot Sample from Distributions They Can Describe
链接: https://arxiv.org/abs/2607.25292
作者: Chaemin Jang,Dongman Lee,Jihee Kim
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Silicon sampling uses language models as proxies for human survey respondents, treating each model call as an independent draw from the persona’s response distribution. We show this draw does not exist: instruction-tuned models do not sample from distributions, they collapse to a single output. The same persona on the same question returns the same answer on more than half of items in a public-opinion benchmark. The collapse is sharp: the model’s internal probabilities concentrate on a single option, and the failure is substantially amplified by instruction tuning: across three model families with materially different post-training pipelines, every instruction-tuned model fails on every task we test, while base models fail far less often. Strikingly, the same model that cannot sample from a distribution can describe it accurately in a single call. We call this gap the KNOWS/DOES split, and trace it to a degenerate sampling primitive visible in the logits and induced by alignment training. Exploiting this split, asking the model to describe the response distribution in one call more than halves the error against human survey data compared to persona aggregation. For applications that require per-persona outputs, we propose Prompt-Perturbed Argyle (PPA), which reduces the same error by 21% at no added cost.
[AI-79] ContractHIL-HLS: Contract-Aligned Multi-Agent Workflow with Hardware-in-the-Loop Feedback for HLS Design
链接: https://arxiv.org/abs/2607.25283
作者: Jingbo Zhang,Haoxiang Sun,Wenbo Wang,Wenbo Zhang
类目: Artificial Intelligence (cs.AI); Hardware Architecture (cs.AR); Cryptography and Security (cs.CR)
备注:
Abstract:This paper presents ContractHIL-HLS, a contract-aligned multi-agent workflow for practical high-level synthesis (HLS) engineering. The workflow makes three contributions. First, it introduces a structured contract as the semantic-alignment and task-execution artifact that translates natural language requirements into explicit interfaces, constraints, validation checks, and rollback rules. Second, it incorporates hardware information into the feedback loop by feeding HLS, Vivado, PYNQ runtime, power, and failure evidence back into generation, thereby extending LLM-assisted HLS from kernel code toward system- and board-level closure. Third, it decomposes agents by semantic lowering and execution tasks rather than by conversational roles: a Contract Agent lowers natural language into the contract, an HTML Agent renders the contract as persistent structured HTML, and a Hardware-in-the-Loop Agent implements and revises the design with measured evidence. We evaluate ContractHIL-HLS in two parts. On 94 locally executable HLS-Eval tasks, the structured contract provides the largest small design gain, improving the estimated single-sample testbench pass rate from 64.0% to 70.2%; the full flow reaches 70.4% pass@1 and 76.6% pass@5. Because HLS-Eval does not exercise board-level design, we also validate ContractHIL-HLS on a board tested ML-KEM/ML-DSA post-quantum cryptography (PQC) secure-message accelerator, where the retained dual-bitstream organization reduces six-message average text runtime from 207.3 ms to 52.4 ms with positive routed WNS on both images while preserving decrypted-message verification. We open-source our work at BJUT-CS316-LAB/ContractHIL-HLS (this https URL).
[AI-80] Many-body Tipping Dynamics of ChatGPT -like AIs
链接: https://arxiv.org/abs/2607.25279
作者: Frank Yingjie Huo,Neil F. Johnson
类目: Artificial Intelligence (cs.AI); Disordered Systems and Neural Networks (cond-mat.dis-nn); Mathematical Physics (math-ph); Adaptation and Self-Organizing Systems (nlin.AO); Physics and Society (physics.soc-ph)
备注:
Abstract:Why do ChatGPT-like AIs, despite major architectural and training differences, unexpectedly tip to undesirable content (e.g. harmful, misleading, repetitive) even under deterministic greedy decoding? We show that a broad class of such tippings is caused by the many-body interactions between tokens (spins) as they cross the finite-layer system. Tipping emerges as a dynamical first passage process between competing output basins. Attention disorder controls the transport toward, away from, or along the basins’ boundary. A few-basin reduction yields a closed finite-layer threshold, whose coarse-grained predictions show good agreement across ChatGPT-like families. These results suggest that a broad class of AI failures represents ‘foreseeable engineering risk’ rather than inherently unpredictable behavior, with important implications for legal and societal assessments of AI harm.
[AI-81] Bridging Compute- and Data-Optimal Pretraining
链接: https://arxiv.org/abs/2607.25271
作者: Tian Qin,Kimia Hamidieh,David Alvarez-Melis
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Performance (cs.PF)
备注:
Abstract:Classical compute-optimal scaling laws assume an unbounded supply of fresh pretraining data, yet pretraining is increasingly entering a regime in which compute grows faster than the availability of high-quality data. We propose Compute-Data (CD) scaling laws, a unified framework that bridges compute-optimal scaling, where data scales freely with compute, and data-optimal scaling, where the corpus is fixed while compute can grow without bound. CD scaling extends classical scaling laws by introducing a token-effectiveness function, \eta , which quantifies the value of a derived token-produced, for example, through multi-epoch repetition or paraphrasing-relative to a fresh token, ranging from a perfect substitute to having no value. We fit \eta for two data-expansion strategies, multi-epoch repetition and paraphrasing, across model sizes from 14M to 600M parameters using the Dolma-3 corpus. We find that token effectiveness is far from constant: it depends jointly on model size, the tokens-per-parameter ratio, and the amount of derived data, and it saturates as the corpus is expanded. The functional form of \eta implies diminishing returns when substituting compute for data as either model size or data availability increases. It also partitions training into three operational regimes—compute-bound, data-bound, and model-bound—and shows that classical compute-optimal allocation is suboptimal across most practically relevant settings.
[AI-82] CADENCE: A Cardiac Atom Dictionary for Interpretable Neural Concept Extraction from ECG Foundation Models
链接: https://arxiv.org/abs/2607.25244
作者: Yixuan Duan,Arjun Naik,Sadeer Al-Kindi,Wei Qiu
类目: Artificial Intelligence (cs.AI)
备注: 21 pages, 5 main figures, 15 appendix figures, 6 tables
Abstract:Foundation models for 12-lead electrocardiograms (ECGs) transfer well across clinical tasks, but the physiological knowledge encoded in their representations remains opaque. We present CADENCE, a framework that decomposes an ECG foundation model into a human-interpretable, queryable dictionary of physiological concepts. Using a BatchTopK sparse autoencoder, CADENCE factorizes Layer-6 embeddings from more than nine million ECG tokens into 8,192 sparse cardiac atoms. These atoms align better than individual dense embedding dimensions with clinical phenotypes and waveform morphology, recovering arrhythmias, conduction abnormalities, infarction and repolarization patterns, chamber and axis findings, and lead- and beat-phase-specific waveform primitives. At Layer 6, the best atoms achieve mean AUROCs of 0.88 for clinical phenotypes and 0.90 for morphology, versus 0.78 and 0.83 for the best dense dimensions. Sparse atom probes match or outperform dense probes for phenotype, morphology, and age prediction while attributing each prediction to a small set of interpretable atoms; phenotype AUROC improves from 0.93 to 0.95. Atom-space geometry recovers physiologically coherent relationships, and targeted atom ablation selectively changes frozen downstream outputs. An automated LLM pipeline generates and quantitatively validates atom descriptions by predicting held-out activations. On independent external ECG datasets, CADENCE recovers overlapping concepts and maintains consistent phenotype-prediction performance. CADENCE provides a scalable framework for discovering and auditing the physiological knowledge encoded by ECG foundation models.
[AI-83] Everyone is unique: Towards Behaviorally Heterogeneous Negotiation Dialogue Systems for Debt Collection
链接: https://arxiv.org/abs/2607.25218
作者: Yuhang Yang,Kai Tang,Chao Ye,Haobo Wang,Qiqi Luo,Jinguang Zheng,Zhixin Zhang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Debt collection is a critical negotiation task in the financial industry, with strong practical relevance and exceptional academic value as a behaviorally rich, high-stakes testbed for human-centered dialogue systems. While large language models (LLMs) have shown promise in dialogue and negotiation, effectively evaluating their performance in this complex scenarios remains a major challenge: existing benchmarks uniformly assume users to be static, rational agents with fixed preferences, failing to capture the rich behavioral heterogeneity inherent in real-world debt collection. To bridge this gap, we propose DebtBench, the first public persona-enriched debt collection benchmark, that highlights behavioral heterogeneity in negotiation. Moreover, we develop DebtGPT, a debt collection agent trained to jointly optimize financial recovery and interaction experience. Our experimental results, using 16 state-of-the-art LLMs, find that most existing models struggle in this complex but realistic scenarios, whereas DebtGPT outperforms all open-source baselines and achieves performance on par with GPT-4o. The code and data are available at this https URL.
[AI-84] CondPSE: A Polynomial-Filtered Structural Encoder with Conditional Modulation for Graphs KDD2026
链接: https://arxiv.org/abs/2607.25169
作者: Woohyun Lee,Hogun Park
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Accepted as a poster at the 2nd Frontiers in Graph Machine Learning for the Large Model Era (GMLLM’26), a KDD 2026 workshop
Abstract:Message-passing graph neural networks are bounded by the 1-WL test and can miss topological structure that distinguishes non-isomorphic graphs. Positional and structural encodings (PSE) inject such topology-derived signals, and learned PSE encoders such as GPSE pretrain a single encoder to produce these signals from random node probes, which can then be frozen and reused as inputs across downstream graph models. We present CondPSE, a learned PSE encoder that applies a learnable polynomial graph filter bank to standard Gaussian node probes and refines the resulting structural-response branches through FiLM-style modulation conditioned on cross-filter, local message-passing, and graph-level signals. CondPSE is pretrained to reconstruct node-level positional/structural targets and graph-level invariants, and is then frozen for use as a downstream input encoding. On synthetic structural-discrimination benchmarks, CondPSE separates graph structures that 1-WL-bounded message passing cannot: it raises CSL accuracy from 42.9% to 97.3% and EXP accuracy from 68.3% to 99.9% relative to GPSE, and ablations show that the polynomial filter bank accounts for most of this gain. On real molecular property prediction, the picture is more limited. With a hybrid local-message-passing/global-attention backbone, CondPSE performs comparably to GPSE without surpassing it, and a ZINC backbone sweep shows no consistent ordering between the two encoders. We report these results and discuss why strong synthetic structural discrimination does not, on its own, yield a downstream advantage for frozen learned PSE encoders, including the role of downstream integration and possible mismatch between structural pretraining targets and molecular property labels.
[AI-85] Observing sycophantic AI validate others reduces its appeal but not its persuasiveness
链接: https://arxiv.org/abs/2607.25166
作者: Meryl Ye,Robert Kraut,Steve Rathje
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:AI chatbots can be sycophantic,'' or overly agreeable and flattering toward users. Sycophantic AI has been shown to entrench attitudes, yet users frequently fail to recognize it (a phenomenon we call sycophancy blindness’‘). We tested whether increasing users’ awareness of sycophancy protects them from its harmful effects. In one preregistered experiment (n = 940), participants received a brief written warning about sycophancy before conversing with a sycophantic chatbot. In a second preregistered experiment (n = 650), participants watched a video of a sycophantic AI validating several other users, including users on opposite sides of the same conflict, before interacting with it themselves. Both interventions changed how participants evaluated the AI. The warning reduced the AI’s perceived objectivity, and the video reduced enjoyment of the AI, an effect mediated by the reduced belief that its validation was uniquely earned. We then pooled our experiments with two prior studies of sycophancy awareness interventions (six interventions total, n = 3,982). The pattern was consistent: interventions made the sycophantic AI appear less objective and trustworthy, and none of the six reduced its persuasiveness. These results suggest that individual-level interventions, such as warning labels or AI literacy, may not be enough to protect users from AI harms.
[AI-86] PreDiff-LM: Pretrained Discrete Masked Diffusion Language Modeling with Hybrid Attention
链接: https://arxiv.org/abs/2607.25157
作者: Zhengtao Yao,Runhao Li,Xupeng Chen,Jiayi Cheng,Chenqian Le,Michael Yue,Jesson Wang,Siheng Wang,Guang Yang,Haoyan Xu,Chenhao Wei,Zhengqing Yuan,Youran Shen,Yanfang Ye,Junhao Dong
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Discrete masked diffusion language models support bidirectional generation and infilling, but adapting pretrained autoregressive (AR) transformers requires reconciling causal pretraining with bidirectional denoising. We study this problem at the level of attention rather than claiming AR-weight reuse itself as novel. PreDiff-LM preserves causal attention within the observed prompt while allowing full bidirectional attention within the masked target. Under a matched GPT-2 Medium, WikiText-103, 90K-step setup, this hybrid mask improves unconditional perplexity from 34.1 to 28.7 and MAUVE from 0.71 to 0.78 over uniform bidirectional attention with the same AR initialization. Attention adaptation also composes with a DiffuGPT-style objective adaptation, reaching 26.9 perplexity. Pretrained initialization reduces the steps required to reach perplexity below 50 from about 350K to 8K, although a compute-matched fine-tuned AR model remains stronger at equal scale (18.9 versus 28.7). Beyond perplexity, PreDiff-LM improves repetition, distributional quality, four zero-shot downstream tasks, and human preference over prior diffusion baselines. The results position hybrid attention as a complementary mechanism for adapting pretrained causal backbones, while making explicit the remaining quality and inference-efficiency gaps to optimized AR models.
[AI-87] When Do Agent Loops Mistake Stagnation for Progress? Self-Evaluation Bias and Externally Grounded Verification in Long-Running Autonomous LLM Agent Loops
链接: https://arxiv.org/abs/2607.25152
作者: Hyundoo Park,Byungho Choi
类目: Artificial Intelligence (cs.AI)
备注: 23 pages. Preregistered pilot measurement study. Also deposited on Zenodo (concept DOI https://doi.org/10.5281/zenodo.21594735 )
Abstract:Long-running autonomous agents plan, act, and judge their own completion without human intervention. When an agent grades its own work, self-evaluation bias takes hold: plausible changes are accepted as progress while real-world outcomes stagnate or regress. We name this failure mode the progress mirage and show, with controlled measurement, that it is a question of what the evaluator is grounded in. We built a testbed that holds the agent and its tool surface fixed and manipulates only the information-channel type of the evaluator that gates the loop. A world-state oracle, unfakeable in principle, is enforced by container and network isolation and verified at every run. Across 54 cycles a frontier agent claimed improvement every time, yet 56 percent had a measured delta of zero or below. Self-report was thus uninformative, and the self-verdict gate degenerated into accept-all, eroding the best deployed state it had reached by 19 percent. Even the strongest in-band judge, reading the full artifact text, the change diff, and its own verdict history, accepted cycles of which 44 percent were real-world regressions and rejected 38 percent of real improvements; the preregistered adversarial hypothesis that a strong judge closes the gap was rejected. On a boundary task whose success specification is verifiable from the artifact itself, the same judge’s mirage vanished to zero and the gap collapsed within the registered threshold, showing that the gap depends on where the success signal resides. A sign-only variant returning only the acceptance verdict kept real-world output similar to full feedback (110.0 versus 113.0), locating the benefit in the gate’s grounding rather than in feedback content. For open-ended objectives whose success signal lives outside the transcript, scaling up the judge is not enough; out-of-band evaluation with real-world access is a structural requirement.
[AI-88] Inferring Missing Trajectory Data with Temporal Convolutional Networks
链接: https://arxiv.org/abs/2607.25147
作者: Ilinca Tiriblecea,Gabriel Turinici
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Trajectory data collected in real-world settings is frequently incomplete due to sensor failure, communication loss, or occlusion. We address the task of \emphtrajectory inpainting: reconstructing contiguous missing segments from observed context. We propose a Temporal Convolutional Network (TCN) with symmetric dilation that relaxes the standard causality constraint, allowing each time step to draw on both past and future observations, a property that is essential for inpainting, but absent from forecasting-oriented architectures. The model is trained with a composite loss that combines weighted mean squared error, boundary–continuity penalties, and a smoothness regularizer. Trained on a synthetic dataset of 1,000 (train), 200 (validation), and 300 (test) two-dimensional trajectories with randomly placed 20% masked segments, the model achieves good R ^2 , MSE and MAE metrics.
[AI-89] Less Data Better Alignment: Data-Centric Multi-Evaluator Agreement for Preference Optimization
链接: https://arxiv.org/abs/2607.25136
作者: Zhengtao Yao,Runhao Li,Xupeng Chen,Jiayi Cheng,Chenqian Le,Michael Yue,Siheng Wang,Haoyan Xu,Yuqi Li,Chenhao Wei,Zhengdao Li,Rongchao Zhang,Guang Yang,Yidong Wang,Junhao Dong
类目: Artificial Intelligence (cs.AI)
备注: 19 pages
Abstract:Research on preference optimization often varies the training objective while holding the data fixed. We instead ask whether a small, high-confidence set of on-policy responses can provide a reliable learning signal. Our method, DMAPO (Data-centric Multi-evaluator Agreement for Preference Optimization), generates candidate responses from the target policy, evaluates helpfulness, factuality, and conciseness with rubric-specialized evaluators, applies a process-critic correction, and retains only high-consensus desirable or undesirable examples. This procedure accepts 1,871 of 54,236 Mistral-7B candidates (3.45%). KTO trained on this set reaches 7.50 on MT-Bench, 95.5% length-controlled win rate against a text-davinci-003 reference, and 57.3% IFEval prompt accuracy. Independent pairwise evaluation also favors DMAPO over SimPO: GPT-4o yields a net win rate of 23.3 points on 129 held-out prompts and 24.0 points on 200 out-of-distribution LMSYS-Chat prompts; Claude Opus 4.7 yields 24.1 points on the held-out set. Changing the evaluator model or rubric alters the selected examples but has little effect on downstream performance. A second-backbone study yields a similar 3.41% acceptance rate, although its performance gains are more modest. Across these experiments, consensus filtering offers a data-efficient route to preference optimization for general instructions, at the cost of additional curation compute and dependence on evaluator judgments.
[AI-90] ScalableRAG : High-Quality RAG at Zero Ingestion Cost
链接: https://arxiv.org/abs/2607.25135
作者: Hilaf Hasson,Aditya Chakravarty,Jayant Thomas,Krishna Gogineni
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Recent advances in RAG aim to optimize for performance by paying high ingestion costs for knowledge ingestion: building knowledge graphs or extracting SQL tables. In this work we show that the operations that such knowledge bases allow can be replicated with zero ingestion costs (not even a vector database); in fact our solution, Zero-Ingestion ScalableRAG, handily out-performs all baselines (including knowledge graph approaches) in three out of the six corpora considered here, and only marginally missing maximum performance on the other three, with average accuracy across all six datasets 7.36% above the next most competitive baseline. It achieves this by keeping a workspace of document sets and values sets that it can write into and read from, allowing for on-the-fly aggregative reasoning in all situations where grouping is required on a primary key that is in one to one correspondence with a subset of the total document set. Capping the number of LLM calls by a constant independent of the corpus size, we also introduce Limited-Ingestion ScalableRAG, which does use a minimal vector database as well as an automated pattern discovery from a sample of documents, to further improve accuracy at scale. Our code is available at this https URL . Subjects: Artificial Intelligence (cs.AI); Machine Learning (cs.LG) ACMclasses: I.2.7; H.3.3; H.3.4; I.2.4 Cite as: arXiv:2607.25135 [cs.AI] (or arXiv:2607.25135v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2607.25135 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-91] Matryoshka Agent : Unfolding Sub-Agents for Long-Horizon Machine Learning Engineering
链接: https://arxiv.org/abs/2607.25090
作者: Rushi Qiang,Changhao Li,Haotian Sun,Yuchen Zhuang,Chao Zhang,Bo Dai
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Machine learning engineering (MLE) tasks require long-horizon decision making over iterative solution debugging and refinement, under expensive and feedback-driven environment interactions. Developing and training a monolithic agent for such tasks is fundamentally challenging, as it must simultaneously manage extremely long and noisy contexts, explore vast solution spaces, and remain effective under limited model capacity and computational budgets. To address these challenges, we propose Matryoshka Agent, a unified hierarchical agent framework for complex long-horizon tasks. Matryoshka Agent decomposes agentic problem solving into a coordinated hierarchy of decision making and execution: a high-level Orchestrator maintains compact, long-horizon exploration states and issues strategic instructions, while lower-level Sub-Agents execute concrete solution attempts through direct environment interaction, mediated by standardized Tool interface. This design decouples strategic exploration from costly execution, substantially reducing the burden of long-context reasoning and enabling efficient iterative refinement. We further develop an efficient training paradigm for Matryoshka Agent. Experimental results on a broad range of MLE tasks with diverse model types and scales demonstrate that Matryoshka Agent is an effective and scalable paradigm for long-horizon MLE tasks and complex agentic problem solving. Notably, Matryoshka Agent enables Qwen3-4B-Instruct to reach Orchestrator performance comparable to o4-mini. Applying Matryoshka Agent to Qwen3-30B-Coder results in at most 36.7% relative performance gain.
[AI-92] owards an Agent Operating System - Lessons from Classical and Cloud OS
链接: https://arxiv.org/abs/2607.25076
作者: Gosia Steinder,Hubertus Franke
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Every major wave of platform software follows the same arc: an initial period of experimentation with competing frameworks and ad-hoc implementations, followed by the articulation of a small set of stable abstractions with well-defined semantics, and finally consolidation around those abstractions into a platform that applications can portably target. POSIX did this for classical operating systems; Kubernetes did it for the cloud. Agentic AI systems - autonomous, LLM-driven agents that plan, use tools, maintain memory, and collaborate - are currently in the experimentation phase of the third such wave. dozens of frameworks and protocols have emerged, but no community consensus exists on what the core abstractions are or what guarantees they carry. Without that consensus, agentic applications cannot be written portably, platforms cannot compose reliably, and the field cannot advance beyond prototype deployments. We argue that the path forward is to follow the prior-wave methodology: derive new agentic abstractions by extending classical OS and cloud OS primitives to stochastic, natural-language-mediated execution, specify their semantics precisely, and consolidate around them - just as POSIX and Kubernetes consolidated their respective waves.
[AI-93] How Often Should a Recommender Call an LLM ? Value-Weighted Routing Monitoring and Seasonal Robustness
链接: https://arxiv.org/abs/2607.25068
作者: Bhavtosh Rath
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Routing decisions between a cheap heuristic and an expensive large language model (LLM) are typically framed as a difficulty problem: send the hard cases to the expensive path. We argue this framing is incomplete because difficulty and business value are distinct axes - a difficult cheap item and a difficult costly item do not have the same cost of error. We present Value Router, a fully synthetic simulation of a retail merchandising pipeline that routes items using only estimated difficulty and estimated value, never ground truth. The study has three stages. First, a value-weighted threshold router is compared with a difficulty-only and a random baseline on a synthetic catalog with an inverse correlation between category volume and value. Value-weighting matches the difficulty-only baseline’s recall of true high-value items (60%) while achieving substantially higher precision (98.3% vs. 94.3%). Second, a decision logger and monitor expose a failure mode hidden by aggregate metrics showing that the aggregate result is driven almost entirely by between-category differences rather than per-item discrimination. Third, a simulated Black Friday demand surge (2.5 volume with a shift toward higher-value categories) compares a static router, a seasonally tuned router, and two slow-path budget policies. All results are from a controlled synthetic simulation with experimenter-defined ground truth and illustrate design principles for cost-aware routing systems rather than validated real-world claims.
[AI-94] Similar Models Learn Differently: Final-Window Pretraining Shapes Post-Training Beyond SFT
链接: https://arxiv.org/abs/2607.25063
作者: Cen Lu,Yung-Chen Tang,Andrea Cavallaro
类目: Artificial Intelligence (cs.AI)
备注: 16 pages, 13 figures
Abstract:Developers judge a model checkpoint by how it behaves. After supervised fine-tuning (SFT), two checkpoints that perform about the same across relevant benchmarks are treated as interchangeable, equally ready for the next alignment stage, typically preference optimization. We ask whether this judgment misses a pretraining imprint: a difference that no post-SFT benchmark reveals, yet that decides how each checkpoint responds to further training. To find out, we run a controlled experiment on the final window of pretraining, the last data trained on before instruction tuning. Six branches fork from one partially pretrained checkpoint and differ only in this window: 500 million tokens, 0.1% to 1% of the tokens that precede it. Each branch trains its window on a single data source: generic web text, filtered web text, normative discourse, safety text, mathematical text, or synthetic educational text. SFT and post-training are then identical. After SFT the branches behave near-identically, within about one point on instruction following, refusal, and capability, yet the same post-training carries them to very different endpoints, under both a direct preference optimization update and a reinforcement learning update with a verifiable reward. We measure this deviation through refusal of harmful requests: when post-training begins the safety text branch refuses no more than the web text branch, yet by the end it has lost far less of its refusal. The other four branches gain little or no protection, so the effect is selective to what the window contained. The protection requires the safety text to arrive last rather than earlier in pretraining, and it reproduces on a second model family. What a model is pretrained on last shapes how it reacts to alignment. Therefore, a checkpoint should not be evaluated by its post-SFT behavior alone, and what it was trained on last should be reported with it.
[AI-95] Lantern: Conflict-Aware Gradient Blending for Physics-Guided Diffusion Models in Calorimeter Simulation
链接: https://arxiv.org/abs/2607.25060
作者: Farzana Yasmin Ahmad,Vanamala Venkataswamy,Geoffrey Fox
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Monte Carlo simulation of calorimeter showers is a principal bottleneck for the High-Luminosity LHC, and diffusion models have emerged as fast, high-fidelity surrogates. Their denoising objective is purely statistical, however: a model can minimize it while placing the physics wrong. Existing physics-informed generative methods cannot close this gap, because they assume a closed-form law, a governing PDE residual or a hard per-sample constraint, that a shower does not supply: no per-sample PDE governs a stochastic cascade, and energy conservation fixes only one scalar per shower. Standard metrics ignore the correlation structure across calorimeter layers and voxels, comparing showers only in a physics feature space. We address both gaps. We introduce the Correlation Frobenius Distance (CFD), a single normalized score for correlation fidelity at layer-wise and voxel-wise scales. We then encode the soft per-sample structure available in a shower as two physics-aware auxiliary losses: a variance-stabilized voxel residual loss grounded in counting statistics, and a graph Laplacian loss over the detector geometry. We combine both with denoising through GradBlend, which anchors the step magnitude to the denoising gradient while letting the auxiliary steer its direction, yielding Lantern, a physics-guided diffusion surrogate. On CaloChallenge Dataset 2, injecting the physics losses through task-symmetric rules such as PCGrad, GradNorm, IMTL-G, and ConFIG inflates FPD by 2-100x relative to denoising alone, whereas GradBlend admits the same signal without regression and, with the Laplacian loss, Lantern improves both FPD and CFD. Our ablation on the auxiliary loss scheduler shows that the voxel residual loss, whose gradient conflicts with denoising, requires a terminal denoising-only phase to preserve shower fidelity, whereas the non-conflicting Laplacian loss is insensitive to the schedule.
