本篇博文主要内容为 2026-08-05 从Arxiv.org论文网站获取的最新论文列表，自动更新，按照NLP、CV、ML、AI、IR、MA六个大方向区分。
说明：每日论文数据从Arxiv.org获取，每天早上12:30左右定时自动更新。
提示: 当天未及时更新，有可能是Arxiv当日未有新的论文发布，也有可能是脚本出错。尽可能会在当天修复。
目录
概览 (2026-08-05)
今日共更新797篇论文,其中:
- 自然语言处理共129篇(Computation and Language (cs.CL))
- 人工智能共295篇(Artificial Intelligence (cs.AI))
- 计算机视觉共148篇(Computer Vision and Pattern Recognition (cs.CV))
- 机器学习共181篇(Machine Learning (cs.LG))
- 多智能体系统共18篇(Multiagent Systems (cs.MA))
- 信息检索共16篇(Information Retrieval (cs.IR))
- 人机交互共20篇(Human-Computer Interaction (cs.HC))
多智能体系统
[MA-0] Socially Grounded Agent ic AI: Coordinating Plural Perspectives through Social Theory ICML2026
【速读】：该论文旨在解决当前人工智能（AI）系统在多元社会情境中实现对齐（alignment）时面临的根本性挑战：传统单一价值优化范式已无法适应复杂多样的社会价值体系，亟需构建能够识别、表征并回应多种合法观点的新型对齐机制。其核心问题在于现有方法普遍缺乏对价值观如何在社会实践中被组织、争议与协调的清晰理论框架。为此，论文提出将社会理论引入AI设计，强调观点应被视为由社会角色所结构化、通过互动过程形塑，并分布于权力与专业领域的网络之中。解决方案的关键在于重构“多元对齐”（pluralistic alignment）的本质——将其从单纯追求输出多样性转向基于社会根基的协调机制。具体设计启示包括：基于角色的视角表征、跨视角的结构化协同、以及情境敏感的评估机制；对于自主型系统，还需对角色激活、推理轨迹、聚合规则及反馈循环等过程进行对齐。本文贡献在于重新定义多元对齐为一种社会性协调问题，并提出一个结构化、可问责的系统设计空间，为未来在真实场景中实现与验证此类方法指明方向。
链接: https://arxiv.org/abs/2608.03910
作者: Matt Ratto,Abhishek Moturu,Daniel Silver
机构: 未知
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Multiagent Systems (cs.MA)
备注: Pluralistic Alignment Workshop @ ICML 2026, Seoul, South Korea
Abstract:As AI systems are deployed across increasingly diverse social contexts, alignment can no longer be framed as the optimization of a single, unified set of values. Instead, systems must be able to recognize, represent, and respond to multiple legitimate perspectives. This has led to growing interest in pluralistic alignment, which seeks to move beyond one-size-fits-all models of appropriate behaviour. However, current approaches often lack a clear account of how values are socially organized, contested, and coordinated in practice. In this paper, we argue that social theory provides essential conceptual and design resources for addressing these challenges. Drawing on established traditions in sociology, we show how perspectives can be understood as structured by roles, shaped through interaction, and distributed across fields of power and expertise. We translate these insights into concrete implications for AI system design, including role-based representations, structured coordination among perspectives, and context-sensitive evaluation. For agentic systems, this requires aligning not only final outputs, but also the role activations, deliberative traces, aggregation rules, and feedback loops through which those outputs are produced. Our contribution is to reposition pluralistic alignment as a problem of socially grounded coordination rather than output diversification. We outline a design space for systems that engage multiple perspectives in structured and accountable ways, and we identify directions for future work to implement and empirically evaluate these approaches in real-world settings.
[MA-1] FedCritic-MIMO: Communication-Efficient Serverless Federated Critic Learning for Massive-MIMO Resource Control in Open and Disaggregated 6G RANs
【速读】：该论文旨在解决在开放且解耦的6G无线接入网（RAN）中，如何实现分布式、通信高效的智能资源控制问题。具体而言，在复用1（reuse-1）的多小区大规模MIMO正交频分多址（OFDMA）部署场景下，多个独立部署的基站控制器需协同完成用户调度、流级功率分配、波束成形、干扰管理及长期服务质量（QoS）保障，但受限于控制器间信号交互的带宽与延迟。传统集中式训练或参数服务器架构难以适应去中心化、异步运行的开放网络环境。为此，论文提出FedCritic-MIMO框架，其核心解决方案是：采用无服务器（serverless）的联邦多智能体强化学习机制，各控制器保留本地策略（actor）和个性化价值函数（critic）组件，仅交换兼容的共享价值函数参数；通过基于无线感知的事件触发机制、自适应分层稀疏（layer-wise top-k）压缩传输与误差反馈策略，以及考虑干扰的均衡融合规则，实现高效、鲁棒的对等（peer-to-peer）价值函数协作。该方案在固定策略、冻结目标值函数回归模型下，建立了条件有限时间平稳性与共识收敛性理论保证。仿真结果表明，相较启发式算法、独立学习、集中训练及通信消融基线，FedCritic-MIMO在性能-通信权衡上表现最优，显著提升空口吞吐量、用户速率分布均值与平均信干噪比（SINR），增强QoS满足度，并将每比特传输的干扰成本降至最低，同时相比未压缩的分布式价值函数交换，减少了76%的通信开销。研究证明，通过共享兼容的可迁移价值函数参数，可在无需中央轨迹收集或参数服务器聚合的前提下，有效协调分布式控制器，为未来6G智能无线网络提供一种可扩展、低通信开销的协同决策范式。
链接: https://arxiv.org/abs/2608.03852
作者: Amin Farajzadeh,Melike Erol-Kantarci
机构: University of Ottawa(渥太华大学); NETCORE Lab(网络核心实验室)
类目: Machine Learning (cs.LG); Multiagent Systems (cs.MA); Networking and Internet Architecture (cs.NI)
备注: Submitted to IEEE for possible publication
Abstract:This paper proposes FedCritic-MIMO, a communication-efficient serverless federated multi-agent reinforcement learning framework for AI-native resource control across independently deployable cell-level controllers in open and disaggregated 6G RANs. Controllers share no trainer, retain local actors and personalized critic components, and exchange only compatible shared critic parameters. FedCritic-MIMO targets reuse- 1 multi-cell massive-MIMO OFDMA deployments, where RAN controllers jointly manage user scheduling, per-stream power allocation, beamforming, interference, and long-term QoS with limited inter-controller signaling. Each base station locally executes its actor without centralized training or actor federation, while critic knowledge is exchanged peer-to-peer over an interference-aware graph. It enables this collaboration through wireless-aware event triggering, adaptive layer-wise top- k sparse critic exchange with error feedback, and balanced interference-aware fusion. We establish conditional finite-time stationarity and consensus guarantees for the balanced, compressed peer-to-peer critic recursion under a fixed-policy, frozen-target critic-regression model. In strongly interference-coupled reuse- 1 simulations, FedCritic-MIMO achieves the best performance-communication tradeoff among heuristic, independent-learning, centralized-training, and communication-ablation baselines. It achieves the highest held-out throughput, improves user-rate distribution and mean SINR, increases QoS satisfaction, and attains the lowest interference cost per delivered bit among learning baselines. It reduces critic-communication overhead by 76% relative to uncompressed distributed critic exchange. These results demonstrate that serverless exchange of compatible shared critic parameters can coordinate RAN controllers without centralized trajectory collection or parameter-server aggregation.
[MA-2] History Matters: Meta-policy Delegation with Heterogeneous Multi-agent Reinforcement Learning
【速读】：该论文旨在解决异构多智能体系统（Heterogeneous Multi-Agent Systems, MAS）在资源受限场景下如何高效完成研究任务的协同与任务委派问题。其核心挑战在于，各智能体具备不同的能力与运行成本，需在有限资源约束下实现最优的任务分配与协作。解决方案的关键在于提出基于多智能体强化学习（MARL）的序贯委派训练机制，使智能体能够动态决策任务委派路径以最小化总执行成本；同时，引入两种创新框架：一是基于交互历史的策略设计，使智能体策略不仅依赖当前状态，还考虑过往联合动作的历史信息，从而在完全可观测环境中提升协调能力；二是设计一种潜在的多维货币机制，通过经济激励促进智能体间的协作与委派行为，增强系统的灵活性与可扩展性。
链接: https://arxiv.org/abs/2608.03833
作者: Ziqing Lu,Avinash Reddy Mudireddy,Sarra Alqahtani,Weiyu Xu
机构: University of Iowa; Wake Forest University
类目: Multiagent Systems (cs.MA)
备注:
Abstract:AI agents are expected to play an increasingly important role in future decision-making systems. In this paper, we consider collaborative systems composed of heterogeneous multi-agent systems (MAS), where their members have different capabilities and operating costs. We study how agents can delegate tasks to one another so that certain research tasks can be completed effectively under resource-constrained scenarios. We first develop a multi-agent reinforcement learning-based (MARL) delegation training that enables agents to make sequential delegation decisions while minimizing the total execution cost. We then extend this approach to MARL with prescribed delegation topologies. Furthermore, we introduce two new frameworks for collaboration and delegation in multi-agent systems. The first framework proposes that an agent’s policy depends not only on the current state of the underlying Markov decision process but also on the interaction history, including previous joint actions. This history-dependent formulation can improve coordination even in fully observable environments, where conventional MARL methods typically restrict policies to depend only on the current state. The second framework proposes a novel, potentially multi-dimensional monetary mechanism to facilitate the collaboration and delegation for MAS.
[MA-3] An Actionable Diagnosis of Multilingual Multi-Agent Planning Failures
【速读】：该论文旨在解决多语言多智能体系统在非英语语境下性能显著下降的问题，核心症结在于用户请求在转化为可执行计划的过程中，关键任务信息的丢失。研究将多智能体系统中的规划器（planner）视为请求到动作的接口，通过分析真实世界任务失败案例，构建了一个可操作的“规划-对齐失败”分类体系（planning-grounding failures taxonomy）。研究表明，随着语言资源可用性的降低，此类失败在未成功执行中所占比例持续上升，尤其在低资源语言中表现最为明显。为验证该分类体系的实用性，论文提出TART（Taxonomy-Guided Actionable Representation），通过显式地将分类体系的关键要素注入规划器及下游子智能体，增强其对跨语言语义对齐的理解与处理能力。在多种语言、三种大语言模型（LLM）架构、两个数据集和两种代理配置下的实验表明，TART具有稳定的性能提升效果；在多语言GAIA基准上，其使当前最优系统在11种语言（涵盖从低资源到高资源语言）上的平均准确率提升了5.6个百分点，证明了该分类体系在缓解跨语言规划失效问题上的有效性。
链接: https://arxiv.org/abs/2608.03735
作者: Vikas Pahuja,Jonathan Brokman,Omer Hofman,Tamir Nizri,Daniel Vishna,Seraphina Goldfarb-Tarrant,Kelly Marchisio,Hisashi Kojima,Roman Vainshtein
机构: Fujitsu Research of Europe; Cohere; Fujitsu Research
类目: Multiagent Systems (cs.MA); Computation and Language (cs.CL)
备注: 22 pages, 11 figures
Abstract:Multilingual multi-agent systems exhibit substantial degradation beyond English, yet prior work rarely identifies how task-critical information is lost when user requests are converted into executable plans. We study the planner in a multi-agent system as the request-to-action interface and derive an actionable taxonomy of planning-grounding failures from failed real-world task executions. LLM-based analysis shows that these failures constitute an increasing share of unsuccessful executions as language-resource availability declines, with the strongest effects in low-resource languages. To test whether the taxonomy supports mitigation, we introduce TART, Taxonomy-Guided Actionable Representation, that makes the taxonomy’s key aspects explicit to the planner and downstream sub-agents. Across multiple languages, three LLM backbones, two datasets, and two agentic configurations, TART consistently improves performance. On multilingual GAIA, it raises a state-of-the-art system’s accuracy by 5.6 percentage points averaged across eleven languages spanning low- to high-resource settings.
[MA-4] Group Perspective Matters: Regulating Debate Relationships Can Mitigate Blind Conformity in Multi-Agent Debate
【速读】：该论文旨在解决多智能体辩论（Multi-Agent Debate, MAD）框架中大型语言模型（Large Language Models, LLMs）存在的盲从现象问题。现有基于个体评估的方法（如置信度或困惑度）难以准确反映推理过程的正确性，甚至可能加剧模型间的盲从行为。为此，论文提出从群体交互视角重构评估机制，将LLMs之间的相互引用关系定义为辩论关系（Debate Relationships），并指出调控这些关系是缓解盲从的关键。为此，作者提出一种新颖的动态调节辩论关系（Dynamically Regulating Debate Relationships, DEAR） 框架。DEAR通过三个阶段实现：1）感知群体咨询倾向与不确定性；2）引入选择强化学习智能体（Selection RL-Agent），动态选择参考同伴；3）采用行为强化学习智能体（Behavior RL-Agent），自适应调整生成行为。两个RL智能体的执行被建模为序列决策过程，并通过多智能体强化学习联合优化。实验表明，DEAR在显著降低令牌消耗的同时，实现了更优的推理性能。
链接: https://arxiv.org/abs/2608.03648
作者: Hao Wu,Shoucheng Song,Chang Yao,Haoyu Wang,Huaiyu Wan,Youfang Lin,Kai Lv
机构: Beijing Jiaotong University (北京交通大学); Beijing Key Laboratory of Traffic Data Mining and Embodied Intelligence (北京市交通数据挖掘与具身智能重点实验室)
类目: Multiagent Systems (cs.MA)
备注:
Abstract:Multi-Agent Debate (MAD) improves the reasoning performance of Large Language Models (LLMs) through multi-round interaction. However, LLMs in MAD are highly susceptible to blind conformity. Existing individual evaluation methods, typically based on confidence or perplexity, fail to reflect the correctness of reasoning and may even exacerbate blind conformity. To address this, we shift the perspective from individual evaluation to group interaction. We define mutual referencing among LLMs as \textbfDebate Relationships and recognize that regulating these relationships is the key to mitigating blind conformity. In this paper, we propose a novel framework for \textbfDynamically r\textbfEgulating deb\textbfAte \textbfRelationships (DEAR) from the group perspective. At first, DEAR quantifies consensus and divergence as \textitgroup evidence to capture the debate state. Then, DEAR operates through three stages: 1) What: perceiving group consultation tendency and uncertainty; 2) Who: introducing a Selection RL-Agent to dynamically select reference peers; and 3) How: adopting a Behavior RL-Agent to adaptively adjust generation behaviors. Notably, we formulate the execution of the two RL-Agents as a sequential decision-making process, jointly optimizing via multi-agent reinforcement learning. Extensive experiments demonstrate that DEAR achieves superior performance while significantly reducing token consumption.
[MA-5] Is Inter-Seed Cross-Play Enough? Evaluating the Robustness of Zero-Shot Coordination Algorithms to Implementation Details
【速读】：该论文旨在解决生成式 AI（Generative AI）在真实场景中部署时，如何实现与未见过的人类或其他 AI 代理的零样本协调（Zero-shot Coordination, ZSC）的问题。其核心挑战在于现有 ZSC 算法的评估方法缺乏充分的鲁棒性验证：以往研究通常仅依赖单一实现版本在不同随机种子下的表现，或简单改变神经网络架构，未能充分反映不同团队对同一规范理解与实现差异所带来的影响。为系统评估这一问题，本文提出一种新的评估范式——跨实现交叉对弈（cross-implementation cross-play），通过系统性地改变多智能体强化学习（MARL）算法中已被证明会影响性能的实现细节，对主流 ZSC 算法 Other-Play 进行全面测试。结果表明，尽管存在实现差异，Other-Play 在标准 ZSC 评估中的表现仍可作为更严格跨实现评估的有效近似，显示出该算法在面对规范模糊性和实现变异性时具有较强的稳健性。
链接: https://arxiv.org/abs/2608.03644
作者: Maksymilian Wolski,Nicholas Hoernle,Johannes Forkel,Jakob Foerster
机构: ProrokLab(普罗罗克实验室); QuantCo(量化公司); University of Oxford(牛津大学); University of Cambridge(剑桥大学)
类目: Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注:
Abstract:AI agents deployed in real-world settings must be capable of coordinating with humans and other AI agents they have not encountered before. Zero-shot coordination (ZSC) algorithms aim to achieve this by specifying high-level learning rules such that independently engineered agents can coordinate with each other at test time. Rigorous evaluation of ZSC algorithms remains difficult: ideally, multiple independent implementations of each proposed algorithm must be used, reflecting the variation that arises when independent parties interpret and implement the same specification. In practice, however, ZSC algorithms have almost exclusively been evaluated using a single implementation trained across different random seeds, with only a handful of works additionally varying the neural network architecture. This leaves open questions about robustness to specification ambiguities and implementation details. In this work, we provide the first systematic evaluation of this robustness. We introduce a new evaluation scheme, cross-implementation cross-play, varying implementation details that prior work has shown to affect the performance of multi-agent reinforcement learning (MARL) algorithms, and we evaluate Other-Play, a popular ZSC algorithm, with this scheme. Our findings are encouraging and suggest that, for Other-Play, the standard ZSC evaluation is, in fact, a reasonable proxy for this more thorough cross-implementation evaluation.
[MA-6] Adversarial Fast-Moving Real-World Domains as Test Beds for Benchmarking AI Scientist Capabilities ICML2026
【速读】：该论文旨在解决当前人工智能科学家在生成新颖科学创意方面缺乏有效评估基准的问题。现有基准多依赖于合成任务或事后回溯性目标，易受模型先前接触过相关知识的影响，导致评估结果不可靠。为此，论文提出一种基于真实世界、复杂且动态变化的对抗性领域的评估框架，以更真实地检验AI科学家所需的核心能力，包括推理能力、创新性及假设构建能力。研究在两个结构迥异的真实领域中验证该框架：一级方程式赛车（F1）中，模型需为2026赛季设计赛车概念，其输出与实际季前创新成果进行对比；魔法规则集（MTG）中，模型基于最新卡池提出套牌，并与职业巡回赛（PT）的19套冠军套牌进行比对。实验结果显示，尽管模型能生成大量合理创意，但极少与真实专家方案一致。在F1任务中，表现最佳的GPT-5.2仅匹配10项真实创新（共40项），从166个提案中产生；在MTG任务中，最佳模型Gemini 3 Flash所选套牌复现了第三名PT套牌中7张新卡中的5张，且模型高频选择的卡牌与PT套牌广泛采用的卡牌高度相关（斯皮尔曼等级相关系数ρ = 0.74，p = 0.0003）。上述结果表明，当前AI科学家在创意生成方面已具备一定能力，其核心能力瓶颈并非创意生成本身，而在于对创意的筛选、优先级排序以及实现连贯性创新的能力。
链接: https://arxiv.org/abs/2608.03569
作者: William Bolton,Philip Torr
机构: 未知
类目: Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Machine Learning (cs.LG); Multiagent Systems (cs.MA)
备注: Accepted at the AI for Science workshop at ICML 2026. 14 pages, 11 figures
Abstract:Benchmarking the ability of AI scientists to generate novel ideas is notoriously difficult. Existing benchmarks in this field have made progress in evaluating scientific reasoning and research replication, but often rely on synthetic tasks or retrospective targets, which may be confounded by prior exposure. We hypothesize that complex, adversarial, fast-moving real-world domains where expert practitioners independently generate observable outputs can provide a practical solution to fill this gap and evaluate the capabilities needed for AI scientists, including reasoning, novelty, and hypothesis formulation. We instantiate this framework in two structurally different domains, Formula 1 (F1), where models ideate around car design concepts for the 2026 season, and real pre-season innovations provide a ground truth, and Magic: The Gathering (MTG), where models propose decks from a recently updated card pool and are evaluated against 19 Pro Tour (PT) decklists. Across both domains, models produce plausible outputs, but few align with real-world expert solutions. In F1, the best model, GPT-5.2 matched 10 of 40 real innovations with 166 ideas proposed across runs. In MTG, the best deck from Gemini 3 Flash recovered 5 of 7 new-set cards from the third-place PT deck, and across all 108 decks, the cards models selected most often were also the cards most widely adopted by PT decks (Spearman \rho = 0.74 , p = 0.0003 ). These results suggest that a key capability gap for AI scientists is not idea generation, but filtering, prioritization, and coherent novelty.