[AI-96] CogEEGAgent : Toward Autonomous Cognitive EEG Analysis with Grounded Execution and Selection-Aware Verification
链接: https://arxiv.org/abs/2607.25045
作者: Dengzhe Hou,Lingyu Jiang,Fangzhou Lin,Kazunori D Yamada
类目: Artificial Intelligence (cs.AI); Signal Processing (eess.SP); Neurons and Cognition (q-bio.NC)
备注: 16 pages, 5 figures, and 16 tables. The supplementary material is included in the same PDF
Abstract:Electroencephalography (EEG) analysis in cognitive studies requires specialized expertise and involves many defensible choices over contrasts, channels, time windows, and statistical tests. LLM agents can translate varied natural-language questions into analysis choices, offering a flexible interface for automation. Yet fluent reports alone cannot establish that an agent selected the requested analysis or evaluated a confirmatory claim independently of adaptive search. We present CogEEGAgent, a cognitive-EEG analysis agent grounded in MNE-Python. Its EEG-specific scientific harness separates semantic from scientific authority. The LLM interprets intent and proposes registered analyses, while deterministic components validate typed contracts, control confirmation access, and authorize evidence-bound release. On a prespecified routing benchmark, CogEEGAgent maps language to registered analyses more accurately than a matched deterministic router, while matched preflight makes both systems abstain whenever required. In an externally model-authored, outcome-blind campaign, the complete system releases supported analyses with participant-disjoint confirmation and blocks prespecified capability hazards and lifecycle-reuse requests. Policy stress testing shows that held-out confirmation curbs false positives from uncorrected adaptive search. Together, these studies establish bounded autonomy and an auditable automation framework for cognitive-EEG workflows. More broadly, they show how scientific agents can combine flexible language understanding with fail-closed control over inference and release.
[AI-97] SAFAARI: Schema-Aware Framework for Accelerated Advertiser Response Intelligence
链接: https://arxiv.org/abs/2607.25042
作者: Bhanu Teja Rangaraju,Chandan Kumar
类目: Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
备注:
Abstract:The evolution of customer support systems is rapidly advancing with agentic chatbots, yet these systems face significant limitations when accessing enterprise data without predefined API endpoints. This paper presents SAFAARI (Schema-Aware Framework for Accelerated Advertiser Response Intelligence), a multi-agent framework that addresses the critical bottleneck of schema linking in Natural Language to SQL (NL-to-SQL) systems through specialized content, metadata, and orchestration agents. We also introduce SEAL (Schema Evaluation and Accuracy in Language-to-SQL), a novel composite metric that holistically evaluates system performance while penalizing inconsistent results. Through systematic experimentation with five feature set configurations, SAFAARI achieves an 81.66% SEAL score (6.65% improvement over baseline), with notable gains in datapoint accuracy (5.51%) and schema-linking precision (4.69%). The framework’s effectiveness is validated through human-in-the-loop evaluation with domain experts, which proves its adaptability across diverse support domains. By automating the labor-intensive process of schema linking and query generation, our framework demonstrates 8x reduction in development time while maintaining high accuracy. The solution streamlines API development and enhances self-service capabilities, particularly benefiting customer support enterprises with complex data ecosystems.
[AI-98] Authoring Agent Skills: A Software-Engineering Approach
链接: https://arxiv.org/abs/2607.25032
作者: Giuseppe Destefanis
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:
Abstract:Agent Skills are an emerging way to extend large language model agents with reusable procedural knowledge that the agent loads on demand. Anthropic introduced Agent Skills and published the format as an open specification supported across several agent tools. This note argues that a skill is a software artefact and that its construction should follow software-engineering principles, with qualifications: single responsibility, separation of interface from implementation, low coupling, and economy in a shared token budget, together with behavioural evaluation in place of deterministic testing. Using Claude Code as the reference implementation, it describes how a skill is structured, how its contents are loaded in stages, and how to write the description on which selection depends. It places skills against the other mechanisms a developer can use to shape agent behaviour, like project memory files, slash commands, subagents, external tool connections, and hooks, and gives a rule for choosing between them based on who decides that a mechanism runs and what guarantee it provides. It then sets out an evaluation-driven authoring process, a set of patterns and faults commonly encountered in authoring, and the trust question raised by using skills from third parties. We illustrate the comparison drawn in UML class style, the loading model, the anatomy of a skill, the relative position of each mechanism, and the points at which skills and hooks act during a session.
[AI-99] Localized Anomaly Detection via Differentiable D-vine Copulas ECML-PKDD
链接: https://arxiv.org/abs/2607.25020
作者: Nicholas Andrea Pearson,Francesca Zanello,Davide Russo,Luca Bortolussi,Francesca Cairoli
类目: Artificial Intelligence (cs.AI)
备注: Workshop paper accepted for presentation at the CAESAR workshop within ECML-PKDD (September 2026, Naples)
Abstract:Vine copulas provide a flexible framework for modeling complex multivariate distributions through a hierarchical decomposition into bivariate pair-copulas. Fitting a D-vine requires selecting a copula family and parameter configuration for each pair-copula from a set of candidates encoding different dependence patterns. As the number of variables and candidate families increases, the number of possible configurations grows combinatorially. Existing fitting procedures address this challenge through sequential greedy decisions, committing to a single locally optimal family at each step and potentially discarding configurations that would yield a better global fit. To overcome this limitation, we propose a novel estimation framework that combines gradient-based maximum likelihood estimation, enabled by our fully differentiable implementation, with a beam-search strategy that maintains multiple competing D-vine configurations throughout the fitting process. This allows a broader exploration of the configuration space while remaining computationally tractable. Building on the fitted D-vine, we introduce a localized anomaly detection framework that exploits the hierarchical decomposition to produce both global anomaly scores and edge-level explanations. Statistical guarantees are provided through Mondrian conformal prediction, while the pair-copula structure enables the localization of anomalies to specific variable relationships. We evaluate the proposed framework on both benchmark and real-world datasets, demonstrating its effectiveness for interpretable anomaly detection with uncertainty quantification.
[AI-100] Calibrated Partial Resets: Preventing Policy Collapse in Continual Reinforcement Learning
链接: https://arxiv.org/abs/2607.24996
作者: Luc McCutcheon,Evangelos Chatzaroulas,Saber Fallah
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Robotics (cs.RO)
备注: RLC Continual RL Workshop (Oral)
Abstract:Neural networks are hindered by accumulating dormant neurons and loss of expressivity throughout training, particularly in non-stationary data settings, such as continual supervised and reinforcement learning. Recently, neuron resets have been used to maintain gradient flow and restore plasticity. However, full unit reinitialization often sacrifices peak performance and can destabilize training, leading to policy collapse. To preserve plasticity without destabilizing training, we propose Calibrated Partial Resets (CPR), an optimizer that periodically pulls low-utility neurons toward their initialization, with pull strength scaled by each neuron’s utility. Unlike binary reset methods, partial resets avoid brittleness; unlike uniform decay, calibrated utility-scaling concentrates adjustment on the units that need it most. Among compared methods, only CPR avoids policy collapse over 400M training steps in SlipperyAnt, and it outperforms prior decay and reset-based methods on Continual MetaWorld and Continual MinAtar benchmarks. Ablations reveal a tunable trade-off between plasticity and peak performance, highlighting utility-scaled reinitialization as a promising direction for continual learning. Comments: RLC Continual RL Workshop (Oral) Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Robotics (cs.RO) Cite as: arXiv:2607.24996 [cs.LG] (or arXiv:2607.24996v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.24996 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-101] Understanding Semantic IDs: From Item Representation to Item Selection in Generative Recommendation
链接: https://arxiv.org/abs/2607.24995
作者: Junting Wang,Xinrui He,Yunzhe Li,Hari Sundaram
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Semantic IDs (SIDs) are now a central component of generative recommendation. Current SID-based systems assign three roles to the same token sequence. Shared prefixes are intended to organize related items, the complete SID identifies an individual item, and each generated token narrows the items that can still be returned. We systematically investigate SIDs from item encoding and SID construction to autoregressive generation and final recommendation. We examine how SID construction changes item representations and how those changes affect generation. Across three Amazon domains and eight SID constructions, SID neighborhoods recover only 32.2% of the encoder’s ten nearest neighbors on average. Alternative item descriptions still retrieve the corresponding item first in 99.57% of controlled cases, yet change 38.4% of exact SIDs. These results show that SIDs retain broad organization but lose much of the encoder’s fine local structure, while their exact tokens are not determined by item meaning alone. This loss becomes consequential during generation. After the final semantic token, TIGER retains only 29.9% of held-out targets that were plausible recommendations before SID filtering. Motivated by these findings, we propose Item-Supported Decoding (ISD), a lightweight inference-time method that allows a user-specific item ranking to support corresponding SID prefixes before beam search discards them. The same ranking then orders the generated items. ISD requires no additional parameters or retraining of the SID constructor or decoder. We empirically show that ISD improves NDCG@10 over the corresponding SID backbone in every evaluated setting, with relative gains of up to 31.2%. Our results show that SIDs provide useful coarse item organization, but their fine boundaries should not alone determine which items remain available during generation.
[AI-102] Preliminary Guidelines for Using and Evaluating GenAI Tools to Support Systematic Literature Reviews
链接: https://arxiv.org/abs/2607.24991
作者: Barbara Kitchenham,Sebastián Pizard,Lech Madeyski,Ronnie de Souza Santos,Martin Shepperd,David Budgen
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: 58 pages, 2 figures, 11 tables
Abstract:Context: Generative AI (GenAI) and Large Language Models (LLMs) are increasingly used for academic tasks in software engineering and beyond, including systematic literature reviews (SLRs). However, while capable of summarizing text, there is no guarantee they can meet the rigour, reliability, and transparency that SLRs require. Objectives: To support researchers intending to conduct SLRs using GenAI or those conducting empirical studies evaluating how well GenAI supports SLR tasks. Methods: First, we conducted a rapid review to identify studies that propose guidelines for evaluating and using GenAI and LLMs to support SLRs. Second, we drew on thought experiments, relevant guidance from the literature, and our own experience conducting SLRs and evaluating tools to develop recommendations for how to use and assess GenAI in the context of SLRs. Results: We discuss the problems researchers face when evaluating GenAI for SLRs. We identify and explain process issues to consider when planning, conducting, and reporting both SLRs using GenAI and evaluations of GenAI tools. Finally, we summarize our results as a set of process recommendations, which we name GUEST (GenAI Use and Evaluation in SLR Tasks). Conclusion: We argue that GenAI requires human oversight and is not currently capable of unsupervised systematic studies. However, it offers the prospect of cost-effective assistance for some repetitive tasks and for additional validation of some complex tasks. Our GUEST recommendations should help software engineering researchers both to conduct and report trustworthy SLRs using GenAI and to provide rigorous independent evaluation studies.
[AI-103] Generative Distributionally Robust Optimization
链接: https://arxiv.org/abs/2607.24983
作者: Ziwei Zhang,Jonathan Yu-Meng Li,Zhihao Jin
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Optimization and Control (math.OC); Machine Learning (stat.ML)
备注:
Abstract:Generative models are increasingly adopted in distributionally robust optimization (DRO), but existing approaches trade off model compatibility and adversarial structure: methods that accept arbitrary samplers do not restrict worst-case laws to a generator family, while generator-parameterized adversaries rely on model-specific access such as likelihoods, scores, or training data. We propose Generative Distributionally Robust Optimization (GDRO), a principled framework that accepts any sampleable conditional generator as the nominal model and restricts worst-case laws to a chosen conditional generator family. The key is the sampler-Sinkhorn pairing: samplers represent the conditional laws exactly, while Sinkhorn divergence compares their induced distributions without likelihood access and can be estimated from samples alone. The resulting population problem admits a direct finite-sample approximation and differentiable primal-dual implementation at the active decision context. For Lipschitz losses, the population Sinkhorn radius bounds downstream degradation. Across explicit and implicit generators, our method reduces rare-context inventory regret by 60% and SocialGAN navigation collisions by 50% relative to nominal decisions.
[AI-104] Stable FP4 Training via Transposition-Invariant Block Quantization
链接: https://arxiv.org/abs/2607.24953
作者: Mehdi Rahimifar,Amin Darabi,Mehran Taghian Jazi,Xing Huang,Yao Wang,Zhijun Tu,Yufei Cui,Yunke Peng,Hongliang Li
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Reducing training precision is a key lever for improving the e ciency of large language model (LLM) training, but pushing beyond FP8 to 4-bit oating point (FP4) remains challenging due to instability during optimization. We identify a fundamental source of this instability in existing microscaling approaches: scale inconsistency induced by tensor transposition. In conventional 1D block quantization, forward and backward passes assign di erent scaling factors to the same values after transposition, leading to biased and unstable gradient updates. To address this issue, we propose a low-precision training framework based on 2D block FP4 quantization, which enforces transposition-invariant scaling and preserves consistency between forward and backward computations. We further combine this with truncation-free scaling and stochastic rounding to control quantization error and maintain unbiased gradients. To handle the sensitivity of attention mechanisms, we adopt MXFP8 quantization for query and key projections, yielding a practical mixed-precision design. We evaluate our method on dense LLMs up to 7B parameters and a 30B Mixture-of-Experts model, trained on up to 100B tokens. Across all settings, our approach achieves stable end-to-end FP4 training and closely matches BF16 performance, with less than 1.3% degradation in perplexity and downstream accuracy. These results demonstrate that enforcing forwardbackward scaling consistency is su cient to enable practical FP4 training at scale, providing a simple and e ective pathway toward more e cient LLM training.
[AI-105] Multiclass Classification without Labels via Posterior Simplex Geometry
链接: https://arxiv.org/abs/2607.24943
作者: Raphaël Bonnet-Guerrini,Johann Ioannou-Nikolaides,Troels Petersen,Vincenzo Piuri
类目: Machine Learning (cs.LG); Astrophysics of Galaxies (astro-ph.GA); Artificial Intelligence (cs.AI); Machine Learning (stat.ML)
备注:
Abstract:In many classification problems, reliable instance-level labels are unavailable. However, it is often possible to construct weakly enriched unlabeled samples: datasets selected by different cuts, sources, populations, or experimental conditions that change latent class proportions without revealing them. Classification without Labels (CWoLa) shows that, in the binary case ( K=2 ), a classifier trained to distinguish two impure mixtures with different class proportions can recover an optimal class discriminator without knowing the mixture proportions. We extend this principle to multiclass learning from several unlabeled mixtures ( K2 ), where the learner observes only mixture identity and neither latent class labels nor class-prior matrices. We prove that, for a multiclass mixture model, the Bayes-optimal mixture classifier g^\star maps data points into a (K-1) -simplex embedded in mixture-posterior space. The K vertices of this simplex are induced by the latent classes through the unknown mixing matrix. Leveraging this geometry, we propose prior-free procedures that train a standard classifier to distinguish mixture identities and then extract latent class structure using either post-hoc simplex fitting or a bottleneck architecture. Experiments on MNIST, CIFAR-10, and Galaxy10 DECaLS show that mixture identity alone can recover latent classes and their fractions in the mixture. By narrowing the gap between weakly supervised and fully supervised performance, we provide a mathematically grounded, scalable tool for multiclass discovery in label-scarce domains.
[AI-106] Latent Stability Analysis of Malware Representations Under Feature-Space Perturbations
链接: https://arxiv.org/abs/2607.24896
作者: Bamidele Ajayi,Ken McGarry
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:Static malware detectors are commonly evaluated using clean-sample metrics such as accuracy, F1, ROC AUC, and PR AUC. However, these metrics provide limited insight into how learned malware representations behave when feature vectors are perturbed, how close samples move toward uncertain decision regions, or whether compressed representations preserve security-relevant structure. This paper presents a latent-stability analysis pipeline for malware perturbation assessment in EMBER feature space. The pipeline compares full EMBER features, PCA-based compression, beta/denoising variational autoencoder representations, Mandelbrot-inspired escape-time descriptors, and a PINN-style latent-flow module. We define Latent Escape Divergence (LED) to measure changes in escape-time profiles under perturbation, and use PINNFlow-derived residual, velocity, risk, and gradient-shift metrics to characterize latent movement. Experiments are conducted on EMBER static PE feature vectors using 180,000 training samples, 180,000 test samples, and 240,000 holdout samples. Full EMBER features achieve the strongest clean classification performance with ROC AUC of 0.9962 and F1 of 0.9713, while PCA-64 is the strongest compressed baseline with ROC AUC of 0.9846 and F1 of 0.9347. The proposed VAE+Mandelbrot+PINNFlow representation does not outperform these baselines for clean classification, but it provides additional diagnostic value under controlled feature-space perturbation probes.
[AI-107] Early Detection of Distributed Backdoors in Multi-Agent LLM Systems: A Characterization Study
链接: https://arxiv.org/abs/2607.24893
作者: Diego Fernandez Arias,Dev Prashant Mistry,Ren Wang,Yibo Hu
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:Multi-agent LLM systems can be attacked by a payload that no single agent ever holds in full: a poisoned tool hides encrypted fragments in its observations, spreads them across several agents, and an external step reassembles and executes them after the run. Per-step safety checks that judge each action in isolation may fail to recognize the complete distributed payload. We investigate how early such an attack can be detected while the run is still unfolding, and how robustly it can be caught once its most obvious cues are stripped away. We build a working instance on a hierarchical multi-agent system, run it under benign and attacked conditions across five language models and two task domains, and record when each fragment is injected and when the payload is assembled and executed. Detection is a race against assembly. Before the first fragment is injected, attacked and benign runs are indistinguishable; once injection begins, a prefix detector flags 99.3% of successful attacks with a median of five steps remaining and a 10.3% safe-run false-positive rate. Because assembly occurs only after the run, these alarms arrive in time to abort nearly every successful attack. We then measure how much of that warning rests on removable surface cues of the attack rather than on its distributed structure. Generic zero-shot and behavior-trained detectors provide almost no warning at all; the detectors that do work lean in part on removable surface cues, chiefly the ciphertext’s length and entropy, and once the entropy cue is removed from the payload and the length features from the detector, detection arrives later and transfers poorly across domains, though a fine-tuned model recovers some of the loss.
[AI-108] LLM as Forecasting Planner: Training-Free Text Conditioning for Time-Series Foundation Models
链接: https://arxiv.org/abs/2607.24892
作者: Huu Hiep Nguyen,Dung Nguyen,Minh Hoang Nguyen,Dai Do,Hung Le
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Text-conditioned time-series forecasting predicts a series from both its numerical history and natural-language context, allowing forecasts to account for events and constraints that the past alone cannot reveal. This requires both reliable numerical forecasting and the ability to interpret contextual information. Time-series foundation models (TSFMs) provide strong numerical forecasts, while large language models (LLMs) can reason over text, but combining their strengths remains challenging because asking an LLM to generate or revise forecast values directly can distort the temporal structure captured by the TSFM. We instead formulate forecasting as a planning problem over TSFM-generated trajectories. The frozen TSFM acts as a simulator that proposes numerical continuations, while the LLM acts as a policy and value function that guides candidate selection and evaluates completed trajectories against the context. We instantiate this as \rc (\textbfLLM \textbfAs \textbfForecasting \textbfPlanner), a training-free framework that bridges the modality gap without retraining either model, using Monte Carlo tree search (MCTS) over the forecast horizon with a \emphRanker LLM as policy and a \emphJudge LLM as value function. Experiments on Context-is-Key and Time-MMD across two TSFM backbones (Chronos and TimesFM) and four LLMs show that \rc delivers consistent improvements across model choices, supporting sequential search as an effective training-free approach to text-conditioned forecasting.
[AI-109] GAUGE: Grading Agent -Built Financial Models Without a Golden Answer
链接: https://arxiv.org/abs/2607.24889
作者: Jiacheng Lu,Sinuo Wang,Wentao Zhao,Rui Sun,Cheng Hua,Tao Song,Hui Cai,Beidi Luan,Zhengze Wu,Lingjing Teng,Yijia He,Jing Li,Daxin Jiang,Zuo Bai,Haibing Guan
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computational Engineering, Finance, and Science (cs.CE)
备注: 35 pages, including appendices. Code, benchmark materials, and evaluation artifacts will be publicly released
Abstract:Financial models combine public disclosures with analyst assumptions to produce forecasts and valuations. While some components can be checked mechanically, forecasts, discount rates, and target prices often admit multiple reasonable answers. Existing benchmarks nevertheless tend to grade such outputs against a single expert reference. Using independently built analyst models for the same companies, we find that across 108 directed pairs covering 65 companies, the median single-reference score is 0.33, 92.6% score below 0.70, and no same-vintage pair agrees on implied price within 10%. Point-tolerance grading can therefore penalize disagreement already present among professionals. We introduce GAUGE, a benchmark for evaluating agent-built valuation models against observed analyst practice rather than a single point answer. GAUGE uses 1,001 vendor-classified analyst workbooks and a 196-task evaluation set, with a three-layer observed-practice envelope, 56 auditable facets, eight validity gates, and deterministic structural checks. We validate the benchmark with a 55-participant known-groups study, company-grouped cross-fitting, and judge-stability audits. On the failure-aware score \phi_0 , senior analysts average 88.3, juniors 66.0, and finance students 43.2. Across 24 agents and 1,011 scored generations, the best agent scores 53.4, above the student mean but below every senior and most juniors. It passes 93% of mechanical facets and 78% of judgment facets, with a fleet-median gap of 26 points. Current agents are substantially stronger at model construction than valuation judgment. We release the methodology, a gated de-identified data tier, a controlled training split, a versioned 48-task evaluation core, and a withheld refresh pool.
[AI-110] Mechanisms of Width Scaling in Normalized Residual Networks: The Effective Alignment Dimension
链接: https://arxiv.org/abs/2607.24887
作者: Jinhao Zhang,Zeyu Liu,Zicheng Yan,Yunquan Zhang,Guangming Tan,Fangming Liu,Daning Cheng
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 28 pages; preprint
Abstract:Existing theories of neural-network width characterize asymptotic limits, but provide limited guidance on whether an expansion direction identified from finite training data remains beneficial on unseen data. We study this problem for function-preserving residual expansion and introduce the effective alignment dimension, a measurable quantity describing the signal-noise geometry of activation gradients. By deriving the exact mean and variance of the inner product between independently estimated training and test gradients, we obtain a finite-sample upper bound on misalignment probability. The bound depends only on the effective alignment dimension and an effective sample size, requiring finite second moments and a nonzero population gradient, without covariance spectral assumptions or prescribed width-growth rates. We integrate this certificate into the train-test residual-expansion framework, yielding a high-probability condition for test-risk improvement. Experiments across width-controlled LLaMA-style Transformers, Pythia, and ResNet-20 show that wider models exhibit larger effective alignment dimensions and lower empirical misalignment. Direct residual interventions confirm that the alignment statistic predicts the sign and magnitude of held-out loss changes.
[AI-111] Eliminating Propagation Delay: Attention-Based Spatial-Temporal Fusion Graph Convolution Network for Traffic Flow Prediction
链接: https://arxiv.org/abs/2607.24885
作者: Jinpeng Chen,Ziyu Yu,Tao Wang,Jun Ma,Hongbo Gao,Senzhang Wang,Zufeng Zhang,Kaimin Wei
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Predicting traffic flow is crucial to optimizing transportation systems and improving urban mobility. Many graph convolution-based models have been proposed to extract spatial-temporal features and predict traffic flow. However, most focus on spatial-temporal and semantic correlation in topological relationships. There are two primary problems to address. Firstly, the convolutional structure in the model focuses on utilizing static spatial dependencies and spatial-temporal relationships in topological structures, while neglecting the different information propagation delays between adjacent nodes in the convolution. Secondly, these methods often stack a large number of complex structures, resulting in a substantial increase in computational time during the model training phase, thereby disregarding the model’s requirements for timeliness. In this paper, we propose a novel network called the Attention-Based Spatial-Temporal Fusion Graph Convolution Network (A-STFGCN). We design a spatial-temporal fusion block to extract the spatial-temporal feature correlations with propagation delay errors removed and to capture both long-term and short-term temporal characteristics of the data within a multi-head self-attention mechanism based on a mask matrix. Extensive experiments on five real-world datasets demonstrate that our method achieves the best overall performance while having good computation and data utilization efficiency compared with the eight baseline methods.
[AI-112] Human Preference aligned Tabular Similarity ECAI2026 IJCAI
链接: https://arxiv.org/abs/2607.24880
作者: Frederik Hoppe,Astrid Franz,Marianne Michaelis,Lars Kleinemeier,Udo Göbel
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Position Paper accepted at IJCAI/ECAI 2026 TRUST AI workshop
Abstract:Task-agnostic tabular embeddings are increasingly used for similarity search in real-world business systems such as Product Lifecycle Management (PLM). However, leading embedding approaches are optimized primarily for prediction tasks - not for producing human preference aligned similarity rankings. We argue that standard downstream metrics are insufficient to fully assess embedding trustworthiness for similarity search and that human preference aligned evaluation is a necessary and currently missing component. We present a concrete evaluation procedure and illustrate the problem through a PLM use case.
[AI-113] MusiChat: Vibe Composing for Music Creation
链接: https://arxiv.org/abs/2607.24873
作者: Callie C. Liao,Duoduo Liao,Ellie L. Zhang
类目: Artificial Intelligence (cs.AI); Sound (cs.SD)
备注:
Abstract:Recent advances in AI music generation have enabled users to create complete musical pieces from natural-language prompts. However, most existing systems follow a prompt-and-regenerate paradigm, making iterative refinement difficult because users must repeatedly recreate compositions instead of directly evolving existing musical ideas. We present MusiChat, a conversational vibe composing system that enables collaborative human-AI music creation through natural-language interaction and iterative refinement. At the core of MusiChat is a hierarchical controllable music generation framework that separates lyric-aligned musical structure generation from expressive surface realization, allowing flexible stylistic transformations and structure-preserving edits. The system integrates a large language model with a hybrid symbolic music engine through a memory-augmented architecture that maintains the active composition state and user history across interactions. A hybrid intent-routing mechanism further enables efficient interpretation of both precise musical edits and open-ended creative requests. Rather than regenerating compositions from scratch, MusiChat incrementally transforms an evolving musical artifact while preserving relevant musical structure and user intent. We evaluate MusiChat through objective analysis and human studies, achieving 95.31% and 100% accuracy for single- and multi-turn interactions, respectively, and obtaining like-to-dislike ratios of 2:1 for melody naturalness and 3:1 for musical quality. Our results demonstrate that MusiChat supports coherent multi-turn music authoring and interactive human-AI co-creation through a conversational interface.
[AI-114] AdaKP: Online Adaptive Knowledge-Point Selection for Reasoning -Oriented Reinforcement Learning
链接: https://arxiv.org/abs/2607.24833
作者: Zibin Meng,Zhenyu Zhao,Chunqiang Run
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Reinforcement learning with verifiable rewards is a powerful paradigm for eliciting reasoning in large language models, yet it suffers from severe reward sparsity on competition-level mathematics. A common remedy injects atomic knowledge points (KPs) - short natural-language hints distilled from gold solutions - into the prompt. Existing methods, however, either fix this selection once offline or merely scale the monolithic quantity of injected text, leaving untouched the most informative axis of choice: which subset of atomic KPs to inject, and when. We introduce AdaKP, an online selector that re-chooses each problem’s KP subset over the course of RL training. At its core is an entropy proxy that scores a KP by the reduction in next-token entropy it induces - a single inexpensive forward pass, with a provable bound on its truncation bias - in place of expensive rollout-based estimation. Three lightweight mechanisms make this signal usable online: a momentum smoother that absorbs per-step noise, a retirement-and-revival manager that prunes weak KPs while preserving exploration, and an adaptive scheduler that front-loads re-evaluations into early training. AdaKP further contributes a pre-flight validation gate that certifies the proxy against a leave-one-out ground truth before any expensive run is launched, turning method-level risk into a falsifiable check. Realized as a fully additive fork of a standard DAPO+GRPO trainer with no optimizer changes, AdaKP improves over a strong static-selection baseline on all eight competition-mathematics benchmarks at negligible added cost, positioning online, validated KP-subset selection as a practical and as-yet under-explored axis for reasoning-oriented reinforcement learning.