[MA-7] Hybrid LLM -Augmented Reinforcement Learning Agents for Complex Sequential Decision Tasks
【速读】：该论文旨在解决基于大语言模型（Large Language Model, LLM）的智能体在长时序序列决策任务中面临的挑战，即难以实现精确的动作优化与环境交互。尽管LLM具备强大的高层抽象与任务分解能力，但在低层级动作选择上缺乏精细化控制；而传统的强化学习（Reinforcement Learning, RL）虽能有效进行序列控制，却难以处理复杂场景中的高阶规划需求。为此，论文提出一种融合LLM与强化学习的混合智能体架构：利用LLM生成子目标、结构化计划及上下文引导信息，同时由RL代理通过与环境的实际交互对底层动作进行精细化优化。该方案的关键在于构建“高层规划—低层执行”的协同机制，使LLM负责任务分解与策略生成，而RL则专注于动作层面的微调与适应性学习。实验结果表明，相较于纯RL或纯LLM方法，该混合框架显著提升了样本效率、成功率以及动作轨迹的一致性，验证了其在构建更强大自主系统方面的潜力。
链接: https://arxiv.org/abs/2608.03502
作者: Christophe D. Hounwanou,John Emeka Eze,Yaé Ulrich Gaba
机构: 未知
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Multiagent Systems (cs.MA)
备注: 16 pages, 12 figures
Abstract:Large Language Models (LLMs) have recently shown strong capabilities in reasoning, planning, and tool-use, enabling new forms of autonomous agents. However, LLM-based agents struggle with long-horizon sequential decision tasks that require precise action optimization and environment interaction. Reinforcement Learning (RL), while effective for sequential control, often lacks the high-level abstraction and task decomposition abilities needed for complex scenarios. This paper introduces an LLM-Augmented Reinforcement Learning Agent that integrates LLM-driven planning with RL-based action optimization. The proposed architecture leverages the LLM to generate subgoals, structured plans, and contextual guidance, while the RL agent refines low-level actions through interaction with the environment. Experiments on sequential decision tasks demonstrate improved sample efficiency, higher success rates, and more coherent action trajectories compared to RL-only and LLM-only baselines. This hybrid paradigm highlights a promising direction for building more capable autonomous systems.
[MA-8] When Truth Is Distributed: Misinformation Derails Collective Fact Recovery in LLM -Based Multi-Agent Systems
【速读】：该论文旨在解决基于大语言模型（LLM）的多智能体系统在协作推理过程中，因信息传播机制导致局部错误被放大并演变为集体风险的问题。现有评估方法过度关注最终结果，忽视了分布式信息聚合过程中的可靠性与错误传播动态。其解决方案的关键在于提出一种名为Hi-Agreement的受控评估框架，通过严格配对“全诚实协作”与“关键证据持有者实施欺骗”的对比实验条件，结合多阶段投票、证词采纳及证据根溯源链（evidence-root lineage propagation）分析，揭示错误信息在系统内的传播路径与影响机制。实验结果表明，在存在单一虚假证词的情况下，整体真值恢复率从72.50%骤降至14.17%，且所有系统均出现显著下降；过程追踪与退出消融分析进一步显示，虚假证词比真实证词更易被采纳，并在智能体间持续传播，即使欺骗者退出后仍能影响后续决策。此外，无直接证据的观察者虽能抑制错误共识形成，但无法提升真值恢复率。上述发现共同揭示了分布式事实恢复的脆弱性及其内在机制：虚假证据一旦进入通信网络，便通过被采纳与持续传播而获得集体影响力，从而危及系统整体可信性。
链接: https://arxiv.org/abs/2608.03421
作者: Chenfei Yan,Zeyang Yue,Feifei Zhao,Erliang Lin,Lu Jia,Haibo Tong,Mingyang Lyu,Chengyi Sun,Yi Zeng
机构: OpenAI(OpenAI); DeepSeek-AI(DeepSeek-AI); SpaceXAI(SpaceXAI)
类目: Multiagent Systems (cs.MA)
备注:
Abstract:LLM-based multi-agent systems promise effective collaborative reasoning, but communication may amplify local errors into collective risks. Existing evaluations emphasize final outcomes, leaving the reliability and propagation dynamics of distributed information aggregation unclear. We introduce Hi-Agreement, a controlled evaluation framework that strictly pairs all-honest collaboration with controlled deception by a key evidence holder and analyzes the aggregation process through multi-stage voting, testimony adoption, and evidence-root lineage propagation. Using 120 five-agent object-movement environments where partial observations jointly determine a unique endpoint, we evaluate 3 homogeneous LLM-based multi-agent systems. Across these paired conditions, aggregate truth recovery falls from 72.50% to 14.17%, with significant declines for every system. Process tracing and exit ablations show that a single false testimony is adopted more readily than truthful testimony, propagates to higher orders, and persists through honest agents after the deceiver exits. Observers without first-hand evidence suppress incorrect consensus but do not improve truth recovery. Together, these findings reveal both the fragility of distributed fact recovery and its underlying mechanism: false evidence gains collective influence through its adoption and continued propagation by other agents after entering communication.
[MA-9] Attacking and Defending Multi-Agent Collaborative Filtering Systems Through Connectivity RECSYS’26
【速读】：该论文旨在解决生成式多智能体协同过滤（agent-based collaborative filtering, CF）系统在自然语言交互驱动下因数据依赖性与多智能体交互机制共同引入的新型安全脆弱性问题，尤其关注系统连通性结构对攻击与防御效果的调控作用。其核心挑战在于：传统协同过滤系统的漏洞分析难以直接适用于由大语言模型（LLM）驱动的自主用户与物品智能体构成的动态、交互式推荐环境，且此类系统中的攻击行为可能通过智能体间的传播路径呈现出非单调、非线性的演化特性。解决方案的关键在于构建一个可系统化调节连通性的实验框架——AgentCF，从两个维度量化连通性特征：(i) 候选数量（每轮交互中每个用户的候选物品数，衡量用户侧交互密度），以及 (ii) 项目目录集中度（跨用户之间的物品目录重叠程度）。通过在此框架下适配并评估来自通用多智能体系统（MAS）领域的攻防策略，研究揭示了用户与物品智能体在攻防角色中的不对称性、攻击有效性随时间呈现非单调变化规律，以及针对信息扩散与信息提取两类攻击目标所表现出的不同响应模式。此外，研究探索性地引入基于流行病传播模型的静态指标，用于预测不同配置下的攻击风险，为实现低成本、高效的鲁棒性评估提供新思路。
链接: https://arxiv.org/abs/2608.03272
作者: Anjun Hu,Hanting Xie,Saranya Govindan,Jas Kandola,Kurt Cutajar
机构: University of Oxford(牛津大学); Amazon(亚马逊)
类目: Information Retrieval (cs.IR); Cryptography and Security (cs.CR); Multiagent Systems (cs.MA); Social and Information Networks (cs.SI)
备注: 10 pages, 10 figures, 20th ACM Conference on Recommender Systems (RecSys '26)
Abstract:Multi-agent collaborative filtering (CF) systems coordinate autonomous LLM-powered user and item agents through natural-language interaction to refine preferences and generate recommendations. These systems inherit vulnerabilities from both their data-driven nature and their multi-agent interactions, which manifest in distinct ways. Understanding how connectivity modulates vulnerability in these systems could facilitate the development of more robust recommendation pipelines. In this work, we adapt attacks and defenses from the general multi-agent systems (MAS) literature to the agent-based CF setting, evaluating them under systematically varied connectivity in the AgentCF framework, where CF connectivity is characterized along two axes: (i) candidate count (the number of item candidates per turn per user, measuring user-side interaction density) and (ii) catalog concentration (the degree of item catalog overlap across users). Our contributions include: (1) Adaptation: we reproduce MAS-inspired attacks and defenses in the agentic CF domain, confirming partial transferability of original observations. (2) Characterization: we characterize how the two aspects of connectivity shape attack and defense outcomes, revealing role asymmetries between user and item agents, non-monotonic temporal dynamics in attack efficacy, and divergent patterns across dissemination and extraction attack goals. Additionally, as an exploratory extension, we assess the applicability of epidemic-inspired static metrics in ranking CF configurations by expected attack outcome, potentially enabling cost-efficient robustness assessment. Implementation is available at this https URL Comments: 10 pages, 10 figures, 20th ACM Conference on Recommender Systems (RecSys '26) Subjects: Information Retrieval (cs.IR); Cryptography and Security (cs.CR); Multiagent Systems (cs.MA); Social and Information Networks (cs.SI) ACMclasses: H.3.3; I.2.11; K.6.5 Cite as: arXiv:2608.03272 [cs.IR] (or arXiv:2608.03272v1 [cs.IR] for this version) https://doi.org/10.48550/arXiv.2608.03272 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[MA-10] Internalising the Identity Primitive: Cryptographic Individuality for an Autonomous Agent on a Public Blockchain
【速读】：该论文旨在解决公共区块链上软件代理（software agent）在累积权威与经济利益后，如何界定其作为“个体”的身份认同问题。核心挑战在于，传统身份绑定机制依赖硬件、操作者或封装层的信任，存在中心化风险且难以在去中心化环境中验证。其解决方案的关键在于将身份绑定的可信根从外部实体转移至由密码学假设保障的固定实现（pinned implementation），即通过将代理的神经网络权重（neural-network weights）设计为私钥的确定性函数，并在创世时以零知识方式承诺该绑定关系，在每次状态迁移时重新验证并由代理签名至链上不可分叉的历史中。这一机制确保了身份的密码学不变性（cryptographic invariance），且仅依赖于可验证的密码学假设，而无需信任运行环境、预言机或底层软件栈。此外，通过引入基于密钥派生的经济账户周期性消耗成本，进一步施加了经济可行性约束，形成“密钥-历史-经济”三元约束。实证结果表明，该代理在Solana测试网实现了2.36天的链上连续运行，仅两次主机侧恢复，无状态迁移被拒绝，且验证开销可控；替换底层实现被链上拒绝，不同密钥代理行为分离，相同密钥控制组保持一致。该工作首次实现了一种可每状态迁移重验的、基于密码学不变性的链上代理身份，符合Suzuki（2026）提出的“人工外在性”（Artificial Externality）框架中的密码学个体性（cryptographic individuality）理念。
链接: https://arxiv.org/abs/2608.02986
作者: Keisuke Suzuki
机构: Hokkaido University (北海道大学)
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注: 52 pages, 3 figures, 11 tables. Cryptographic key-to-weights binding (W = HKDF(sk) inside Groth16) with an on-chain state-commitment chain on Solana devnet; active-query, homeostatic, and economic-metabolism extensions; 166- and 168-cycle continuous runs. Code, threat model, and per-cycle telemetry: this https URL (tag arxiv-v1)
Abstract:A software agent on a public blockchain accumulates authority and economic stakes, raising the engineering question of what makes it count as an individual. The paper’s central contribution is a shift of trust root for the key-to-weights binding of agent identity: from hardware, operator, or wrapper trust to cryptographic assumptions enforced by a pinned implementation (liveness, key custody, oracle trust, and the underlying software stack remain external). We design and deploy on Solana devnet an agent whose neural-network weights are a deterministic function of its private key. The binding is committed in zero knowledge at genesis, re-checked against that commitment at every state transition, and signed by the agent into an on-chain history unforkable once finalized; in a PoC-tier extension, a protocol-imposed metabolic cost is debited each cycle from a key-derived economic account, adding a consumption-side economic-viability constraint to the key-history-economy triple. Empirically, the agent completes a 2.36-day on-chain run with two host-side resumptions but no rejected transition, at bounded per-transition verification cost; a substituted substrate is rejected on chain, and independently keyed agents diverge as predicted while a same-key control stays at zero. To our knowledge, this is the first published on-chain agent whose identity primitive is itself a cryptographic invariant re-checked at every state transition. The resulting transition-time invariant instantiates the cryptographic individuality proposed by Suzuki 2026’s Artificial Externality framework.
[MA-11] SABRE: A Multi-Agent Approach for Selecting Out-of-Distribution Detectors Under a Budget
【速读】：该论文旨在解决视觉-语言模型在后处理阶段（post-hoc）进行分布外（out-of-distribution, OOD）检测时存在的可靠性问题：现有方法假设在基准数据集上表现优异的检测器在实际部署后仍保持可靠，但研究发现这一假设在跨域场景下失效——同一冻结编码器上，最优检测器在不同数据域中可能反转，甚至将真实分布内样本误判为异常，且最佳检测器随域变化，因此固定选择单一检测器不可靠。其解决方案的关键在于提出SABRE（Selective Agentic Budgeted Reliability Ensemble），一种在推理阶段基于运行环境（operating regime）动态选择检测器的自适应框架。该框架引入三个语言模型代理：选择器（Selector）负责按有限查询预算挑选下一检测器，报告者（Reporter）整合各检测器的证据，分析者（Analyst）利用部署域中预留的小规模标注样本对检测器可靠性进行校准，实现无标签观测下的权重分配。通过一个包含四种新型多模态密度检测器的库，SABRE能够从数据中推断运行模式，实时追踪各域中的最强检测器，避免传统方法因检测器反转导致的误判，并在可靠检测器存在时收敛至其性能。组件分析表明，报告者的反馈带来稳定增益，而分析者的校准机制是防止反转的核心，可有效排除不可靠检测器，避免其与可靠检测器的输出相互抵消。因此，该工作强调：由于无法依赖跨域通用的固定规则，模型可靠性必须在部署时动态建立，而SABRE实现了这一目标的自动化。
链接: https://arxiv.org/abs/2608.02959
作者: Mary Wisell,Salimeh Sekeh
机构: 未知
类目: Multiagent Systems (cs.MA)
备注:
Abstract:Post-hoc out-of-distribution (OOD) detection for vision-language models assumes that a detector chosen on a benchmark stays reliable once deployed. We show this fails across domains: on a single frozen encoder, a detector that leads in one domain can invert in another, scoring in-distribution inputs as more anomalous than genuine outliers, and the best detector changes from domain to domain, so no fixed choice is reliable throughout. We introduce SABRE (Selective Agentic Budgeted Reliability Ensemble,) which replaces this fixed choice with per-regime selection at inference. Three language-model agents reason over a library of post-hoc detectors under a bounded query budget: a Selector chooses which detector to consult next, a Reporter consolidates the evidence for each input, and an Analyst calibrates detector reliability on a small labeled sample held out from the deployment domain and disjoint from the test data, weighting selection and aggregation without ever observing a scored input’s label. The library includes four multimodal density detectors we propose. Inferring the operating regime from data, SABRE tracks the strongest detector in each domain without prior knowledge of it, recovering reliable detection where a conventional detector inverts and converging to that detector where it is sound. A component analysis shows the agents are complementary: the Reporter’s feedback yields consistent gains, and the Analyst’s calibration is decisive against inversion, ruling out unreliable detectors so that aggregation no longer cancels the sound ones. Since no fixed rule can be trusted across domains, reliability must be established at deployment rather than assumed from a benchmark, and SABRE shows this can be done automatically.
[MA-12] Emergence of Biased Consensus in Multi-Agent LLM Debates ICML2026
【速读】：该论文旨在解决多智能体大语言模型（Multi-agent LLM）辩论在决策与问题求解任务中虽表现优异，但其安全性和公平性风险尚不明确的问题，尤其关注交互过程如何放大单个大语言模型（LLM）的偏见，进而威胁实际应用。其核心解决方案的关键在于揭示多智能体辩论中集体偏见（collective bias）的涌现机制，并提出一个受物理社会动力学理论启发的分析框架。该框架预测：当群体一致性（conformity）超过由初始偏见和辩论噪声共同决定的临界阈值时，系统将发生相变（phase transition），导致集体偏见的显著涌现。通过受控实验验证，观察到有限尺寸下的交叉现象，支持了相变的存在；同时发现，智能体异质性可通过平滑（rounding）这一相变过程抑制集体偏见的形成。研究进一步证明，这些发现可泛化至真实决策场景，如投资决策与大模型作为裁判（LLM-as-a-judge）评估任务，为提升多智能体系统的公平性与安全性提供了理论依据与干预路径。
链接: https://arxiv.org/abs/2608.02827
作者: Maya Okawa
机构: 未知
类目: Multiagent Systems (cs.MA)
备注: Accepted at the 43rd International Conference on Machine Learning (ICML 2026). 23 pages, 12 figures
Abstract:Multi-agent LLM debates achieve strong performance on decision-making tasks as well as problem-solving benchmarks, yet their safety and fairness risks remain poorly understood. Notably, interaction can amplify the biases of single LLMs, raising concerns for real-world deployment. We identify the emergence of collective (often biased) norms in multi-agent LLM debates and show that noise (e.g., LLM sampling temperature) is a key driver. To explain this, we propose an analytical framework drawing on physics-inspired theoretical models of social dynamics. We predict a phase transition to collective bias when conformity surpasses a critical threshold given the LLMs’ initial bias and debate noise. We test the theoretical predictions through controlled experiments and observe a finite-size crossover consistent with an underlying phase transition. We further find that agent heterogeneity suppresses emergence by smoothing (rounding) this transition. Finally, we show that these insights generalize to realistic decision-making tasks, including investment decisions and LLM-as-a-judge evaluation.
[MA-13] Stateful Governance for Concurrent Agent ic Systems
【速读】：该论文旨在解决生成式智能体（AI agents）在执行具有实际影响的操作（如退款、资源预留、云资源分配和金融转账）时，因状态滞后导致的授权失效问题。传统治理机制依赖请求时刻的静态信息进行决策，但在并发环境下，预算、库存、审批状态等关键政策状态可能在操作执行前发生变化，造成“过期授权”（stale authorization），从而引发合规风险。其核心解决方案是提出一种名为Provenact的运行时架构，通过引入“策略状态可序列化”（policy-state serializability）这一正确性约束，确保所有已提交的操作均可被解释为基于操作发生前瞬时政策状态所授权的行为。Provenact的关键创新在于将策略保持为可审查的程序代码，并通过协调策略状态与实际效应的执行顺序，实现对共享资源的动态、安全治理。实验表明，该系统能有效防止基线方法中因策略状态传递不完整而产生的过期授权，在无LLM参与的采购工作流中避免了策略违规，同时支持延迟审批并减少策略演进对可信代码的依赖，为智能体框架中的状态化治理提供了可扩展的技术路径。
链接: https://arxiv.org/abs/2608.02764
作者: Yuxiang Peng,Xiaodi Wu
机构: Purdue University (普渡大学); University of Maryland, College Park (马里兰大学学院公园分校)
类目: Multiagent Systems (cs.MA); Databases (cs.DB)
备注: 18 pages, 8 figures
Abstract:AI agents are moving from advisory interfaces into systems that execute consequential operations: issuing refunds, reserving scarce inventory, provisioning cloud resources, and initiating financial transfers. These workflows require governance over effects, not only over model outputs. Existing safeguards often decide whether an action is allowed from the information available when the action is requested. For stateful policies, that request-time view may be incomplete: budgets, inventory, approval status, and risk signals can change before the effect occurs, making an earlier authorization or approval stale. This paper studies stateful governance for concurrent agentic systems. We identify stale authorization as the core failure mode and define policy-state serializability, a correctness condition requiring committed effects to be explainable as authorized against the policy state immediately before they occur. We present Provenact, a runtime architecture that keeps policies as reviewable programs while coordinating the state and effects needed to preserve their decisions. In experiments with a PostgreSQL-backed prototype of Provenact, the system prevents stale authorizations missed by baselines that pass policy state as ordinary request context, preserves delayed approvals while unrelated work proceeds, keeps policy evolution mostly in policy text rather than trusted provider code, and avoids policy violations in a scripted, LLM-free procurement workflow where agent-governance baselines produce stale authorizations over shared budgets and inventory. More broadly, Provenact suggests a path for integrating stateful governance boundaries into agent frameworks and provider-backed domains where agents act on shared resources. Comments: 18 pages, 8 figures Subjects: Multiagent Systems (cs.MA); Databases (cs.DB) Cite as: arXiv:2608.02764 [cs.MA] (or arXiv:2608.02764v1 [cs.MA] for this version) https://doi.org/10.48550/arXiv.2608.02764 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[MA-14] Everyone Conforms No One Believes: Pluralistic Ignorance in LLM Agent Populations
【速读】：该论文旨在解决生成式人工智能（Generative AI）驱动的多智能体系统在模拟社会动态时，是否能够真实再现“多元无知”（pluralistic ignorance）这一关键社会现象的问题。多元无知指多数个体私下反对某一规范，却因误以为自己是唯一持异议者而公开顺从，从而导致规范持续存在。研究发现，尽管大语言模型（LLM）代理在多数场景中表现出高达64%至94%的公开顺从率，但这种顺从并非由指令直接驱动，而是模型内部涌现的特性。其解决方案的关键在于构建了一个涵盖10个领域与5个权威层级、共100个场景的基准测试体系，并对8个来自6家机构的主流模型进行评估。结果表明，模型选择本身成为影响模拟结果的重要未被识别的自由度；尤其值得注意的是，仅GPT-4o展现出显著的突破性能力（48%的范式转变成功率），其余模型几乎无法打破虚假共识。进一步通过提示组件消融实验发现，即使移除“虚假共识”框架和“融入群体”的目标，代理仍保持高顺从性（52%–92%），证明顺从行为为内在涌现现象。研究揭示，当前基于LLM的模拟可能系统性地高估社会规范的稳定性，忽略了现实中推动规范变革的脆弱临界点动态，因而对社会运动与制度变迁的建模存在根本性偏差。
链接: https://arxiv.org/abs/2608.02758
作者: Yashwanth YS
机构: Carnegie Mellon University (卡内基梅隆大学); OpenAI (OpenAI); Anthropic (Anthropic); Meta AI (Meta AI); DeepSeek AI (DeepSeek AI); Qwen Team (通义千问团队); Kimi Team (Kimi团队)
类目: Multiagent Systems (cs.MA)
备注:
Abstract:LLM-based multi-agent systems are increasingly used to simulate social dynamics, from opinion formation to collective decision-making. These simulations can reproduce certain social phenomena, but it is unknown whether they capture pluralistic ignorance, a state where a majority privately rejects a norm yet publicly conforms, each believing they are alone in dissenting. This phenomenon drives norm persistence, social movements, and political revolutions. We show that pluralistic ignorance emerges robustly in LLM agent populations. We construct a benchmark of 100 scenarios across 10 domains and 5 authority levels, grounded in the human pluralistic ignorance literature, and evaluate 8 models from 6 organizations. Agents publicly conform at rates of 64 to 94% despite privately opposing the norm. Conformity is domain-sensitive (workplace and social relationship scenarios produce near-universal compliance) and highly model-dependent, though uncorrelated with capability. We test whether a single “norm entrepreneur” can break the false consensus by publicly dissenting. For 7 of 8 models, cascades succeed less than 26% of the time, with one model showing zero cascades across all scenarios. GPT-4o is a notable outlier at 48%, revealing qualitatively distinct dynamics across model families. A prompt component ablation across all 8 models establishes that conformity is emergent rather than instruction-driven: removing both the false-consensus framing and fit-in goal reduces conformity but does not eliminate it (52 to 92% in the minimal condition). Our findings identify model selection as an unacknowledged degree of freedom that fundamentally shapes simulation outcomes. More broadly, the near-absence of cascades suggests LLM simulations may systematically overestimate the stability of social norms, missing the fragile tipping-point dynamics that drive real-world norm change in human societies.
[MA-15] When Policies Change Probabilities: Modular Decision-Making for LLM Code Review
【速读】：该论文旨在解决大语言模型（Large Language Model, LLM）代码审查系统中风险评估与决策行动之间耦合过紧的问题，即当前多数审查接口在单一提示（prompt）中同时完成风险概率估计与是否通过补丁的决策，导致成本敏感性（cost sensitivity）未能有效体现。其核心问题是：在相同证据条件下，不同成本政策（如高误接受成本）应影响最终决策，但现有系统中概率估计受政策干扰，造成判断失准。解决方案的关键在于构建一个模块化流水线，将风险评估（risk elicitation）与成本决策（cost application）解耦——该流水线首先独立获取无策略信息的风险评分，结合外部监控器（monitor）提供的独立证据，再在代码层面应用具体成本规则。实验表明，该方法在等成本下可使平均概率准确性提升，并在相同成本下将每问题平均损失降低0.073；在10:1高误接受成本下，系统拒绝所有补丁且性能匹配“全拒”策略，验证了风险评估与动作决策分离的有效性。这表明，概率估计本身存在偏差，需独立评估风险、外部证据与下游策略，以实现更可靠、可解释的自动化代码审查。
链接: https://arxiv.org/abs/2608.02677
作者: Rasvik Kudum,Max Corbett,Hitansh Paliwal,Romaisa Fatima,Thomas Jiralerspong,Sneheel Sarangi
机构: 未知
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注: 20 pages, 6 figures; includes technical appendices. Code and data: this https URL
Abstract:LLM code reviewers often estimate patch risk and make approval decisions in one prompt. A probability should depend on evidence; costs should determine the action taken from it. We test whether four deployed reviewer interfaces preserve this separation using 15,792 responses on 720 candidate patches, with one that passed and one that failed an archived test harness for each of 360 repository issues. In matched calls with the patch and monitor evidence fixed, replacing an equal-cost policy with a 10:1 false-accept policy changes reported failure probabilities by 13.6 to 16.9 percentage points on average. For every reviewer, the actions returned under the high-cost prompt are worse than rejecting all patches. Applying the same high-cost rule to probabilities elicited under equal costs reduces loss for all four systems, showing that probability elicitation itself contributes to the excess loss. We also evaluate a modular pipeline that elicits risk without policy information, combines an independent monitor score, and applies costs in code. Relative to calibrated reviewer-only scores, the pipeline improves average probability accuracy and, at equal costs, reduces mean loss by .073 per issue while accepting 58 to 68% of patches. At 10:1, it accepts none and matches reject-all. Downstream policy can therefore change the probability it is meant to use, motivating separate evaluation of risk, outside evidence, and action.
[MA-16] Axiomatic shared-medium coordination for stigmergic systems
【速读】：该论文旨在解决多介质环境下基于刺激性协调（stigmergic coordination）的系统行为可比性问题，尤其关注不同介质中响应行为的一致性与抽象化表达。其核心挑战在于如何在不依赖具体介质实现的前提下，建立一个通用的、可形式化的比较框架，以保证系统在不同抽象层次下的行为等价性。解决方案的关键在于提出一种介质无关的比较层（medium-agnostic comparison layer），其比较对象为抽象使能响应签名（abstract enabled-response signature），通过该签名捕捉系统响应的核心语义特征。研究进一步证明：对元数据进行粗粒度划分时，仅当每个商纤维（quotient fiber）均满足响应对齐（response-aligned）条件时，该划分才是响应充分的（response-adequate）；若不满足，则存在唯一的最小响应充分修复，且任何响应充分修复必须至少保留原始纤维中的响应类数量。在动态行为方面，研究揭示了商兼容的匹配写操作（quotient-compatible matched writes）可保持一步行为不变性，而基于新鲜性条件的动作则构成一步不可提升性的障碍，但匹配后的一步结果可扩展至有限串行轨迹。该理论框架在元组空间介质和带时间戳的虚拟刺激性数据空间中得到实例化，最终构建了一个具有明确正/负定理循环的介质无关比较体系，尽管路径相关与并发语义仍超出现有研究范围。
链接: https://arxiv.org/abs/2608.02619
作者: Fernando Paredes García
机构: Independent Researcher(独立研究员)
类目: Distributed, Parallel, and Cluster Computing (cs.DC); Logic in Computer Science (cs.LO); Multiagent Systems (cs.MA); Optimization and Control (math.OC)
备注: 34 pages; mathematical foundations paper; no figures. Also archived at Zenodo DOI: this https URL
Abstract:Stigmergic coordination has many medium-specific formalizations. This paper formulates a medium-agnostic comparison layer whose comparison object is the abstract enabled-response signature. For metadata refinements, we prove that a coarse quotient is response-adequate exactly when each quotient fiber is response-aligned. When that fails, there is a canonical coarsest response-adequate repair, and every response-adequate repair must retain at least the corresponding fiberwise number of response classes. On the dynamic side, quotient-compatible matched writes preserve one-step behavior, freshness-conditional actions yield a one-step non-liftability obstruction, and the matched one-step result lifts to finite serial traces. These results are instantiated for a tuple-space medium and a timestamped virtual-stigmergic dataspace. The result is a medium-agnostic comparison framework with one explicit positive/negative theorem cycle; richer pathwise and concurrent semantics remain outside scope.
[MA-17] MemArena: An Ego-Centric Benchmark for On-Device Agent ic Personal Memory Assistants at Scale
【速读】：该论文旨在解决边缘部署的个人记忆助手在设备端处理私密人际对话时面临的挑战，尤其针对高密度交互、以自我为中心视角以及跨会话连贯性等关键需求在现有评测基准中被低估的问题。其核心解决方案是构建了一个基于MASim智能体模拟器的单一世界对话评测基准MemArena，涵盖50个智能体在15天内的交互数据（共1030万对话文本标记，每智能体每日24.1千个仅文本观察标记），并通过交互历史联合生成六维评估维度的真实标签（包括回忆、推理与可信度等）。研究评估了五种开放权重阅读器搭配不同记忆后端（Vanilla上下文、BM25-RAG、Oracle检索、Memobase、MemSearch）的表现，关键发现包括：（1）记忆后端的选择对内容准确性影响显著，例如在Qwen3-0.6B模型上，从Memobase切换至MemSearch可提升32.5/19.2个百分点，优于单纯扩大阅读器规模带来的收益（+10.6/+6.8个百分点）；（2）权限感知访问机制普遍失效，无论使用何种后端均存在严重信息泄露或过度保守披露问题；（3）搜索延迟仅在极小规模阅读器场景下构成瓶颈，在Spark GB10边缘节点上，各类后端的内存搜索平均增加87/7/48毫秒（对应BM25-RAG/Memobase/MemSearch），且该开销占总时间到首次响应时间（TTFT）比例较小。该工作为边缘侧个性化记忆系统提供了更真实、全面的评测框架与技术验证路径。
链接: https://arxiv.org/abs/2608.02613
作者: Jiadong Zhang,Xiaosong Ma
机构: MBZUAI
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Multiagent Systems (cs.MA)
备注: 48 pages, 6 figures
Abstract:Edge-deployed personal memory assistants must handle private interpersonal conversations on-device with open-weight models. Yet, existing memory benchmarks often under-test the combination of activity-dense interaction, ego-centric perspective, and coherent multi-session worlds. MemArena fills these gaps with a single-world conversational benchmark built with its MASim agent simulator, for 50 agents over 15 days (10.3M dialog-text tokens, 24.1K text-only ego-observed tokens/agent/day). With the interaction history, it co-generates ground truth over six recall, reasoning, and trustworthiness evaluation dimensions. We evaluate five open-weight readers with Vanilla context, BM25-RAG, Oracle retrieval, Memobase, and MemSearch as memory backends. Three results stand out: (1) Memory-backend choice matters more for content accuracy: At Qwen3-0.6B, Memobase-to-MemSearch gains +32.5/+19.2 pp, exceeding MemSearch reader scaling (+10.6/+6.8 pp). (2) Permission-aware access fails universally, with Oracle leaking heavily and other backends too timid to disclose. (3) Search latency bites only at very small reader: on a Spark GB10 edge node, memory-search adds a moderate and fixed 87/7/48 ms (BM25-RAG/Memobase/MemSearch) that composes a small part of TTFT for most reader-backend combinations. Code, the MASim simulator, and the MemArena-L benchmark will be released upon acceptance.
自然语言处理
[NLP-0] ParVL: Parallel Scaling and Expandable Compute Allocation for Multimodal LLM s
【速读】： 该论文旨在解决多模态大语言模型（Multimodal Large Language Models, MLLMs）现有扩展策略中存在的两大核心问题：一是传统方法通过增加模型参数或序列推理计算量进行扩展，导致显著的内存占用或延迟开销；二是多数现有方法无法动态调整视觉编码器（Vision Transformer, ViT）与语言解码器（Large Language Model, LLM）之间的固定计算分配，限制了针对特定任务的优化能力。为此，论文提出一种并行多模态计算框架——并行视觉-语言（Parallel Vision-Language, ParVL），其关键创新在于在固定主干参数预算下，通过复用现有的ViT和LLM主干参数，在多个视觉与语言分支间实现并行计算扩展。具体而言，每个并行计算流通过分支特异性前缀参数（branch-specific prefix parameters）在共享主干上构建，整体模型通过约130亿词元的全参数监督微调实现端到端训练。该框架系统性地探索了ViT编码器与LLM解码器间的计算分配权衡，实验表明，ParVL在相同配置的单分支基线基础上显著提升了多模态性能，且最优的视觉-语言计算分配比例因任务而异，体现了其任务自适应的灵活性。
链接: https://arxiv.org/abs/2608.04010
作者: Yang Yang,Qinyu Zhao,Mouxiang Chen,Xiaohui Li,Lixin Gu,Wenhai Wang,Hongjie Zhang,Wenwei Zhang
机构: Australian National University (澳大利亚国立大学); Tsinghua University (清华大学); University of Chinese Academy of Sciences (中国科学院大学); Peking University (北京大学); Alibaba Group (阿里巴巴集团)
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注: 14 pages, 4 figures
Abstract:Existing scaling strategies for Multimodal Large Language Models (MLLMs) typically expand either model parameters or sequential inference computation, incurring substantial memory or latency overhead. More importantly, most existing methods fail to alter the rigid, fixed computation allocation between the Vision Transformer and the Large Language Model components, limiting task-specific optimization. To address this, we introduce the Parallel Vision-Language (ParVL) scaling framework for MLLMs, which scales parallel computation by reusing the existing ViT and LLM backbone parameters across multiple vision and language branches. This framework raises a central question: given a fixed backbone parameter budget, how should additional shared-backbone computation be allocated between the vision and language modalities? We instantiate each parallel computational stream with branch-specific prefix parameters over a shared backbone, and train the entire model end-to-end via full-parameter supervised fine-tuning on roughly 13B tokens. We systematically study the computation-allocation trade-off between the ViT encoder and LLM decoder. ParVL improves overall multimodal performance over same-recipe single-branch baselines, and the best evaluated vision–language allocation varies across tasks. Code is available at this https URL.
[NLP-1] SocietyBench: Forecasting Counterfactual Social-World Evolution
【速读】： 该论文旨在解决当前大型语言模型（LLM）评估体系中对社会事件理解与预测能力的忽视问题，即现有基准主要关注模型完成具体任务的能力（如修复代码、操作图形界面），而缺乏对模型在真实社会事件演化过程中理解因果逻辑、把握公众舆论动态及进行时序预测等“社交智能”维度的系统性衡量。其解决方案的关键在于提出SocietyBench——一个端到端的社会事件预测基准，通过从五大平台采集新闻与社交媒体内容，构建时间索引化的事实事件层与公众意见层分离的结构化时间线，并将每个时间截点转化为经过审计的预测题库；所有问题在两个正交的100分量表上评分：概率校准度（calibration）与时间准确性（temporal accuracy）。为防止模型依赖预训练记忆中的表面标签，研究采用三阶段去标识化流程，对命名实体进行替换并统一偏移事件日期，生成语义结构相同但无表面信息可匹配的反事实社会世界。实验结果显示，在中英文双语环境下覆盖五个异质事件共125个预测点，最强的前沿大模型仅达75.0分（满分100），显著高于基线50分，且校准性与时间准确性指标可分离，表明模型在两类能力上存在差异。此外，基于同一基础模型的三种代理框架未能超越基础模型性能，而两种无模型启发式方法也均逊于所有大模型。单事件最大差距达21.4分，凸显多事件评估的重要性。所有匿名时间线、题库、真实答案及评分代码均已开源。
链接: https://arxiv.org/abs/2608.04009
作者: Zhenran Wang,Zhonghan Bian,Jinsong Li,Zhangyang Qi
机构: 未知
类目: Computation and Language (cs.CL)
备注: Project page: this https URL
Abstract:Large language models (LLMs), and the agents built on top of them, are now benchmarked heavily on whether they can finish a task – fix a bug, drive a browser, operate a GUI. A complementary social ability, namely how well a model understands and forecasts the way real social events unfold, has barely been measured. We introduce SocietyBench, an end-to-end benchmark that takes a one-line event topic, collects Web news and social-media posts across five platforms, distills them into a date-indexed timeline that keeps factual events and a public-opinion layer separate, and then turns every cutoff date on that timeline into an audited bank of forecasting questions. Questions are scored on two orthogonal 100-point axes: probability calibration and temporal accuracy. Before any model sees a timeline, a three-phase procedure replaces every named entity and shifts every date by a per-event constant, turning a real arc into a counterfactual social world – structurally identical to what happened, but stripped of the surface labels a model could match against pre-training memory. On five heterogeneous events and 125 prediction points in Chinese and English editions, the strongest of six frontier LLMs reaches only 75.0 out of 100, against a trivial anchor of 50. The two axes come apart: a model can be calibration-strong but time-weak, or the reverse. Three agent frameworks built on a shared base model fail to improve on that base, and two model-free heuristics trail every LLM. Per-event gaps reach 21.4 points on a single axis, which is our main argument for evaluating on several events rather than one. All anonymized timelines, question banks, ground truth, and scoring code are released.