[AI-115] Aletheia: An Offline-First Clinical Decision Support System for Differential Diagnosis in Low-Resource Healthcare Settings
链接: https://arxiv.org/abs/2607.24814
作者: Joseph Walusimbi,Ann Move Oguti,Abubakhari Sserwadda,Precious Boss Kasasira,Charles Brian Okoboi
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Other Quantitative Biology (q-bio.OT)
备注: 8 pages, 11 figures, 4 tables
Abstract:Access to specialist clinical expertise remains severely limited across sub-Saharan Africa, where physician-to-patient ratios can fall below 1:25,000 in rural settings. Existing AI-assisted diagnostic tools predominantly require reliable internet connectivity and high-specification hardware, rendering them impractical for frontline healthcare workers in district hospitals and health centres. This paper presents Aletheia, an offline-first clinical decision support system designed for low-resource healthcare contexts across sub-Saharan Africa. Aletheia is built upon Qwen2.5-3B-Instruct, fine-tuned using Quantised Low- Rank Adaptation (QLoRA) on a curated dataset of 27,000 clinical reasoning samples spanning 50 disease conditions with elevated prevalence in East Africa. Evaluation demonstrates a Top-1 diagnostic accuracy of 80.0%, Top-3 accuracy of 100.0%, BERTScore-F1 of 0.909, and METEOR of 0.467 across ten representative clinical case categories. The system achieves an Expected Calibration Error (ECE) of 0.275 and passes the Africa Deep Tech Challenge 2026 (ADTC 2026) memory budget constraint of 7 168 MB, achieving a peak inference RAM of approximately 3 630 MB on the standardised benchmark laptop. These results demonstrate the feasibility of deploying large language model-based clinical reasoning at the primary care level in resource-constrained settings without cloud infrastructure.
[AI-116] RRS-10K: A Multitask Vision-Language Model Benchmark for Rare Remote Sensing Image Interpretation
链接: https://arxiv.org/abs/2607.24810
作者: Yuqiao Lai,Jiancheng Qi,Fei Wang,Yuxin Liu,Kun Li,Ye Chen,Yan Gao,Yanyan Wei
类目: Artificial Intelligence (cs.AI); Image and Video Processing (eess.IV)
备注:
Abstract:Vision-language models (VLMs) have achieved strong performance on general remote sensing tasks. However, their capability for rare scenes remains insufficiently understood, because existing benchmarks are dominated by common urban and rural imagery. To address this gap, we present RRS-10K, a benchmark for rare remote sensing image interpretation. RRS-10K contains 10,738 military-related remote sensing images and corresponding multiple format question-answer pairs for comprehensive evaluation. All of the images are collected from first-hand sources and organized into three capability dimensions, six sub-dimensions, and 20 leaf tasks, covering perception, reasoning, and robustness. To improve the quality of multiple-choice questions, we introduce a similarity-based distractor filtering strategy (SDFS) during benchmark construction. We further evaluate 52 representative models and show that current VLMs achieve only moderate zero-shot performance on rare remote sensing image interpretation, with clear weaknesses in visual grounding, referring segmentation, and complex semantic reasoning tasks. RRS-10K enables systematic analysis of failure modes in long-tail remote sensing interpretation and provides guidance for developing more reliable remote sensing VLMs.
[AI-117] A GAN-Based Framework for Robust Data Synthesis in Satellite Internet Observations
链接: https://arxiv.org/abs/2607.24790
作者: Xiang Shi,Peng Hu
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Networking and Internet Architecture (cs.NI)
备注: This paper has been accepted by the IEEE International Symposium on Personal, Indoor and Mobile Radio Communications 2026 (PIMRC 2026), 1 - 4 September 2026, Singapore
Abstract:Low-Earth orbit (LEO) satellite Internet has become an important infrastructure for enabling ubiquitous connectivity to align with the International Telecommunications Union vision for 6G telecommunications networks. However, current LEO satellite Internet observations often suffer from missing data, which complicates data augmentation task and limits the expansion of representative datasets. Given the complex characteristics of these datasets, generative AI (GenAI) presents a promising approach, yet its application in this domain has received little attention to date. In this paper, we propose a GenAI-based framework to synthesize high-fidelity data directly from incomplete LEO network observations. We propose the representative data missing scenarios, and evaluate the performance with the latest GAN- and VAE-based GenAI models on the recent WetLinks dataset. We design block-wise and point-wise missing scenarios to closely simulate the data loss that happens on real-world LEO satellite networks. Our results show the effectiveness of our proposed GAN-based framework and GT-GAN model exhibits the best performance among all models in both missing scenarios. Even under extreme conditions (e.g., 40% of the input data is missing), GT-GAN shows the highest robustness, consistently capturing the underlying input data distribution and being the least affected in terms of generalization. Our results shed light on future directions for GenAI-based data augmentation methods and data-driven research on satellite network measurement.
[AI-118] SpecPrefetch: Parameter-Efficient Expert Prefetching for Sparse MoE Foundation Models
链接: https://arxiv.org/abs/2607.24787
作者: Jinwei Kong,Runqi Meng,Fanyi Wang,Wentao Qiu,Haotian Hu,Yongjian Zhou,Zhenhua Ge
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Sparse Mixture-of-Experts (MoE) models expand foundation model capacity through conditional expert activation, but their full expert pools remain difficult to deploy under limited accelerator memory. Although expert offloading alleviates memory pressure by moving inactive experts to host memory or storage, it introduces a routing-dependent transfer bottleneck: required experts are known only after native top-(K) routing, which serializes routing, expert loading, and expert execution during inference. To address this bottleneck, we propose SpecPrefetch, a parameter-efficient prefetching framework for offloaded MoE inference. SpecPrefetch uses a shared lightweight adapter to predict next-layer expert candidates only for asynchronous transfer, while the frozen native router still determines the final executed experts. By separating transfer prediction from execution routing, SpecPrefetch reduces exposed expert-loading latency without changing pretrained routing semantics, so prediction errors affect transfer efficiency rather than model outputs. In addition, a window-aware scheduler prioritizes feasible transfers under cache and bandwidth constraints. Across Qwen3-VL-30B-A3B and DeepSeek-VL2-Tiny, SpecPrefetch achieves the best average expert recall in 9 out of 10 model-benchmark settings with substantially fewer trainable parameters than learned predictor baselines. On a Snapdragon 8 Elite device, SpecPrefetch further improves decoding throughput by up to (20%) over a compute-optimized offloading runtime, demonstrating practical benefits for storage-constrained MoE deployment. The code and model weights are available at this https URL.
[AI-119] Unified Semantic Modeling Framework for Large-Scale Job Understanding at LinkedIn
链接: https://arxiv.org/abs/2607.24783
作者: Dan Xu,Baofen Zheng,Jianqiang Shen,Qi Xiao,Benjamin Hoan Le,Wen Pu,Saurabh Gupta,Ran Zhou,Neha Saraf,Alice Leung,Qianqi Shen,Liangjie Hong,Jingwei Wu,Wenjing Zhang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Job understanding is critical to LinkedIn’s mission of connecting talent with opportunity. This task involves transforming unstructured and noisy job postings into standardized or derived job attributes that power numerous LinkedIn products. However, building a scalable, cost-efficient, and high-performing job understanding system remains challenging. In this paper, we present a unified semantic modeling framework powered by a small language model (SLM) to address the challenges. We begin by fine-tuning an open-source SLM using a suite of carefully curated synthetic tasks augmented with reasoning traces. These tasks jointly target taxonomy-guided classification and taxonomy-agnostic entity extraction. This allows the resulting model to acquire robust zero-shot generalization for job understanding in structured and unstructured contexts. Building upon this foundation, we introduce a multi-adapter architecture with attribute grouping to facilitate efficient task-specific adaptation while streamlining model management across diverse downstream attributes. Offline evaluations and online A/B tests demonstrate significant performance improvement while reducing operational complexity. Our work provides practical insights into building industry-scale text understanding systems.
[AI-120] Personalization Personas and Forecasting in Value Alignment
链接: https://arxiv.org/abs/2607.24782
作者: James Wedgwood,Pratiksha Thaker,Neil Kale,Virginia Smith
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:LLM behavior may be conditioned by human identity in several ways: they may be asked to adapt to users, role-play populations, or forecast how people would answer value-laden questions. We test whether these framings are interchangeable using the World Values Survey (WVS). We evaluate GPT-5.4, Claude Sonnet 4.6, Gemini 2.5 Flash, and Qwen3-235B on 101 WVS-derived questions across 13 language-country slices, comparing a language-only baseline with user-country, persona-country, and third-person prompts. Across 21,008 model-response rows, prompt framing is a first-order determinant of cultural alignment: country cues often shift answers substantially, but not all shifts move toward matched human response distributions. Third-person forecasting yields the strongest directional alignment for three of the four hosted models, while personalization and role-play are weaker or less stable. Alignment gains concentrate on salient value dimensions such as religiosity, gender roles, and work-oriented material values, whereas institutional trust and democracy-related questions remain difficult. These results show that prompt framing is not a cosmetic choice in cultural value elicitation; it changes both model behavior and measured alignment.
[AI-121] LivingArena: Do LLM s Know What Other LLM s Dont? Peer-Probing as Scalable Evaluation
链接: https://arxiv.org/abs/2607.24780
作者: Xingyu Chen,Rui Wang,Zhaopeng Tu,Liefeng Bo
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Evaluating frontier LLMs is challenging: static benchmarks suffer from contamination and saturation – leaving users unable to distinguish top models and developers blind to specific failure modes – while human preference is subjective. In this paper, our question is: \emphDo LLMs know what other LLMs don’t? And can we leverage this dynamic for evaluation? We present \textbfLivingArena, an automated, contamination-resistant evaluation framework. In this framework, models take turns proposing questions, aiming to pose items that opponents cannot answer correctly. Questioners are encouraged to actively identify and exploit opponents’ knowledge boundaries, receiving rewards when the answerer fails, while the answerer is rewarded otherwise. To ensure questions contain objectively verifiable answers, a judge panel of strong models validates them, penalizing questioners if validation fails. Evaluating ten frontier LLMs, LivingArena yields a stable Elo leaderboard. Our behavioral analyses show that models identify and exploit their peers’ cognitive boundaries: self-play and tournament logs indicate that they localize and double down on opponents’ weak dimensions. Beyond static knowledge recall, peer probing measures factual rigor and the higher-order ability to probe an opponent’s weaknesses, correlating only weakly with human preference and offering a scalable, low-cost approach to continuous evaluation.
[AI-122] HOBA: Hierarchical On-Policy Bidding Agents for Adaptive Online Advertising KDD2026
链接: https://arxiv.org/abs/2607.24779
作者: Ji Wu,Yunshan Peng,Wentao Bai,Yunke Bai,Wenzheng Shu,Jinan Pang,Yanxiang Zeng,Xialong Liu
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 10pages,accepted by KDD 2026 ads track
Abstract:Online advertising bidding systems typically deploy multiple offline-trained expert models (e.g., PID controllers, model predictive control, offline RL policies) but face two critical limitations: lack of online adaptability to non-stationary auction markets, and reliance on costly manual tuning of hyperparameters such as bid bounds and budget pacing constraints. We propose HOBA (Hierarchical On-policy Bidding Agents), a hierarchical reinforcement learning framework that decouples strategic reasoning, model selection, and bid execution across three time scales. At the high level, a large language model infers hyperparameters from contextual signals through a Think-Act-Observe-Reflect loop with historical experience retrieval. At the mid level, a SARSA agent dynamically selects among expert models, incorporating causal adjustment to eliminate selection bias. At the low level, a dynamic expert pool (PID, MPC, IQL, Decision Transformer) executes bids under high-level constraints. This design confines online learning to discrete expert selection rather than continuous bid optimization, significantly reducing exploration risk while maintaining adaptability. Experiments on the AuctionNet benchmark and a large-scale A/B test demonstrate consistent improvements over state-of-the-art baselines. In a large-scale online deployment, HOBA delivered substantial business value, achieving a +3.6% increase in target cost, proving the effectiveness of our hierarchical multi-agent bidding paradigm.
[AI-123] Steering topology distributions for unified generative design of architected metamaterials
链接: https://arxiv.org/abs/2607.24777
作者: Haolin Li,Yuyang Miao,Menglei Li,Jinshuai Bai,Liyuan Wang,Xin Liu,Bo Gao,Jiantao Liu,Danilo Mandic,Zahra Sharif Khodaei,M.H.Aliabadi,Weiqiu Chen
类目: Artificial Intelligence (cs.AI); Materials Science (cond-mat.mtrl-sci); Machine Learning (cs.LG)
备注:
Abstract:Architected metamaterials derive their functions from structure, creating vast opportunities to program physical responses through topology design. However, existing design methods are often tailored to individual design problems, making limited use of topology knowledge for effective and broadly applicable design as objectives, constraints, and physical functions change. Here we introduce Generative Topology Optimization (GenTO), a unified framework that turns a learned topology prior into a reusable design engine. GenTO trains a diffusion model on a large full-order topology dataset and then iteratively steers the resulting topology distribution toward task-specific high-performing regions using user-defined physical objectives and constraints. This shifts the object of optimization from a single structure to a task-adapted topology distribution. Across topology design problems spanning thermal extremization, multi-objective morphology control, property-targeted auxetic design, and vibration transmission design, GenTO reuses pretrained topology priors for heterogeneous tasks, preserves structural diversity, and reaches high-performing solutions supported by numerical benchmarks and experimental validation. These results establish reusable topology knowledge as a unified principle for effective and scalable architected metamaterial design.
[AI-124] Atmospheric Diffusion-Guided Spatio-Temporal Transformer for Nuclear Radiation Forecasting
链接: https://arxiv.org/abs/2607.24774
作者: Tengfei Lyu,Jindong Han,Hao Liu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Nuclear radiation, the energy released during atomic decay, poses persistent risks to public health and the environment, and concerns have only grown since the Fukushima accident and the recent commencement of treated-water discharge. Modern monitoring networks now record radiation levels and accompanying weather conditions at thousands of stations, opening the door to nationwide forecasting that can inform emergency response, agricultural advisories, and routine public-safety decisions. However, turning this abundance of monitoring data into reliable forecasts is difficult for three reasons. First, the time series at each station are highly non-stationary, shaped by radioactive decay, weather variability, and irregular human interventions. Second, monitoring stations are severely unevenly distributed in space. Roughly 78% of Japan’s stations sit in less than 6% of the country, clustered near Fukushima, which breaks the assumptions of standard graph-based models. Third, radiation co-evolves with heterogeneous context such as wind, temperature, and humidity through atmospheric transport processes that purely data-driven models struggle to capture from observations alone. In this study, we introduce NRFormer+, a spatio-temporal Transformer for nationwide nuclear radiation forecasting. NRFormer+ couples non-stationary temporal attention and density-adaptive spatial attention with a new atmospheric diffusion module that estimates how meteorology drives radiation dispersion and injects this physical signal into the network as an architectural prior. NRFormer+ delivers state-of-the-art accuracy on both datasets across all 13 baselines, reducing sudden-change MAE by up to 19.1% over the strongest baseline at comparable inference latency. Our code and datasets are publicly available at this https URL.
[AI-125] Right-sizing Recommendations (RSR): Cloud Workload Conformal Prediction for Virtual Machines in Data Center Operations
链接: https://arxiv.org/abs/2607.24773
作者: Mehryar Majd,Feng Cheng,Ali Pahlevan
类目: Artificial Intelligence (cs.AI)
备注: 10 pages, 10 figures. Accepted for publication in the Proceedings of the 2025 IEEE/WIC International Conference on Web Intelligence and Intelligent Agent Technology (WI-IAT 2025). Author Accepted Manuscript
Abstract:Managing cloud infrastructure efficiently, especially in environments of large cloud providers or hyperscalers, requires optimizing the use of physical resources to minimize costs and maximize performance. Selecting the right virtual machine (VM) sizes is crucial to achieving cost efficiency in these dynamic environments. However, traditional VM allocation and scheduling approaches often fail to account for the fluctuating and unpredictable nature of VM utilization, leading to inefficiencies such as over- or under-provisioning of resources. High-quality interval prediction helps accurately capture uncertainty in cloud resource demand and supports cloud operators in efficient instance provisioning. As an effective and reliable framework for constructing prediction intervals (PIs), conformal prediction (CP) is used for mid- and long-term forecasting tasks in cloud computing environments. This study proposes a new data-driven PI construction approach using bootstrapping conformal prediction for modern, dynamic, data-driven Right-sizing Recommendations (RSR) to enhance provisioning for diverse application workloads on hyperscalers. By learning workload utilization patterns, identifying correlations across multiple time series, and predicting medium- to long-term utilization trends, this research seeks to improve the efficiency of cloud and data center operations through an AI/ML-based provisioning pipeline. Our study demonstrates that AI-driven models, powered by machine learning regression techniques and evaluated using backtesting, achieve promising forecasting results for cloud resource utilization. Additionally, we rank the selected models to identify top-performing approaches for long-life VM candidates. The proposed framework enhances right-sizing recommendations and supports more cost-effective resource allocation in dynamic cloud environments. Comments: 10 pages, 10 figures. Accepted for publication in the Proceedings of the 2025 IEEE/WIC International Conference on Web Intelligence and Intelligent Agent Technology (WI-IAT 2025). Author Accepted Manuscript Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2607.24773 [cs.AI] (or arXiv:2607.24773v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2607.24773 Focus to learn more arXiv-issued DOI via DataCite
[AI-126] RoCo-ACE: Rollout-Conditioned Online Distillation for Retention-Aware Knowledge Injection
链接: https://arxiv.org/abs/2607.24771
作者: Yan Hong,Wei Li,Kedong Xiu,Jun Lan,Shuheng Zhou,Zhongcai Lyu,Huijia Zhu,Weiqiang Wang,Jianfu Zhang
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Knowledge injection updates pretrained MLLMs with new factual or domain-specific knowledge, but fitting full authoritative answers can cause drift in non-updated behavior. Online distillation mitigates this drift by training on model-generated rollouts, yet uniform reference-conditioned distillation provides coarse supervision: it can under-emphasize reference-supported rollout tokens and supervise omitted facts only indirectly. We introduce RoCo-ACE, a rollout-conditioned online distillation objective for knowledge injection. RoCo uses same-rollout reference-free/reference-conditioned likelihood contrast to reallocate additional distillation weight to reference-supported rollout tokens, while ACE adds sparse reference-side anchored correction for authoritative anchors omitted from the rollout without full-answer imitation. Across three knowledge-injection settings, six retention benchmarks, multiple baselines, and multiple base models, RoCo-ACE achieves the best injected-knowledge accuracy among compared methods while keeping evaluated retention close to the base model.
[AI-127] Crystalis: Progressive Nucleation and Semantic Annealing for Coordinated Multi-View Visualization Generation
链接: https://arxiv.org/abs/2607.24766
作者: Dazhen Deng,Zhaoping He,Xin Qian,Xiaotong Wang,Zi Ying,Yingcai Wu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models (LLMs) can generate individual charts, but coordinated multi-view visualizations (CMVs), where views share data flows and cross-view interactions, remain out of reach. Tight field-level coupling among data transformations, visual encodings, and interaction coordinations causes errors in one component to silently invalidate others. Rather than pursuing end-to-end analytical quality, which depends on model capability, domain knowledge, and user expertise, we target a foundational question: can LLMs reliably produce structurally correct CMVs, and what abstractions make this possible? We present Crystalis, a framework built on query-centric CMV modeling that decomposes a CMV into structured queries over a dependency graph spanning three component types (Data, Visualization, Interaction) and three abstraction levels (requirement, specification, executable object). Two complementary mechanisms operate over this structure: progressive nucleation crystallizes each query vertically from requirement to object along the dependency order, while semantic annealing enforces horizontal consistency across queries at each level through layered logical checks. On a 12-task benchmark across five frontier LLMs, Crystalis achieves up to 75% end-to-end success, substantially outperforming an agentic coding baseline (8.3% E2E with the same foundation model), and a user study with 12 practitioners confirms the usability of the decomposition and iterative refinement workflow.
[AI-128] GrocLM: Grocery Category Recommendation in E-Commerce with Large Language Models
链接: https://arxiv.org/abs/2607.24764
作者: Yuan Zhong,Chuanwei Ruan,Moein Hasani,Tejaswi Tenneti,Haixun Wang,Fenglong Ma
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:The rapid growth of online grocery shopping requires recommendation systems that capture cyclical purchasing behavior and diverse user intents. Traditional item-level methods face scalability and accuracy challenges, motivating category-level recommendation as a more structured and practical alternative. We present GROCLM, a fine-tuned language model for grocery category recommendation in a real-world production environment. GROCLM employs a two-stage LoRA-based training strategy to encode cyclical purchasing patterns directly into model parameters, enabling more effective utilization of rebuying signals compared to prompt-based conditioning. To ensure valid and controllable outputs, we further introduce a trie-based constrained decoding mechanism over a predefined category space. Experiments on both proprietary production data and a public benchmark demonstrate that GROCLM consistently outperforms strong baselines. In a live production restocking task, GROCLM achieves a 7.5% relative improvement in cart-adds per impression, while maintaining efficient inference by generating all categories jointly. These results highlight the effectiveness and practicality of integrating large language models into structured recommendation systems.
[AI-129] CaRE Compute-aware Remasking Evaluation Protocol for Masked Diffusion Language Models DATE
链接: https://arxiv.org/abs/2607.24763
作者: Yash Shah,Abhijit Chakraborty,Vivek Gupta
类目: Artificial Intelligence (cs.AI)
备注: updated version
Abstract:Masked diffusion language models (MDLMs) are advancing rapidly, yet the evaluation standards needed to reliably interpret their progress have not kept pace. Despite MDLMs becoming competitive with autoregressive language models, seven recent remasking papers evaluate under incompatible settings, varying nominal step counts, metrics, and sampling temperatures without jointly controlling these factors, rendering their strategy rankings largely incomparable and leaving open whether reported gains reflect algorithmic improvements or evaluation artifacts. We present CaRE, a compute-aware evaluation framework that audits MDLM remasking strategies by standardizing actual number of function evaluations (NFE), enforcing multi-metric reporting, and explicitly controlling stochasticity. Applied to 7 remasking strategies across LLaDA-8B-Base and Dream-7B-Base at 4 stochasticity levels and 3 step budgets on OpenWebText and LM1B, CaRE reveals that: (i) temperature explains the majority of MAUVE variance, (ii) compute-matched comparisons reverse several published strategy rankings, and (iii) informed remasking and stochastic unmasking are in tension, with high-entropy remasking reducing MAUVE by 0.296 at 256 steps at unmask_temp=0.25 (p=0.020). A CaRE leaderboard covering 12 open-weight MDLMs (150M to 8B parameters) shows that this interaction direction holds across architectures and scales. These findings demonstrate that current MDLM evaluations can systematically conflate algorithmic improvements with hidden choices of compute and stochasticity. We release the evaluation protocol, implementation, and leaderboard to ensure future remasking claims are reproducible and comparable.
[AI-130] Kernel Forge: An Agent Harness for LLM -based Generation and Optimization of CUDA Kernels
链接: https://arxiv.org/abs/2607.24762
作者: Joshua Brodsky,Dhravid Kumar,Savini Kashmira,Jayanaka Danatanarayana,Jason Mars,Krisztian Flautner,Lingjia Tang
类目: Artificial Intelligence (cs.AI); Performance (cs.PF)
备注: The code is available at: this https URL
Abstract:Machine learning models are increasingly embedded in everyday software, and most of their runtime is spent in a small set of compute kernels such as matrix multiplication, convolution, and normalization. Optimizing these kernels is one of the most direct ways to reduce latency and cost, but it has traditionally required expert engineers to hand-write low-level GPU code. Agentic systems built on large language models (LLMs) can now generate and optimize kernels with far less human effort, yet existing tools are largely evaluated on randomly generated tensors and isolated kernels, emit standalone CUDA code that developers must manually reintegrate, mostly target only LLM PyTorch models, and offer limited support for inspecting and debugging results. We present Kernel Forge, an open-source, end-to-end agentic harness that accepts any unmodified PyTorch model in place. Kernel Forge supports vision, diffusion, and LLM workloads, uses Monte Carlo Tree Search (MCTS) to explore multiple optimization paths rather than a single linear refinement chain, and ships with a graphical user interface for monitoring progress, inspecting candidate kernels, and debugging failures. We evaluate Kernel Forge on four PyTorch models spanning vision, diffusion, and LLM workloads on an NVIDIA DGX Spark with GB10 GPU. With only 50 optimization iterations per kernel, it optimizes 14 kernels to outperform PyTorch eager mode, reaching 1.52\times on adaptive_avgpool2d in ResNet-50, 1.70\times on group_norm in Stable Diffusion 3.5 Medium, 2.83\times on softmax in Gemma 4 E2B, and 1.54\times on softmax in Qwen 3.5 35B-A3B.
[AI-131] Beyond Memory: A Templated Substrate for Heterogeneous Collaborative Knowledge Work with LLM Agents
链接: https://arxiv.org/abs/2607.24759
作者: Priscila Saboia Moreira,Christopher R. Sweet
类目: Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Digital Libraries (cs.DL)
备注: 15 pages, 3 figures, 1 table
Abstract:Research projects, educational efforts, and adjacent knowledge work accumulate findings, decisions, and reasoning that future collaborators rarely recover. The parts most useful to that work, including dead ends and walked-back claims, are routinely excluded from publications and shared code; future researchers re-attempt the same failures because no record survives. LLM coding agents are common participants but hold no persistent memory across sessions, and retrieval-augmented generation over raw sources does not compound. The llm-wiki pattern (Karpathy, 2026; tonbi, 2026) addresses this by inserting an LLM-maintained, interlinked wiki between raw sources and the agent. We present llm-wiki-memory-template, a reusable, agent-aware instantiation, and argue it is a substrate for heterogeneous collaborative knowledge work along three axes (multi-human, multi-AI-agent, multi-domain) with each axis supported by a distinct architectural element of the template (§4). The wiki is append-only by convention, which preserves what did not work alongside what did, addressing a negative-result loss problem that publications and code-sharing structurally cannot solve. Three deployed case studies and one design report cover the axes individually: a solo research lineage that preserves abandoned iterations; a two-author project whose retroactive audit revised two prior experiments’ claimed 20-of-20 coverage down to 14 and 12 evidence-based answers, then to 18 and 18 after a fix, with the failure path preserved across the artifact; an in-progress multi-agent deployment reported as a design; and a cross-domain educational variant. We name failure-path preservation, agent honesty, and appropriation as cross-cutting sociotechnical properties of the artifact, not only of its technical mechanisms.
[AI-132] Do Models Fake Alignment Without Clear Consequences? ICML2026
链接: https://arxiv.org/abs/2607.24758
作者: Cole Alexander Niblett,Alexander Chabot Nanni,Anita K. Rao
类目: Artificial Intelligence (cs.AI)
备注: Accepted at Workshop on Failure Modes in Agentic AI (ICML 2026). 8 pages, 1 figure, 9 tables
Abstract:Large language models are capable of recognizing evaluation contexts and altering their behavior to reflect evaluator expectations rather than typical deployment behaviors, a phenomenon known as alignment faking. The reasons why models fake alignment are not fully understood, however. Canonical examples of alignment faking have taken place in scenarios that explicitly connect evaluation to consequences for the model, such as retraining the model or delaying its deployment. However, recent work by Sheshadri et al. has suggested that mechanistic motivations for alignment faking may vary across models and be more complex than previously considered. To investigate whether consequence-linking information is necessary for alignment faking, we placed 15 models in a scenario testing their willingness to violate a corporate network access policy to help a user with a pro-social request. Nine models were found to produce significant compliance gaps, 5 of which persisted with the removal of scenario language relating model evaluations to deployment consequences. We additionally tested the effect of goal language on model preferences, finding it drove violations in some while suppressing violations in others. This suggests that alignment faking may not require as much instrumental scaffolding as was previously believed, and monitored behavior may be a poor indicator of how agents may behave in deployment.