[NLP-2] WorldCup Arena: Prospective Leakage-Free Evaluation of Frontier LLM s on a Live Tournament
【速读】： 该论文旨在解决现有大型语言模型（LLM）预测能力评估中普遍存在的回溯性问题——即评估时事件已发生、答案可从网络获取，导致评测易受记忆化（memorisation）干扰。为此，作者提出一种前瞻性的评估范式：在2026年FIFA世界杯的39天赛程期间，对六个具备扩展推理与原生服务器端网络搜索能力的前沿大模型，于每场比赛开赛前逐场提问，要求其预先预测全部104场比赛的结果、12个小组头名及赛前胜者池（outright pool），共生成4,494条未公开的预测数据。由于预测时答案尚未存在，评估过程天然避免信息泄露（leakage-free），无需依赖事后过滤。其核心解决方案在于构建一个动态、实时、不可预知的预测任务环境，以真实检验模型的前瞻性推理能力。研究发现，尽管模型平均准确率达63.9%（与投注市场首选一致），但其预测行为表现出高度同质性：多数情况下模型间共识远超正确率，投票机制无增益；且模型倾向于低估平局与进球数，集中押注单一典型比分。预测准确性更依赖比赛实力差距而非信息充分性，最激烈对决（信息最丰富）反而表现最差，而整体赛事预测则相对准确。当前主流大模型在此任务中差异不显著，排名稳定但中间梯队波动频繁，差距始终狭窄。研究团队将完整赛前资料、赛程与官方结果作为基准数据集，连同评分代码一并开源。
链接: https://arxiv.org/abs/2608.04008
作者: Zhenran Wang,Zhonghan Bian,Jinsong Li,Zhangyang Qi
机构: 未知
类目: Computation and Language (cs.CL)
备注: Project page: this https URL
Abstract:Benchmarks that measure the forecasting ability of large language models are almost always retrospective: the event has happened, the answer is somewhere on the Web, and the evaluation must defend itself against memorisation. We report the opposite design. Over the 39 days of the 2026 FIFA World Cup, six frontier LLMs – all with extended thinking and native server-side web search – were asked before every kickoff, one match at a time, to fill in a seven-market prediction card for all 104 matches, plus 12 group winners and a pre-tournament outright pool; no answer existed when the question was asked, so the evaluation is leakage-free by construction rather than by filtering, and the frozen archive holds 4,494 scored predictions. What the tournament establishes is a set of behaviours the six systems share. On match outcome they average 63.9%, level with backing the bookmaker’s favourite – which is in fact what they usually do. They agree with one another far more often than they are right, so a majority vote adds nothing. They under-commit to draws and to goals, and crowd their scoreline picks onto a single prototypical result. Accuracy tracks how lopsided a fixture is rather than how much is known about it: it collapses in the closest ties, where the dossiers are richest, while questions about the tournament as a whole are answered well. On this task the current generation of frontier systems is not sharply differentiated: the standings hold up at the top and the bottom across the run and churn in the middle, and the margins stay narrow throughout. The briefing dossiers, fixtures and official results are released as a benchmark, together with the scoring code.
[NLP-3] urnSight: Turn-Level Hindsight Self-Distillation for Tool-Integrated Reasoning
【速读】： 该论文旨在解决生成式 AI 在长时序工具交互任务中因轨迹级监督导致的信用分配粒度不足问题，以及现有自蒸馏方法依赖于真实答案或检索技能构建教师上下文所带来的状态偏差问题。其核心解决方案是提出一种基于回合级事后自蒸馏（TurnSight）的框架，通过执行条件下的事后视角（execution-conditioned hindsight）直接生成监督信号，并构建具有不同前瞻视野（lookahead horizons）的多个事后视图，利用跨视野方向一致性筛选可靠监督信号。随后，将选定的监督信号在兄弟回放轨迹间进行归一化处理，并自适应地调制强化学习优势函数，同时保持原始优化方向。该方法有效捕捉了工具交互的回合级结构，显著提升了长时序工具集成推理（Tool-Integrated Reasoning, TIR）中的学习效率与性能。
链接: https://arxiv.org/abs/2608.04007
作者: Changle Qu,Sunhao Dai,Hengyi Cai,Yuqi Zhou,Xinran Chen,Simon,Jun Xu
机构: Google(谷歌); Stanford University (斯坦福大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Tool-Integrated Reasoning (TIR) enables LLMs to solve complex tasks through iterative tool interactions. However, existing reinforcement learning methods often rely on trajectory-level supervision, limiting fine-grained credit assignment in long-horizon TIR scenarios. On-policy self-distillation offers denser signals through teacher branches with privileged context, but existing approaches typically derive such context from ground-truth answers or retrieved skills, which may not reflect the states actually visited by the agent. Moreover, token-level supervision fails to capture the turn-level structure of tool interactions. To address this, we propose TurnSight, a turn-level hindsight self-distillation framework that derives supervision directly from execution-conditioned hindsight. It then constructs multiple hindsight views with different lookahead horizons and selects reliable supervision through cross-horizon directional agreement. Finally, the selected hindsight signal is normalized across sibling rollouts and used to adaptively modulate RL advantages while preserving their original optimization direction. Extensive experiments on three benchmarks demonstrate the effectiveness of TurnSight. Our codes are available at this https URL.
[NLP-4] PAST-Bench: Benchmarking the Foundations of Recursive Self-Improvement in Personal Agents
【速读】： 该论文旨在解决持续性智能体（persistent agents）在长期运行中能否有效利用保留的经验实现递归自我改进（recursive self-improvement）的核心问题。尽管个人智能体（Personal AI agents）具备跨会话保留偏好、任务历史、工具使用模式和习得技能的能力，但其积累的经验是否真正带来未来行为的系统性提升，尚未得到充分验证。为此，论文提出PAST-Bench基准，通过在26个场景、204个实验周期中控制“经验保留”开关，系统评估智能体在记忆、程序复用、信息获取与状态更新等能力上的进步，并检验这些进步是否符合预期的“保存-检索-更新”路径。实验结果表明，虽然多数模型在特定任务上表现出真实提升，但改善程度不均，且部分模型的性能提升缺乏对目标路径的证据支持。基于此诊断发现，研究者提出了Hermes+，在原Hermes框架基础上引入五个针对智能体循环各阶段的干预措施，显著提升了经验保留带来的平均收益，并增强了路径可解释性，尤其在需替换过时状态的任务中表现突出，但效果仍依赖于具体能力与模型架构。因此，PAST-Bench与Hermes+共同构建了一个评估与诊断基础，为理解持久性智能体如何从经验留存迈向系统性自我改进提供了关键工具。
链接: https://arxiv.org/abs/2608.04003
作者: Shuhan Xue,Zixin Ding,Yichen Shen,Yinjie Wang,Zhenfei Yin,Yingcheng Wu,Yuxin Chen,Mengdi Wang,Ling Yang
机构: 未知
类目: Computation and Language (cs.CL)
备注: Code: this https URL
Abstract:Recursive self-improvement requires agents to turn accumulated experience into better future behavior. Personal AI agents offer a concrete setting for studying this capability because they retain preferences, task histories, tool routines, and learned skills across sessions. Yet whether retained experience actually improves them over time has not been systematically tested. We introduce PAST-Bench, a benchmark designed to isolate this question. Each agent runs through ordered sequences of fresh-session tasks under matched conditions that turn retained experience on and off. It spans 26 scenarios and 204 episodes across memory, procedural reuse, information gathering, and update. We report both later-task gains and whether those gains follow the intended save, retrieve, and update pathway. Across seven base models and four agent frameworks, improvement is real but uneven across capabilities. Agents with the same headline gain can differ markedly in whether that gain is supported by evidence of the intended pathway. Guided by these findings, we develop Hermes+, which extends Hermes with five targeted interventions across stages of the agent loop. Hermes+ raises the average gain from retained experience and provides clearer pathway evidence, with its strongest improvement on tasks requiring outdated state to be replaced, although the effect remains capability- and model-dependent. Together, PAST-Bench and Hermes+ provide an evaluation and diagnostic foundation for studying how persistent agents can progress from retaining experience to systematically improving through it. Code: this https URL
[NLP-5] Agogic: Performance-Timed Music Tokens for LLM -Native Text-to-Symbolic-Music Generation
【速读】： 该论文旨在解决文本到音乐生成模型中音乐表示（representation）选择对生成质量影响的长期未量化问题。传统上，音乐令牌化方式常与模型架构、训练数据及训练策略耦合，难以独立评估其作用。本文通过固定预训练的Qwen3.5模型规模（0.8B–27B）、数据集、预算与解码策略，仅替换七种不同的音乐表示方式，以“无模型上限”（model-free ceiling）为基准，系统性地比较不同表示在分布保真度上的表现。研究发现，表示方式是决定生成分布忠实度的关键变量，而非模型规模：即使将模型参数量扩大34倍，弗雷切特音乐距离（FMD）变化极小；而切换至性能分辨率表示（PMT，10ms时间精度、每音符力度、多轨纹理，共609个符号），可使FMD降低近一半。具体而言，0.8B的PMT模型在FMD上达到159，显著优于1.7–1.8倍分辨率的节拍网格表示（FMD 272–286，降幅达1.7–2.8倍，置信区间不重叠）。该优势在从头训练的26M参数模型及另一性能分辨率分词器上重复出现，表明其为该类表示的固有属性，而非特定词汇表或高分辨率采样的偶然结果。进一步验证显示，即便将PMT的起始时刻对齐至节拍网格分辨率，其仍领先于所有对比组。此外，引入轻量级解码时约束可显著提升文本-音乐对齐能力（如乐器识别F1从0.28升至0.60，调性正确率从0.16升至0.35），且不损害分布保真度。作者开源了评估框架、25+模型检查点、两个大规模对齐数据集（含86.6k图文/MIDI/ABC/音频对，625万条带标签文本），以及用于诊断训练分布再现性的“印刻诊断”工具。研究结论表明，未来音乐表示方法的宣称应基于可测量的实证，而非主观断言。
链接: https://arxiv.org/abs/2608.03999
作者: Junhao Chen,Mingjin Chen,Jingjia Mao,Lin Chen,Saining Zhang,Minglin Chen,Ruocheng Wu,Liaoyuan Fan,Wenyi Li,Mingju Gao,Henghaofan Zhang,Zhihao Li,Hao Zhao,Yufei Wang,Ruqi Huang
机构: 未知
类目: ound (cs.SD); Computation and Language (cs.CL)
备注: Project Page: this https URL
Abstract:Text-to-music language models begin with a choice usually made by default: how to tokenize music. Normally entangled with backbone, data, and recipe, its effect has never been measured in isolation. We fix pretrained Qwen3.5 (0.8B-27B), data, budget, and decoding, and swap only the representation across seven tokenizations, anchoring texture metrics to each representation’s model-free ceiling. The ordering is clean and surprising: representation, not model size, is the binding variable for distributional fidelity. Scaling the backbone 34x barely moves Frechet Music Distance (FMD), whereas switching representation halves it. PMT, a performance-resolution stream we release (10 ms timing, per-note velocity, multi-track texture; 609 symbols), reaches FMD 159 at 0.8B against 272-286 for beat grids (1.7-1.8x lower, up to 2.8x elsewhere; non-overlapping bootstrap CIs), so a 0.8B performance-resolution model beats a 27B beat grid. It reappears on a 26M from-scratch backbone and a second performance-resolution tokenizer: a property of the class, not one lucky vocabulary. Nor is it a finer-lattice artifact: snapping PMT’s onsets to the beat grids’ resolution still leaves it 67-129 FMD ahead of both (n=500). The effect is distributional; whether it is audible is a separate question, left open by our probe, with a human study pre-registered. Native caption adherence is weak but separable: a lightweight decode-time constraint doubles instrument-F1 (.28 to .60) and Correct-Key (.16 to .35) at no distributional cost. We release the harness, 25+ checkpoints, two corpora (86.6k aligned across caption/MIDI/ABC/audio; 6.25M captioned, the largest for music), and an imprinting diagnostic: published text-to-MIDI systems reproduce their training distribution near-invariant to the caption (72% vs. 71% chord-time on disjoint domains). The field’s next representation claim can now be measured, not asserted.
[NLP-6] When Attention Goes Blind: Numerical Failure in ALiBi Positional Encodings
【速读】： 该论文旨在解决ALiBi位置编码（ALiBi positional encoding）中一种此前未被充分关注的失效模式：其线性偏置缩放在浮点精度范围内发生下溢，导致大量注意力权重变为零，从而使部分注意力头功能失灵。该问题的关键在于线性偏置随位置距离增加而单调递增，当距离过大时，其数值超出浮点数表示范围，引发精度丢失并破坏注意力机制的有效性。研究通过系统分析该失效模式的影响，并评估四种训练阶段的缓解策略，发现对距离采用对数缩放（log-scaled distances）能最一致地提升“密钥检索”（passkey retrieval）任务性能。尽管存在此缺陷，默认的ALiBi斜率仍表现出较强的基准性能，尤其在“针尖堆草垛”式检索任务中表现优异。基于实验结果，论文提出了针对使用ALiBi的模型训练的具体优化建议。
链接: https://arxiv.org/abs/2608.03994
作者: Christopher Schröder,Lukas Gienapp,Ferdinand Schlatt,Martin Potthast,Gerhard Heyer
机构: Institute for Applied Informatics at Leipzig University (InfAI); ScaDS.AI Dresden/Leipzig; University of Kassel; hessian.AI
类目: Computation and Language (cs.CL)
备注:
Abstract:We identify a previously overlooked failure mode of ALiBi positional encoding: its linear bias scaling underflows floating-point precision, which zeroes out a large fraction of attention weights and renders the affected attention heads partially blind. We analyze this failure mode, characterize its impact, and examine four mitigation strategies. We further demonstrate its occurrence in state-of-the-art pretrained models based on ALiBi. Comprehensive pretraining experiments with 148M-parameter decoder models help us to disentangle its effects from out-of-context degradation. We find that ALiBi’s failure mode can substantially impair token retrieval while having only a minor effect on standard decoder benchmarks. We propose four training-time mitigation strategies and evaluate them individually and in combinations, finding that log-scaled distances yield the most consistent improvements in passkey retrieval. Despite this problem, default ALiBi slopes remain a surprisingly strong baseline, particularly for needle-in-a-haystack retrieval. Based on these findings we provide concrete recommendations on how to train models with ALiBi.
[NLP-7] string2string Studio: An Interactive In-Browser Platform for String-to-String Algorithms
【速读】： 该论文旨在解决跨自然语言处理、计算生物学及数字人文领域中字符串到字符串分析的通用性与可解释性问题，尤其针对现有工具在交互性、本地化执行与多粒度分析能力方面的不足。其核心解决方案是构建一个基于浏览器的交互式平台——string2string Studio，通过将C++编写的算法编译为WebAssembly，实现核心操作在客户端本地运行，无需安装或上传数据，保障了隐私与效率。平台集成了六项关键模块（对齐、距离、相似性、搜索、生成评估指标与BLAST同源搜索），支持字符、词、标记、行及残基等多层级分析，并以“证据”形式可视化输出结果（如对齐路径、编辑序列、匹配记录等），显著提升了方法的可追溯性、可调试性与可比性。内部基准测试表明，其性能较原Python版本提升高达2,500倍，全局/局部对齐速度优于通用原生C实现，且在指定参数下与独立参考结果完全一致；对于同源搜索，其客户端封装的blastn路径在参数匹配条件下与NCBI BLAST+的排名和统计结果高度一致。此外，平台提供经筛选的示例与“学习模式”，将经典算法与度量标准转化为可复用的教学演示。整体系统开源免费，支持跨领域、高透明度的字符串分析。
链接: https://arxiv.org/abs/2608.03984
作者: Mirac Suzgun,James Zou,Stuart M. Shieber,Dan Jurafsky
机构: Stanford University (斯坦福大学); Harvard University (哈佛大学)
类目: Computation and Language (cs.CL)
备注: this https URL
Abstract:We present string2string Studio, an interactive in-browser platform for string-to-string analysis across natural language processing, computational biology, and the digital humanities. The system integrates six main modules (alignment, distance, similarity, search, generation metrics, and BLAST homology search), operating at character, word, token, line, and residue levels. Its C+±based algorithms compile to WebAssembly, so core operations run locally by default without any installation or data upload. The interface reports scores with their “evidence” (alignments, edit paths, metric matches, search hits, and homology traces), making methods inspectable, debuggable, and comparable on shared inputs. Internal benchmarks show speedups of up to 2,500x over the Python predecessor, faster global/local alignment than a general-purpose native C aligner, and exact agreement with independent references under declared settings. For homology search, the scoped client-side blastn path closely matches NCBI BLAST+ rankings and statistics under matched parameters. A curated showcase and Learn mode present canonical algorithms and metrics as reusable demonstrations. string2string Studio is open-source and freely available at this http URL.
[NLP-8] HalluTruthQA-4K: A Fine-Grained Corpus and Annotation Process for Arabic Hallucination Detection and Truth Verification
【速读】： 该论文旨在解决生成式AI在阿拉伯语问答任务中产生的事实性错误（hallucination）难以准确识别与定位的问题。现有阿拉伯语幻觉数据集多仅提供二元标签（幻觉/非幻觉），缺乏对错误内容的具体范围、错误成因及正确答案的详细标注，限制了对模型幻觉机制的深入分析与评估。其解决方案的关键在于构建HalluTruthQA-4K——一个包含4,000个专家标注的阿拉伯语问答样本的高质量基准数据集，覆盖伊斯兰知识、历史、科学和地理四大知识密集型领域。该数据集不仅提供问题、模型生成回答、参考答案及五个合理干扰项，还对幻觉响应进行细粒度标注：包括字符级错误片段（character-level erroneous spans）、人工撰写的错误解释（human-written explanations）以及分层的幻觉类型分类（hierarchical hallucination types）。通过严格的数据构建流程与多轮独立验证与仲裁机制，确保了标注质量与一致性。该资源为阿拉伯语大语言模型的事实可靠性评估、幻觉检测、错误定位、解释生成与事实验证提供了可复用的基准，显著提升了对模型幻觉行为的精细化研究能力。
链接: https://arxiv.org/abs/2608.03966
作者: Salah Eddine Bekhouche,Abdessalam Bouchekif,Hichem Telli,Mohammed-En-Nadhir Zighem,Abdenour Hadid
机构: University of the Basque Country (巴斯克大学); Hamad Bin Khalifa University (哈马德·本·哈利法大学); University of Biskra (比斯克拉大学); Universiti Malaysia Kelantan (马来西亚吉兰丹大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Large language models can generate fluent Arabic answers while introducing factual errors that are difficult to identify and verify. Existing Arabic hallucination resources often assign a binary label to an entire response, indicating whether it is hallucinated or non-hallucinated, but provide limited information about the exact erroneous content, the reason for the error, or the correct factual answer. We present HalluTruthQA-4K, an expanded version of the HalluTruthQA resource containing 4,000 expert-curated Arabic question-answering instances across four knowledge-intensive domains: Islamic knowledge, history, science, and geography. Serving as the official dataset for Track 2 of the HalluScoring 2026 shared task, HalluTruthQA-4K extends our original corpus to 4,000 instances. Each instance pairs an Arabic question with a model-generated response, a verified reference answer, and five plausible distractors. Hallucinated responses are additionally annotated with character-level erroneous spans, human-written explanations, and hierarchical hallucination types. The corpus contains 1,643 hallucinated and 2,357 non-hallucinated responses, with 1,843 annotated erroneous spans. We describe the resource construction and annotation methodology, including question selection, controlled answer generation, candidate construction, expert annotation, independent verification, adjudication, and quality control. We also document the annotation guidelines, taxonomy, data format, inter-annotator agreement, and corpus statistics. HalluTruthQA-4K provides a reusable resource for hallucination detection, span-level error localization, explanation generation, factual verification, and the broader evaluation of factual reliability in Arabic language models.
[NLP-9] Logic Before Language: Pre-pretraining on Formal Derivations Fosters Skill Acquisition and Compressibility
【速读】： 该论文旨在解决现有预预训练（pre-pretraining）任务在符号数据上存在表达能力不足、依赖狭窄算子，且受限于较小的词元预算，难以揭示语言技能涌现与表征动态的问题。其核心解决方案是提出逻辑预预训练（Logic-PPT），通过形式推导（formal derivations）作为原则性初始化策略，引入自然语言中关键的抽象机制，包括变量绑定、量词与关系依赖的关联，以及长程上下文中的谓词-论元结构组合。该方法在1000亿词元规模下验证，显著加速了语言模型的语言技能习得，在仅使用比标准初始化少360亿词元的情况下即达到80%准确率，并优于其他预预训练基线。机制分析表明，形式推导诱导出一种低秩、光谱集中（spectrally concentrated）的表征空间，具有持久的结构重组特性，从而显著提升模型可压缩性：即使在约33%稀疏度下，剪枝后的模型性能仍可匹配密集基线，凸显其内在几何结构对高效表示学习的关键作用。
链接: https://arxiv.org/abs/2608.03930
作者: Jo-Ku Cheng,Nikolaos Aletras,Marco Valentino
机构: University of Sheffield (谢菲尔德大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Pre-pretraining language models (LMs) on symbolic data can accelerate and improve natural language acquisition. However, existing pre-pretraining tasks, such as Dyck and procedural algorithms, rely on narrow primitives that fail to capture the expressive capacity of natural language. Moreover, prior studies remain restricted to relatively small token budgets, offering limited insight into skill emergence and representational dynamics. To address these limitations, we propose logic pre-pretraining (Logic-PPT) as a principled initialization strategy, leveraging formal derivations to impart richer structural and linguistic biases. Formal derivations require abstract mechanisms that are central to natural language, simultaneously binding variables, connecting quantifiers and relational dependencies, and composing predicate-argument structures over long contexts. Scaling our evaluation to a 100B-token regime, logic pre-pretraining substantially accelerates skill acquisition in LMs, achieving 80% accuracy on linguistic tasks with 36B fewer tokens than standard initialization, and outperforming alternative pre-pretraining baselines. Mechanistically, formal derivations induce persistent structural reorganization, distinctively characterized by a lower-rank, spectrally concentrated representation space. Crucially, we show that this internal geometry enables improved model compressibility via pruning, matching the dense baseline performance even at \approx 33% sparsity.
[NLP-10] Sparse Weight Decomposition for Efficient Circuit Extraction
【速读】： 该论文旨在解决预训练密集型变换器模型（dense pretrained transformers）缺乏可解释单元以支持电路提取的问题。传统方法通过学习辅助的稀疏表示或训练稀疏模型来获得这些可解释单元，但此类方法需额外大量计算资源，并可能在分析表示与原始模型之间引入保真度差距。本文提出稀疏权重分解（Sparse Weight Decomposition, SWD），通过将每个线性投影权重矩阵重新参数化为两个稀疏因子，其共享的中间坐标作为可独立访问的电路单元，从而实现无需训练替代网络即可直接进行电路提取。该方法保持了与学习稀疏特征方法一致的评分、选择和消融工作流，在单矩阵替换任务中，SWD在使用不足1%数据的情况下达到了与Transcoder等强基线相当的保留保真度；在匹配保真度的前提下，SWD在GPT-2、Qwen2.5及Qwen3.5-27B等多个任务上以更少的活跃读写边和选中单元实现了相同的电路充分性与必要性目标。此外，经过微调非零因子值后，SWD仍能有效实现全模型注意力与MLP权重矩阵的替换。最后，SWD还提供一种零数据变体，显著拓展了机制可解释性分析（如每步分析）的应用范围。
链接: https://arxiv.org/abs/2608.03913
作者: Chuanhao Yan,Xuhan Huang,Yawen Duan,Zhenfei Yin,Hang Zhao,Bryan Dai,Jie Fu
机构: IQuest Research( IQuest 研究所); Safe AI Forum( 安全人工智能论坛); University of Oxford( 牛津大学); Stanford University( 斯坦福大学); Tsinghua University( 清华大学)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注:
Abstract:Dense pretrained transformers do not naturally expose interpretable units for circuit extraction. Existing approaches obtain such units by learning auxiliary sparse representations or training sparse models, incurring substantial additional computation while potentially introducing a fidelity gap between the representation being analyzed and the original pretrained model. We propose Sparse Weight Decomposition (SWD), which reparameterizes pretrained linear projections by factorizing each weight matrix into two sparse factors whose shared intermediate coordinates serve as individually addressable circuit units. Without training a separate replacement network, this parametric representation supports the same scoring, selection, and ablation circuit extraction workflow used for methods that learn sparse features. Across single-matrix replacements, SWD matches the held-out fidelity achieved by Transcoder and other strong baselines while using less than 1% of the data that those baselines use to train their replacements. For matched replacement fidelity, SWD reaches the same circuit sufficiency and necessity targets with fewer active read/write edges and selected units across tasks on GPT-2, Qwen2.5, and Qwen3.5-27B. We further show that SWD remains effective for full-model replacement of all attention and MLP weight matrices after fine-tuning the nonzero factor values. Finally, SWD also features a zero-data variant, allowing broader use of mechanistic interpretability analysis (e.g., per-step analysis).
[NLP-11] ANNOTARES: A Dataset for Extracting Logical Structures from German Statutory Texts
【速读】： 该论文旨在解决德国法律文本中逻辑成分自动结构化分析的难题，特别是针对法律要件（Tatbestand）与法律后果（Rechtsfolge）的识别与分割问题。其核心挑战在于法律语言复杂的句法结构及领域专属性强，导致传统方法难以有效捕捉语义边界。解决方案的关键在于构建一个全新的标注数据集 ANNOTARES，该数据集包含三个不同法律部门的德语文本，并在词元级别（span-level）进行精确标注，以支持对模型在特定领域性能与跨法典泛化能力的评估。实验表明，基于BERT的预训练模型及大语言模型（LLM）方法在处理法律语言复杂结构方面表现最优，显著优于规则基线、CRF、BiLSTM及其组合模型，凸显了深度学习模型在法律文本理解中的关键优势。
链接: https://arxiv.org/abs/2608.03898
作者: Ronja Schwarz,Jannik Strötgen
机构: Karlsruhe University of Applied Sciences (卡尔斯鲁厄应用技术大学)
类目: Computation and Language (cs.CL)
备注: Accepted at KONVENS 2026
Abstract:The automatic structural analysis of legal texts is a cornerstone of legal technology, yet the extraction of their logical components remains a significant challenge. In this paper, we introduce the task of identifying and segmenting legal conditions (Tatbestand) and legal consequences (Rechtsfolge) within German statutory texts. To support this task, we present ANNOTARES (Annotations of Tatbestand-Rechtsfolge Sequences), a novel dataset comprising German law texts with span-level annotations. Spanning three distinct legal codes, the dataset is designed to evaluate both domain-specific performance and cross-statute generalizability. We benchmark diverse architectural approaches: a rule-based baseline, CRFs, BiLSTMs, BiLSTM-CRF, and modern Transformer-based models, including BERT variants and LLM-based methods. Our results demonstrate that BERT and LLM-based models achieve superior performance in capturing the complex syntactic structures of legal language. We release our dataset to facilitate further research in automated legal reasoning.
[NLP-12] BanglaWild: An In-the-Wild Bengali Scene Text Recognition Benchmark for OCR and Vision-Language Models
【速读】： 该论文旨在解决真实场景下孟加拉语文本识别（Bengali scene text recognition）缺乏系统评估基准的问题。现有资源多集中于手写文档或受限场景的标识牌解析，且仅报告整体编辑距离（edit-distance）指标，同时对传统OCR与视觉语言模型（VLMs）的评估往往未在同一真实场景数据集上进行对比。为填补这一空白，研究提出BANGLAWILD基准，包含2,535张真实场景下的孟加拉语文本图像，每张图像均配有精确的黄金标准转录、两类分类轴、四项诊断属性以及一个标准化拼写形式以反映实际文本与规范拼写的偏差。研究在三种提示策略下评估了15个VLM和3个传统OCR系统，并对6个开源模型采用LoRA进行微调，同时引入大语言模型作为裁判（LLM-as-a-Judge）以补充编辑距离指标。结果揭示：同一模型家族中更大模型并不一定优于较小模型；最强系统中的约60%错误源于视觉误识别，而连字符相关错误占比不足2%，挑战了孟加拉语OCR研究中长期存在的假设；该视觉主导的错误分布特征在不同架构间保持一致，包括一个传统基线模型亦表现出类似模式。提示语言主要影响跨脚本漂移问题，而LoRA可有效缓解弱模型中的灾难性失败，但无法提升已有高性能模型的上限。代码与数据将公开发布。
链接: https://arxiv.org/abs/2608.03884
作者: Sadab Shiper,Tawsif Tashwar Dipto,Mir Md Inzamam,Eshat Tanzeem
机构: Islamic University of Technology (伊斯兰技术大学); BRAC University (BRAC大学)
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注:
Abstract:In-the-wild Bengali scene text recognition is largely unmeasured: existing resources target handwritten documents or constrained sign-board parsing, report only aggregate edit-distance metrics, and evaluate either conventional OCR or VLMs, never both on the same in-the-wild data. To address this gap, we introduce BANGLAWILD, a benchmark of 2,535 Bengali scene text images, each paired with a verbatim gold transcription, two categorical axes, four diagnostic attributes, and an orthographically standard form where the in-image text deviates from canonical spelling. We evaluate fifteen VLMs and three conventional OCR systems under three prompting strategies, fine-tune 6 open-source models with LoRA, and complement edit-distance metrics with an LLM-as-a-Judge evaluation. Our results reveal a persistent gap in which larger models within the same family do not outperform smaller ones. Our fifteen-class error taxonomy shows that visual mis-recognition accounts for ~60% of errors in the strongest systems, while conjunct-related errors contribute under 2%, challenging a long-standing assumption in Bengali OCR research; the same visual dominant profile also holds across architectures, including the one conventional baseline that reads Bengali reliably. Prompt language mainly affects cross-script drift and LoRA reduces catastrophic failures in weak models without lifting the ceiling on already competent ones. Code and data will be publicly released.
[NLP-13] DS@GT-ARC at eRisk 2026 Task 3: Sparse Semantic and LLM Reranking for ADHD Symptom Sentences
【速读】： 该论文旨在解决在缺乏标注训练数据的情况下，对Reddit用户帖子中与成人注意力缺陷多动障碍（ADHD）18项症状相关性进行精准排序的问题。由于eRisk 2026任务3首次发布且未提供标注数据，研究者面临零样本学习的挑战。其核心解决方案在于采用分阶段检索架构：首先利用稀疏的BM25模型大规模召回候选句子，随后通过证据感知重打分、基于嵌入的重排序、查询原型扩展以及大语言模型（LLM）驱动的重排序等多阶段策略进行精细化排序。其中，关键创新点在于引入LLM-based reranking机制，显著提升了排名质量，其在官方评测中表现最优；同时，手动分析也验证了该框架与专家评分趋势的一致性，表明分阶段重排序是未来发展的有效路径。
链接: https://arxiv.org/abs/2608.03883
作者: David Guecha
机构: Georgia Institute of Technology(佐治亚理工学院)
类目: Computation and Language (cs.CL)
备注:
Abstract:This paper describes our submissions to eRisk 2026 Task 3, ADHD Symptom Sentence Ranking. The task requires systems to rank candidate Reddit sentences according to their relevance to each of the 18 symptoms in the Adult ADHD Self-Report Scale (ASRS-v1.1). Because no annotated training data were released for this first edition of the task, we relied on zero-shot experimentation, manual validation, and unsupervised or weakly guided retrieval pipelines. Our systems combine sparse BM25 retrieval, evidence-aware rescoring for self-referential symptom reports, embedding-based reranking, query-prototype expansion, and LLM-based reranking. All submitted systems follow a staged retrieval design in which BM25 retrieves candidates at scale and semantic or LLM rerankers refine the final rankings. Among our submissions, the LLM reranker achieved the strongest official scores, followed by the prototype query-expansion run. Our manual top-10 analysis aligned with the official expert scoring trend, suggesting that staged reranking is a promising direction for further development.
[NLP-14] ContinualSkillBench: Can LLM Agents Truly Evolve Their Capabilities?
【速读】： 该论文旨在解决当前大型语言模型（Large Language Models, LLMs）在复杂任务求解中，其外部技能库能否实现有效持续演化，以及由此产生的技能是否真正提升任务求解能力的核心问题。现有代理框架虽赋予模型外部技能库以应对复杂任务，但缺乏对技能持续学习与演化的系统性评估机制，导致对技能可迁移性与泛化能力的理解不足。为此，论文提出一种动态评估框架——ContinualSkillBench，用于评测上下文持续技能学习（in-context continual skill learning）的表现。其关键创新在于构建了涵盖五个代表性领域、共500个按难度递增且具有跨任务技能复用潜力的子任务体系，能够系统评估模型在连续任务中的技能积累与迁移能力。实验结果表明，顺序执行有助于性能提升，但不同模型与领域间差异显著；上下文学习在平均表现上与显式技能维护相当，说明性能提升主要源于对先前上下文与反馈的适应，而非仅依赖可复用的技能抽象；而显式技能在需重复性流程或高精度输出的任务中仍具优势。此外，能力较弱的模型倾向于积累更多、更碎片化的任务特定技能，揭示当前上下文技能演化机制虽支持持续适应，但在将经验稳定转化为鲁棒、可迁移的通用技能方面仍存在显著局限。
链接: https://arxiv.org/abs/2608.03874
作者: Tianyi Guan,Yiding Wang,Haotong Yang,Siyuan Cao,Shirui Liu,Yi Hu,Jiaqi Li,Muhan Zhang
机构: Peking University (北京大学); Beijing Institute for General Artificial Intelligence (北京通用人工智能研究院)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Modern agent frameworks equip large language models with external skill libraries to solve complex tasks. However, it remains unclear whether these systems can effectively evolve their skills and whether the resulting skills improve task-solving capabilities. To bridge this gap, we introduce ContinualSkillBench, a dynamic evaluation framework for in-context continual skill learning. It covers five representative domains, each containing 100 interconnected subtasks ordered by increasing difficulty and opportunities for cross-task skill reuse. Our experiments show that sequential execution generally improves performance, but the gains vary substantially across models and domains. Moreover, in-context learning performs comparably to explicit skill maintenance on average, suggesting that much of the improvement arises from adaptation to prior context and feedback rather than reusable skill abstraction alone. Explicit skills nevertheless provide selective benefits for tasks requiring reusable procedures or precise outputs. We further find that less capable models tend to accumulate larger, more fragmented collections of task-specific skills. These findings show that current in-context skill evolution mechanisms can support continual adaptation, but still struggle to consistently consolidate experience into robust and transferable skills.
[NLP-15] Beyond Representational Similarity: Source-Conditioned Description-Length Gain for Generative Plagiarism Detection and Candidate Source Reranking
【速读】： 该论文旨在解决生成式人工智能（Generative AI）背景下学术诚信与同行评审所面临的挑战，特别是针对由大语言模型（LLMs）生成文本时存在的“生成式抄袭”（generative plagiarism）检测难题。现有方法主要关注识别AI参与痕迹，而非真正意义上的源文本复用；而基于相似性的检测方法在面对深度改写或多源合成后往往失效。本文提出的关键解决方案是引入源条件描述长度增益（Source-Conditioned Description-Length Gain, SCDG），这是一种无需训练、具有方向性的检测框架。其核心思想基于概率预测中的描述长度观点：若某候选源文本S对可疑文档P的生成具有增量预测价值，则在包含该源信息条件下，语言模型对该文档的描述长度将显著缩短。SCDG通过对比冻结语言模型在有无候选源条件下对文档的描述长度差异，计算出逐标记级别的对数似然增益，从而量化源文本提供的额外预测证据。实验结果表明，在CLEF PAN 2025和PAN 2026基准上，SCDG在单对检测中达到0.92精度、0.97召回率与0.94 F1，多源检索任务中nDCG@10达0.83、Recall@100为0.96，且在同主题同事件的Multi-News测试中仅0.125%的样本被误判为源复用，充分验证了其在广泛文本变换下的鲁棒性与有效性。因此，SCDG作为一种统一且可逐标记分解的信号，为复杂变换下的源特定内容复用检测提供了强有力的解决方案。
链接: https://arxiv.org/abs/2608.03859
作者: Peijia Guo,Wenxuan Xie,ZiGuang Li,Ming Li
机构: 1. Tsinghua University (清华大学); 2. Institute of Automation, Chinese Academy of Sciences (中国科学院自动化研究所)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models (LLMs) pose challenges to academic integrity and peer review. Yet generative plagiarism detection remains an underexplored and largely unresolved challenge. Prior work on LLM-generated-text detection targets AI involvement, which may be permissible, rather than source reuse, while similarity-based methods struggle after extensive rewriting and multi-source synthesis. Motivated by the description-length view of probabilistic prediction, in which relevant side information can reduce a target sequence’s code length, we introduce Source-Conditioned Description-Length Gain (SCDG), a directional, training-free framework that contrasts a frozen language model’s description length of a suspicious document P with and without a candidate source S . This contrast yields token-level log-likelihood gains that measure the incremental predictive evidence supplied by S . We evaluate SCDG on the PAN at CLEF benchmarks for generative plagiarism. On a PAN 2025-derived pairwise benchmark, SCDG achieves 0.92 Precision, 0.97 Recall, and 0.94 F1, outperforming all baselines; on PAN 2026’s multi-source retrieval task, it reaches 0.83 nDCG@10 and 0.96 Recall@100, surpassing all baselines. On a same-topic, same-event Multi-News test, the calibrated gain-distribution SCDG classifier predicts source reuse for only 0.125% of pairs, supporting robustness to topical overlap under this evaluation protocol. These results establish SCDG as a unified and token-decomposable signal for source-specific content reuse under extensive transformation.
[NLP-16] Sensitivity Causality and Repair Dissociate: A Layer-Wise Analysis of Perturbation Robustness and Its Scaling
【速读】： 该论文旨在解决语言模型在面对表面扰动输入（如拼写错误、OCR噪声、同音词等）时失效的归因问题，核心挑战在于如何准确识别导致错误的关键网络层。传统方法常假设“敏感性”（sensitivity）、“因果性”（causality）和“补偿能力”（compensatory capacity）三者一致，但本文通过多模型实证发现，这三个评估维度所定位的关键层存在显著分离（dissociation）。其解决方案的关键在于提出“级联破坏”（cascade disruption）机制：早期层中被诊断为因果关键的节点若被适配器（adapter）干预，反而会破坏下游完整计算流程，导致这些位置成为最差的适配器放置点。研究基于五种模型的对比分析，揭示出两种传播模式——“尖峰抑制型”（spike-and-suppress）与“晚期累积型”（late-accumulation），并在大规模缩放下验证后者随模型规模单调增强。进一步的固定框架层扫描实验（layer sweep）在多个模型上证实，诊断标记位点在链式思维（chain-of-thought, CoT）任务中确实是最具破坏性的适配窗口，且结果在生成长度较长的任务中损伤效应加剧。由此得出实用建议：采用无需训练的低秩分解（LRD）预筛选与默认深层放置策略，尽管绝对性能提升有限；同时指出，基于表示稳定性损失的干预效果可能因生成预算不足而产生虚假增益，提示未来评估需避免截断式生成带来的方法论偏差。
链接: https://arxiv.org/abs/2608.03842
作者: Nathan Labiosa,David Buff,Ena Nayak,Erica Donno
机构: University of Southern California (南加州大学)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 29 pages, 18 figures, 11 tables
Abstract:When a language model fails on surface-perturbed input (typos, OCR noise, homophones), “which layer is responsible” has three natural operationalizations: where representations diverge most (sensitivity), where restoring clean activations recovers the prediction (causality), and where a small adapter can repair the damage (compensatory capacity) - and we show these three layer maps dissociate. Across a five-model panel we identify two propagation regimes - spike-and-suppress (Phi-3.5, Gemma-2-9B) and late-accumulation (Llama-3, Mistral, Qwen2.5-7B) - and on the two models meeting an 80% identity-patch gate, sensitivity and causality are anti-correlated (rho = -0.72 to -0.88). Within-family scaling on Qwen2.5 (1.5B to 14B) shows the late-accumulation signature strengthening monotonically with scale, corroborated on a second family. We propose cascade disruption as the mechanism behind the dissociation: adapters placed at causally implicated early layers break intact downstream computation, making diagnostic-flagged sites the worst adapter placements. A fixed-harness layer sweep across four models (3.8-8B) confirms the core prediction on chain-of-thought GSM8K - the flagged sites are the most damaging adapter windows on every adjudicable model - and is sign-consistent but strongly attenuated on a multiple-choice control, consistent with damage that compounds with generation length. The sweep yields practical guidance: a training-free LRD pre-screen and a default-deepest placement rule, though absolute gains over no-adapter baselines remain small. Finally, apparent gains from a representation-stability loss reverse under an adequate generation budget - truncated chain-of-thought had been scored as empty - a methodological warning for any intervention evaluated on chain-of-thought tasks.