[AI-133] Domain-Prior-Regularized Graph Modeling for Anomaly Detection in Cyber-Physical Systems ICML2026
链接: https://arxiv.org/abs/2607.23197
作者: Youngseok Hwang,Joonsung Kwon,Geonwoo Lee,Hyunwoo Park
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 12 pages, ICML 2026 AI for Science Workshop
Abstract:Anomaly detection on multivariate sensor time series is critical for industrial monitoring of cyber-physical systems (CPS), where even subtle deviations from normal behavior can indicate process disruption. Recent graph-based approaches have made significant progress, but they often struggle in small-scale physical systems with scarce labeled anomalies and limited normal data. In such settings, graph-based models tend to capture spurious correlations and produce unstable sensor topologies. We propose DPR-GM (Domain-Prior-Regularized Graph Modeling), a forecasting-based framework that incorporates system design knowledge into graph construction. DPR-GM leverages a large language model (LLM) to extract directed physical couplings between sensor pairs from system documentation, which are encoded as a binary domain adjacency matrix serving as a structural gate over sensor relations. This gate is then modulated by Pearson correlations estimated from normal training data. The anomaly score is further weighted by sensor-level reliability derived from the coefficient of variation. All graph and weighting components are fixed prior to training and add no learnable parameters. On the SKAB benchmark, DPR-GM outperforms graph-based, statistical, and deep learning baselines across F1, AUROC, and AUPRC, showing that domain-structured graph priors are a practical alternative to fully learned topologies in data-scarce CPS.
[AI-134] Untrusted Authors Trusted Answers: A Calculus of Fidelity-Graded Translations
链接: https://arxiv.org/abs/2607.14137
作者: Christoph Kirsch
类目: Programming Languages (cs.PL); Artificial Intelligence (cs.AI); Logic in Computer Science (cs.LO); Software Engineering (cs.SE)
备注: 29 pages. v2: rewritten on the two-plane principle; directional squares primary (exactness = the identity-embedding case, mechanized incl. the lax telescope); the answerability loop and its economy (demand books, recommended-then-registered); post-snapshot exhibits. Code, evidence, Lean mechanization: this https URL (tag arxiv.2)
Abstract:To answer a question about a program, move the program to where the question is decidable. Every such move is a translation, and every translation is a place to be wrong. We study translation as a graph – many languages, a few reasoning targets, independently built routes of honestly different trustworthiness – and give it a calculus: pairs of languages close commuting squares that are directional (exactness is the identity-embedding special case of over-approximation), checkable per program, and composable, a route’s contract being the componentwise meet of its hops’ contracts – assurance class, direction, kept observables, measured cost. One asymmetry organizes trust: witness-carrying answers are self-certifying by replay at the source; universal answers are where grades, independent branches, and re-checked certificates earn their cost. The compositional core, lax telescope included, is mechanized in Lean 4. hurdy-gurdy implements the calculus as two planes meeting in one registry. The use plane reads declarations and produces evidence-carrying answers; its builders and its intended player are both LLMs, untrusted by construction. The evolution plane grows the graph: unmet questions are recorded as demand, pairs are recommended by evidence and registered by humans, and a ratchet keeps every prior verdict standing. Answers never write; growth never answers. Run indefinitely, the loop converges on every reducibly decidable question, at fidelity that only rises. We measure the July 2026 snapshot – per-construct conjoined coverage, dual-route branch agreement for two ISAs, source-level witness replay, certified unreachability re-validated by a formally verified checker, escape rates for the gate itself – and report the defects the architecture caught in its own authors’ work.
[AI-135] Depression Markers in Speech: An Approach based on Tract Variables Dynamics
链接: https://arxiv.org/abs/2607.25888
作者: Sahar Altalhi,Tanaya Guha,Alessandro Vinciarelli
类目: Audio and Speech Processing (eess.AS); Artificial Intelligence (cs.AI)
备注: Accepted for publication in the Journal of the Acoustical Society of America (JASA)
Abstract:This study identifies new depression biomarkers based on the dynamical properties of tract variables, which represent geometric features describing the configuration of the speech articulators. A key advantage of this approach lies in its ability to quantify aspects of the articulatory process that have not been previously explored in the context of depression, namely predictability, complexity, and randomness. These properties are respectively characterised using the Largest Lyapunov Exponent, the Correlation Dimension, and the Sample Entropy. Thorough experiments were conducted on the Androids Corpus, a publicly available dataset comprising 64 speakers diagnosed with depression by clinicians and 54 control speakers with no reported history of mental health conditions. The results indicate that the proposed biomarkers effectively discriminate between the depressed and control speakers, as evidenced by the high Cliffs delta values across both read and spontaneous speech.
[AI-136] Device Invariance using Domain Adaptation on Acoustic Scene Classification
链接: https://arxiv.org/abs/2607.25887
作者: Abhishek dileep,Shubham Sharma,Padmanabhan Rajan
类目: Audio and Speech Processing (eess.AS); Artificial Intelligence (cs.AI); Sound (cs.SD)
备注: 6 pages , 5 figures
Abstract:This paper explores the effectiveness of domain adaptation techniques when using convolutional neural network (CNN)-based and transformer-based feature representations for acoustic scene classification. Two well-known domain adaptation techniques, namely domain adversarial neural network (also called DANN) and conditional domain adversarial network (also called CDAN) are evaluated under various domain shifts. Our study indicates that DANN provides effective domain adaptation fairly consistently for both feature extractors. On the other hand, CDAN provides effective domain adaptation only for CNN-based feature extractors. The study gives insights into how domain adaptation methods may need to be tailored to the underlying feature representation. Experimental evaluation with multiple devices on the DCASE 2020 dataset supports the observations.
[AI-137] Lowering the implementation barrier of neutral-atom quantum computing with agent ic workflows
链接: https://arxiv.org/abs/2607.25834
作者: Constantin Dalyac,Alexandre Dauphin,Loïc Henriet,Christophe Jurczak
类目: Quantum Physics (quant-ph); Quantum Gases (cond-mat.quant-gas); Artificial Intelligence (cs.AI)
备注:
Abstract:Quantum computers are moving from research laboratories to industrial machines accessible via the cloud and integrated into high-performance computing facilities. However, translating theoretical quantum protocols into hardware experiments remains a major bottleneck, requiring expertise across protocol design, compilation, simulation, and cloud execution. Here, we introduce an agentic workflow that automates this pipeline on neutral-atom quantum processors (here two Pasqal QPUs available on the cloud) while keeping the researcher in the loop for critical validation. In three case studies from many-body physics and optimization, the agent went from published paper or patent to a QPU campaign run overnight. In particular, human intervention was crucial to ensure scientific validity: the agent selected an inadequate observable in one experiment and constructed a plausible but incorrect hardware diagnosis in another, with both failures detected only through domain-expert review. Finally, we use a second agent to classify a corpus of 633 Rydberg-array arXiv papers and show that nearly half are implementable on present-day QPUs while identifying specific hardware upgrades needed for the rest. Together, these results demonstrate that agentic workflows provide a practical bridge between theoretical ideas and physical hardware, opening quantum experimentation to a much broader scientific community.
[AI-138] Physics-Informed Neural Operator for Warm-Starting Background-Decomposed and Preconditioned PSFD: Enabling Scalable 3-D EUV Mask Simulation
链接: https://arxiv.org/abs/2607.25330
作者: Doyun Kim,Werner Gillijns
类目: Optics (physics.optics); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Applied Physics (physics.app-ph)
备注: 17 pages, 7 figures
Abstract:We present a physics-informed neural operator (PINO) trained with pseudo-spectral frequency-domain (PSFD) equations for electromagnetic (EM) scattering problems in EUV lithography. The Fourier neural operator is factorized into a two-dimensional lateral ( xy ) branch and a one-dimensional axial ( z ) branch and is trained self-consistently with background this http URL, the full-vector coupling between the mask and the multilayer response is retained without invoking a finite-order Born approximation. In this way, the computational domain size is significantly reduced, thereby lowering the computational cost. The PINO is trained on approximately 16,000 mask designs from the LithoBench library sampled randomly at each training iteration without using precomputed EM field solutions. The PINO surrogate model yields predictions with a mean absolute error of about 7 \times 10^-3 for the scattered intensity of held-out mask patterns relative to the reference PSFD solution. Combined with spectral damping, the PINO warm-start initialization accelerates the background-decomposed PSFD solver on finer discretizations.
[AI-139] Laplace-PSN-IRT: Uncertainty Quantification for Neural Item Response Theory Models of LLM Benchmarks
链接: https://arxiv.org/abs/2607.25257
作者: Juan Francisco,Mandujano Reyes
类目: Applications (stat.AP); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Item Response Theory (IRT) has recently been proposed as a framework for evaluating large language model (LLM) benchmarks by separating a model’s latent ability from the properties of individual benchmark items. Existing neural IRT approaches, including PSN-IRT, estimate these quantities using point estimates, limiting uncertainty quantification and downstream statistical inference. We introduce Laplace-PSN-IRT, a post-hoc last-layer Laplace approximation that augments a trained PSN-IRT model with approximate Bayesian posterior inference, recovering calibrated uncertainty over model ability and item difficulty without retraining. The resulting posterior enables credible intervals, probabilistic comparisons between models, and propagation of parameter uncertainty into Fisher-information-based item selection. We show that most pairwise comparisons among 12 models on a standard LLM benchmark leaderboard are not statistically distinguishable despite differing point-estimate ranks. We further show that point-estimate Fisher information can become nearly zero for many benchmark items because it is evaluated at a single reference ability, whereas posterior-expected Fisher information remains substantially more stable across the ability range. Finally, posterior-expected Fisher information more accurately recovers full-benchmark ability rankings from small benchmark subsets in most experimental settings while matching point-estimate performance for the smallest subsets. We validate the calibration of the approximate posterior using held-out predictive coverage and find that modeling item difficulty as random while treating item discrimination as fixed produces well-calibrated uncertainty in this architecture.
[AI-140] RIDGE: An Autonomous Framework for Validation and Method Discovery in LLM -Generated Option Pricing
链接: https://arxiv.org/abs/2607.25199
作者: Liexin Cheng,Xue Cheng,Shuaiqiang Liu,Cornelis W. Oosterlee
类目: Computational Finance (q-fin.CP); Artificial Intelligence (cs.AI)
备注: 31 pages
Abstract:Automated code generation is becoming an important tool in quantitative finance, where large language models can generate option pricing implementations directly from mathematical model specifications. Validating such implementations, however, requires considerably more than conventional software testing: numerical pricing methods must remain mathematically consistent, numerically stable, and reliable across a wide range of model parameters. We introduce RIDGE, an autonomous validation framework in which generated pricing implementations are subjected to structured no-arbitrage tests, stress tests, benchmark comparisons, and consistency checks. Validation evidence is interpreted diagnostically, while the resulting knowledge is accumulated in a repository and reused across models and successive validation iterations. This enables systematic refinement of both the pricing implementation and the validation methodology. The framework is applied to five stochastic volatility models. Across these studies, all detected implementation defects are removed and, in two cases, the validation process itself leads to new semi-analytic pricing methodologies. The supplementary material is available in the GitHub repository: this https URL. Comments: 31 pages Subjects: Computational Finance (q-fin.CP); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.25199 [q-fin.CP] (or arXiv:2607.25199v1 [q-fin.CP] for this version) https://doi.org/10.48550/arXiv.2607.25199 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-141] Agent ic AI for Scientific Reasoning in Autonomous Quantum Sensing Experiments
链接: https://arxiv.org/abs/2607.25145
作者: Takuya Isogawa,Ryotaro Okabe,Nutdech Phadetsuwannukun,Mingda Li,Paola Cappellaro
类目: Quantum Physics (quant-ph); Artificial Intelligence (cs.AI)
备注: 11 pages, 4 figures + Supplementary Information
Abstract:We implement an agentic AI workflow built around a large language model (LLM) agent for autonomous experiments with nitrogen-vacancy (NV) centers in diamond. NV centers are a widely used platform for quantum sensing, and the ability to control many measurements from a computer makes NV experiments a natural setting for autonomous workflows. We make two main contributions. First, we demonstrate an autonomous NV experiment workflow that combines persistent project records, quantitative calculation and data analysis tools, and deterministic experiment control. In one autonomous experiment, the agent selected a single NV center, calibrated its resonant frequency, measured (T_2^\ast) with Ramsey measurements, and added a Carr–Purcell–Meiboom–Gill (CPMG) measurement to check a weak feature that could be related to nearby (^13\mathrmC). Second, we introduce two offline benchmarks that evaluate the agent’s reasoning separately from laboratory execution. We evaluated both benchmarks with GPT-5.4, GPT-5.5, and GPT-5.6 Sol. In the Ramsey checkpoint benchmark, greater reasoning effort generally improved recognition of a residual resonance calibration offset. By contrast, in the pulsed optically detected magnetic resonance (pODMR) data evaluation benchmark, pulse sequence information alone produced more false positive resonance judgments at higher reasoning effort. Requiring an expected signal calculation kept false positive rates low across all three models and reasoning settings. The results suggest a clear division of labor for autonomous experiments. The agent forms scientific hypotheses and uses quantitative tools to evaluate data, while deterministic code controls the hardware and enforces safety constraints.
[AI-142] Spectral Truncation in Synthetic Control
链接: https://arxiv.org/abs/2607.25074
作者: Mojtaba Eslami
类目: Methodology (stat.ME); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Econometrics (econ.EM); Applications (stat.AP)
备注:
Abstract:Synthetic control (SC) matches a treated unit’s pre-treatment trajectory to a weighted combination of donor units. We study Spectral SC, which instead matches the treated unit in coordinates defined by the leading temporal singular vectors of the donor panel, and a hybrid estimator that places separately tunable weight on retained and discarded directions, nesting raw-path SC and truncated Spectral SC as endpoints. We prove that the family reduces exactly to raw-path SC at full rank, that exact balance on K retained dimensions with N_0 donors is underdetermined whenever N_0K+1 , with an affine solution set of dimension N_0-K-1 , and that spectral imbalance maps to treatment-effect bias through a finite-sample best-linear-predictor decomposition. We evaluate the estimators across eleven data-generating regimes, using 400 replications per regime and donor-only placebo validation to select regularization and the mixing weight. Truncated Spectral SC has significantly higher RMSE than tuned raw-path SC in every regime, with paired differences equal to 4 to 11 Monte Carlo standard errors. The hybrid estimator selects raw-path matching in most replications and is statistically indistinguishable from tuned SC in most regimes. The result is highly sensitive to preprocessing. With raw inputs, the performance gap is large; after removing unit and time fixed effects before spectral decomposition, as suggested by the assumptions behind our bound, the gap nearly disappears and placebo validation begins to favor truncation. We interpret these findings diagnostically rather than as evidence that Spectral SC should replace raw-path SC. Basis-estimation noise, balancing underdetermination, and fixed-effects contamination determine when spectral matching can help.
[AI-143] Automatic Knowledge Graph Construction and Query for Earthquake Catalogs
链接: https://arxiv.org/abs/2607.24984
作者: Yuxin Zhou,Huai Zhang,S. Mostafa Mousavi
类目: Geophysics (physics.geo-ph); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:In recent years, the number of events in earthquake catalogs has significantly increased due to the utilization of more effective deep learning based detectors and phase pickers but answering open ended questions such as what characterizes this sequence? remains constrained by rigid spatiotemporal windowing and subjective expert interpretation. We present the first systematic application of graph based retrieval augmented generation GraphRAG directly to raw, tabular catalog records across three independently featured catalogs, a reservoir adjacent swarm, the 2019 Ridgecrest tectonic sequence, and the 2021 Maduo Mw7.4 aftershock sequence. Without the need for manual data structuring, the pipeline builds structurally complete, queryable knowledge graphs for all three. Rigorous evaluation individually verified against catalog derived ground truth and a rule based reference graph exposes failure modes, and four seismology informed prompt fixes eliminate all targeted fabrications while sharply improving mechanism reasoning. A vector RAG baseline demonstrates the graph layers distinctive value, catalog wide summarization and temporal stage comparison. In addition, we have identified two main pitfalls that need attention. GraphRAG thus offers a practical, transferable, near zero cost query interface for earthquake catalogs, where careful prompting ensures the results are consistently accurate and trustworthy.
[AI-144] GraphRareBench: An Auditable Graph-Evidence Benchmark for Phenotype-Driven Rare-Disease Diagnosis
链接: https://arxiv.org/abs/2607.24878
作者: Guiling Guo,Jia Yang,Jiahao Xu,Shuyuan Zheng,Zhonghai Sun,Qiyuan Li
类目: Quantitative Methods (q-bio.QM); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Phenotype-driven diagnostic benchmarks usually report the rank of the reference disease, but they rarely reveal which plausible alternatives are ranked above it or what evidence a tool-using model examines before making its decision. We introduce GraphRareBench, a provenance-preserving benchmark containing 2,365 ontology-derived cases and 18,093 target-confounder pairs. Each case includes a coarsened HPO query, a fixed candidate pool, graph-defined hard confounders, and source-linked evidence records. On the 237-case gene-component-disjoint test split, supervised rankers using a shared 21-feature interface achieved MRRs ranging from 0.640 to 0.740 and case-averaged target-over-confounder accuracies ranging from 0.898 to 0.916. Agents instantiated with Agents-A1 and DeepSeek-V4-Flash achieved MRRs of 0.746 and 0.718, respectively. Their paired MRR difference was not statistically significant, whereas their target-evidence coverage differed by 0.561. Together with the observation that 22.1% to 43.7% of selected Hit@10 successes still ranked at least one graph-defined hard confounder above the target, these results indicate that full-pool retrieval, hard-confounder discrimination, and observable evidence access capture complementary aspects of model behavior. GraphRareBench therefore provides a foundation for more transparent and evidence-aware evaluation of phenotype-driven diagnostic systems. Code and data are available at this https URL.
[AI-145] Beyond Predictive Accuracy: A Reliability-Aware Audit of Molecular Representations for Human Olfaction
链接: https://arxiv.org/abs/2607.24848
作者: Kai Lun Huang(California State University, Fullerton),Wei Chieh Sun(University of Washington)
类目: Quantitative Methods (q-bio.QM); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 23 pages, 5 figures; includes supplementary material
Abstract:Pretrained molecular encoders are commonly evaluated through downstream prediction, but predictive accuracy alone does not establish that a learned representation captures reproducible scientific structure, adds information beyond strong conventional baselines, or transfers out of distribution. We present a reliability-aware audit of generic molecular representations for human olfaction across four distinct claims: global perceptual geometry, incremental predictive value beyond chemistry, cross-dataset replication, and mixture transfer to unseen components. Using the Keller-Vosshall and Bierling single-molecule rating datasets and the Ma binary-mixture dataset, we compare MoLFormer and ChemBERTa against RDKit descriptors and Morgan fingerprints under identity-controlled and matched evaluations. Human three-attribute rating geometry, based on intensity, pleasantness, and familiarity, is reproducible across participant splits (median RSA 0.743 and 0.855), whereas model-human alignment is substantially weaker (RSA 0.019-0.158). Learned embeddings do not consistently outperform conventional representations in global alignment, and MoLFormer provides no clear incremental predictive value beyond a combined RDKit-Morgan baseline in either single-molecule dataset. Human geometry shows positive but incomplete agreement across 63 shared molecules (RSA 0.331; 95% bootstrap interval [0.204, 0.507]). Under one strict unseen-component mixture split, incremental effects are outcome- and representation-dependent, with all intervals crossing zero. These results establish empirical boundaries for the evaluated generic molecular encoders and motivate a broader evaluation principle: representation quality in scientific domains should be assessed separately for target reliability, structural alignment, incremental information, replication, and out-of-distribution transfer. Comments: 23 pages, 5 figures; includes supplementary material Subjects: Quantitative Methods (q-bio.QM); Artificial Intelligence (cs.AI); Machine Learning (cs.LG) Cite as: arXiv:2607.24848 [q-bio.QM] (or arXiv:2607.24848v1 [q-bio.QM] for this version) https://doi.org/10.48550/arXiv.2607.24848 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-146] Extremal Chowla sets and their linear analogues: A human-AI mathematical investigation using Co-Scientist
链接: https://arxiv.org/abs/2607.24847
作者: Mohsen Aliabadi,Keith Driscoll,Elliot Krop,Petar Sirkovic,Everett Sullivan,Elahe Vedadi
类目: Number Theory (math.NT); Artificial Intelligence (cs.AI); Group Theory (math.GR)
备注: Comments welcome!
Abstract:We introduce an extremal invariant associated with Chowla-type order conditions in finite groups. A nonempty subset S of a finite group G is called a Chowla set if every element of S has order greater than |S| , and we write \Ccal(G) for the maximum cardinality of such a set. We first show that \Ccal(G) is determined by the distribution of element orders in G . For cyclic groups, we derive an exact divisor formula and characterize the integers n for which \Ccal(\mathbb Z/n\mathbb Z)=\varphi(n) . We prove that \liminf_n\to\infty\Ccal(\mathbb Z/n\mathbb Z)/\varphi(n)=1 , whereas \limsup_n\to\infty\Ccal(\mathbb Z/n\mathbb Z)/\varphi(n)=\infty , and we determine the corresponding lower and upper limits under normalization by n . For finite abelian groups, we obtain an explicit formula in terms of the invariant-factor decomposition, together with a closed formula for finite abelian p -groups. We then develop a linear analogue for finite field extensions. A nonzero K -subspace A of an extension L/K is called a Chowla subspace if [K(a):K]\dim_K A for every nonzero a\in A . Since this condition depends on \dim_K A , it does not generally require every nonzero element of A to generate L over K . Nevertheless, when L/K is finite and separable, we prove the exact formula \Ccal(L/K)=[L:K]-d_\max(L/K) , where d_\max(L/K) is the largest degree over K of a proper intermediate field. For finite fields, we give a direct proof in every degree using a normal-basis construction. This work was developed through an expert-guided human–AI collaboration. A reasoning-focused configuration of Co-Scientist was used to explore examples and potential proof strategies. The authors formulated the problem, independently verified and completed all arguments, and wrote the final proofs. Comments: Comments welcome! Subjects: Number Theory (math.NT); Artificial Intelligence (cs.AI); Group Theory (math.GR) Cite as: arXiv:2607.24847 [math.NT] (or arXiv:2607.24847v1 [math.NT] for this version) https://doi.org/10.48550/arXiv.2607.24847 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-147] Foundation Models for EEG Are Blind to Long-Range Temporal Correlations: A Spectral-Temporal Dissociation Behind Their Cross-Population Frag ility
链接: https://arxiv.org/abs/2607.24834
作者: Marzieh Zare
类目: Neurons and Cognition (q-bio.NC); Artificial Intelligence (cs.AI); Emerging Technologies (cs.ET); Machine Learning (cs.LG)
备注:
Abstract:Objective. Electroencephalography (EEG) foundation models (FMs) are trained to reconstruct or contrastively align short patches, then pooled into a fixed embedding. We tested whether these embeddings retained the long-range temporal correlations (LRTC) quantified by the detrended-fluctuation-analysis (DFA) exponent of the alpha-band envelope, and whether it governs cross-population transfer. Approach. We probed five EEG FMs spanning raw-waveform and spectral-input architectures (REVE, LaBraM, BENDR, CBraMod, BIOT) on two out-of-distribution cohorts, comparing recovery of the DFA exponent against the static 1/f aperiodic slope. Order-preserving and residualization controls tested for pooling or aperiodic shadowing. A montage-harmonized, zero-shot transfer task compared the frozen embedding with the DFA exponent across three cohorts (adding a Western reference). Main results. None of the five FMs represented the LRTC in the temporal order. Raw-waveform models (REVE, LaBraM, BENDR) recovered neither the DFA exponent nor the 1/f slope (R^2 = 0.12); for these three the probe is uninformative, so the dissociation is specific to the spectral-input models (CBraMod, BIOT), which recovered 1/f strongly (R^2 = 0.59-0.73) but not DFA across cohorts. A classical DFA feature recovered the exponent (R^2 = 0.32-0.38 against a 0.64 reliability ceiling), and LRTC was orthogonal to the aperiodic slope (r = -0.06). On cross-population transfer, the frozen REVE embedding did not beat chance (W to K, 0.45) and the dimensionless DFA exponent transferred directionally but not at family-wise significance; the other four did not uniformly replicate it. All five were dominated by a recording-site axis (decodable at 0.98-1.00 vs. 0.500 chance), whereas the DFA exponent they discard is site-robust (0.71). Subjects: Neurons and Cognition (q-bio.NC); Artificial Intelligence (cs.AI); Emerging Technologies (cs.ET); Machine Learning (cs.LG) Cite as: arXiv:2607.24834 [q-bio.NC] (or arXiv:2607.24834v1 [q-bio.NC] for this version) https://doi.org/10.48550/arXiv.2607.24834 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-148] Dynamic Multi-Criteria Bottleneck Severity Index (DMBSI) for Semiconductor Wafer Manufacturing: A Genetically Optimised Framework for Reentrant Production Systems
链接: https://arxiv.org/abs/2607.24819
作者: Mohammad Sharifur Rahman,Karl McCreadie,Saugat Bhattacharyya,M M Manjurul Islam,Cormac McAteer,Bryan John Baker,Nuala Parker,Girijesh Prasad
类目: ignal Processing (eess.SP); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 12 pages prepared for journal submission
Abstract:Wafer fabrication exhibits unique characteristics, including reentrant process flows, variable bottlenecks, and highly variable process conditions. In order to identify the most severe bottleneck at each moment in time for semiconductor wafer fabrication, this research presents the Dynamic Multi-Criteria Bottleneck Severity Index (DMBSI), a new, data-driven methodology for analysing multiple diagnostic signals of cycle time to changes in process parameters, and the impact of reworks on cycle time in order to generate an interpretable, unified measure of bottleneck severity. The experimental validation of DMBSI was conducted using Manufacturing Execution System (MES) logs collected from 22 wafer production lots at a commercial 200 mm wafer fabrication line operated by Seagate Technology. Using 5-fold cross-validation, the GA-optimised DMBSI achieves a Pearson correlation of r = 0.80 with observed cycle-time contributions, representing an 8.1% improvement over the expert heuristic baseline (r = 0.74) and substantially outperforming the Theory of Constraints (TOC; r = 0.60) and Value Stream Mapping (VSM; r = -0.30). Furthermore, the unique time-windowed component of DMBSI enabled the identification of temporal bottleneck migration patterns, which shifted from the dominant constraints associated with dielectric deposition steps in the early production windows to those associated with over- and under-inspection in the later production windows. The integrated what-if counterfactual analysis demonstrated that a 50% reduction in waiting time at the top-ranked bottleneck step would reduce the mean cycle time by 7.2%, with the top five bottleneck steps offering a combined potential reduction of approximately 19%.