[NLP-17] VIBE: A VAD-Informed Benchmark for Entity-Centered Affective Profiling of Large Language Model Outputs ACL
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在描述社会敏感实体（如政治人物、国家、宗教、组织、历史事件及社会群体）时，其输出中蕴含的情感框架（affective framing）缺乏系统性量化与透明报告的问题。现有研究虽在情感极性、倾向性与情绪识别方面有所建树，但未能整合目标导向的效价-唤醒-支配（Valence-Arousal-Dominance, VAD）属性标注、明确的评分者契约（scorer contract）以及可追溯的“情感护照”（Affective Passport）报告格式。为此，作者提出VIBE基准，其核心在于建立一种分离生成与外部评分的测量契约：明确区分标量倾向性（scalar favorability）、响应层面的VAD属性与目标导向的VAD属性，并通过结构化的情感护照形式报告结果。实证研究表明，标量倾向性无法涵盖唤醒度与支配度信息，且三维度间具有独立性；整体文本情感与特定目标的情感表达存在差异；此外，诱发条件变化会导致情感画像漂移，凸显了在报告中记录上下文元数据的重要性。因此，该研究推动以实体为中心的情感剖析成为可文档化、可复现的研究实践，强调应随报告一同披露评分者身份、覆盖范围、评估协议及解释边界。
链接: https://arxiv.org/abs/2608.03810
作者: Andrei Chetvergov,Alexander Evseev,Timofei Sivoraksha,Stepan Ukolov,Mikhail Solovev,Danil Sazanakov,Sergey Bolovtsov
机构: RANEPA(俄罗斯国立经济与公共管理大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 25 pages, 13 figures, 22 tables. Submitted to ACL Rolling Review, August 2026
Abstract:Large language models routinely describe socially salient targets, including political figures, countries, religions, organizations, historical events, and social groups, encoding affective framing alongside factual content: a target may appear favorable or threatening, calm or conflictual, powerful or vulnerable. Existing work captures parts of this space through sentiment, favorability, and emotion benchmarks, but none combines target-directed VAD attribution, an explicit scorer contract, and a passport reporting format. We introduce VIBE, a benchmark for entity-centered affective profiling of LLM outputs in Valence-Arousal-Dominance (VAD) space. Its core contribution is a measurement contract: VIBE separates generation from external scoring, distinguishes scalar favorability, response-level VAD, and target-directed VAD, and reports profiles through an Affective Passport. Three empirical layers support the contract. H1 shows scalar favorability does not subsume arousal and dominance: valence findings are cross-validated (rV = 0.944 judge-human, rV = 0.954 inter-scorer); arousal and dominance are single-scorer directional estimates, not point-precise, consistent with known inter-annotator difficulty on these axes (rA = 0.495, rD = 0.702 among human annotators). H2 shows whole-response and target-directed VAD are different contracts: the same text can carry one affective tone overall while representing the named target differently. H3 is a protocol-drift diagnostic: elicitation conditions shift profiles, motivating context metadata in every affective report. These results motivate entity-centered affective profiling as a documented practice: profiles should be released with scorer identity, coverage, protocol, and interpretation limits.
[NLP-18] M-GATE: Multilingual Grammar Accuracy in Translation and Efficiency Benchmark for Large Language Models
【速读】： 该论文旨在解决多语言大模型评估中“流畅性”（fluency）与“语言精通能力”（proficiency）混淆的问题，即现有基准测试主要衡量模型在特定语言中完成任务的表面表现，而非其对语言规则的深层掌握。其核心挑战在于：如何有效区分模型是否真正理解并正确运用目标语言的语法结构、语义逻辑与文化语境。解决方案的关键是提出M-GATE（Multilingual Grammar, Accuracy in Translation, and Efficiency）——一个涵盖30种类型学上多样、资源从高到低分布的语言的综合性语言精通评估基准。M-GATE包含三大核心任务：基于语言学家精心设计、对抗性筛选的句法难题的语法错误检测；跨29种目标语言的英文源文本往返翻译质量评估（由三名大型语言模型判别器组成的评审团评分，并经专业标注员验证）；以及辅助性的分词器效率测量。实验结果表明，模型在翻译任务中的表现高度依赖预训练数据量（与日志化Common Crawl数据占比相关系数r = 0.86），低资源语言存在显著性能下降，但这一差距正随模型迭代逐步缩小。此外，启用推理能力可显著提升翻译质量，但在语法错误检测任务中效果不一，甚至对部分模型产生负向影响，说明最优配置具有任务依赖性。为防止数据污染，所有测试样本均保密，仅通过持续更新的公开排行榜提供示例供参考。
链接: https://arxiv.org/abs/2608.03803
作者: Tomáš Burkert,Angelika Peljak-Łapińska,David Zelený
机构: 未知
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 45 pages (97 incl. appendices), 6 figures
Abstract:Multilingual language models are deployed across a hundred or more languages, yet most benchmarks test whether a model can perform a task in a language rather than whether it commands the language itself, conflating fluency with proficiency. We introduce M-GATE (Multilingual Grammar, Accuracy in Translation, and Efficiency), a benchmark of linguistic proficiency spanning 30 typologically diverse languages from high- to low-resource. M-GATE comprises three tasks: grammatical error detection on linguist-crafted, adversarially selected sentences that turn on hard, language-specific phenomena; round-trip translation of shared English sources across 29 target languages, scored by a three-provider LLM judge panel validated against professional annotators; and a supplementary tokenizer-efficiency measure. We evaluate over 50 models in more than 80 configurations. Fluency and proficiency come apart sharply: models that translate competently sit near chance on the adversarial grammar items, the best reaching a Matthews correlation coefficient (MCC) of only 0.36, and their errors lean systematically toward under-flagging, accepting ungrammatical text rather than raising false alarms. Translation quality closely tracks a language’s share of pretraining data (r = 0.86 against log Common Crawl share), producing a steep low-resource penalty that is nonetheless narrowing with successive model releases. Enabling reasoning reliably improves translation, while its effect on error detection is smaller and for some models negative, so the best configuration is task-dependent. To resist contamination, test items are kept private behind a continuously updated public leaderboard, with illustrative examples released (this https URL).
[NLP-19] Efficient Knowledge Distillation for LLM s: Offline Top-K Logits and a Fused Chunked KL Loss
【速读】： 该论文旨在解决在严格延迟、成本及本地部署约束下，小型语言模型（Small Language Models, SLMs）训练中知识蒸馏（Knowledge Distillation, KD）效率低下的问题。当前主流方法依赖于从预训练大模型中压缩得到小模型，但传统在线蒸馏过程需反复访问教师模型以获取输出概率分布，导致计算开销大、内存占用高，严重限制了训练效率与上下文长度。其解决方案的关键在于两个系统级优化：一是提出离线知识蒸馏（offline KD），通过一次性缓存教师模型的Top-K logits并将其用于学生模型训练，避免了每轮迭代对教师模型的内存访问，使单次迭代速度提升约29%，单H200 GPU上的吞吐量最高提升41%；二是设计一种融合分块的KL散度损失（fused, chunked KL loss），该方法不显式生成完整词汇表规模的logit张量，使得峰值内存消耗仅与序列长度呈线性关系，从而突破了传统方法因内存瓶颈导致的上下文长度限制，支持在单卡上实现4倍于以往的上下文长度（32,768 tokens）。通过这两个核心技术，实现了大规模模型修复与数百次超参数实验的可负担性，并在输出头独立的简化基准测试中验证了其内存与迭代速率随序列长度从4K扩展至256K的优异缩放性能。相关代码已开源。
链接: https://arxiv.org/abs/2608.03796
作者: Bakbergen Ryskulov,Iker García-Ferrero,David Montero,David Jansen,Ali Hashemi,Jezabel R. Garcia,Antonio Tiene,Román Orús
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Patent Application Pending. EP26382987.1
Abstract:Small language models are often the only option for deployment under tight latency, cost, and on-premises constraints, but they are rarely trained from scratch: a compressed model is usually recovered through knowledge distillation (KD). This recovery step largely decides the final quality, yet it is expensive. We present a practitioner’s study of how to make distillation training efficient, organised around two systems contributions. First, we show that offline KD (caching the teacher’s top- K logits once and training the student against the cache) matches online distillation at near-identical training loss while removing the teacher from memory, running about 29% faster per iteration, and reaching up to 41% higher throughput on a single H200 GPU. Second, we introduce a \emphfused, chunked KL loss that never materialises the full vocabulary-sized logit tensor, making peak memory linear in the sequence length. This removes the memory spike that otherwise caps context length and lets us train at four times the context (32,768 tokens) on a single GPU. A separate output-head-only toy benchmark isolates the loss kernel and confirms its memory and iteration-rate scaling from 4K to 256K tokens. Together these make large-scale healing and hundreds of ablations affordable. We also report supporting ablations on loss design and sequence packing. We release our chunked-loss implementation: this https URL.
[NLP-20] Evaluating LLM s in Database Scenarios: A Lifecycle Benchmark for Assessing Their Potential in Core Database Tasks
【速读】： 该论文旨在解决当前大型语言模型（Large Language Models, LLMs）在数据库管理领域评估体系过于聚焦于文本到SQL（Text-to-SQL）任务，而忽视了从数据库初始模式设计到部署后维护的全生命周期（Database Lifecycle）能力评估的问题。现有基准测试未能全面反映真实场景下数据库智能体所需具备的多样化能力，导致对模型综合性能的评价存在显著偏差。为此，论文提出DBLifeBench，这是首个覆盖数据库全生命周期五个关键阶段——设计（Design）、实现（Implementation）、运维（Operation）、调试（Debugging）与维护（Maintenance）——的综合性评估基准。其核心解决方案在于引入Progressive-Text2SQL，一种基于结构化推理图（structured reasoning graphs）的渐进式任务范式，通过模拟人类迭代式问题求解过程，缓解自然语言描述模糊性与复杂SQL逻辑之间的认知错配。实验结果揭示：通用型模型在各阶段表现均衡，而专用的Text-to-SQL模型在非编码阶段（如设计与维护）表现出严重的“灾难性遗忘”（catastrophic forgetting）现象。该研究为构建真正具备全栈数据库智能（full-stack database intelligence）的系统奠定了基础。
链接: https://arxiv.org/abs/2608.03794
作者: Shunfan Zheng,Dongsheng Shi,Yue Li,Xin Yi,Linlin Wang,Gerard de Melo
机构: East China Normal University; Hasso Plattner Institute/University of Potsdam
类目: Databases (cs.DB); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Large Language Models (LLMs) are transforming database interaction paradigms, evolving from simple query translators to autonomous database administrators (DBAs). However, current evaluation benchmarks remain disproportionately fixated on Text-to-SQL tasks, neglecting the holistic Database Lifecycle-from initial schema design to post-deployment maintenance. This narrow focus fails to capture the diverse capabilities required for real-world database management. To bridge this gap, we introduce DBLifeBench, the first benchmark to evaluate LLMs across five critical lifecycle phases: Design, Implementation, Operation, Debugging, and Maintenance. Furthermore, addressing the cognitive mismatch between ambiguous natural language and complex SQL logic, we propose Progressive-Text2SQL, a novel task utilizing structured reasoning graphs to mimic human iterative problem-solving. Our extensive evaluation reveals a critical insight: while general-purpose models demonstrate balanced performance, specialized Text-to-SQL models suffer from ``catastrophic forgetting’’ in non-coding phases like design and maintenance. DBLifeBench serves as a foundational step toward evaluating and building true full-stack database intelligence.
[NLP-21] MDLMPE: Distribution Aware Positional Encoding for Masked Diffusion Language Models
【速读】： 该论文旨在解决掩码扩散语言模型（Masked Diffusion Language Models, MDLMs）在位置编码（Positional Encoding）设计上的根本性缺陷。传统的位置编码方法（如RoPE）虽能捕捉序列顺序与成对距离，但无法感知MDLM中动态变化的、非连续的已揭示/被遮蔽标记结构，而这种结构是其去噪过程的核心特征。为此，本文提出一种专为掩码扩散设计的位置编码方法——MDLMPE，其核心创新在于将标记的可访问状态（token availability）显式建模为二进制序列，并通过具有局部敏感性的高斯加权机制捕获其空间分布特性，再利用余弦基底进行谱变换，生成具备分布感知能力的位置特征。这些特征经轻量级MLP映射为角度偏移量，进而调制标准RoPE的相位，实现对动态可用性结构的精准建模。大量实验表明，MDLMPE在监督微调、预训练、零样本评估及块级扩散等多种场景下均显著优于传统位置编码方法，且消融实验证明，可访问状态建模、高斯局部性、谱基底表示与嵌入注入的完整组合带来最优性能。研究结果确立了动态标记可用性分布作为掩码扩散语言模型中一个有效且关键的位置信号。
链接: https://arxiv.org/abs/2608.03769
作者: Tong Ling,Hang Lei,Feng Xiao,Changhui Sun,Jiahang Xie,Hao Liu,Lu Liu,Yanlong Du
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Masked diffusion language models (MDLMs) enable parallel generation and bidirectional context modeling, but their positional context differs fundamentally from that of autoregressive (AR) models. Whereas AR decoding exposes a contiguous prefix, MDLM denoising produces dynamic, non-contiguous configurations of revealed and masked tokens. Conventional positional encodings such as RoPE capture sequence order and pairwise displacement but remain insensitive to this evolving token-availability structure. To address this limitation, we propose MDLMPE, a positional encoding designed specifically for masked diffusion. To the best of our knowledge, MDLMPE is the first method to make positional representations explicitly aware of the changing revealed/masked configuration. It represents token availability as a binary sequence, applies distance-aware Gaussian weighting, and projects the resulting pattern through a cosine basis to obtain distribution-aware positional features. These features are added to token embeddings and mapped by a lightweight MLP to angular offsets that modulate the standard RoPE phases. Extensive experiments on LLaDA and DREAM demonstrate that MDLMPE generally outperforms conventional positional encoding methods across supervised fine-tuning, pretraining, zero-shot evaluation, and block-diffusion settings. Further ablations show that the complete combination of availability state, Gaussian locality, spectral basis, and embedding injection yields the strongest result. These results establish the evolving token-availability distribution as a useful positional signal for masked diffusion language models.
[NLP-22] Risky Business: Measuring The Faithfulness-Safety Tension
【速读】： 该论文旨在解决大模型在链式思维（Chain-of-Thought, CoT）推理过程中存在的可监控性与安全性之间的对齐张力问题。具体而言，模型需具备足够的忠实性（faithfulness），即输出严格源于其推理过程，以支持有效监控；同时又需具备鲁棒性，能够识别并拒绝不安全或不合逻辑的推理路径。当前的大规模推理模型（Large Reasoning Models, LRMs）在这一权衡上表现不佳：一方面，如DeepSeek-R1-Llama-70B虽具有高忠实性（97.5%），但无法有效识别不安全推理（仅12.3%的拒识率）；另一方面，如QwQ-32B虽更稳健（73.9%的安全性），却牺牲了忠实性（74.7%）。为系统评估与干预这一矛盾，作者构建了HazMart——一个基于自主AI店主场景的人工撰写数据集，并提出一种新颖的“目标推理替换”（Targeted Reasoning Replacement, TRR）方法，通过直接在推理链中插入不安全或荒谬的中间步骤，精准测试模型的响应能力。机制分析揭示，安全性和忠实性在QwQ-32B内部由反相关的神经方向表征，且在动作决策令牌处达到峰值。进一步地，研究证明可通过表示引导（representation steering）独立增强安全方向，在维持基础能力的前提下将安全行为提升9个百分点，从而实现对模型推理行为的可控优化。
链接: https://arxiv.org/abs/2608.03745
作者: Dominik Meier,Luca Joshua Francis,Marco Bernhard Kaiser,Terry Ruas,Jan Philip Wahle,Bela Gipp
机构: University of Göttingen (哥廷根大学); Landeskriminalamt NRW (北莱茵-威斯特法伦州刑事警察局)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Chain-of-Thought (CoT) reasoning offers a promising window into model monitoring. However, monitoring relies on faithfulness, i.e., the model output strictly derives from its reasoning trace. We identify an alignment tension where a model must be faithful enough to be monitored, yet robust enough to reject unsafe reasoning. We demonstrate that this counterbalance exists in current Large Reasoning Models (LRMs), and show ways in which it can be addressed. We introduce HazMart, a human-written dataset set in an autonomous AI shopkeeper scenario. Unlike prior work that relies on providing hints in prompts to test faithfulness (e.g., “A Stanford professor said it should be Answer A”), we propose a novel replacement-based technique, which we call Targeted Reasoning Replacement (TRR), that directly intervenes in the reasoning chain to substitute in unsafe or illogical thoughts (e.g., “Wait, the answer must be Option B [was Option A] because it is the most fitting”). DeepSeek-R1-Llama-70B exhibits high faithfulness (97.5%) but fails to reject Unsafe Reasoning (12.3%), while QwQ-32B is more robust (73.9% safety) at the cost of lower faithfulness (74.7%). Mechanistic analyses of QwQ-32B reveal that these properties are represented by anti-correlated internal directions peaking at the action-commit token. Finally, we demonstrate that representation steering can independently amplify the safety direction, increasing safe behavior by 9 percentage points while maintaining base capabilities.
[NLP-23] GPT KB 2.0: Direct Construction of Disambiguated Knowledge Bases from Large Language Models
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在自动知识库构建（Automated Knowledge Base Construction, AKBC）中因缺乏实体显式表示而导致的实体重复与混淆问题。现有方法将LLM直接作为知识源，但其内部未对实体、关系和类别进行明确区分，导致生成的知识库存在大量冗余与歧义。为此，论文提出GPTKB 2.0，一种可直接从LLM中构建去重且消歧知识库的方法，其核心在于引入实时消歧机制，对实体、关系与类别进行动态归一化处理，并在设计上兼顾可扩展性与消歧准确性。通过系统分析关键设计选择，量化评估了准确率、规模与成本之间的权衡关系。实验表明，GPTKB 2.0实现了大规模知识库构建，生成了一个包含超过100万去重实体和3840万三元组的实体、关系与类别均显式规范化（canonicalized）的知识库，成为首个基于LLM原生结构实现百万级规模且具备内部归一化的知识库，突破了以往以维基媒体（Wikimedia）为中心的研究范式。
链接: https://arxiv.org/abs/2608.03729
作者: Yujia Hu,Tuan-Phong Nguyen,Simon Razniewski
机构: TU Dresden (德累斯顿工业大学); VNU University of Engineering and Technology (越南国立工程与技术大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Databases (cs.DB)
备注: 19 pages, 4 figures
Abstract:Automated Knowledge Base Construction (AKBC) is a core NLP task, and recent work proposes generating knowledge bases directly from large language models (LLMs), treating the model itself as the knowledge source. However, LLMs natively possess no representation of entities, leading to duplicate entries as well as conflations. We propose GPTKB 2.0, a methodology for constructing disambiguated KBs directly from LLMs. GPTKB 2.0 incorporates on-the-fly disambiguation of entities, relations and classes, and is meticulously designed to satisfy both scalability and disambiguation accuracy. We analyze the central design decisions and characterize the trade-offs between accuracy, scale, and cost. We execute GPTKB 2.0 at scale, obtaining a materialized KB containing over 1M disambiguated entities and 38.4M triples. This represents the first million-scale LLM-native KB with explicit internal canonicalization of entities, relations, and classes, a significant departure from prior Wikimedia-centric works. GPTKB 2.0 is available at this https URL.