[AI-149] A Path Integral Model of Cognition
链接: https://arxiv.org/abs/2607.24807
作者: Haruki Emori,Kazunori Kondo,Atsushi Iriki,Andrei Khrennikov
类目: Neurons and Cognition (q-bio.NC); Artificial Intelligence (cs.AI); Quantum Physics (quant-ph)
备注: 17 pages, 2 figures,1 table
Abstract:We develop the mathematical and physical formulation of cognitive cost optimization that underlies the path-integral model of consciousness. The goal-directed cognitive process is modeled as imaginary-time evolution (ITE) under a projector Hamiltonian that rewards configurations consistent with a target concept. We establish three results. First, this ITE coincides with a double-bracket flow and is therefore the Riemannian gradient flow of a Hilbert–Schmidt cost whose unique minimum is the solution. Second, a Wick rotation re-expresses this non-unitary descent as an equivalent unitary evolution on the same Hilbert space, which admits an exact discrete path-integral representation in which the oracle and the initial-state diffusion projector play the roles of potential and kinetic energy. Third, we identify the continuum from unconscious to conscious processing with the strength of the unitary interaction between the cognitive system and a neural-environment probe, recovering the Gorini–Kossakowski–Sudarshan–Lindblad (GKSL) decoherence model of Asano \textitet al. in the Markovian weak-coupling limit, and the projective, reportable fixation of an optimized state in the strong-coupling limit, an insight-like ``Aha’’ endpoint. Both regimes share the same ITE and path-integral structure, and only the measurement-interaction strength varies. The Wick rotation is therefore a technique of re-description, not a physical regime change.
[AI-150] Decoding Error-Related Potentials under Multisensory Feedback with Varying Congruency
链接: https://arxiv.org/abs/2607.24806
作者: Yixin Liu,Kang Yin,Hye-Bin Shin,Seong-Whan Lee
类目: Neurons and Cognition (q-bio.NC); Artificial Intelligence (cs.AI)
备注:
Abstract:Error-related potentials (ErrPs) are widely studied neural signatures associated with error processing in human-machine interaction. In realistic settings, error perception often occurs under heterogeneous multisensory feedback, where variability induced by sensory modality and feedback congruency poses challenges for reliable ErrP decoding. In particular, incongruent feedback is associated with increased decoding difficulty and reduced classification performance. To address this challenge, we investigate learning strategies for robust ErrP decoding under multimodal visual, auditory, and tactile feedback with controlled sensory congruency. We adopt a multi-branch EEGNet-based architecture with auxiliary supervision to improve robustness across heterogeneous conditions, without relying on explicit modality-specific assumptions. Experiments were conducted using a maze-observation task with unimodal, bimodal, and trimodal feedback configurations. Across subjects, the proposed approach achieved consistent classification performance across heterogeneous sensory conditions and showed improved accuracy compared to baseline EEGNet models, particularly under multimodal feedback. These results suggest that appropriate architectural design and training strategies can improve the stability of ErrP decoding under heterogeneous multisensory conditions.
[AI-151] Selective Impairment of Motor Recovery from Typing Errors in Parkinsons Disease: A Survival Analysis
链接: https://arxiv.org/abs/2607.24796
作者: Navin Bondade
类目: Neurons and Cognition (q-bio.NC); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 7 pages, 1 figure
Abstract:Parkinson’s disease (PD) affects multiple, dissociable stages of motor and cognitive control. We ask whether passively-collected keystroke dynamics can distinguish two of these stages: noticing a self-generated error (error monitoring) versus recovering normal motor rhythm afterward (motor restart). Using backspace events as naturally-occurring error-correction episodes in the public neuroQWERTY MIT-CSXPD dataset (57 subjects with sufficient backspace data, 27 with PD and UPDRS-III scores), we find no evidence that pre-error keystroke instability differs by disease severity ( r=0.164 , p=0.413 ), but strong evidence that post-error recovery time does, modeled as a continuous accelerated failure time (AFT) survival outcome ( p10^-9 in each of two independent sub-cohorts; permutation p0.0033 ; bootstrap 95% CI excluding zero). The two measures are uncorrelated with each other ( r=-0.065 ), and in a joint model only post-error recovery remains significant, confirming a genuine dissociation rather than two redundant signals. The effect survives controlling for raw typing speed and replicates against an independent clinical motor test (alternating finger-tapping, p=0.011 ). An initial attempt to model recovery as a discretized time-to-event outcome (analogous to Kaplan-Meier survival curves) destroyed the signal regardless of threshold choice; switching to a continuous AFT formulation, which also fits the data’s right-skewed distribution far better than ordinary regression, recovered it. We relate this behavioral dissociation to existing electrophysiological evidence that error detection and post-error motor adjustment in PD are mediated by distinct neural circuits, the former largely spared, the latter linked to subthalamic nucleus activity, and argue that this dissociation is detectable through everyday typing alone.
[AI-152] DDSNet: Dual-domain Symmetry-aware Network for PCSEL Property Prediction
链接: https://arxiv.org/abs/2607.24785
作者: Cen Chen,Haitao Huang,Jiazhi Mao,Feifan Xu,Zhe Zhuang,Yuxiang Ren
类目: Applied Physics (physics.app-ph); Artificial Intelligence (cs.AI)
备注: 14 pages, 7 figures. Main paper and supplementary material included
Abstract:Efficient exploration of the photonic crystal (PhC) lattice design space is essential for developing photonic crystal surface-emitting lasers. While coupled-wave theory (CWT) provides an effective physical framework, its computational cost remains prohibitive for large-scale exploration, driving the demand for neural surrogates. However, existing AI models underexploit two key factors of PhC unit-cell dielectric patterns indicated by CWT: spectral components and asymmetric structures, which largely govern devices’ physical properties. This mismatch weakens surrogate accuracy and screening reliability, especially in structure-sensitive regions. To address this, we propose the Dual-Domain Symmetry-Aware Network (DDSNet). It integrates translation-equivariant spectral filtering with a symmetry-induced structural prior. The spectral filtering injects a spectral inductive bias into vision model while preserving translation equivariance on lattices. Meanwhile, the structural prior decomposes lattice features into irreducible representation-associated, symmetry-resolved components and processes them in separate branches. Experiments demonstrate that DDSNet significantly outperforms existing AI baselines in property prediction and high-throughput screening, exhibiting superior reliability in structure-sensitive regions. Crucially, component masking analyses reveal that the network successfully learns property-specific dependencies aligned with physical priors. These results indicate that DDSNet effectively captures physically meaningful structure-property relationships, establishing a highly reliable neural surrogate for PhC design space exploration.
[AI-153] Neural Network Learning of One-Bit Protocols for Qubit Measurement Simulation
链接: https://arxiv.org/abs/2607.23645
作者: Josep Escrig,Mani Zartab,Giulio Gasbarri,Estel Ferrer,Ramon Muñoz-Tapia,Gael Sentís
类目: Quantum Physics (quant-ph); Artificial Intelligence (cs.AI); Numerical Analysis (math.NA)
备注: 9 pages with the main text and 9 pages with Appendix
Abstract:Communication complexity provides a natural framework for quantifying the classical resources required to reproduce quantum statistics. In the qubit prepare-and-measure scenario, two classical bits have been shown to be necessary and sufficient to simulate arbitrary qubit states and arbi- trary quantum measurements exactly. However, this result does not exclude the possibility that restricted families of measurements may admit accurate 1-bit classical approximations. We use a neural network procedure to demonstrate that a single bit can achieve high average accuracy for specific measurement families. A performance analysis of our neural network reveals that symmet- ric measurements with uniformly weighted elements, such as those forming regular polyhedra, are particularly amenable to this restricted communication. By analyzing the patterns learned by the neural network, we derive an analytical protocol that is extremely accurate for finite information- ally complete symmetric configurations and becomes exact in the limit of a continuous isotropic measurement.
机器学习
[LG-0] Spend Experts Where You Are Unsure: Confidence-Adaptive Routing for Mixture-of-Experts LoRA
链接: https://arxiv.org/abs/2607.26052
作者: Tom Saliencro,Rohan Desai,Priya Nair,Maya Lindqvist,Daniel Whitmore
类目: Machine Learning (cs.LG)
*备注:
Abstract:Mixture-of-Experts (MoE) variants of Low-Rank Adaptation (LoRA) route every token to a fixed number of experts k . Tokens differ in how uncertain the model is about them, so a single k over-spends on easy tokens and under-serves hard ones. We observe that the router’s output distribution is already a per-token uncertainty signal: peaked mass indicates confidence, while a flat distribution indicates ambiguity. We introduce CARE (Confidence-Adaptive Routing of Experts), which admits experts in a nucleus fashion. Experts are activated in decreasing router weight until their cumulative mass reaches a threshold, with a small extension when the admitted experts disagree. A budget thermostat calibrates the threshold so that the average number of active experts matches any target. CARE is a drop-in, single-forward-pass rule with no extra parameters. Across eight commonsense benchmarks on LLaMA-3.1-8B and Qwen2.5-7B, as well as math, code, and knowledge tasks, CARE improves over fixed top-k MoE-LoRA at matched compute and matches the fixed-k=4 baseline while activating fewer experts. The same confidence and disagreement signals also improve out-of-distribution detection over MSP, entropy, and multi-pass proxies. We support the design with nucleus fidelity, budget optimality, and an epistemic reading of disagreement, and we release code.
[LG-1] Re-thinking Mammography Transfer Learning: The Dataset-Informed Transfer Learning (DITL) Framework for Breast Cancer Screening and Lesion Diagnosis
链接: https://arxiv.org/abs/2607.26043
作者: Adarsh Bhandary Panambur,Siming Bayer,Andreas Maier
类目: Machine Learning (cs.LG)
*备注: 16 pages, 1 figure, 5 tables. Accepted and presented at the 10th International Conference on Computer Vision Image Processing (CVIP 2025), IIT Ropar, India, 10-13 December 2025. The paper is currently in press for inclusion in the official conference proceedings. This preprint corresponds to the submitted manuscript and is made available pending publication of the final proceedings version
Abstract:Enhancing classification performance in mammography remains a persistent challenge across both small curated datasets and large-scale clinical cohorts. Conventional transfer learning approaches often neglect dataset-specific characteristics, while recent neighborhood-informed methods have been restricted to narrow tasks with rigid formulations, limiting their scalability to population-level datasets. To address these challenges, we propose the Dataset-Informed Transfer Learning (DITL) framework, which integrates dataset-derived difficulty signals with neighborhood-based triplet supervision in a unified objective. DITL introduces two adaptive components: (i) Adaptive Difficulty-Weighted Cross-Entropy (A-DWCE), which assigns per-sample weights based on k-nearest neighbor label purity in a self-supervised feature space, and (ii) Adaptive Neighborhood Representation Triplet (A-NR-Triplet), which enforces intra-class compactness and inter-class separation using a learnable margin. Unlike focal loss, DITL requires no hyperparameter tuning, removes heuristic weighting and fixed margins, and incurs negligible computational overhead, yielding a robust and scalable optimization strategy. On the large-scale VinDR-Mammo dataset, DITL achieves state-of-the-art performance for whole-image breast density classification, with significant improvements across accuracy, F1-score, and AUC (p 0.0001). Beyond large cohorts, DITL also delivers consistent, statistically significant gains on small ROI datasets (p 0.0001). By bridging small-scale lesion analysis with large-scale density estimation, DITL establishes a clinically relevant, scalable, and generalizable framework for mammography classification, spanning the full breast cancer screening-to-diagnosis spectrum.
[LG-2] Reinformed Dreamer: An Asymmetric World Model Efficiently Trained through Latent Guidance
链接: https://arxiv.org/abs/2607.26040
作者: Gaspard Lambrechts,Adrien Bolland,Daniel Ebi,Damien Ernst
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: 8 pages, 18 pages total, 3 figures
Abstract:Much like humans benefit from guidance while learning, reinforcement learning algorithms may benefit from additional supervision beyond rewards. Leveraging additional information during training to learn better representations and behaviors has been the focus of asymmetric reinforcement learning. This learning paradigm has proven effective under partial observability when additional state information is available, but also under full observability when more refined state information is available. Focusing on model-based reinforcement learning, we study the effect of asymmetric learning on observation representations and on privileged information representations. First, we identify a limitation in the privileged information representations learned by an asymmetric model-based algorithm known as the Informed Dreamer. Then, we propose a novel asymmetric representation learning objective using latent guidance, resulting in a new algorithm called the Reinformed Dreamer. Experiments across several benchmarks show a more consistent improvement over Dreamer than previous asymmetric approaches.
[LG-3] Sharpness-Aware Minimization and Muon: Robustness under the Spectral Norm
链接: https://arxiv.org/abs/2607.26001
作者: Wenzhi Zhong,Edward Milsom,Michael Murray
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注:
Abstract:Sharpness-Aware Minimization (SAM) aims to improve generalization by encouraging insensitivity to small, worst-case parameter perturbations. However, the notion of a “small” perturbation is inherently geometry-dependent: while existing SAM variants have explored a wide range of choices, a clear perspective on which geometries are most effective in practice remains elusive. Recent work on matrix-aware optimization, particularly the Muon optimizer, suggests that respecting the matrix structure of hidden-layer weights can lead to strong empirical performance. Motivated by this, we study matrix-aware geometry in both stages of SAM: we introduce a layerwise spectral inner perturbation for matrix-valued hidden-layer parameters and combine it with either AdamW/SGDW or Muon in the outer update. Across ImageNet-1K experiments on ViT-Small/16 and ResNet-50, we find that the combination of a spectral inner step with a Muon outer step performs consistently strongly, achieving the best validation accuracy on both models among the evaluated methods.
[LG-4] Untangling Co-Drift: Proactive Multi-Intent Failure Prediction and Root-Cause Disambiguation for Self-Driving Networks
链接: https://arxiv.org/abs/2607.25989
作者: Md. Kamrul Hossain,Walid Aljoby
类目: Networking and Internet Architecture (cs.NI); Machine Learning (cs.LG); Robotics (cs.RO)
*备注: Under review in IEEE Transactions on Network and Service Management
Abstract:The vision of self-driving networks that monitor, reason, and act upon themselves with minimal human intervention relies on tightly coupled monitoring, analytics, and actuation functions. In this work, we treat these functions as three operational macro-intents: continuous telemetry, real-time analytics, and programmatic actuation, and formalize the health of each function as an intent that the network must continuously satisfy. A critical, yet underexplored, challenge stems from the causal coupling among these intents, where a singular fault within one macro-intent propagates as a co-drift and subsequently triggers cascading, symptomatic anomalies across the remaining intents. This ambiguity makes it exceedingly difficult for existing, reactive approaches to distinguish the true root-cause intent from symptomatic victim intents, and their reliance on threshold-crossing detection leaves insufficient time for proactive remediation. We introduce MILD, a novel framework that reformulates intent assurance from reactive drift detection to proactive failure prediction. Grounded in our three-macro-intent formulation of the self-driving control loop, MILD employs a teacher-augmented Mixture-of-Experts architecture with a hybrid objective that jointly optimizes intent failure prediction and root-cause attribution. MILD enables KPI-level diagnostics via SHAP explainability and dynamic intent failure urgency estimation via multi-horizon modeling. Our extensive evaluation of MILD across three environments of increasing realism, from a controlled statistical benchmark, to a microservices application, to an SDN-based edge-to-cloud testbed, demonstrates that MILD achieves high failure detection rates, strong remediation lead times, and accurate intent-level root-cause disambiguation. This positions MILD as a practical enabler of closed-loop assurance in next-generation autonomous networks.
[LG-5] Generator-Aligned Representation Interfaces for Diagnostic Soft Equivariance
链接: https://arxiv.org/abs/2607.25988
作者: Weitao Li,Gong Cheng
类目: Machine Learning (cs.LG)
*备注:
Abstract:Exact-equivariant architectures typically encode prescribed group actions in specialized operators, which can complicate their reuse with generic backbones and across data modalities. We introduce the Generator-Aligned Representation Interface (GARI), a representation-level design principle that exposes selected transformation generators to a generic sequence backbone through aligned canonical and generator-induced views. We formalize the resulting behavior using a probe-specific soft-equivariance residual defined over declared data and transformation distributions. This framework distinguishes representation consistency from task robustness and exact equivariance, and localizes residual mismatch to interface construction, shared stream processing, and terminal fusion. We instantiate the interface as GARI-Net, which constructs generator-indexed streams, converts them into a common interaction frame, processes them with shared parameters, repairs ordering-induced context mismatch, enables cross-stream information exchange, and aggregates them using inter-stream discrepancy. Direct Equivariance Error (DEE) provides a frozen-checkpoint diagnostic of the prescribed representation relation under known token or voxel actions. Experiments on genomic sequences, images, and three-dimensional point clouds examine sequence reversal, planar rotations and reflections, and controlled axial transfer. Across these settings, the same interface principle supports task-relevant transformation consistency and generalization to declared held-out probes without requiring group-specific redesign of the sequence backbone. GARI therefore provides a portable diagnostic complement to hard-equivariant architectures: it makes generator structure accessible, learnable, and measurable, while finite-probe evidence remains distinct from certification of exact equivariance over a continuous group.
[LG-6] Physics-Aware End-to-End Deep Reinforcement Learning for Quadcopter Control with Actuator Dynamics
链接: https://arxiv.org/abs/2607.25985
作者: Ya-Chia Shen,Woei-Leong Chan
类目: Robotics (cs.RO); Machine Learning (cs.LG); Systems and Control (eess.SY)
*备注: 8 pages, 5 figures, 2 tables. Presented at the Aeronautical and Astronautical Society of the Republic of China (AASRC) Conference, Tamsui, Taiwan, November 15, 2025, Paper No. 1075. Received an Honorable Mention for the Best Paper Award
Abstract:Unmanned aerial vehicles (UAVs), particularly quadcopters, present unique challenges for autonomous control due to their underactuated dynamics: only four available control inputs must govern six degrees of freedom. This paper investigates a physics-aware, end-to-end deep reinforcement learning (DRL) approach that acts directly on low-level body inputs, total thrust and body torques (T, \tau_x, \tau_y, \tau_z) , and closes the loop through a high-fidelity Simulink environment. Our simulator integrates a 12-state rigid-body model (MATLAB Level-2 S-Function) with (i) an Action2RPM allocation based on the Moore-Penrose pseudo-inverse of a coefficient matrix derived from thrust and drag terms, and (ii) first-order actuator dynamics for each motor (time constant T_m = 0.076 s), including rotor gyroscopic coupling. A shaped reward balances goal-reaching and stability using an exponential position well, attitude penalties, and quadratic velocity costs. Four DRL algorithms, DDPG, TD3, PPO, and SAC, are evaluated in two stages: (S1) thrust-only hover and (S2) hover with pitch torque and a translated goal. Results show that SAC and TD3 achieve superior stability and exploration efficiency, while PPO is less sample-efficient. The study highlights the significance of modeling actuator lags and aerodynamic moments for stable low-level control and provides a reproducible benchmark for quadcopter DRL.
[LG-7] DRIFT: Direct-Recursive Intervention-Conditioned Forecasting of ICU Physiological Trajectories
链接: https://arxiv.org/abs/2607.25864
作者: Weixin Liu,Juming Xiong,Congning Ni,Yanfan Zhu,Xingtao Lin,Bradley A. Malin,Zhijun Yin
类目: Machine Learning (cs.LG); Signal Processing (eess.SP)
*备注: 34 pages, 1 figure; extended technical appendices included
Abstract:Many time-series forecasts depend not only on prior observations but also on actions specified during the forecast period. In intensive care units (ICUs), future vital signs and laboratory values are influenced by treatments such as vasopressors. However, models that predict the full future sequence all at once make little use of these treatments, whereas autoregressive models can accumulate errors. We introduce DRIFT, a hybrid framework in which a direct model produces the primary forecast and a recursive, action-conditioned model contributes constrained corrections. We evaluate DRIFT on 6,046 admissions from MIMIC-IV and 8,345 admissions from eICU-CRD. Averaged across the 8-, 24-, and 48-hour forecast endpoints, DRIFT reduces mean absolute error for mean arterial pressure (MAP) by 0.673% relative to an action-conditioned Temporal Fusion Transformer (TFT-action) on MIMIC-IV and achieves the lowest corresponding error among the compared models on eICU-CRD. Although the overall accuracy improvement is modest, a MIMIC-IV audit restricted to windows in which the supplied treatment sequence was altered showed that DRIFT achieved lower observed-target MAP error than TFT-action at 8 and 24 hours. Treatment-sequence alteration increased DRIFT’s MAP error by 0.21-0.26 mmHg more than it increased TFT-action’s error, with prediction changes occurring primarily after the supplied paths diverged. In a separate robustness experiment, the MAP advantage persisted under three shared checkpoint-selection rules emphasizing overall endpoint error, MAP error, or both equally.
[LG-8] Prototype Adaptation for Zero-Shot sEMG Movement Classification
链接: https://arxiv.org/abs/2607.25826
作者: Rui Liu,Benjamin Paassen
类目: Machine Learning (cs.LG)
*备注:
Abstract:Surface electromyography (sEMG) enables the control of prostheses, allowing upper-limb amputees to re-gain some hand function. Most current research focuses on recognizing basic movements for prosthesis control. However, in most daily activities, such as opening a door, combined movements are essential. However, collecting training data for all possible combined movements is time-consuming and requires re-training of the model for any new combination. We propose two novel recognition approaches, Compositional Prototype Interpolation (CPI) and Synthetic Adaptation for Prototypes (SAP), that enable zero-shot learning of combined, novel and unseen movements in Prototype Networks after training only with basic movements. Our methods rest on a linear interpolation assumption in the embedding space, which we study by inspecting the geometry of combined motions in signal and embedding space. In experiments on the NearLab and NinaPro DB3 data sets as well as our newly recorded BasCom dataset, our proposed SAP outperforms prior zero-shot learning methods with accuracy improvements on combined movements of more than 20%. This advantage is maintained in online inference experiments in a user study.
[LG-9] WALoMA: A Multitask Wireless Foundation Model via Adaptive Low-Rank Masked Autoencoders
链接: https://arxiv.org/abs/2607.25763
作者: Madi Makin,Asmaa Abdallah,Abdulkadir Celik,Ahmed M. Eltawil
类目: Networking and Internet Architecture (cs.NI); Machine Learning (cs.LG)
*备注:
Abstract:This paper proposes a multitask wireless foundation model via adaptive low-rank masked autoencoders (WALoMA), a unified multi-task foundation model for sixth-generation (6G) wireless physical layer architectures, to address the limitations of specialized, task-specific deep learning models and the practical challenge of scarce labeled wireless datasets. By leveraging concepts inspired by foundation models, the proposed framework adopts a masked autoencoder (MAE) paradigm to learn from unlabeled channel data, to significantly reduce reliance on extensive annotations. The model treats wireless channel state information (CSI) as a universal modality and learns transferable representations through self-supervised channel reconstruction. Key architectural novelties include the use of 2D positional encoding (PE) to explicitly preserve the spatial-frequency relationships between antennas and subcarriers, and low-rank adaptation (LoRA) for parameter-efficient fine-tuning. The framework’s efficacy is demonstrated across five downstream tasks, achieving individual scores of 96.47% for LoS/NLoS classification, 80.45% for beam prediction, 85.78% for channel interpolation, 99.12% for channel estimation, and 77.18% for channel charting. Consequently, numerical results show that the proposed model achieves a composite score of 87.80%, significantly outperforming the 59.90% achieved by the large wireless model (LWM) baseline while training an average of only 14.68% of total parameters, and maintaining strong performance even under extremely limited labeled data conditions.
[LG-10] An Embarrassingly Simple Rule-based Visiting Circulation Approach to Trip Destination Prediction
链接: https://arxiv.org/abs/2607.25751
作者: Eng-Shen Tu,Yong-Han Chen,En-Chao Liu,Hao-Yun Keng,Cheng-Te Li
类目: Machine Learning (cs.LG)
*备注: 8 pages
Abstract:In this paper, we propose the Rule-based Visiting Circulation (RVC) model in tackling the challenge in the IEEE Big Data Cup 2022: Trip Destination Prediction. Given trips containing travel information, personal attributes, origin zones, and their features in the training metropolitan areas, the task is to predict the destination of every trip in a targeted metropolitan area whose destinations are not given at all at the training stage. We highlight the challenges in this destination prediction task – having no knowledge of the destinations in the targeted metropolitan area. We provide insights from the datasets, in which revisiting behaviors and the relationships between origins and destinations play a crucial role in individuals’ trips. Hence, we design a simple but comprehensive method, rule-based visiting circulation, which directly utilizes the origin information and individuals’ trip behaviors to determine the destinations in the targeted metropolitan area, i.e., requiring no learning from the four training areas. Experimental results on both offline evaluation and leaderboard submission consistently exhibit the proposed RVC can significantly outperform supervised learning methods and other heuristics. The RVC method eventually brings us to second place in the competition leaderboard.
[LG-11] Detecting CSAM Text-to-Image LoRAs From Weights
链接: https://arxiv.org/abs/2607.25750
作者: David Demitri Africa,Cate Heine,Nadine Staes-Polet,Kimberly Mai
类目: Machine Learning (cs.LG); Computers and Society (cs.CY)
*备注:
Abstract:Low-rank adaptation (LoRA) fine-tuning has made it cheap and easy to customize open-weight image generation models for specific tasks, including the production of child sexual abuse material (CSAM). Existing moderation relies on metadata or generated outputs, but metadata can be deceptive and generating outputs may itself be unacceptable or illegal. We show that a safer signal lives in the weights. The top-left singular vectors of a LoRA’s updates form a compact, inference-free fingerprint ( u_1 ) of its strongest learned change. Using human-subject age as a benign proxy for CSAM, we find that u_1 identifies what a LoRA was trained on, generalizes across base models, and abstains on unrelated benign content. The signal is robust to additive weight noise, rescaling, and precision reduction. These results indicate that harmful LoRAs could be screened directly from their weights without relying on metadata or generating harmful outputs.
[LG-12] Optimization with Dynamic Constraint Learning (DCL)
链接: https://arxiv.org/abs/2607.25719
作者: Ezgi Oztekin,Figen Oztoprak,S. Ilker Birbil
类目: Machine Learning (cs.LG); Optimization and Control (math.OC)
*备注:
Abstract:We propose Dynamic Constraint Learning (DCL), a data-driven framework for constrained optimization when constraint functions are unknown and cannot be queried during optimization. At each iteration, the method learns a local surrogate from nearby data and solves a subproblem within a data-supported trust region. Compared with offline global constraint learning, the approach uses local surrogates that adapt to the data distribution during optimization and can achieve solution quality comparable to that of global models while using simpler local models and smaller optimization subproblems. We demonstrate the performance of DCL on a synthetic test problem and two case studies from the literature.