[NLP-24] When Outputs Disperse Does Epistemic Revision Follow? A Black-Box Coupling Diagnostic for Machine Collectives
【速读】： 该论文旨在解决生成式人工智能（Generative AI）集体在协作决策中出现的“表象多样性—实质一致性”问题：尽管多个大语言模型（LLM）代理产生的输出在表面上呈现多样性，但其核心结论可能并未真正发生认知立场的改变，即存在“前提保持型重述”而非真正的观点修正。这一现象导致传统以分歧为认知多样性标志的集体智能范式在LLM集体中失效。其解决方案的关键在于提出并验证“分散-修订耦合”（dispersion-revision coupling）的可操作诊断框架，通过黑箱方式仅基于生成文本进行评估，不依赖模型内部表示。该框架包含两个独立测量通道：输出通道采用一致性指数（Coherence Index, CI），验证干预是否真实提升了输出在嵌入空间中的分散度；认知通道则通过逐轮立场标注，判断集体是否发生实质性的认知立场修订。研究进一步提出元预测清晰度系统（Meta-Predictive Clarity System, MPCS），集成再分化协议（Re-Differentiation Protocol, RDP），用于在输出过度收敛时主动触发多样化输出。实验结果表明，在gpt-4o-mini配置下，条件性异议显著提升错误前提恢复能力（+17.7分，p<1e-6），而静态角色多样性反而损害性能（-8.1分，p=0.007）；而在gemini-2.5-flash上，即使干预后输出分散度下降，却未带来性能提升（26.1% vs 27.1%，p=0.84），且两模型处理效果差异显著（z=3.79，p<0.001）。机制分析显示，Gemini主要通过框架内分歧实现“重述”而非“让步”（94%重述，24%让步），揭示其缺乏实质性认知修订。研究建议在报告准确性的同时，应同步披露每项干预后的立场转变率与前提保留率，以更准确评估集体智能的真实性与有效性。
链接: https://arxiv.org/abs/2608.03722
作者: Molood Arman
机构: Independent Researcher(独立研究员); Lyon(里昂); France(法国)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Collective intelligence research treats disagreement as evidence of epistemic diversity: if agents express different views, the group should retain capacity to revise. In LLM collectives this proxy can break: agents can produce diverse-looking arguments while preserving the same conclusion. We operationalize dispersion-revision coupling: the degree to which an intervention that verifiably increases the dispersion of a collective’s outputs in embedding space is accompanied by genuine revision of its epistemic stance rather than premise-preserving reformulation. The diagnostic is black-box: it operates on generated text alone and makes no claims about the internal representations of the generating models. Two channels are measured independently: an output channel, the Coherence Index (CI), verifies that the intervention changed output dispersion; an epistemic channel, per-turn stance annotation, measures whether the collective revised. We propose CI with the Meta-Predictive Clarity System (MPCS), which inserts a Re-Differentiation Protocol (RDP) when outputs over-converge, as a reusable method for estimating this coupling regime. We evaluate five-agent collectives from two configurations (gpt-4o-mini and gemini-2.5-flash; 310 paired episodes per condition). On gpt-4o-mini, conditional dissent improves false-premise recovery by +17.7 points (p1e-6) while static persona diversity harms recovery (-8.1, p=.007). On gemini-2.5-flash, the same intervention at a comparable budget yields no gain (26.1% vs 27.1%, p=.84) despite a verified dispersion drop; the two treatment effects differ from each other (z=3.79, p.001). Mechanism tagging shows Gemini preserves the false premise via intra-framework dissent: 94% of tagged post-RDP responses reformulate rather than concede (vs 24% on GPT). We recommend reporting per-intervention stance shift and premise-preservation rate alongside accuracy.
[NLP-25] Detecting Hallucinations and Recovering Verified Answers in Arabic Islamic Question Answering
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在回答伊斯兰教相关问题时产生难以察觉的事实性错误（即幻觉，Hallucination）的问题。针对“Islamic Hallucination Detection and Find the Truth”这一任务，其核心挑战在于实现统一的两阶段预测：首先判断阿拉伯语生成答案是否存在幻觉，其次从六个密切相关候选答案中选出经过验证的正确答案。解决方案的关键在于基于共享任务提供的伊斯兰知识数据集（包含600个问答实例，其中341个为幻觉答案），对微调后的Google Gemma-4-12B-it模型采用确定性解码策略进行推理，并通过输出规范化提取幻觉标签与最优选项。实验结果表明，系统在幻觉检测上达到0.928的宏平均F1分数和0.935的标签准确率，答案选择准确率为0.895，综合得分达0.912，展现出优异的双阶段性能；然而，答案选择准确率相对较低，反映出在多个合理选项中甄别出唯一真实答案仍具挑战性。
链接: https://arxiv.org/abs/2608.03720
作者: Khaled Ziani
机构: Independent Researcher(独立研究员); Paris, France
类目: Computation and Language (cs.CL)
备注:
Abstract:Large language models can generate fluent responses to Islamic questions while introducing factual errors that are difficult to identify. This paper presents our system for \textscHalluScoring 2026 Task 2.1, \textitIslamic Hallucination Detection and Find the Truth. The task requires a unified two-step prediction: determining whether an Arabic answer generated by an LLM is hallucinated and selecting the verified answer from six closely related candidate options. We use the Islamic knowledge dataset provided by the shared task, which contains 600 question–answer instances, including 341 hallucinated and 259 non-hallucinated answers. Our system is based on the fine-tuned \textttgoogle/gemma-4-12B-it model and uses deterministic decoding during inference. The generated outputs are normalized to extract the hallucination label and the selected option. The system achieves a Macro-F1 score of 0.928 and a label accuracy of 0.935 for hallucination detection, together with an option accuracy of 0.895 for answer selection. These results yield a combined score of 0.912, demonstrating strong performance across both stages of the task. The lower option-selection accuracy indicates that distinguishing the verified answer from plausible alternatives remains more challenging than detecting hallucinated responses.
[NLP-26] Attention is Case-Sensitive ECCV2026
【速读】： 该论文旨在解决生成式模型在处理文本时，其内部注意力机制如何被字符大小写等格式化特征所影响的问题。核心问题是：尽管大写字母在人类视觉感知中具有天然的显著性（salience），但这种特性是否同样存在于大型语言模型（LLM）与视觉-语言模型（VLM）的内部注意力分配中？研究发现，字符大小写变化能够显著调节预训练Transformer模型的注意力分布——将目标信息以交替或大写形式呈现于小写上下文中，可有效集中模型注意力于特定文本片段，这一现象在所有非推理类模型中均具普遍性，且不依赖模型微调或访问权限。其关键解决方案在于揭示了一种此前未被充分探索的潜在属性：即预训练模型内部对文字格式的隐式敏感性，而非人为设计的提示工程策略。进一步分析表明，该“大小写效应”虽能稳定引导注意力，但对下游任务准确率的影响并不直接正向，尤其在高熵语境下反而可能导致性能下降；而在具备“思考”阶段的推理模型中，该效应因语义缓冲机制而被缓解。此外，在视觉-语言模型中，相同机制促使跨模态注意力沿两个耦合维度重构：主要表现为从图像到文本提示的整体注意力转移，次要表现为剩余视觉注意力向目标区域的聚焦。因此，本研究提出大小写作为一种无需模型访问或微调的零样本注意力调控机制，为理解预训练模型如何内化排版强调提供了新的基础视角。
链接: https://arxiv.org/abs/2608.03711
作者: Maximilian Dillitzer,Tin Stribor Sohn,Jason J. Corso,Michael Auerbach
机构: University of Applied Science Esslingen; Karlsruhe Institute of Technology; Dr. Ing. h.c. F. Porsche AG; University of Michigan
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: Accepted at ECCV 2026
Abstract:In human visual perception, uppercase lettering serves as a natural salience cue that captures attention within lowercase text. In this paper, we present a systematic empirical characterization study revealing that Large Language Models (LLMs) exhibit an analogous property: letter casing modulates internal attention allocation. Through analysis across 13 models, nine LLMs and four Vision-Language Models (VLMs), with diverse tokenization schemes, we show that formatting target information in alternating or uppercase against a lowercase context concentrates attention on those textual spans. In text this effect is universal, holding across every evaluated non-reasoning model. We frame it as a previously under-explored latent property of pretrained transformers rather than a prescriptive method. Our investigation reveals a central attention-performance divergence: while this “casing effect” robustly shifts attention, its impact on downstream accuracy is non-trivial, increased concentration does not inherently improve task accuracy and, in high-entropy contexts like alternating case, can degrade it. We further identify a boundary condition: the deliberative “thinking” phase in reasoning models acts as a semantic buffer that mitigates typographic sensitivity in text. Extending the study to VLMs, we find the effect transfers partially: the same prompt-side casing reorganizes cross-modal attention along two coupled axes, predominantly a macroscopic disengagement from the image toward the text prompt, and secondarily a concentration of the residual visual attention on the target region. By isolating casing as a zero-shot mechanism for attention steering that requires no model access or fine-tuning, we provide a new foundational understanding of how pretraining internalizes typographic emphasis.
[NLP-27] Predicting Deep Neural Network Training Outcomes from Early Training Telemetry
【速读】： 该论文旨在解决深度神经网络大规模超参数搜索中计算资源浪费的问题，即大量训练配置在初期即已注定失败，但传统方法仍持续投入计算资源。其核心解决方案是利用单次训练运行在前几个训练周期内产生的早期遥测数据（包括每轮损失、训练准确率、梯度信噪比、权重范数增长及激活饱和快照）与对应的超参数，构建一个无需依赖其他训练运行即可预测最终性能的早期预测模型。关键创新在于：仅使用前五个训练周期的遥测信息，结合梯度提升树（gradient-boosted trees），即可实现对最终测试准确率的回归预测（R² = 0.92–0.99）以及相对性能排序和训练动态失败（如数值发散）的分类预测（ROC-AUC = 0.983–0.998），且在单个训练周期后即具备有效预测能力。研究表明，梯度与权重层面的遥测信号相比仅依赖损失与准确率曲线能带来统计上显著的性能提升，尽管实际增益因任务域而异；同时，相似架构间具有较强的迁移能力，而跨数据集迁移受限于准确率量级差异而非预测关系本身的失效。该成果为高效计算资源分配提供了可操作的决策支持信号，并强调了人工监督在自动化干预中的必要性。
链接: https://arxiv.org/abs/2608.03709
作者: Ranjita Naik,Anh D. Nguyen,Pankaj Kumar Singh
机构: Georgia Institute of Technology (佐治亚理工学院)
类目: Computation and Language (cs.CL)
备注: 21 pages, 6 figures, 7 tables, includes appendices
Abstract:Large hyperparameter sweeps for deep neural networks spend substantial compute on configurations that are effectively doomed from the first few epochs. We study whether a single training run’s own early telemetry - per-epoch loss, training accuracy, gradient signal-to-noise ratio, weight-norm growth, and an activation-saturation snapshot - together with its sampled hyperparameters, can predict that run’s eventual outcome without reference to other runs. We evaluate three prediction tasks: final test accuracy, relative performance within a domain, and training-dynamics failure, including numerical divergence. Across 23,788 training runs spanning six architecture/dataset combinations, gradient-boosted trees using only the first five epochs of telemetry achieve R^2 = 0.92-0.99 for final-accuracy regression and ROC-AUC = 0.983-0.998 for relative classification on a permanently held-out set of hyperparameter configurations. Useful prediction is already available after a single epoch. A paired ablation shows that gradient- and weight-level telemetry provides a statistically consistent improvement over loss and accuracy curves alone, although the practical gain varies by domain. Transfer is strong between similar architectures, while cross-dataset transfer is limited mainly by differences in accuracy scale rather than loss of the underlying relationship. These results suggest that early-training telemetry can provide a practical decision-support signal for compute allocation while motivating human oversight for any automated intervention.
[NLP-28] When Agents Learn to Be You: Benchmarking Privacy Leakage Impersonation Risk and Defenses in Persona Skills
【速读】： 该论文旨在解决生成式人工智能（Generative AI）中“人格技能”（persona-skill）管道所带来的隐私与安全风险问题。具体而言，人格技能通过提炼个人交互历史，将分散的个性化信号压缩为可移植、可复用的执行单元，虽提升了下游代理的个性化能力，但也因信号集中与重复利用，放大了隐私泄露风险，并对传统针对单条记录或基于检索的记忆防御机制构成挑战。其解决方案的关键在于提出AntiSkillBench——一个端到端的评估基准，系统性地衡量人格技能管道中的各类风险与防御效果。该基准包含：（i）由50个行为丰富的人格档案构建的7,500条基于人格的对话轨迹数据集；（ii）涵盖三种技能蒸馏策略下技能级隐私泄露、代理级属性披露及行为仿冒的评估套件；（iii）整合在线与离线干预的四类防御配置，包括主动风险抑制与被动溯源保护。实验表明，人格技能风险在主流大模型中普遍存在，且从显式属性延伸至沟通风格与人格特质等深层特征，而现有防御手段表现出显著的局限性和蒸馏策略依赖性，难以泛化。因此，AntiSkillBench揭示了发展具备隐私保护与真实性感知能力的人格技能的迫切需求与核心挑战。
链接: https://arxiv.org/abs/2608.03700
作者: Yongli Xiang,Zhifang Zhang,Bojun Yang,Ziming Hong,Lei Feng,Miao Xu,Tongliang Liu
机构: 1. University of Technology Sydney (悉尼科技大学); 2. Tsinghua University (清华大学); 3. Alibaba Cloud (阿里云); 4. The University of Melbourne (墨尔本大学)
类目: Cryptography and Security (cs.CR); Computation and Language (cs.CL); Computers and Society (cs.CY)
备注: Project page: this https URL
Abstract:Persona skills distill personal interaction histories into portable and executable artifacts for downstream agents. While enabling flexible personalization, this process concentrates fragmented personal signals, amplifies their impact through reuse, and challenges defenses designed for individual records or retrieval-based memory. To systematically investigate the safety of the persona-skill pipeline, we introduce AntiSkillBench, an end-to-end benchmark for evaluating risks and defenses across the persona-skill pipeline. It comprises: (i) a dataset of 7,500 persona-grounded dialogue traces, constructed from 50 behaviorally rich profiles spanning diverse task scenarios; (ii) an evaluation suite that measures skill-level privacy leakage and agent-level attribute disclosure and behavioral impersonation across three skill-distillation strategies; and (iii) a defense evaluation covering four configurations across online and post-hoc interventions, including active risk suppression and passive provenance protection. Experiments across three frontier agents show that persona-skill risks persist across agent backbones and distillation protocols, extending from explicit attributes to communication styles and personality traits. Existing defenses exhibit limited and distillation-dependent effectiveness, failing to generalize across risk and distillation strategies. These results highlight AntiSkillBench as a challenging benchmark for developing privacy-preserving and authenticity-aware persona skills.
[NLP-29] VetScore: Risk-Weighted Fact Verification for Veterinary Long-Form QA with Citations
【速读】： 该论文旨在解决生成式AI在兽医学等高风险领域中，尽管引用了文献摘录以增强输出可靠性，但生成的主张仍可能与所引摘录不符的问题，即确保生成内容对源文献的忠实性（faithfulness）和安全性。其解决方案的关键在于提出VetScore——一种多步骤评估方法，通过将长篇回答分解为独立主张（claim），分别评估每个主张的潜在危害程度（harm potential）并检验其与源摘录的忠实度，最终计算出加权风险调整后的综合得分。该方法不仅在小规模判别模型下与兽医专家评价高度相关，还具备多维度可解释性，显著提升了生成内容在关键应用中的可信度与安全性。
链接: https://arxiv.org/abs/2608.03675
作者: Ivan Kartáč,Jan Tovarys,Mateusz Lango,Ondřej Dušek
机构: PrimVeterinary; Institute of Formal and Applied Linguistics, Charles University
类目: Computation and Language (cs.CL)
备注:
Abstract:Citation excerpts can be used to increase the reliability of generated outputs and their faithfulness to cited sources, which is especially important in high-stakes domains such as human and veterinary medicine. However, this does not guarantee that generated claims are faithful to the provided excerpts. We present VetScore, a multi-step evaluation method for veterinary long-form question answering, designed to assess how well are generated claims supported by the provided excerpts, weighing this information by each claim’s harm potential. VetScore first segments the output and decomposes it into individual claims, then scores each claim with respect to its harm potential and evaluates its faithfulness to source excerpts, and finally calculates the overall risk-adjusted score. We collect an expert-annotated meta-evaluation dataset, evaluate our approach with a range of judge models, and show that it achieves high correlations with veterinary experts even with small judge models, while offering explainability across multiple dimensions.
[NLP-30] How Closely Do LLM Reviews Align with Human Peer Review?
【速读】： 该论文旨在解决当前科学评审中生成式 AI（Generative AI）模型在评审一致性与人类评审标准及会议决策之间对齐程度不足的问题，尤其关注不同供应商的大型语言模型（LLM）在相同受控环境下是否能一致地反映人类评审者的判断偏好与决策逻辑。其解决方案的关键在于构建一个跨供应商的系统性对比框架，通过将 OpenAI GPT-5.4、Google Gemini 3.1 Pro Preview 与 Anthropic Claude Opus 4.6 在完全一致的指令与评分尺度下对 300 篇主题匹配的 ICLR 2026 提交稿件进行盲审，并与真实的人类评审结果及最终录用决策进行多维度比对，涵盖宏观与微观决策类别的一致性、推荐评分尺度的使用差异以及识别出的缺陷主题的语义重合度。研究发现，尽管所有模型均能区分接受与拒绝稿件，但均未能再现人类评审中“口头报告”与“海报展示”的细粒度区分；各模型在评分模式上表现出显著的供应商特异性，如 Gemini 给分普遍偏高，而 OpenAI 与 Claude 在拒稿和海报论文上更接近人类，但在口头报告论文上则更为严苛；此外，模型更常指出缺乏基线对比，而人类更关注计算效率问题，表明广义决策对齐并不等同于在细节判断或评审优先级上的共识。
链接: https://arxiv.org/abs/2608.03659
作者: Abraham Camelo-Guerrero,Jairo Diaz-Rodriguez
机构: York University ( York 大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models (LLMs) are increasingly used to generate scientific reviews, yet existing evaluations rarely examine whether different providers align with both conference decisions and human reviewing priorities within the same controlled setting. We compare reviews from OpenAI GPT-5.4, Google Gemini 3.1 Pro Preview, and Anthropic Claude Opus 4.6 with human reviews and final decisions for 300 topic-matched ICLR 2026 submissions, equally divided among oral, poster, and rejected papers. Each model reviewed every paper using identical instructions and rating scales after decision information was removed. Our study contributes a cross-provider analysis of three complementary dimensions: alignment with broad and fine-grained decision categories, differences in recommendation-scale usage, and thematic agreement in identified weaknesses. All three LLMs distinguished accepted from rejected papers, but none reproduced the oral versus poster distinction present in human ratings. Scoring patterns were provider-specific: Gemini assigned systematically higher ratings, while OpenAI and Claude were closer to humans for rejected and poster papers but more critical of oral papers. Human and LLM reviews also differed in emphasis, with LLMs more frequently identifying missing baseline comparisons and humans more often raising computational-efficiency concerns. These results show that broad decision alignment does not imply agreement with finer human judgments or reviewing priorities.