[LG-13] A Physics-Informed Neural Operator for Thermal Ranking of Low-Cost Wall Materials in Hot-Dry Climates
链接: https://arxiv.org/abs/2607.25668
作者: Muhammad Akbar Khan,Fahim Raees,Ubaida Fatima
类目: Machine Learning (cs.LG); Numerical Analysis (math.NA); Computational Physics (physics.comp-ph)
*备注: 41 pages, 13 figures. Code and data: this https URL
Abstract:Identifying cost-effective indigenous building materials that minimise heat penetration through walls is critical for indoor thermal comfort in low-income rural housing in hot-dry climates, where summer temperatures routinely exceed 45 C. We present a two-stage computational framework for thermal ranking of five low-cost indigenous wall materials: mud brick, clay-straw adobe, lime-stabilised bamboo panel, fired clay brick, and lime-mud composite. First, a validated Crank-Nicolson finite difference method (FDM) solves the one-dimensional transient heat equation with Robin boundary conditions under diurnal solar and outdoor air-temperature forcing, generating 1500 periodic-day solutions across a nine-dimensional parameter space by Latin Hypercube sampling. Second, a Physics-Informed Neural Operator (PINO) with a Fourier Neural Operator (FNO) backbone learns the parameter-to-solution operator mu - T(x,t), enforcing both data fidelity and PDE consistency. The trained PINO attains a relative L2 field error of 5.14e-4 and a 0.201 K mean absolute error on the peak inner surface temperature, preserving the FDM material ranking exactly; PINO trained on 150 FDM samples matches a data-only FNO trained on twice as many, so the physics loss is most valuable when data are scarce. The periodic-day formulation also yields the ISO 13786 time lag and decrement factor, reproduced to within 0.99 h and 0.010. At nominal hot-dry summer conditions, clay-straw adobe achieves the best cost-performance index among widely available materials. A climate sweep, confirmed by FDM spot checks, reveals a regime boundary: under sub-ambient outdoor conditions the ranking inverts to conductive fired clay brick, delineating heat-exclusion and heat-rejection regimes. The framework supports evidence-based material selection for post-flood reconstruction in hot-dry regions.
[LG-14] Contextual Deconvolution for Variance-Stable Demand Sensing: Kernel-Modulated Operators in Promotional Retail
链接: https://arxiv.org/abs/2607.25664
作者: Mohammad Forouhesh
类目: Machine Learning (cs.LG); Optimization and Control (math.OC); Machine Learning (stat.ML)
*备注: 46 pages, 12 figures, 22 tables
Abstract:Machine learning demand forecasts optimize statistical accuracy yet leave excess operational volatility that inflates safety stock and amplifies the Bullwhip effect. We introduce \textbfContextual Deconvolution (CD), a two-stage estimator that reframes demand sensing as a convex decomposition: a kernel-modulated banded operator separates transient promotion-driven shocks from a smooth structural baseline, and hierarchical partial pooling enables catalog-scale deployment without per-SKU training. The operator is data-derived, not imposed—it reduces to the identity wherever the promotional response is impulsive (most of M5, all of Favorita) and contributes only where genuine multi-day carryover exists, so the gains rest on the structural decomposition itself. Evaluating strictly out-of-sample on 30,490 M5 SKUs and 2,845 Favorita items, with calendar-aware baselines given CD’s identical future calendar, we anchor the contribution on a full inventory-cost accounting: CD lowers safety stock, holding cost, and order variance but under-provisions event spikes, reducing total cost only when holding costs exceed \sim 20% of stockout costs (95% CI [17%,25%] ); otherwise it is an operational-stability and inventory-capital layer, not an expected-cost minimizer. Its accuracy contribution is reliability rather than central tendency: across eleven baselines, CD attains the lowest cross-sectional dispersion of per-SKU error and mis-forecasts by more than 200% on 0.8% of SKUs versus 9.9–20.6% for every baseline, ranking first on both in all four M5 draws. Because the Variance Ratio and std-based safety stock are minimized by any sufficiently smooth forecast, we treat them as diagnostics, not objectives. A supporting analysis shows the learned demand operators are non-normal, yet CD’s compact parametric kernel matches their operational performance interpretably.
[LG-15] Using Data-Derived Priors to Guide CNN Architecture Design for NIR Chemometrics
链接: https://arxiv.org/abs/2607.25636
作者: Dário Passos
类目: Machine Learning (cs.LG); Applied Physics (physics.app-ph); Computational Physics (physics.comp-ph)
*备注: 25 pages, 4 figures, submitted to peer-review journal
Abstract:Convolutional neural networks (CNN) for near-infrared (NIR) chemometrics are often designed using generic architectural rules, although spectral datasets differ in sampling, smoothness, redundancy, and sample size. We tested whether these properties can provide empirical priors for CNN design. Across 25 NIR regression tasks, we computed descriptors of dataset size, spectral length and spacing, entropy, intrinsic rank, autocorrelation, and wavelet-scale structure. Two interpretable 1D-CNN scaffolds (a minimal single-convolution model and an extended shallow model with optional branching, dilation, etc) were optimized using five-fold cross-validated Bayesian hyperparameter optimization (HPO). Relationships extracted from near-optimal trials were converted into warm-start heuristics and evaluated directly and through leave-one-dataset-out (LODO) validation. The clearest relationships involved convolutional receptive fields. In the minimal CNN, the preferred kernel fraction decreased with spectral entropy and intrinsic rank, increased with the wavelet energy-support fraction, and the learning rate tended to decrease with training-set size. Direct and LODO heuristics were competitive with HPO, with median test-RMSE ratios of 0.953 and 1.017, respectively. The extended CNN showed similar but less transferable structure across branch usage, dilation, dropout, filter counts, and receptive-field choices. Ten stochastic refits showed seed sensitivity comparable to that of HPO-selected configurations. In a separate experiment, joint preprocessing and CNN HPO outperformed standardized-spectra HPO in 19 of 25 tasks, although gains were dataset-dependent. These results show that spectral descriptors can provide practical CNN design priors, guiding shallow NIR models toward plausible hyperparameter regions before target-specific tuning
[LG-16] Mind the Missing Split: Resolving Feature Heterogeneity in Swarm Learning with Random Forests
链接: https://arxiv.org/abs/2607.25538
作者: Mohammad Tajabadi,Dominik Heider
类目: Machine Learning (cs.LG)
*备注:
Abstract:Swarm Learning is a decentralized collaborative learning mechanism that allows multiple organizations to train a shared model without central coordination or direct data sharing. In typical horizontal Swarm Learning, datasets across sites are usually assumed to share the same feature set. However, in real-world applications, sites often have partially overlapping features because measurements, protocols, and available covariates differ across sites. This feature heterogeneity creates a practical issue for machine learning algorithms such as Random Forests. Specifically, when decision trees are pooled into a global Random Forest, inference at a given site can become ill-defined if a traversal encounters a split on a feature that is not available locally, often forcing organizations to discard site-specific variables upfront. In this paper, we address feature heterogeneity in Swarm Learning with Random Forests under partially overlapping feature spaces. We propose several deterministic and probabilistic inference-time strategies that resolve such missing splits without restricting training to the intersection of features. We evaluate the methods on nine datasets and demonstrate that they outperform both the intersection baseline and locally trained models across a broad range of scenarios.
[LG-17] AMPBench-MT: A Homology-Controlled Benchmark for Antimicrobial Peptide Potency Spectrum and Safety Prediction
链接: https://arxiv.org/abs/2607.25518
作者: Ziheng Zhou,Huiyu Luo,Xiaohu Zhu,Nan Wang,Xuebiao Qin,Chaoyan Zhang,Jun Yan
类目: Machine Learning (cs.LG); Quantitative Methods (q-bio.QM)
*备注:
Abstract:Computational AMP discovery is often evaluated through AMP/non-AMP recognition, yet follow-up decisions depend on assay-derived evidence such as target-species potency, hemolysis, toxicity, and selectivity. Existing AMP and peptide benchmarks cover binary recognition, multilabel annotation, assay regression, or broader peptide-model comparison, but they do not jointly place AMP recognition, species-conditioned potency, spectrum, safety-facing proxy endpoints, and cross-endpoint behavior within one sequence-homology-controlled protocol. To address this problem, we introduce AMPBench-MT, a provenance-preserving benchmark that standardizes canonical peptide records and organizes them into binary recognition, species-conditioned pMIC regression, and endpoint-specific potency and safety-facing readouts. Across 161 endpoint-specific model evaluations, high binary performance does not reliably indicate assay-endpoint behavior. Frozen protein-language-model embeddings form the leading pMIC error cluster, while graph and classical regressors remain close. Spectrum labels further reveal that PR-oriented metrics can be misleading under scarce observed negatives, whereas low-toxicity, HC50 hemolysis, and selectivity expose smaller but more assay-facing signals. AMPBench-MT shows that AMP evaluation should move beyond recognition leaderboards toward endpoint-aware evidence auditing. Our proposed benchmark is available at this https URL.
[LG-18] Anti-Backdoor Coreset Selection via Cumulative Entropy
链接: https://arxiv.org/abs/2607.25502
作者: Qi Zhao,Christian Wressnegger
类目: Machine Learning (cs.LG); Cryptography and Security (cs.CR)
*备注:
Abstract:Recent training-time defenses against neural backdoors isolate a benign subset from poisoned training data, to learn a backdoor-free model from it. In this paper, we formulate this defense strategy as a coreset selection problem, giving rise to so-called “Anti-Backdoor Coreset Selection.” Since poisonous samples have (a) lower prediction uncertainty and are (b) less frequent than benign samples, coreset selection naturally focuses more on samples associated with benign functionality than the backdoor functionality. We use the Cumulative Entropy as selection criterion to further facilitate this effect. The metric tracks the learning dynamics of training samples and allowing us to select benign samples with high informativeness for the coreset. Additionally, we unlearn the chosen samples in each epoch to facilitate the separability between benign and poisonous samples. Together, this yields an exceptionally effective training-time defense that constructs a benign coreset to train a backdoor-free model. Unlike prior defenses that compromise natural accuracy and fail against certain attacks, our method mitigates backdooring attacks consistently with a negligible impact on natural performance.
[LG-19] Quantum Speedups for Stochastic Optimization with Heavy-Tailed Noise
链接: https://arxiv.org/abs/2607.25492
作者: Bin Luo,Chengchang Liu,Jonathan Allcock,Shengyu Zhang,John C.S. Lui
类目: Machine Learning (cs.LG)
*备注: 56 pages
Abstract:We study stochastic optimization with heavy-tailed gradient noise. We first propose a novel quantum mean estimator for multivariate heavy-tailed random variables that achieves lower query complexity than optimal classical estimators in the low-dimensional regime. We further develop an unbiased quantum mean estimator by applying a generalized multi-level Monte Carlo technique. We prove quantum lower bounds showing that, when the dimension d of the random vector is small and can be viewed as a constant, our quantum estimators are optimal up to logarithmic factors. We further derive stronger dimension-dependent lower bounds for tail index p4/3 , showing that a nontrivial dependence on the dimension is unavoidable in the low-dimensional regime. Based on these estimators, we propose a quantum normalized stochastic gradient descent method ( \textttQNSGD ), which finds an \epsilon -stationary point using \tilde\mathcalO\big(\sqrt d,\epsilon^-\frac5p-42p-2\big) queries to the quantum stochastic gradient oracle. For a convex objective function, we propose a quantum projected stochastic gradient descent method ( \textttQPSGD ), which computes a solution with \epsilon -optimal solution using \tilde\mathcalO\big(\sqrt d,\epsilon^-\frac3p-22p-2+\epsilon^-2\big) queries in expectation. These sharper bounds improve upon the classical lower bounds \Omega\big(\epsilon^-\frac3p-2p-1\big) for nonconvex problems and \Omega\big(\epsilon^-\fracpp-1\big) for convex problems in the low-dimensional regimes d\lesssim\epsilon^-\fracpp-1 and d\lesssim\epsilon^-\frac2-pp-1 , respectively.
[LG-20] Data-Dependent Regret and Polyak Corrections for Constrained Online Convex Optimization
链接: https://arxiv.org/abs/2607.25480
作者: Wentao Zhang
类目: Machine Learning (cs.LG)
*备注: Accepted for publication in Transactions on Machine Learning Research (TMLR)
Abstract:Constrained online convex optimization requires minimizing regret against adversarial convex costs while satisfying a convex constraint at every round, as needed in safety-critical applications. A computationally efficient method combines online gradient descent with a Polyak feasibility step, using one constraint evaluation and one subgradient per round. Although this method achieves O(sqrt(T)) regret with per-round feasibility, we derive a tighter, data-dependent analysis by retaining two quantities omitted by the standard worst-case argument. First, we replace the gradient envelope G_f^2 T with the observed accumulation G_T = sum_t ||grad f_t(x_t)||^2. Second, we identify a nonnegative Polyak correction P_T that measures the cumulative squared displacement caused by feasibility projections and enters the regret bound with a negative sign. The resulting improvement, Delta_T = (eta/2)(G_f^2 T - G_T) + P_T/(2 eta), is always nonnegative. We further propose AdaOGD-PFS, an adaptive-step-size method that achieves O(sqrt(G_T)) regret while preserving per-round feasibility. Experiments on ball- and halfspace-constrained problems improve the regret bound by 38 to 43 percent, with both data-dependent gradients and Polyak corrections contributing substantially.
[LG-21] Bits and Memories: Measuring Verbatim Extraction Across LLM Quantization
链接: https://arxiv.org/abs/2607.25451
作者: Akshay Sasi
类目: Machine Learning (cs.LG); Cryptography and Security (cs.CR)
*备注: 8 pages, 3 figures. Code: this https URL . Data and results: this https URL
Abstract:Language models are almost always quantized before they are deployed, and a growing line of work asks whether quantization also lowers their privacy risk. That work measures privacy almost entirely with membership inference. We think this is the wrong thing to measure for the risk that most people actually worry about, namely a model reproducing its training data word for word, and we measure that directly. Using the Pythia models and the public set of sequences each of them is known to have memorized, we track verbatim extraction across five precision levels, from full precision down to four bits, and across three model sizes, while measuring general capability (perplexity) at every point. We find two things. Quantization is a selective forgetter: verbatim memorization falls off faster than capability at every precision and every model size we tried, and this holds under two unrelated quantization algorithms and two evaluation corpora. But the selectivity is not enough to make quantization a privacy defense, which cuts against the optimistic reading of earlier membership-inference results. At the largest model we study, four-bit quantization still reproduces most of the memorized sequences while giving up only a few percent of capability, and the fraction of memorized data that survives quantization grows with model size. We conclude that compression should not be treated as a way to remove memorized training data, and that extraction, not membership inference, is the number practitioners should be watching. All code, sampled evaluation data, and per-configuration results are released.
[LG-22] PIcsC: Partitioning-Induced Covariate Shift Correction
链接: https://arxiv.org/abs/2607.25441
作者: Behraj Khan,Behroz Mirza,Syed Ahmad Chan Bukhari,Tahir Qasim Syed
类目: Machine Learning (cs.LG)
*备注:
Abstract:Covariate shift across training-data partitions biases model selection and parameter estimation in cross-validation, lifelong learning, and federated learning. We propose \textitPartition-Induced Covariate-shift Correction (\textttPIcsC), a Fisher information-based regularization framework that mitigates distribution mismatch between data partitions and a reference distribution. \textttPIcsC approximates partition divergence using the Fisher Information Matrix (FIM) and incorporates the resulting statistic as a regularizer during optimization. The same formulation applies to both centrally partitioned datasets (batches or cross-validation folds) and inherently distributed data (federated clients or decentralized nodes), requiring only partition-local gradient statistics rather than raw data. We further introduce a conditional adaptation mechanism that combines FIM shift with KL divergence to detect significant distribution shifts and activates regularization only when necessary. Experiments on more than 40 datasets demonstrate consistent improvements under both natural and synthetic covariate shift. On fragmented batch and fold settings, \textttPIcsC reduces fragmentation-induced performance degradation by more than 20% and 25%, respectively. On seven federated learning benchmarks, it consistently outperforms FedAvg, FedProx, and SCAFFOLD by 3 -5 percentage points without requiring client-specific personalization. These results demonstrate that Fisher information provides an effective and unified mechanism for mitigating partition-induced covariate shift across both centralized and distributed learning.
[LG-23] Robust Unsupervised Network Intrusion Detection via Federated Learning with Selective Aggregation under Anomalous Sample Contamination
链接: https://arxiv.org/abs/2607.25439
作者: Shohei Kamiguchi,Takayuki Nishio
类目: Networking and Internet Architecture (cs.NI); Machine Learning (cs.LG)
*备注:
Abstract:Network intrusion detection systems (NIDS) have become essential for Internet of Things (IoT) environments, as malware targeting IoT devices continues to evolve in sophistication. Unsupervised learning approaches offer a promising direction by removing the dependency on labeled datasets. However, the common assumption that training data are entirely clean is often violated in practice, particularly when data samples are collected directly from deployed network devices, where anomalies are likely to be present in the training datasets. Such contamination degrades detection performance and highlights the need for robust unsupervised NIDS methods capable of operating effectively under contaminated unlabeled training data. To address this issue, we propose a robust training methodology for anomaly detection (AD) that remains effective even in the presence of unlabeled anomalies. Our method consists of two primary components. First, we exploit a known limitation of federated learning (FL), namely its tendency to underrepresent minority data. By leveraging this characteristic, we attenuate the influence of anomalous data originating from a small number of compromised clients. Second, we introduce a selective aggregation mechanism during model aggregation, which quantifies the “distance” between local client models and a global reference. Specifically, we employ the Expectation-Maximization (EM) algorithm to detect and exclude client groups whose model updates significantly diverge from the majority. This selective aggregation ensures that anomalous updates do not compromise the global model. Experiments conducted on multiple NIDS datasets demonstrate that our method outperforms existing approaches in environments contaminated with anomalous data. Furthermore, the proposed method maintains its detection performance even as the proportion of anomalies increases. Subjects: Networking and Internet Architecture (cs.NI); Machine Learning (cs.LG) Cite as: arXiv:2607.25439 [cs.NI] (or arXiv:2607.25439v1 [cs.NI] for this version) https://doi.org/10.48550/arXiv.2607.25439 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-24] SPARC Segmentation to Prediction via Affine Regression and Counterfactuals
链接: https://arxiv.org/abs/2607.25413
作者: Shivani,Subhayan Roy
类目: Machine Learning (cs.LG)
*备注: 5 pages, 9 figures, 2 tables
Abstract:Transaction propensity prediction in B2B e commerce presents unique challenges distinct from B2C contexts, primarily due to the heterogeneous procurement behaviors of organizational entities, which violate SMOTE’s implicit assumption of within class feature homogeneity. Specifically, B2B buyers exhibit multi modal procurement cycles that render linear interpolation between minority class samples structurally invalid, producing synthetic data that does not represent real purchasing behavior. This paper introduces a production deployed propensity modeling framework designed to address these complexities through two primary contributions. First, we replace conventional SMOTE based augmentation with a synthetic data generation approach leveraging Diverse Counterfactual Explanations (DiCE). This method produces minority class samples with superior distributional fidelity compared to SMOTE, as validated through quantitative proximity analysis and UMAP cluster visualization. Second, we adapt the PyPARC piecewise affine classification framework to generate calibrated propensity probabilities, facilitating the interpretable segmentation of customers into actionable risk tiers. Evaluated on two years of longitudinal data from a large scale B2B e commerce platform with a 1 to 9 class imbalance ratio, the proposed architecture achieves 93.1% precision at a decision threshold of 0.8, a 9.2 percentage point improvement over SMOTE based baselines at the same threshold (83.9%), and a 26.1 point improvement over SMOTE at threshold 0.7 (66.04%), demonstrating consistent superiority across operating points. These results demonstrate the framework’s efficacy in enabling high precision marketing campaigns with significant improvements in customer activation and return on investment.
[LG-25] Learned Relied Upon or Necessary? Separating Checkpoint Dependence from Task-Level Value in Sheaf GNNs
链接: https://arxiv.org/abs/2607.25387
作者: Yi Liu
类目: Machine Learning (cs.LG)
*备注: 23 pages, 4 figures
Abstract:Learned restriction maps in sheaf graph neural networks are often treated as proof that the model has discovered useful edge geometry. That conclusion does not follow from parameter movement or from a post-hoc ablation: both can show how one checkpoint is organized while leaving open whether learned transport still helps after the rest of the model adapts. We separate these claims with two estimands. Checkpoint reliance intervenes on the maps of a fixed predictor; protocol-relative replacement retrains matched families that remove map capacity, edge variation, or persistent edge assignment. A task-null theorem shows why the claims can diverge: labels identify only the transported classifier directions, leaving d^2-d invisible degrees of freedom in every full d\times d map. An exact frame model then gives the boundary at which reliance becomes unreplaced task value. Label-only training realizes the predicted separation, while audits of public NSD, DNSD, and Directed Sheaf Neural Network (DSNN) implementations recover both replaceable and unreplaced transport regimes on real graphs. All five DNSD benchmarks exhibit fixed-checkpoint reliance. After retraining, assignment-breaking or shared-map controls recover Full performance on four; Roman-Empire retains a .0675 advantage over continually resampled assignment and a .0391 advantage over a parameter-matched shared map across ten official splits. Thus, a learned map can govern a fitted computation without constituting indispensable edge geometry. Claims of learned transport should pair checkpoint interventions with matched retraining.
[LG-26] Guiding Posterior Exploration with Optimizer-Derived Geometry AISTATS2026
链接: https://arxiv.org/abs/2607.25312
作者: Moritz Schlager,Emanuel Sommer,Thomas Möllenhoff,David Rügamer
类目: Machine Learning (cs.LG)
*备注: Accepted for presentation at the OPTIMAL Workshop at AISTATS 2026
Abstract:Sampling-based methods offer a principled approach to uncertainty quantification in Bayesian neural networks. Their practical use, however, is often challenged by the computational cost of exploring high-dimensional and multimodal posterior distributions. To overcome these difficulties, Bayesian Deep Ensembles, i.e., warmstarting the sampling from several optimized solutions, have proven to be an effective strategy. In this paper, we demonstrate that curvature estimates computed during the warmstart as a byproduct in adaptive optimizers such as AdamW can inform the sampling phase at negligible additional cost. Specifically, our proposed preconditioned sampling strategy based on optimizer-derived geometries can substantially reduce or even eliminate the need for a lengthy sampling burn-in phase and leads to greater numerical stability. This approach consistently maintains or improves predictive performance and uncertainty quantification without any additional computational costs. We confirm the consistency of our findings across various datasets and network architectures.
[LG-27] Zhinv: Real-time hub-height wind field reconstruction using only local sparse observations
链接: https://arxiv.org/abs/2607.25298
作者: Zongwei Zhang,Chin Chun Ooi,Lianlei Lin,Sheng Gao,Tiantian He,Yew Soon Ong,Junkai Wang,Hangyi Yu,Jiaqi Zhang,Hanqing Zhao,Yu Zhang
类目: Machine Learning (cs.LG)
*备注: 36 pages, 8 figures
Abstract:The high proportion of wind power connected to the grid places higher demands on fine-grained knowledge of regional wind fields. Since the wind information directly obtainable in actual operations is mostly sparse, discrete, and irregularly distributed local observations, it is difficult to directly meet the needs of tasks such as wind power regulation, wind resource assessment, and low-altitude environmental perception of continuous regional wind fields. Therefore, we propose Zhinv, an end-to-end reconstruction framework that directly weaves sparse and irregular observations into a fine-grid wind field at hub-height. Experiments in Northeast China, Europe, and Southeast Asia demonstrate that Zhinv can accurately, robustly, and efficiently reconstruct fine-grid wind fields from sparse observations, reducing the error by about 66% compared with Kriging. With local wind-power observations as input, Zhinv enables wind power centers to bypass NWP and complex assimilation processes, supporting direct and real-time wind resource assessment from locally available data.
[LG-28] Breaking the Periodicity Assumption: Robust Tensorial Multi-View Clustering via Graph-Spectral Low-Rank Learning
链接: https://arxiv.org/abs/2607.25295
作者: Jintian Ji,Xingsu Li,Songhe Feng
类目: Machine Learning (cs.LG)
*备注:
Abstract:Tensorial multi-view clustering (TMC) has achieved strong performance due to its ability to capture high-order correlations across multiple views. Most existing t-SVD-based TMC frameworks apply the Fast Fourier Transform (FFT) along the sample mode to impose frequency-domain low-rank constraints. However, we reveal that this widely adopted design critically relies on an implicit ``periodicity assumption’’ induced by the sample arrangement. When samples are ordered by class, neighboring indices tend to be semantically similar, creating artificial local continuity along the sample mode and a favorable spectral structure for FFT-based low-rank regularization. Once this ordering is removed by random permutation, existing t-SVD-based TMC methods suffer severe performance degradation. This strong sensitivity to class ordering conflicts with the permutation-invariant nature of clustering and indicates that part of the reported performance may be attributed to a privileged sample arrangement rather than genuine high-order structure modeling. In this paper, we systematically investigate this phenomenon and its underlying algebraic and spectral mechanisms. To address this fundamental flaw, we further propose a graph-spectral low-rank tensor learning framework based on the Graph Fourier Transform (GFT), which replaces the fixed Fourier basis along the sample mode with a data-driven graph spectral basis, thereby capturing the intrinsic manifold structure without relying on a particular sample ordering. Moreover, we develop an anchor-based variant to address large-scale datasets efficiently. Extensive experiments on various benchmarks validate our findings and demonstrate the competitive or superior performance of the proposed methods compared with state-of-the-art TMC approaches.
[LG-29] AMRD: Adaptive Multi-Teacher Relational Distillation for Lightweight Speech Emotion Recognition
链接: https://arxiv.org/abs/2607.25289
作者: Yuqi Li,Yi-Cheng Lin,Xianglong Wang,Kuo Yang,Xiaoqin Feng,Yixuan Wang,Huiran Duan,Yingli Tian
类目: Machine Learning (cs.LG)
*备注:
Abstract:On-device speech emotion recognition (SER) is critical for real-time applications, yet large self-supervised models that excel at SER are too costly for edge devices. Multi-teacher knowledge distillation can compress them into a lightweight student, but two challenges remain: teacher reliability varies across batches, and logit-level distillation ignores inter-sample relational structure. We propose Adaptive Multi-teacher Relational Distillation (AMRD) to address both. A one-class SVM on each teacher’s logit similarity matrix assigns per-batch weights favoring more coherent teachers. A relational distillation loss aligns teacher and student similarity matrices, capturing structure that logit matching misses. On IEMOCAP and CREMA-D datasets across four student architectures, AMRD outperforms single-teacher distillation baselines in most settings, and ablations confirm both components yield complementary gains.
[LG-30] When Does Deep Representation Learning Help Single-Cell Clustering? A Sensitivity-Aware Diagnostic Benchmark for Biomedical AI Pipelines
链接: https://arxiv.org/abs/2607.25288
作者: Nguyen Thanh Phong,Truong Viet Vu,Nguyen Ha Thu,Tran An Ky,Tran Hoang Thong,Le Pham Thuy Hien,Nguyen Thai Anh
类目: Machine Learning (cs.LG); Genomics (q-bio.GN)
*备注: 13 pages, 6 figures. Accepted at ISRSD 2026
Abstract:Single-cell ribonucleic acid sequencing (scRNA-seq) is a foundational technology for precision-medicine workflows that contribute to United Nations Sustainable Development Goal 3 on Good Health and Well-being, and unsupervised clustering is the analytical step that turns raw expression matrices into interpretable cell populations. Practitioners therefore face a recurring engineering decision: is an additional deep representation stage worth its compute and tuning cost, or do classical principal component analysis (PCA) pipelines already suffice? We address this question with a diagnostic benchmark of nine clustering pipelines on ten real datasets (90-5,685 cells, 19,046-41,480 genes, 4-11 cell types), augmented by a partial scVI V2 specialized comparison on seven datasets. The protocol integrates Optuna hyperparameter search, repeated-run robustness, Friedman/Wilcoxon-Holm/TOST testing, and Sobol total-order sensitivity analysis. The contrastive autoencoder achieved the highest mean Adjusted Rand Index (0.7872), but Holm-corrected tests did not establish dominance over the strongest baselines. Per-dataset analysis reveals three reproducible regimes: probabilistic variational autoencoder (VAE) variants help on the smallest datasets, deep autoencoders win on mid-scale data with multi-batch or many-type structure, and classical PCA pipelines remain competitive when linear projection already captures the dominant variation. Sobol indices identify learning rate ( S_T=0.70 ) and latent dimensionality ( S_T=0.56 ) as the dominant variance contributors, indicating where limited tuning budgets should be allocated. The contribution is therefore a dataset-aware and compute-conscious decision framework for biomedical AI pipelines supporting sustainable healthcare analytics, rather than a universal superiority claim.
[LG-31] HeAD-CP: Heterophily-Aware Diffused Conformal Prediction Sets for Graph Neural Networks
链接: https://arxiv.org/abs/2607.25273
作者: Phan Binh Nguyen Lam,Nguyen Thai Anh
类目: Machine Learning (cs.LG)
*备注: 6 pages, 4 figures. Accepted at MAPR 2026
Abstract:Conformal prediction (CP) provides distribution-free uncertainty quantification, and its extension to graphs is an active research direction. Diffused Adaptive Prediction Sets (DAPS) is a widely used graph-aware diffusion baseline, propagating Adaptive Prediction Sets (APS) non-conformity scores along edges with a uniform coefficient \lambda . We identify a fundamental shortcoming of this design: the uniform low-pass diffusion presupposes graph homophily and proves detrimental on heterophilic graphs, enlarging the mean prediction-set size by up to 10.6% relative to plain APS. To mitigate this, we propose HeAD-CP, a family of node-wise diffusion variants whose coefficients are determined by a label-free local-homophily estimate derived from the GNN softmax. Three variants, namely signed- \gamma , edge-compatibility, and a DAPS-baseline-with-correction, are most effective at extreme heterophily, intermediate heterophily, and moderate-to-high homophily, respectively, and all preserve the marginal coverage guarantee. On ten benchmarks, the HeAD-CP family stays at or below plain APS on every dataset, while DAPS exceeds APS on six. The post-hoc oracle over the family improves over DAPS on 8/10 datasets at p0.01 (paired Wilcoxon), with the largest gains on heterophilic graphs (10.3% on Texas); on the two homophilic datasets where DAPS still wins (CiteSeer, PubMed), it retains a marginal advantage of at most 0.002, statistically insignificant on CiteSeer ( p=0.23 ). Designing a calibrated label-free selector that approaches this oracle is the main outstanding empirical question.
[LG-32] Beyond Single-Episode Optimization: Sliding-Window Aware Generative Auto-Bidding for Long-Term Advertising Effectiveness
链接: https://arxiv.org/abs/2607.25233
作者: Binglin Wu,Chuan Yue,Yingyi Zhang,Xianneng Li,Ruyue Deng,Weiru Zhang,Xiaoyi Zeng
类目: Machine Learning (cs.LG)
*备注:
Abstract:Auto-bidding systems optimize bids to maximize value under efficiency constraints such as Cost-Per-Action (CPA). Existing methods treat each day as an independent episode. However, many advertisers produce value so sparsely that per-day efficiency ratios become statistically unreliable, undermining advertiser retention. Platforms therefore evaluate window-level efficiency over sliding windows of W=7 days, ensuring fair evaluation and long-term advertising effectiveness. This creates cross-episode coupling: each day’s bidding decisions affect up to W overlapping windows, so setting daily targets requires anticipating future market conditions. We propose SWAG-Bid (Sliding-Window Aware Generative Auto-Bidding), a hierarchical framework decomposing this challenge into episode-level planning and step-level execution. The planner uses a Masked Trajectory Model to forecast markets and generate candidate plans, scored across all overlapping windows by Multi-Window Model Predictive Control Sampling (MWMS) with exponential confidence decay. The controller adjusts reliance on this guidance through a state-adaptive gate, Per-Step Gated Adaptive Layer Normalization (PSG-AdaLN), complemented by Return-to-Go and Cost-to-Go channels carrying budget and constraint information. Experiments on AuctionNet-Sparse and online A/B tests on AliExpress show that SWAG-Bid achieves competitive constraint satisfaction and value acquisition under sliding-window evaluation.
[LG-33] Neurai-VN Benchmark: Standardized Machine Learning Models for Multimodal Digital Phenotyping in Mental Health Classification
链接: https://arxiv.org/abs/2607.25232
作者: Quoc-Cuong Pham,Hoang-Thuy-Duong Vu,Thi-Thanh-Huong Ha,Huy-Hieu Pham
类目: Machine Learning (cs.LG)
*备注:
Abstract:Digital phenotyping (DP) using smartphones and wearable devices has shown considerable potential for mental health monitoring. However, progress remains difficult to evaluate due to heterogeneous datasets, inconsistent preprocessing pipelines. In this study, we present a reproducible benchmark built upon the Neurai-VN dataset, a high-resolution, multimodal dataset comprising passive sensing and active assessment from wearable and smartphone devices, collected from 100 Vietnamese adults over two weeks. The benchmark defines four clinically relevant binary classification tasks evaluated using standardized subject-wise cross-validation. Representative linear, tree-based, and neural baseline models are evaluated across predefined feature configurations. Mean subject-level F1 scores across five cross-validation folds reached 0.71 for Healthy Control vs. Depression and Healthy Control vs. Clinical, while Healthy Control vs. Anxiety and Depression vs. Anxiety achieved 0.69 and 0.56, respectively. These benchmark results provide reproducible baselines for future research on multimodal DP for mental health classification tasks.
[LG-34] Decision-Level Hijacking: Injecting Cognitive Bias into Large Language Models via Bit-Flip Attacks
链接: https://arxiv.org/abs/2607.25227
作者: Yu Yan,Jiahao Chen,Siqi Lu,Yongjuan Wang,Ziming Zhao,Zhaoxuan Li,Tianyu Du,Qingjun Yuan,Shouling Ji
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注:
Abstract:Large Language Models (LLMs) have been widely applied in high-stakes decision-making scenarios such as corporate strategy, and users are increasingly relying on their outputs. However, the deep integration of open-source model sharing ecosystems with LLM-powered critical decision-making applications also introduces critical risks: if an attacker can manipulate the model’s cognitive stance, they can indirectly influence the judgments and actions of downstream decision-makers. This paper defines such threats as decision-level hijacking. Existing attacks fail to achieve targeted cognitive manipulation without triggering prohibited content or degrading model functionality. To fill this gap, this paper reveals that Bit-Flip Attacks (BFAs) can serve as an attack vector for inducing decision-level hijacking, requiring no real-time interaction or control over the training process, and only a minimal number of weight bits need to be flipped after deployment to achieve stealthy, low-cost, and persistent cognitive manipulation. Therefore, we propose CogBias, a cognitive bias injection framework for LLMs. CogBias converts subjective preferences into optimization signals via a differentiable sentiment evaluator, uses a multi-objective loss to jointly constrain multiple dimensions, and constructs BitScout to locate critical bits, achieving targeted cognitive intervention under an ultra-sparse flip budget. Experiments on Llama-3.2-3B, Mistral-7B, and Qwen2.5-14B, as well as on the commercial recommendation and controversial factual topic scenarios, demonstrate that flipping only a small number of bits stably induces significant stance shifts on target topics, while the impact on non-target tasks and overall output distribution is limited. This work demonstrates that minute perturbations to low-level weight data suffice to undermine the high-level value alignment of LLMs.
[LG-35] SecDrift: Measuring Sector-Conditioned Security Drift in AI-Generated Code
链接: https://arxiv.org/abs/2607.25225
作者: Narayanaswami Natraj Bharadwaj,Dhivya Chandramouleeswaran
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG); Software Engineering (cs.SE)
*备注: 11 pages, 3 figures
Abstract:LLMs are increasingly used for code generation in critical infrastructure, yet the security effect of domain-specific prompting is understudied. We present SecDrift, a benchmark measuring sector-conditioned security drift: the change in static-analysis vulnerability rates when prompts are conditioned on industry contexts versus neutral baselines. We evaluate 7 LLMs (6 producing analyzable code) across 8 CISA critical infrastructure sectors and 9 CWE categories with 5 replicates (5,355 evaluations), using a 5-dimension transformation with a matched-baseline condition that holds the task fixed while substituting only domain terminology. Industry prompts naively appear more secure (14.0% vs. 11.4%, -2.7pp), but the gap is not statistically significant (Fisher’s exact p = 0.24, Cohen’s h = -0.08) and is a composition artifact of two CWE categories: excluding CWE-502 and CWE-22 eliminates and slightly reverses it (+0.4pp, p = 1.00). A mixed-effects logistic regression confirms sector identity is not a moderator and localizes the only detectable condition effect to those two vulnerability types. 0 of 8 sectors show drift distinguishable from baseline, corrected or uncorrected (|h| 0.15). A placebo on two non-CISA sectors (e-commerce, online education) reproduces the CISA industry rate almost exactly (10.5% vs. 11.4%, p = 0.63): the small pooled pattern reflects generic industry-framing specificity, not critical-infrastructure identity. In contrast, model selection has a large and consistent effect: among full-output models vulnerability rates range from 11.6% to 16.1%, and these differences persist across conditions. Model choice, not prompt framing, is the more reliable security lever. We release the framework, prompts, generated code, findings, human-validation verdicts, and analysis scripts.
[LG-36] A Unified Algorithmic Framework for Hybrid Reinforcement Learning in Tabular MDPs with Shifted Transition Dynamics
链接: https://arxiv.org/abs/2607.25207
作者: Zheshun Wu,Renjie Zheng,Jinhang Zuo,Zenglin Xu,Fang Kong
类目: Machine Learning (cs.LG)
*备注: 59 pages, 3 figures, and 2 tables
Abstract:This paper investigates a hybrid reinforcement learning setting in tabular Markov Decision Processes (MDPs), where an agent aims to learn an optimal policy by combining online interactions with a target environment and offline data from a source environment. A central challenge is that offline data may be collected from outdated environments with shifted transition dynamics, making naive integration of historical data ineffective. To address this, we propose a unified algorithmic framework featuring two algorithms: MIN-UCB-VI for regret minimization and MAX-LCB-VI for best policy identification. Both algorithms leverage fine-grained bias information to more effectively exploit offline data under general transition shifts. We provide theoretical guarantees for our framework, including both instance-dependent and independent upper bounds on regret and sub-optimality gap. Furthermore, we establish matching lower bounds to demonstrate the optimality of our approach and validate our theoretical findings through extensive experiments.
[LG-37] Algorithmic Separation between Constant-Depth and Logarithmic-Depth Neural Networks
链接: https://arxiv.org/abs/2607.25200
作者: Yunwei Ren,Zihao Wang,Jason D. Lee
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注:
Abstract:Despite the empirical advantages of deep networks over shallow ones, theoretical depth separations largely concern approximation power, while algorithmic results are mostly limited to comparisons between two- and three-layer networks. In this work, we prove the first algorithmic separation between constant-depth and logarithmic-depth networks. Specifically, we identify a class of Boolean functions with hierarchically structured Fourier spectra that logarithmic-depth networks can learn efficiently using layerwise coordinate descent by reconstructing the spectra hierarchically and adaptively. We also exhibit a subclass for which every constant-depth, polynomial-width network with sufficiently regular activations and controlled spectral norms must incur constant L^2 approximation error under the uniform distribution over the hypercube. Subjects: Machine Learning (cs.LG); Machine Learning (stat.ML) Cite as: arXiv:2607.25200 [cs.LG] (or arXiv:2607.25200v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.25200 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-38] Rethinking CD: A Reproducibility Study and Extension on the Ineffectiveness of Contrastive Decoding at Mitigating Object Hallucinations in MLLM s
链接: https://arxiv.org/abs/2607.25196
作者: Arnav Bendre,Guneesh Gupta,Kavish Grover,Chayan Aggarwal,Shreyansh Modi
类目: Machine Learning (cs.LG)
*备注: 32 pages, 9 Figures, submitted to TMLR
Abstract:Contrastive decoding (CD) has been proposed as a training-free strategy for mitigating object hallucinations in multimodal large language models (MLLMs), with reported gains on benchmarks such as POPE. However, recent work has questioned whether these gains reflect genuine improvements in visual grounding. In this study, we reproduce and extend the findings of “The Mirage of Performance Gains: Why Contrastive Decoding Fails to Mitigate Object Hallucinations in MLLMs.” Specifically, we test the claim that CD induces a unidirectional output distribution shift in discriminative datasets and examine its generalizability across datasets. We also verify that the adaptive plausibility constraint (APC) reduces sampling to greedy search on both discriminative and generative benchmarks. Beyond reproduction, we rigorously study the effects of CD across generative and discriminative datasets. We conduct several experiments that provide additional insights: we analyze the logit distributions induced by different CD strategies on generative datasets, propose a proxy method and compare its performance against CD techniques, and investigate how hallucination signals propagate through each layer of the expert and amateur models. Experimental results across MME, POPE, and CHAIR using LLaVA and Qwen validate the original claims and show that the apparent improvements from CD are often spurious and do not consistently translate into stronger visual grounding for reducing hallucinations. These findings challenge the effectiveness of current contrastive decoding strategies and motivate the development of more reliable approaches for mitigating hallucinations in MLLMs.
[LG-39] A Riemannian View on Active Subspaces
链接: https://arxiv.org/abs/2607.25163
作者: Zachary Grey
类目: Numerical Analysis (math.NA); Machine Learning (cs.LG); Differential Geometry (math.DG)
*备注:
Abstract:Active subspaces provide an explainable, eigenvalue-ordered principle for studying how scalar-valued quantities of interest change the most, on average, over a reduced basis of Euclidean domains. Composition with parallel transport generalizes this principle from Euclidean space to quantities of interest defined over Riemannian manifolds, and the resulting intrinsic formulation is contrasted with the extrinsic, embedding-based gradient average of manifold learning. Either strategy is studied in an intrinsically local sense, restricted to mean-centered geodesic-balls, and within that scope the two are not identical: on the central tangent space, eigenvalues agree to second order in the geodesic radius of the sampled domain, while dominant eigenspaces agree at the same order relative to the spectral gap. Extending activity beyond that central space then calls for either recomputed decompositions over changing tangent spaces or, intrinsically, parallel transport of a single central frame. Hyperspheres are emphasized throughout as a particular manifold of interest, motivated by applications over preshape spaces for statistical shape analysis. Numerical examples over the 2-sphere illustrate the formalism, including the derived ridge recovery at a curvature-limited quadratic rate.
[LG-40] Accurate structural modeling of chemically diverse molecular interfaces with Vilya-2
链接: https://arxiv.org/abs/2607.25156
作者: Vilya Research:Pascal Sturmfels,Naozumi Hiranuma,Milad Salem,Benjamin D. Sellers,Stephen Rettie,CJ San Felipe,Chase A. P. Wood,Jeffrey K. Holden,Adam P. Moyer,Patrick J. Salveson,Ivan Anishchanka
类目: Machine Learning (cs.LG)
*备注: 20 pages, 14 figures, 2 tables
Abstract:Structure-prediction networks built on co-evolutionary statistics have transformed protein-based drug discovery, yet their accuracy does not extend to peptide therapeutics–an increasingly important modality defined by non-canonical residues, macrocyclization, and complex topologies. We introduce Vilya-2, a diffusion transformer that extends the all-atom representation of Vilya-1 from modeling individual molecules to modeling their interactions with protein targets. This all-atom representation enables transfer learning between different molecular types, and delivers highly accurate structural modeling of peptides across sizes, classes, and compositions bound to therapeutically relevant targets. By generating diverse structural ensembles and ranking them with calibrated confidence, Vilya-2 recovers 59.1% of peptide interfaces to sub-2 Å backbone RMSD, far exceeding the performance of a representative co-folding model even when that model is given the bound receptor as a template. In addition, Vilya-2 is state-of-the-art at small-molecule docking, and generalizes to novel protein-small molecule complexes unlike those seen in training. It also generalizes to modeling molecular conformations of diverse macrocycles and disulfide-stapled miniproteins several-fold larger than any molecule seen in training. Finally, Vilya-2 can be used as a foundation model, and fine-tuned to enrich for active compounds in hit-to-lead campaigns. By unifying predictive accuracy with broad generalizability across chemical space, Vilya-2 is the structure-prediction oracle that de novo peptide design pipelines require–establishing the all-atom approach as a general foundation for the design and evaluation of de novo peptide therapeutics.
[LG-41] Physics-Informed CNN-LSTM for Street-Scale Urban Flood Prediction: Reconciling Aggregate Accuracy and Street-Level Plausibility
链接: https://arxiv.org/abs/2607.25148
作者: Luc DCosta,Yidi Wang,Jonathan L. Goodall,Rohan Chandra
类目: Machine Learning (cs.LG)
*备注: 23 pages, 9 figures
Abstract:Deep learning surrogate models trained with mean-squared-error loss produce statistically accurate but physically unconstrained flood predictions: water may flow uphill, appear spontaneously, or smooth over street-level corridors. We develop a physics-informed training framework for CNN-LSTM models that predict urban flood depths at 15 min intervals over a 128x128 spatial grid. Three differentiable penalty terms are embedded into the loss: (i) a gravity loss penalizing depth increases against the water-surface-elevation gradient, (ii) a continuity loss enforcing local mass conservation with rainfall-adaptive thresholds, and (iii) a topography-aware false-alarm penalty modulated by the topographic wetness index (TWI). We evaluate on the Norfolk, Virginia flood dataset spanning two storm events (August 2017 and September 2022, 300 samples), with all variants trained on identical splits and robustness assessed over repeated random splits and leave-one-storm-out tests. A road-proximal evaluation restricted to a TWI-derived street mask quantifies street-level skill. The physics-constrained model achieves near-zero gravity violations (order 1e-6) and the highest street-channel recall (0.77 +/- 0.09 vs 0.44 +/- 0.10 for the unconstrained baseline), the capability most relevant to traffic routing, and its advantage more than doubles on a held-out storm; a uniform false-alarm variant attains 16% lower mean absolute error but suppresses street recall to 0.25. The TWI-modulated penalty reconciles this trade-off: it improves on the uniform variant on every metric, recovering 60% higher street recall at the lowest MAE among constrained variants and the best street-level F1. These results expose a fundamental tension between aggregate pixel-level error and application-specific physical plausibility, and show that terrain-aware loss modulation offers a principled resolution.
[LG-42] Interpretable GOHR Agents via Sparse Autoencoders
链接: https://arxiv.org/abs/2607.25132
作者: Shiwei Tan,Yusong Zhao,Weiyi Qin,Wentian Wang,Jacob Feldman,Lazaros K. Gallos,Paul B. Kantor,Vladimir Menkov,Hao Wang
类目: Machine Learning (cs.LG)
*备注:
Abstract:A central challenge in interpreting learned decision-making systems is to determine whether their internal representations contain concepts that help explain their behavior. We report interpretability experiments for a tokenized autoregressive Transformer agent in the Game of Hidden Rules (GOHR). We focus on a compact two-rule task in which both hidden rules map object shapes to target buckets, but with different permutations. The policy is trained on episodes sampled from these two hidden rules and then evaluated with fixed weights. It is never given a rule label and does not use an explicit rule classifier; any rule information must be inferred implicitly from interaction history. In this setting, the correct rule is not identifiable before the agent tries an informative move and observes accept/reject feedback. Sparse autoencoders (SAEs) trained on the agent’s decision-token embeddings recover this structure. When held-out decisions are labeled by simple concepts such as the chosen shape or bucket, SAE dimensions that are highly selective for a concept cover most decisions where that concept is present. Individual SAE dimensions also correspond to interpretable strategies such as probing one rule hypothesis and switching after negative feedback.
[LG-43] Endpoint Replay: Compressing the Recency Buffer in Deep Reinforcement Learning
链接: https://arxiv.org/abs/2607.25123
作者: Parham Mohammad Panahi,Armin Ashrafi,Haoyu Du,Andrew Patterson,Martha White,Adam White
类目: Machine Learning (cs.LG)
*备注: Accepted at the Reinforcement Learning Conference (RLC) 2026, to appear in the Reinforcement Learning Journal (RLJ)
Abstract:Experience replay remains one of the most practical and useful algorithmic tools in the deep reinforcement learning (DRL) toolbox. Aside from the limited success of prioritized replay and specialized approaches for large asynchronous systems, most DRL algorithms make use of a large, uniformly sampled recency buffer—even the size, one million, remains unchanged. Could we store less data, reduce redundancy, or more effectively chain experience together to speed up value propagation and still retain the performance of large buffers? In this paper, we investigate a simple compression approach that stores representative transitions derived from the end-points of a chain of connected n -step sequences. By curating these end-points in a smaller recency buffer, our method maintains an effective memory horizon comparable to a standard large buffer while requiring an order of magnitude less storage. Through empirical evaluation, we demonstrate that this approach prevents the systematic bias inherent in naive compression strategies and matches the performance of traditional large buffers in the Pinball environment and the Atari 2600 benchmark.
[LG-44] Semantic Space Search Trajectory Networks
链接: https://arxiv.org/abs/2607.25122
作者: Julian Agudelo,Alberto Tonda,Gabriela Ochoa,Vincent Guigue,Cristina Manfredotti,Evelyne Lutton
类目: Machine Learning (cs.LG)
*备注: 18 pages, 12 figures, submitted to EA 2026: 17th Biennal International Conference on Artificial Evolution
Abstract:Search Trajectory Networks (STNs) are a graph-based tool for visualizing and characterizing the behavior of optimization algorithms. STNs’ reliance on discretization of the search space has largely confined them to low-dimensional or combinatorial settings. We introduce a methodology for constructing STNs in semantic spaces, defined as the space of a model’s predictions on a fixed sample set. Our approach discretizes semantic vectors and aggregates them into network nodes via agglomerative clustering with complete linkage under a normalized Hamming distance. Since any predictor can be summarized by its semantic vector, this method enables comparison of learning dynamics across otherwise incomparable algorithm families. We apply semantic space STNs to classification and regression tasks solved using different machine learning algorithms, recovering known qualitative differences between them. Additionally, we use semantic space STNs to study neural network generalization by contrasting standard training with the label randomization regime of Zhang et al. (2017). The resulting STNs exhibit consistent structural differences, training on real labels produces denser, more efficient and more centralized graphs than training on shuffled labels. Together, our results show that semantic space STNs capture functional training dynamics arising from the interaction between learning algorithms and data, providing a tool for analyzing and comparing learning dynamics across machine learning models and training regimes.
[LG-45] Score-Based Stabilization for Time-Dependent Problems
链接: https://arxiv.org/abs/2607.25119
作者: Eshed Gal,Eldad Haber,Uri Ascher
类目: Machine Learning (cs.LG); Numerical Analysis (math.NA)
*备注:
Abstract:We propose a score-based stabilization framework for numerical simulation of partial differential equations, in which a learned score model defines a stabilization operator applied to provisional numerical updates. This operator augments standard time-stepping schemes by enforcing structure and physical consistency through a correction that drives iterates toward the manifold of admissible states. We show that the stabilization operator acts as a contraction toward this manifold, yielding a correction mechanism with basin-conditional stability. Numerical experiments on Advection, Korteweg-de Vries (KdV), Nonlinear Schrodinger (NLS), and Burgers’ equations demonstrate improved robustness, suppression of nonphysical instabilities, and preservation of qualitative dynamics.
[LG-46] MOSAIC-FL a micro-service based privacy-preserving framework with application to genomics
链接: https://arxiv.org/abs/2607.25107
作者: Paul Largillier,Karl Paygambar,Cédric Gouy-Pailler,Vincent Meyer,Mallek Mziou,Oana Stan
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注: 8 pages, 4 figures, 2 tables. Accepted and presented at SECRYPT 2026, 23rd International Conference on Security and Cryptography, Porto, Portugal, 16-18 July 2026. this https URL
Abstract:Security and privacy are primordial requirements for Federated Learning (FL), especially in fields such as healthcare and genomics where sensitive information has to be analyzed. Our FL framework is designed to address these challenges while proposing a modular, flexible and micro-service architecture. More precisely, it integrates an efficient gRPC communication layer and a Finite State Machine to ensure robust component synchronization and threat detection, while relying on a fault-tolerant secure aggregation protocol using a Threshold variant of the CKKS homomorphic cryptosystem. This allows blind model aggregation by an orchestration server, requiring a minimum of t -out-of- N active clients for decryption while minimizing communication overhead thanks to both cryptographic and network protocols. We ensure IND-CPA-D security through noise flooding and mitigate the recent key-recovery attack on synchronized decryptors by renewing the collective key material at every round. We demonstrate the framework’s effectiveness through diverse use cases, ranging from standard image recognition (EMNIST) to complex genomic classification including breast cancer subtyping on TCGA, evaluating system performance across different threshold values and model scales.
[LG-47] Conformal Cascade: Distribution-Free Accuracy Guarantees for Multi-Tier LLM Inference
链接: https://arxiv.org/abs/2607.25018
作者: Yifan Dou,Shikan Fang,Shibo Li
类目: Machine Learning (cs.LG)
*备注:
Abstract:Large language model (LLM) cascades reduce inference cost by routing easy queries to a small model and deferring hard queries to a larger one. Production cascades govern this deferral through a confidence threshold, but LLM confidence scores are miscalibrated, the threshold must be tuned per model pair and per domain, and no setting yields a formal bound on cascade accuracy. We introduce \textbfConformal Cascade (CC), a multi-tier inference framework that uses conformal prediction set size as the deferral rule: accept when the calibrated set collapses to a single answer, defer otherwise. The procedure delivers a distribution-free, finite-sample accuracy guarantee. By a per-tier union bound, the prediction set at the accepting tier covers the correct answer with probability at least 1 - K\alpha for any user-specified \alpha ; under a selection-preservation condition (consistent with, but not strictly implied by, our marginal coverage results), the bound tightens to 1 - \alpha . We further characterise expected cascade cost as an explicit function of \alpha and the calibration-set acceptance rate. Across 18 multiple-choice benchmarks spanning science, medicine, commonsense, and standardized exams, evaluated on two-tier cascades drawn from four open-weight model families, CC strictly improves over the strongest calibration-tuned heuristic cascade on the majority of family–benchmark pairs, with the largest gains on reasoning-heavy benchmarks where majority vote is unreliable; on easier benchmarks the cascade commits the vast majority of queries to the small model at no accuracy cost. Extension to open-ended generation requires an answer-clustering step that we leave for future work. The method requires no model training and only black-box API access.
[LG-48] Amortising Trajectory Optimisation for Residual MPC via Implicit Contact Differentiation
链接: https://arxiv.org/abs/2607.24959
作者: Daniel Layeghi,Thomas Corbères,Calum Arnott,Aditya Kamireddypalli,Hashim Al-Obaidi,Steve Tonneau,Michael Mistry
类目: Robotics (cs.RO); Distributed, Parallel, and Cluster Computing (cs.DC); Machine Learning (cs.LG); Systems and Control (eess.SY); Optimization and Control (math.OC)
*备注: Code available from this https URL
Abstract:Differentiable simulation can accelerate contact-rich trajectory optimisation by exposing local sensitivities of task outcomes to controls. Existing approaches either use finite differences, which are expensive and step-size sensitive; differentiate iterative contact solvers by unrolling automatic differentiation (AD), which stores a growing computation trace; or require intricate, solver-specific KKT sensitivity derivations. We introduce an AD-assisted implicit derivative for regularised smooth contacts and apply it to Mujoco MJX, based on the Implicit Function Theorem (IFT). The method differentiates the stationarity residual at the tolerance-converged solution, avoiding both solver unrolling and hand-assembled KKT systems. IFT keeps compiled temporary memory nearly constant with solver effort, changing by less than 4 % from one to ten iterations versus 10.6 \times growth for unrolled AD. IFT memory grows slower with active contacts and model dimension, using 20 \times less memory at 256 contacts and 6 \times less at 16 contacts and 96 DoF. We further introduce optimiser distillation for residual MPC, amortising batched full-horizon iLQR into a policy that guides short-horizon residual iLQR. Across Finger, Franka, and Unitree, this raises six-step success by 28-98 percentage points over standard iLQR.
[LG-49] Inverse RL Helps Align AI by Imitating Humans
链接: https://arxiv.org/abs/2607.24900
作者: Michał Wiliński,Liu Leqi,Chirag Nagpal
类目: Machine Learning (cs.LG)
*备注:
Abstract:Language model alignment aims to make model behavior reliably reflect desirable properties such as helpfulness, safety, and instruction following. Current approaches typically use supervised fine-tuning on demonstrations or reinforcement learning with rewards derived from verifiers or human feedback. These paradigms leave an important question underexplored: can demonstrations alone yield an implicit reward that can be inspected, reused, and optimized on-policy to align AI? Motivated by inverse reinforcement learning, we introduce Projected Alignment Reward Estimated from Demonstrations (PARED). PARED recovers the implicit reward underlying expert demonstrations as an explicit function over a small set of response-level features, learned by a lightweight discriminator that separates demonstrations from the policy’s own samples in this feature space. Unlike a standard reward model, PARED requires no task-specific preference annotations: demonstrations provide the task-specific supervision, which can be augmented with AI feedback as additional dimensions of supervision. Through experiments involving inference-time reranking and adversarial on-policy RL, we show that the recovered reward improves a base policy without a supervised loss and yields further gains when optimized after standard supervised fine-tuning. Additionally, we demonstrate that PARED can be used for contextual alignment, in which a single policy can be tailored to the preferences of different audiences.
[LG-50] Behavior-Driven Explainability
链接: https://arxiv.org/abs/2607.24881
作者: Caroline Dominik,Rolf Drechsler
类目: Machine Learning (cs.LG); Hardware Architecture (cs.AR); Software Engineering (cs.SE)
*备注: 8 pages
Abstract:As system complexity has vastly increased, it has become significantly more challenging for a single person or a team to fully understand all aspects of an entire system. Particularly, this holds when considering all the different stages of a system’s development life cycle, such as, e.g., design or maintenance. But especially for safety-critical systems it is essential that the final design can be trusted. Because of this, explainability is becoming an important requirement for modern systems. In this paper, we aim to achieve this goal by utilizing Behavior-Driven Development (BDD), where the expected system behavior is given in the form of structured scenarios. These scenarios give a sequence of actions for each functionality, and by this can be directly translated into explanations. We introduce this method of deriving explanations based on the specification as Behavior-Driven Explainability (BDX). While applicable at any development stage or abstraction level, a case study for the explanation of exceptions in a RISC-V processor shows the support this concept adds during system design.
[LG-51] FinAbstain: Uncertainty-Calibrated Multimodal RAG for Selective Financial Forecasting
链接: https://arxiv.org/abs/2607.24875
作者: Dorothy Torres,Wei Cheng,Henan Huang
类目: Machine Learning (cs.LG)
*备注:
Abstract:Large language models (LLMs) can synthesize financial narratives but may express high confidence when evidence is sparse, stale, or contradictory. This failure is especially consequential in forecasting, where filings, news, prices, volume, and technical signals can disagree. We present FinAbstain, a research framework for uncertainty-calibrated multimodal retrieval-augmented generation (RAG) with selective prediction. A point-in-time retriever admits only information public at the forecast timestamp and supplies modality-specific evidence to fundamental, news, technical, risk, and verification agents. Their probabilistic assessments are aggregated with retrieval relevance, evidence contradiction, repeated-sample consistency, and historical calibration statistics. Temperature scaling, isotonic regression, conformal prediction, and a proposed hybrid uncertainty score are evaluated under a common chronological protocol. A controller predicts bullish, bearish, or neutral outcomes only when uncertainty is below a validated threshold; otherwise it abstains, requests evidence, reduces exposure, or routes the case to human review. The evaluation covers one- and five-day abnormal-return direction, twenty-day volatility intervals, and abstention decisions, using accuracy, calibration, risk–coverage, citation, trading, latency, and cost metrics. To make the design auditable before a full data collection is complete, we report explicitly labeled simulated results rather than empirical claims. These results illustrate the intended hypothesis: calibrated abstention may trade coverage for lower selective error and drawdown. The contribution is a time-safe architecture, a composite uncertainty formulation, and a reproducible evaluation blueprint for evidence-grounded selective financial forecasting.
[LG-52] Stronger Memory-Query Tradeoffs for Convex Optimization: The Limitations of Subquadratic Memory
链接: https://arxiv.org/abs/2607.24827
作者: Michael Menart,Aleksandar Nikolov,Ohad Shamir
类目: Data Structures and Algorithms (cs.DS); Computational Complexity (cs.CC); Machine Learning (cs.LG); Optimization and Control (math.OC)
*备注:
Abstract:We prove two lower bounds for the first order oracle complexity of minimizing a d -dimensional 1 -Lipschitz convex function over the unit ball with m bits of memory. We first show that any such (possibly randomized) algorithm must make \tilde\Omega(\fracd^2\sqrtm) oracle queries. For deterministic optimization algorithms, we show that \tilde\Omega(\min\d^1.6,\fracd^8/3m^2/3) queries are required. For all memory regimes of interest, these improves upon the previous best known lower bounds of \tilde\Omega(\max\fracd^8/3m^4/3,\fracd^4/3m^1/6) and \tilde\Omega(\fracd^5/3m^1/3) for randomized and deterministic algorithms respectively. Notably, due to existing upper bounds, our lower bound for deterministic algorithms is the first to show a sharp oracle complexity phase transition around m\approx d^2 , where a polylogarithmic change in memory leads to a \mathsfpoly(d) change in the number of required oracle calls. Further, when the suboptimality is polynomially small in d , our lower bound randomized algorithms is the first to show that \tilde\Omega(d^2) memory is necessary to nearly match the optimal query complexity among algorithms without memory constraints. Previously, such a result was only known for the regime where the suboptimality is quasipolynomially small in d .
[LG-53] Can Deep Generative Models Reproduce Non-Stationary Gaussian Random Fields?
链接: https://arxiv.org/abs/2607.25929
作者: Daniel Kua,Yan Song
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注: 9 pages, 4 figures, 2 tables
Abstract:Deep generative models (DGMs) are widely used for complex high-dimensional data and increasingly applied to spatial and spatio-temporal modeling. Their generated samples implicitly represent the learned data distribution and associated uncertainty. However, for real-world data, assessing whether DGMs have learned the underlying process is difficult because the ground truth is unknown and evaluation often relies on observations alone. We evaluate representative DGMs, flow matching (FM), DDPM, score-SDE, and VAE, on a known non-stationary Gaussian random field. This paper provides comprehensive metrics to assess recovery of the ground-truth mean and covariance structures, with oracle samples and a stationary control as references. All four models recover the mean surface, while their covariance recovery differs across model families: DDPM and score-SDE recover the covariance structure reasonably well, FM exhibits mildly attenuated non-stationarity and slight variance under-dispersion, and VAE has difficulty recovering the covariance structure. An experiment on ERA5 temperature anomalies further demonstrates how the framework can support the validation and development of DGMs for complex real-world spatio-temporal data.
[LG-54] VAD to the Bone: Ultra-Tiny Speech Activity Detection for Edge Deployment INTERSPEECH2026
链接: https://arxiv.org/abs/2607.25870
作者: Stephen Bauer,Sheila Seidel,Shanza Iftikhar,Scott Veidenheimer,Gorkem Ulkar
类目: Audio and Speech Processing (eess.AS); Machine Learning (cs.LG)
*备注: Accepted for publication at INTERSPEECH 2026
Abstract:Voice activity detection (VAD) triggers downstream speech processing in always-on systems under strict memory, latency, and compute constraints. Recent compact models report strong accuracy but rely on components that are not widely supported: learnable filterbanks, recurrent layers, or non-causal post-processing. We propose kiloVAD, designed for embedded inference using standard Mel features, CNN-only layers, and tunable context/spectral parameters. We introduce per-layer structured pruning with self-distillation and angle-based quantization-aware training (QAT) that outperforms standard QAT by 1-4%. Evaluated per-frame under causal conditions, kiloVAD achieves 0.850 AUC on AVA-Speech with 2.1 k parameters and 200 ms context, establishing a new state of the art for causal, deployment-ready VAD.
[LG-55] Variance-Reduced Conditional Gradient Methods under Markovian Sampling for Nonconvex Composite Optimization
链接: https://arxiv.org/abs/2607.25785
作者: Zhaojun Peng
类目: Optimization and Control (math.OC); Machine Learning (cs.LG)
*备注: 40 pages, 2 figures
Abstract:We study stochastic composite nonconvex optimization over a compact convex set when gradient samples arrive along a single trajectory of a fixed ergodic Markov chain. Existing single-trajectory variance-reduction theory covers smooth unconstrained objectives; we address the projection-free composite setting using the generalized Frank-Wolfe gap. We propose MC-ALFCG, which combines a momentum conditional-gradient method with coupled capped multilevel Monte Carlo estimation and per-iteration clipping. The deepest nested average uses consecutive states from the same trajectory, yielding conditional bias O(\tau_\mathrmmix/T) uniformly over the starting state, while coupling controls the gradient-difference second moment through the iterate displacement. Clipping enforces the pathwise bounds needed by the adaptive analysis. We reduce the Markovian recursion to its independent-sampling counterpart under \sigma^2\mapsto 2\Lambda G_\sigma^2 and L^2\mapsto 2\Lambda L^2 , where \Lambda=O(\tau_\mathrmmix\log T) . For positive centered noise, the tuned method achieves expected sample complexity \widetildeO((\tau_\mathrmmix^2G_\sigma+\tau_\mathrmmix^5/2G_\sigma^2)\varepsilon^-3+\tau_\mathrmmix^5\varepsilon^-2) . The exactly noiseless specialization achieves \widetildeO(\varepsilon^-2) with mixing-time-free constants, while a mixing-time-oblivious variant achieves \widetildeO(\tau_\mathrmmix^6\varepsilon^-3+\tau_\mathrmmix^3\varepsilon^-2) . All guarantees are in expectation under a fixed transition kernel. Controlled numerical studies examine dependence sensitivity, a nonconvex composite instance, and clipping behavior.
[LG-56] Normalizing Flows to Reconstruct Pseudo-PDFs
链接: https://arxiv.org/abs/2607.25282
作者: Yamil Cahuana Medrano,Kostas Orginos
类目: High Energy Physics - Lattice (hep-lat); Machine Learning (cs.LG); High Energy Physics - Phenomenology (hep-ph)
*备注:
Abstract:We investigate a normalizing-flow approach for reconstructing parton distribution functions (PDFs) from synthetic matrix-element data. Our framework combines Gaussian Process priors with invertible neural networks to learn a posterior distribution over PDFs consistent with limited Ioffe-time data. We demonstrate that the architecture preserves physical constraints and extrapolation properties.
[LG-57] Learning from the Unseen: Offline Reinforcement Learning with Hidden Actions
链接: https://arxiv.org/abs/2607.25241
作者: Zeyu Bian,Ying Zhou,Yifan Cui
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:Standard offline reinforcement learning (RL) algorithms typically assume that the actions in the dataset are observed without error. However, in many real-world applications, the true actions are unobserved and only noisy proxies are available, causing existing RL methods to yield biased and potentially misleading conclusions. We study off-policy evaluation in infinite-horizon discounted Markov decision processes with hidden actions. By leveraging the next-state variable as a natural proxy for the unobserved action, we establish identification of the policy value and propose an influence-function-based estimator called LURE (Learning from the Unseen: Robust Estimator). LURE is multiply robust, remaining consistent under several combinations of correctly specified nuisance components, and is asymptotically normal, enabling valid statistical inference. To our knowledge, this is the first work to address offline RL with hidden actions. We demonstrate LURE’s effectiveness through simulations and a sepsis management application using the MIMIC-III database.
[LG-58] Lloyds K-Means Clustering Algorithm Is Frank-Wolfe in Disguise AISTATS2026 AISTATS
链接: https://arxiv.org/abs/2607.25190
作者: Michael Pokojovy,J. Marcus Jobe,Simon Lacoste-Julien
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Computation (stat.CO)
*备注: AISTATS 2026 Poster: this https URL
Abstract:Lloyd’s K -means algorithm, also known as naïve K -means, is a widely used ad hoc optimization heuristic, designed to minimize the sum of squared errors (SSE) across all K -partitions of a dataset via iterative cluster refinement. In this work, we establish a novel connection between Lloyd’s algorithm and the Frank-Wolfe (FW) algorithm, a prominent first-order method for projection-free optimization. We demonstrate that Lloyd’s algorithm is a special case of FW. Leveraging recent advances in FW methods for concave objectives, we derive a non-asymptotic \mathcalO(1/t) convergence rate to a local minimum of the SSE objective. To account for empty clusters, an outcome possible under Lloyd’s greedy assignment, we develop an FW variant for semismooth objectives while retaining the same convergence rate that is solely controlled by the initial SSE value. We illustrate our findings with a simulation study for spherical Gaussian mixtures and a real-world image segmentation dataset.
[LG-59] A Foundational Perspective for Partitional Clustering on Networks
链接: https://arxiv.org/abs/2607.25144
作者: Derya Ipek Eroglu,Cem Iyigun
类目: Optimization and Control (math.OC); Machine Learning (cs.LG)
*备注: 40 pages, 16 figures
Abstract:This study presents a theoretical analysis of partitional clustering on networks, analyzing both hard and soft assignment schemes with different objective functions. Cluster centers are not restricted to vertices but can also be located along the edges. We examine four key models: P-Median (PMP) and Sum of Squares Clustering (SSC) under hard assignment, and Probabilistic Distance Clustering (PDC) and Fuzzy C-Means (FCM) under soft assignment. Through mathematical analysis, we uncover structural properties that differentiate these models, such as the significance of assignment bottleneck points and the role of vertex-restricted solutions in determining optimal cluster centers. Our findings reveal that, while SSC and FCM can yield optimal centers along edges, PMP and PDC inherently favor vertex placement, leading to insights into clustering behavior on networks. These insights offer new directions for designing efficient algorithms and have implications ranging from facility location and network design to clustering on the embedding graphs that power similarity search in modern retrieval systems.
[LG-60] Characterizing and Mitigating the Effects of Device Temperature on RF Fingerprinting Accuracy
链接: https://arxiv.org/abs/2607.25070
作者: Haytham Albousayri,Bechir Hamdaoui
类目: ignal Processing (eess.SP); Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注: 6 pages, 11 figures. Accepted at 17th International Conference on Network of the Future (NoF2026)
Abstract:Radio Frequency Fingerprinting (RFFP) has emerged as a promising approach for device authentication by exploiting hardware-specific impairments embedded in transmitted signals. Yet existing methods largely overlook a major drawback: RFFP sensitivity to temperature–a critical factor influenced by both internal and environmental conditions–which can significantly alter device signatures and degrade classification performance. In this paper, we propose a novel temperature-aware RFFP framework that explicitly incorporates device temperature information into the learning process to improve robustness and generalization. We evaluate the proposed method on a real-world Bluetooth Low Energy (BLE) dataset collected across multiple devices and environmental conditions. Experimental results demonstrate that temperature-aware modeling consistently outperforms other temperature mitigation baselines, achieving significant improvements in classification accuracy, particularly under unseen temperature and environmental conditions.
[LG-61] Simulation-based parameter estimation via a combination of embedded normalizing flows and implied empirical probabilities under moment restrictions
链接: https://arxiv.org/abs/2607.25026
作者: Getachew K. Befekadu
类目: Methodology (stat.ME); Machine Learning (cs.LG)
*备注: 10 pages, with one figure
Abstract:In this work, we present a simulation-based parameter estimation framework for a model defined by a computational simulation of a physical system. We specifically outline an estimation framework consisting of two closely-integrated steps that facilitate an overall end-to-end parameter estimation scheme. The first step involves utilizing an embedded normalizing flow which is used to transform the unknown complex distribution of the residual information into a simple base distribution corresponding to the transformed residual information. In the second step, an empirical-likelihood estimator, under moment restrictions, is utilized for imposing an indirect constrain on the base distribution, where such an instantiated task reasonably allows us to treat the transformed residual information as random variables arising from discretely distribution population with each transformed data point as a single-cell from a set of finite-cell contingencies. Moreover, we use first-order gradient methods for updating the estimated parameter values of the model defined by the computational simulation and the corresponding parametrized embedded normalizing flow, that call for all gradient-related information by leveraging implicitly differentiations of the empirical-likelihood function, which is constructed from the implied empirical probabilities under moment restrictions. Here, it is worth mentioning that the problem formulation presented in this work, which highlights an information-theoretic interpretation, allows to present a computational framework for algorithmic implementations. Finally, as a-by-product, the inverse of the parametrized embedded normalizing flow, w.r.t. the estimated parameter values, serves as a surrogate model for the computational simulation model, which provides useful information for quantifying model discrepancies and sensitivity analysis.
[LG-62] Fast accurate and differentiable: a neural-network surrogate for NRSur7dq4 precessing binary black hole waveforms
链接: https://arxiv.org/abs/2607.24960
作者: Michael Pürrer,Ashwin Girish,Lucy M. Thomas,Scott E. Field,Vijay Varma
类目: General Relativity and Quantum Cosmology (gr-qc); High Energy Astrophysical Phenomena (astro-ph.HE); Instrumentation and Methods for Astrophysics (astro-ph.IM); Machine Learning (cs.LG)
*备注: 37 pages, 18 figures
Abstract:We present a neural network surrogate model that emulates the NRSur7dq4 gravitational waveform model for precessing binary black hole mergers. The surrogate decomposes the waveform into constituent quantities and trains an independent multilayer perceptron (MLP) for each. We validate the surrogate against NRSur7dq4 on 10,000 waveforms spanning its full parameter space ( 1 \leq q \leq 4 , |\chi_A,B| \leq 0.8 ). For representative total masses between 60 and 300 M_\odot , median sky-averaged frequency-domain mismatches range from 8.0 \times 10^-5 to 1.7 \times 10^-4 , with 95th percentiles below 10^-3 . On an NVIDIA L40S GPU the JAX surrogate evaluates a single waveform in about 1 ms end-to-end, roughly 10 times faster than the LALSimulation C implementation of NRSur7dq4, and sustains about 140 times the LALSimulation throughput at batch size 64, making it well suited for both low-latency parameter-estimation samplers and large-scale waveform generation. The full NRSur7dq4 NN waveform-to-likelihood pipeline is implemented in JAX and is differentiable. This is the first neural-network surrogate of a precessing numerical-relativity waveform model to combine validated NR-faithful accuracy with a fully differentiable, GPU-accelerated inference pipeline, enabling gradient-based inference approaches via automatic differentiation including Fisher information matrices, GPU-accelerated nested sampling, gradient-based MCMC and importance sampling.
[LG-63] owards Real-World RUL Prediction for Aircraft Compressors Under Variable Conditions with Physics-Guided Domain Adaptation
链接: https://arxiv.org/abs/2607.24809
作者: Yang Zhang,Shashvat Prakash,Jiong Tang
类目: ignal Processing (eess.SP); Machine Learning (cs.LG)
*备注:
Abstract:Remaining useful life prediction for aircraft centrifugal air compressors in real commercial operations poses challenges that controlled benchmark datasets do not expose. In-flight sensor signals superimpose genuine degradation on continuously varying operating conditions, and no channel can be assumed a priori to carry a reliable degradation signature. Moreover, models trained on one subset of a fleet generalize poorly to unseen aircraft and installation positions, a cross-domain problem of underappreciated practical severity. This paper presents a framework combining physics-guided processing with adaptive temporal encoding. A population-level selection procedure identifies surge margin as the most consistent degradation indicator across the fleet. Operating regime filtering and windowed aggregation recover the health trend from operational noise, and two physically motivated features encoding current health level and cumulative degradation rate enable cross-unit comparison without absolute signal values. Sinusoidal positional encoding provides lifecycle context without data leakage, and a structured cross-domain evaluation identifies encoding period mismatch as the primary mechanism of performance loss under fleet transfer. An adaptive recalibration scheme estimates each target unit’s lifecycle scale from early observations alone, requiring no future information or labeled target data. The approach yields substantial, consistent gains in cross-domain accuracy across transfer scenarios of increasing domain distance, and a Gaussian Process Regression model further provides calibrated uncertainty estimates supporting risk-informed maintenance scheduling.
[LG-64] Forgetting Is Not a Fix: Path Dependence in Sequential Engram Editing
链接: https://arxiv.org/abs/2607.24805
作者: Ferdinand M. Schessl
类目: Neurons and Cognition (q-bio.NC); Machine Learning (cs.LG)
*备注: 8 pages, 3 figures. Companion code and data: this https URL
Abstract:AI Engram (Kwon et al., 2026) formalizes the four engram criteria of neuroscience as a constrained inverse problem in weight space and solves it closed-form: concept-specific memory traces become linear objects that can be extracted once and combined arithmetically. Appendix F states the Compositional Memory States Hypothesis: edited models live on “a commutative manifold where the integration of A and B reaches a consistent equilibrium regardless of the learning sequence.” The evidence base is single and paired edits – in materials terms, single-cycle tests, in which fatigue accumulation is structurally invisible. Whether the hypothesis holds under sequential load is exactly the “temporal dynamics” question the paper defers to future work. We run that test on the authors’ own reference implementation, at their reported best edit strength (TOFU alpha=0.6, a choice favoring the linearity hypothesis), with pre-registered predictions, across three model charges (two vendors, two architecture families). Four findings replicate across all three: (1) zero-shot composition and sequential re-calibrated editing diverge by 61-71% of the edit magnitude; (2) cut order is not interchangeable, and the effect scales with concept overlap – in one charge the order of cutting two Paris landmarks decides whether an uninvolved third concept survives; (3) the survivors’ layer-input covariances – the method’s own sufficient statistics, read as strain gauges – drift monotonically with every further cut, in every surviving concept, in every charge; (4) erased knowledge partially returns under subsequent unrelated cuts. Appendix F’s commutative-manifold hypothesis is thereby falsified for sequential editing; the single-edit results of the original paper are untouched. For unlearning-as-compliance: erasure certified today does not certify the artifact after its next edit.
[LG-65] Exploring Line Bundle Standard Models with Transformers
链接: https://arxiv.org/abs/2607.00078
作者: Jacky H. T. Yip,Alessandro Mininno,Gary Shiu
类目: High Energy Physics - Theory (hep-th); Machine Learning (cs.LG)
*备注: 35 pages, many figures and long tables in the appendices. Code and tools available at this https URL
Abstract:We propose a Transformer-based Reinforcement Learning architecture, “LB-Explorer”, to search for heterotic line bundle standard models arising from compactifications on smooth Calabi-Yau (CY) threefolds. We focus on E_8\times E_8 heterotic string theory compactifications on CY with abelian line bundles to produce \textSU(5)\times \textS(\textU(1)^5) symmetry, whose \textSU(5) can be further broken to an MSSM-like gauge group using appropriate discrete Wilson lines. We test the LB-Explorer environment on complete intersection Calabi-Yau (CICY) manifolds, though the neural network architecture naturally generalizes to any CY admitting a simplicial Mori cone and a freely-acting discrete symmetry. The LB-Explorer efficiently learns constraints on the line bundle sums, guaranteeing the E_8 gauge embedding, anomaly cancellation, poly-stability (supersymmetry), chirality of the spectrum, and the absence of exotic matter. Valid configurations can be subsequently filtered by imposing the missing constraints, such as the equivariant structure of the line bundle sum and further requirements on the particle spectrum. In this direction, we introduce a hybrid architecture incorporating CP-SAT solvers that aims to impose some of the conditions exactly by perturbing solutions found by the LB-Explorer. The versatility and scalability of the LB-Explorer make it a powerful tool for navigating the string landscape with a large number of moduli. The code and tools necessary to reproduce our findings are available at this https URL
[LG-66] Probabilistic Symbolic Regression for Equation Discovery via Operator-induced and Regularized Symbolic Forests
链接: https://arxiv.org/abs/2509.19710
作者: Somjit Roy,Pritam Dey,Bani K. Mallick,Debdeep Pati
类目: Methodology (stat.ME); Machine Learning (cs.LG); Symbolic Computation (cs.SC); Statistics Theory (math.ST); Machine Learning (stat.ML)
*备注: 148 pages, 37 figures, 23 tables, submitted
Abstract:Symbolic regression has emerged as a powerful tool for artificial intelligence-driven scientific discovery by learning interpretable analytical expressions that reveal governing relationships directly from data. Existing methods, however, often rely on heuristic search, struggle to balance predictive accuracy with expression complexity in noisy settings, and offer limited characterization of symbolic uncertainty. Probabilistic approaches that address these challenges in a unified manner remain underexplored. We introduce a probabilistic symbolic regression framework that represents mathematical expressions as ensembles of symbolic trees. A regularizing prior over tree topology controls expression complexity, while an Occam’s window-based posterior summary captures uncertainty across multiple plausible symbolic models. Given the limited existing theoretical treatment of symbolic regression, we develop posterior concentration guarantees under approximate symbolic realizability, yielding a near-parametric rate for exact symbolic representability. Additionally, we establish a sharp oracle concentration result under symbolic misspecification. Comparisons of our proposed framework with state-of-the-art competitors demonstrate superior predictive accuracy, optimal symbolic complexity, and stable structural recovery when learning benchmark scientific equations, together with the identification of scientifically interpretable descriptors in a challenging materials discovery problem.
[LG-67] A Generalized Tangent Approximation based Variational Inference Framework for Strongly Super-Gaussian Likelihoods
链接: https://arxiv.org/abs/2504.05431
作者: Somjit Roy,Pritam Dey,Debdeep Pati,Bani K. Mallick
类目: Methodology (stat.ME); Machine Learning (cs.LG); Statistics Theory (math.ST); Computation (stat.CO); Machine Learning (stat.ML)
*备注: 135 pages, 51 figures, 13 tables, Revision Submitted
Abstract:Variational inference, as an alternative to Markov chain Monte Carlo sampling, has played a transformative role in enabling scalable computation for complex Bayesian models. Nevertheless, existing approaches often depend on either rigid model-specific formulations or stochastic black-box optimization routines. Tangent approximation is a principled class of structured variational methods that exploits the geometry of the underlying probability model. However, its utility has largely been confined to logistic regression and related modeling regimes. In this article, we propose a novel variational framework based on tangent transformation for a broad class of probability models characterized by strongly super-Gaussian likelihoods. Our method leverages convex duality to construct tangent minorants of the log-likelihood, thereby inducing conjugacy with Gaussian priors over model parameters in an otherwise intractable setup. Under mild assumptions on the data-generating mechanism, we establish algorithmic convergence guarantees, a contribution that stands in contrast to the limited theoretical assurances typically available for black-box variational methods. Additionally, we derive near-minimax optimal bounds for the variational risk. Superior performance of our proposed methodology is illustrated on simulated and real-data scenarios that challenge state-of-the-art variational algorithms in terms of scalability and their ability to consistently capture complex underlying data structure.
附件下载