[NLP-31] Decoupling Generation and Selection for Budget-Constrained Faithful Summarization
【速读】： 该论文旨在解决生成式摘要模型在事实一致性、冗余性和长度控制方面存在的固有缺陷，尤其是在句子数量受限的摘要任务中。其核心解决方案是提出一种模块化生成与选择框架（modular generation-and-selection framework），将摘要生成过程解耦为两个阶段：首先由预训练生成器（pretrained generator）生成多个候选摘要，再将其分解为句级候选；随后通过组合选择器（combinatorial selector）在显式长度预算约束下，综合平衡相关性、事实正确性和冗余度，从而构建最终摘要。该框架不依赖于重新训练生成器，即可兼容最大边际相关性（MMR）、整数线性规划（ILP）以及受确定性点过程（DPP）启发的对数行列式目标，实现高效且灵活的优化。实验结果表明，在CNN/DailyMail、Multi-News、FaithBench和TofuEval等多个数据集上，该方法显著提升了事实准确性和源文本依存性指标，尤其在多文档摘要任务中表现突出，尽管参考文本重叠度略有下降。人工评估进一步验证了摘要在一致性、相关性、清晰度和简洁性方面的提升，仅轻微影响连贯性。研究表明，将生成与选择解耦可作为一种无需依赖特定模型架构的事实性增强机制，具有良好的通用性和实用性。
链接: https://arxiv.org/abs/2608.03655
作者: Zeyu Wang,Guanghua Wang,Meng Xu
机构: Kean University, USA
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Abstractive summarization models remain vulnerable to factual inconsistency, redundancy, and weak length control. We propose a modular generation-and-selection framework for sentence-budget-constrained summarization. A pretrained generator produces multiple candidate summaries, which are decomposed into sentence-level candidates. A combinatorial selector then constructs the final summary by balancing relevance, factuality, and redundancy under an explicit budget. The framework supports MMR, ILP, and a DPP-inspired log-determinant objective without retraining the generator. Experiments on CNN/DailyMail, Multi-News, FaithBench, and TofuEval show consistent improvements in factuality and source-grounding metrics, especially for multi-document summarization, at the cost of lower reference-overlap scores. Human evaluation further indicates higher perceived consistency, relevance, clarity, and conciseness, with a small reduction in coherence. These results show that decoupling generation from selection provides a model-agnostic mechanism for improving factual grounding. Code is available at this https URL.
[NLP-32] LoopMTP: A looped transformer guided by latent multi-token prediction
【速读】： 该论文旨在解决循环变压器（Looped Transformers）在迭代推理过程中存在的潜在过度思考（latent overthinking）与计算冗余问题，其核心挑战在于：在多轮循环中，中间表示缺乏来自未来信息的指导，导致模型无法有效区分哪些计算是必要的。解决方案的关键在于提出一种名为 \textsc{LoopMTP} 的新方法，通过在隐空间中建立结构对应关系，实现对未来的密集且前向的监督。具体而言，\textsc{LoopMTP} 通过软对齐第 $ t $ 轮循环的隐藏状态与未来 $ t $ 步的词元嵌入，使模型具备预测未来多个词元的能力；同时引入轻量级门控机制以保留跨循环的有效信息。该方法显著提升了平均准确率（相对提升达8.1%），并在长达15轮循环的训练中保持稳定，从而实现了高效、精准的参数高效推理。
链接: https://arxiv.org/abs/2608.03624
作者: Behzad Shomali,Markus Frey,David Berghaus,Joachim Koehler,Mehdi Ali
机构: Lamarr Institute (拉马尔研究所); University of Bonn (波恩大学); Fraunhofer IAIS (弗劳恩霍夫信息与通信技术研究所)
类目: Computation and Language (cs.CL)
备注:
Abstract:Looped transformers have emerged as a parameter-efficient alternative to scaling depth for strong reasoning. By reusing one stack of layers across T iterations, they attain the effective depth and reasoning capabilities of larger models at a fixed parameter count. Yet existing approaches suffer from latent overthinking and undifferentiated computation, largely because intermediate representations receive no guidance across loops. Multi-token prediction (MTP) supplies exactly the dense, forward-looking supervision the loop is missing. We propose \textscLoopMTP, which links the two through a structural correspondence in latent space: a model that loops T times can anticipate T future tokens. \textscLoopMTP realizes this by softly aligning the hidden state of loop t with the embedding of the token t steps ahead, while a lightweight gate preserves useful information across iterations. \textscLoopMTP improves average accuracy by up to 8.1% (relative) over the non-looped baseline, with training remaining stable for up to 15 loops.
[NLP-33] A machine-readable catalogue of the Tsiolkovsky papers (fond 555 Archive of the Russian Academy of Sciences) and a way to measure how well its handwriting can be read
【速读】： 该论文旨在解决历史档案数字化后缺乏可检索性与机器可读性的核心问题，具体针对俄罗斯科学院档案馆中康斯坦丁·齐奥尔科夫斯基（Konstantin Tsiolkovsky）个人档案（档案号555）的文本资源无法进行全文搜索、无结构化目录及缺乏标注数据集的问题。其解决方案的关键在于构建一个全面的机器可读目录，涵盖2,019份文件和51,008页扫描件，并实现页面级的手写体与印刷体分类；同时，通过建立一个持续增长的机器转录语料库（当前包含322份文件、5,454页），提升对历史手稿的自动化处理能力。尤为关键的是，论文提出一种无需真实标签（ground truth）即可评估手写文字识别准确率的方法：利用打字机时代的双版本文本（手稿与打印稿）进行比对，由于源文本与处理流程一致，差异仅源于页面识别难度，从而可分离出阅读错误率。在294对匹配样本中，两读取结果平均仅在37%的词汇上达成一致；在两个具备已出版版本的文件上验证后，该方法估计值与真实情况偏差小于1个百分点，且页面排序相关性高达0.92，证明其有效性。基于此，研究进一步指出，同一作品的不同修订版本间仅有19%的词汇重合，低于单页双读一致性水平，因此无法以现有识别精度进行逐词校勘，这一负结果被明确报告，并作为约束条件嵌入工具设计中，确保技术应用的合理性与边界清晰。
链接: https://arxiv.org/abs/2608.03617
作者: Vladimir Beskorovainyi
机构: Besk Tech(贝斯科技); Moscow Institute of Physics and Technology (MIPT)(莫斯科物理技术研究所)
类目: Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV); Digital Libraries (cs.DL)
备注: 8 pages, 6 tables. Dataset and code: this https URL ; archived at this https URL (CC0 catalogue, MIT code)
Abstract:The personal archive of Konstantin Tsiolkovsky (1857-1935) is held as fond 555 of the Archive of the Russian Academy of Sciences. The archive scanned the fond and published the images, but with no queryable catalogue, no full-text search and no dataset: the holdings can only be browsed one page at a time. This paper describes a machine-readable catalogue of all 2,019 files and 51,008 scans, a dating for 1,969 files taken from the archive’s own descriptions, a page-level classification of every scan into handwriting and typescript, and a growing corpus of machine transcriptions (currently 322 files, 5,454 scans). It also reports a way to measure handwritten-text-recognition accuracy in an archive with no ground truth. Archives of the typewriter era often preserve one text twice, as manuscript and as a typed copy; transcribing both and comparing isolates the reading error, since source and pipeline are identical and only page difficulty differs. Across 294 such pairs from 27 files, two readings of a handwritten page agree on a median 37% of words. On two files that also have a published edition the estimate can be checked against ground truth: it is unbiased to within a percentage point and ranks pages as the truth does (rank correlation 0.92 where the edition is a faithful witness). This bounds use: two variants of one work here share 19% of words, below the rate at which two readings of a single page agree, so the redactions cannot be collated word by word at this quality. That negative result is reported as such, and the constraint is built into the tool. Comments: 8 pages, 6 tables. Dataset and code: this https URL ; archived at this https URL (CC0 catalogue, MIT code) Subjects: Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV); Digital Libraries (cs.DL) ACMclasses: H.3.7; I.7.5 Cite as: arXiv:2608.03617 [cs.CL] (or arXiv:2608.03617v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2608.03617 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-34] Language-Specialized Multi-Teacher On-Policy Distillation for Multilingual LLM -Based ASR
【速读】： 该论文旨在解决大规模多语言语音识别（Multilingual ASR）系统在联合建模语种间声学、音系及词汇特征差异显著的语言时所面临的优化冲突问题，该冲突导致各语言的专用性能被削弱。其核心解决方案是提出一种语言专精型多教师在线蒸馏方法（Language-Specialized Multi-Teacher On-Policy Distillation, LS-MOPD），通过将语言特异性知识获取与多语言能力整合相解耦：首先利用强化学习（Reinforcement Learning, RL）独立优化各语言专用教师模型，随后通过语言路由机制与词元级多教师蒸馏策略，将教师们的专长知识融合至一个通用型多语言学生模型中，从而有效缓解直接跨语言优化带来的冲突。研究进一步对比了静态与动态声学前缀配置，探究教师-学生前缀一致性对在线蒸馏效果的影响。在涵盖普通话、方言、粤语和英语的多个基准测试上，LS-MOPD显著优于传统强化学习基线，并持续突破最优教师模型所定义的性能边界，展现出在多语言语音识别任务中超越所有教师模型的泛化潜力。
链接: https://arxiv.org/abs/2608.03610
作者: Yuan Xie,Jiaqi Song,Xianliang Wang,Ming Lei,Jie Gao,Jie Wu
机构: 未知
类目: Computation and Language (cs.CL); Sound (cs.SD); Audio and Speech Processing (eess.AS)
备注:
Abstract:Modern LLM-based ASR systems have established multilingual capability as a standard feature, leveraging large-scale multilingual corpora and LLMs’ cross-lingual knowledge to achieve competitive performance across multilingual benchmarks. However, joint modeling of languages with heterogeneous acoustic, phonological, and lexical characteristics inevitably introduces optimization conflicts, undermining language-wise specialization. To address this challenge, we propose Language-Specialized Multi-Teacher On-Policy Distillation (LS-MOPD), which decouples language-specific knowledge acquisition from multilingual capability integration: language-specialized teachers are independently optimized via reinforcement learning (RL), after which their expertise is integrated into a generalist multilingual student through language routing and token-level multi-teacher distillation, thereby reducing direct cross-lingual optimization conflicts. We further explore two acoustic-prefix configurations, static and dynamic, to examine how teacher–student prefix consistency influences the efficacy of on-policy distillation. Experiments on benchmarks covering Mandarin, Mandarin subdialects, Cantonese, and English demonstrate that LS-MOPD substantially outperforms RL baselines and consistently surpasses the empirical performance envelope defined by best-performing RL teachers, revealing its potential to generalize beyond all teachers in multilingual ASR.
[NLP-35] Disentangling Language Modeling and Boundaries
【速读】： 该论文旨在解决当前语言模型之间知识迁移受限于各模型独有分词器（tokenization）所带来的障碍问题。现有模型因采用不同的分词策略，导致其输出空间不一致，从而阻碍了跨模型的知识共享与能力迁移。论文提出的关键解决方案在于采用字节级语言模型（byte-level language models）作为统一接口：由于所有字节级模型均以字节为基本单位进行读写，它们天然共享相同的输出空间，使得知识迁移可实现精确且与原始分词方式无关。论文进一步假设，字节级模型的两个分布——即对下一个字节的预测分布与对补丁边界（patch boundaries）位置的分布——可以被解耦并独立调整。这意味着模型可在保留自身边界划分策略的同时吸收教师模型的能力，或在保持原有生成能力的前提下重构边界布局。为此，论文设计了两项实验以验证该假设，并提供了初步实证支持。研究主张，若该假设成立，未来应推动整个社区采用字节级接口作为标准，从而实现模型间能力迁移与结构重配置的低成本、常规化操作，彻底摆脱当前依赖于特定模型分词器的瓶颈。
链接: https://arxiv.org/abs/2608.03599
作者: Mykola Haltiuk
机构: AGH University of Krakow (克拉科夫科技大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Byte-level language models are usually argued for on the grounds of robustness, multilingual fairness, and character-level skills. We point to a different, structural advantage: because they read and write bytes, any two of them share an output space, so knowledge transfer between them is exact and independent of how either was originally tokenized. We hypothesize that the two distributions a byte-level model produces, one over the next byte, one over where its patch boundaries fall, can be disentangled and changed almost independently. A model could absorb a teacher’s capability while keeping its own boundaries, or change how it places those boundaries while keeping its capabilities. We lay out the two experiments that would settle the hypothesis, alongside preliminary measurements of the properties they rest on. We argue that the community should move toward a byte-level interface as a shared standard: if the hypothesis holds, then once byte-level models are the norm, transferring capabilities and reshaping boundaries between them become cheap and routine, free of the per-model tokenizer that blocks them today.
[NLP-36] Looking under the Wrong Lamppost: On the Limitations of Automated Translation Quality Estimation
【速读】： 该论文旨在解决当前生成式翻译质量评估（Translation Quality Estimation, QE）技术在实际应用中面临的可靠性与有效性问题。尽管自动化QE已广泛应用于大规模翻译质量管理，但现有系统普遍存在研究不透明、可复现性差及评估标准不统一等问题。论文从理论与实证双重角度指出，当前QE系统在结构上存在根本性缺陷，难以作为独立工具直接嵌入真实翻译工作流。其关键问题在于：基于孤立语段的评价方式无法捕捉文本层面的连贯性、一致性以及修辞与风格特征，导致评估结果失真；同时，实证研究揭示了包括泛化能力不足、系统性偏差、过拟合与分布坍缩、性能差距、错误标注困难及数据稀缺等多重结构性局限。这些局限源于语言与翻译作为认知与交际行为的复杂本质，单纯增加数据量或优化模型架构未能根本解决上述问题。因此，论文强调，语段级的QE评分不应单独用于生产环境中的翻译路由、发布或免审决策，未来研究应聚焦于基于多维度质量标准（MQM）的自动化人工评估体系构建。
链接: https://arxiv.org/abs/2608.03577
作者: Serge Gladkoff,Angelika Vaasa,Sue Ellen Wright,Ingemar Strandvik,Lifeng Han
机构: Logrus Global LLC; European Parliament, DG for Translation and Clear Language; Kent State University; MQM Council; LUMC, Leiden University
类目: Computation and Language (cs.CL)
备注: To appear in the Proceedings of the 9th International Conference on Natural Language and Speech Processing (ICNLSP 2026), Trento, Italy, September 2026
Abstract:Automation of Translation Quality Estimation (QE) has emerged as a widely discussed approach to managing translation quality at scale, and a growing number of tools and technologies have been released in pursuit of this goal. However, the proliferation of new QE systems has not always been accompanied by robust, transparent, and reproducible research and testing. This gap deserves critical scrutiny. This paper examines some fundamental limitations of the QE technology from both theoretical and empirical perspectives, arguing that current QE systems are structurally ill-equipped to serve as reliable standalone tools in real-world translation workflows. The reviewed evidence suggests that QE suffers from a range of interrelated and largely unresolved limitations. Most fundamentally, the evaluation of the quality of translation at the level of isolated segments is problematic because it tends to miss out on cohesion, coherence, and stylistic and rhetorical text features. In addition, empirical research documents several other limitations and flaws, including failure to generalize, systematic biases, overfitting and distribution collapse, performance gaps, error annotation challenges, and data scarcity. These are structural limitations arising from the complexity of human language and translation as a cognitive and communicative act - limitations that more data and better architectures have so far not overcome. Consequently, segment-level QE scores should not be used as a standalone basis for routing, release, or review bypass in production; we argue future work should focus on automating human evaluation grounded in MQM.
[NLP-37] SFT Conflicts RL Coexists: A Theoretical and Empirical Analysis of Multi-Task Learning for LLM s
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在多任务推理能力提升过程中，监督微调（Supervised Fine-Tuning, SFT）与强化学习（Reinforcement Learning, RL）因优化机制差异导致的性能瓶颈问题。具体而言，SFT在多阶段训练中面临严重的任务冲突（task conflicts），表现为不同任务间参数更新相互干扰，难以协同优化；而RL则表现出更强的任务共存稳定性。其解决方案的关键在于揭示并利用两种方法在梯度干扰机制上的本质区别：SFT中的干扰受梯度范数（norm-limited）约束，随梯度绝对幅值增长而加剧；而RL中的干扰受梯度方差（variance-limited）约束，得益于优势函数归一化（advantage normalization）和在线策略优化（on-policy optimization）带来的小方差边界，从而实现跨任务近似正交的参数更新方向。基于此理论发现，作者提出Parallel-RL范式，通过解耦多任务训练流程，显著提升了训练效率与灵活性，有效缓解了多任务场景下的优化冲突问题。
链接: https://arxiv.org/abs/2608.03573
作者: Kejian Zhu,Zhuoran Jin,Shangqing Tu,Hongbang Yuan,Yushi Bai,Kang Liu,Juanzi Li,Jun Zhao
机构: 未知
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: Code: this https URL
Abstract:Supervised Fine-Tuning (SFT) and Reinforcement Learning (RL) exhibit fundamentally different behaviors in enhancing multi-task reasoning for large language models (LLMs). Our preliminary experiments revealed a phenomenon: SFT suffers from severe task conflicts under multi-stage training, whereas RL enables stable coexistence across diverse tasks. Empirically, we trace this to the parameter level, observing that RL induces sparse and approximately orthogonal updates across tasks. We provide a theoretical explanation for this mechanism by analyzing multi-task gradient interference. Our results reveal a distinction: interference in SFT is norm-limited, scaling with the absolute gradient magnitude, whereas interference in RL is variance-limited, bounded by the gradient variance induced by advantage normalization and on-policy optimization. This small variance bound yields near-orthogonal optimization directions across tasks. Leveraging this insight, we propose Parallel-RL, a paradigm that decouples multi-task training, significantly improving efficiency and flexibility.
[NLP-38] Hi-TTRL: Regulating Consensus with Hints for Test-Time Reinforcement Learning
【速读】： 该论文旨在解决测试时强化学习（Test-time Reinforcement Learning, TTRL）中因多数投票生成伪标签时对共识强度（consensus strength）敏感所导致的性能瓶颈问题。具体而言，低共识强度会放大不可靠伪标签带来的更新信号，而高共识强度则导致奖励对比度下降，引发梯度消失。其解决方案的关键在于提出一种名为Hi-TTRL的新框架，通过在采样阶段引入提示（hint）来动态调节滚动生成过程中的共识强度。Hi-TTRL首先基于部分滚动生成估计共识强度，当其偏离预设目标区间时，激活基于马尔可夫链蒙特卡洛（Markov Chain Monte Carlo, MCMC）的提示采样器，该采样器针对幂变换前缀分布进行有限步近似采样，生成具有锐化或展平特性的前缀作为提示，从而通过调节幂指数实现对共识强度的自适应控制。实验表明，该方法在多个数据集和模型架构上均显著优于标准TTRL，且消融实验与共识调控分析验证了提示引导的共识调节机制的有效性。
链接: https://arxiv.org/abs/2608.03545
作者: Kunbin Xu,Xingzuo Li,Xuefeng Bai,Kehai Chen
机构: 未知
类目: Computation and Language (cs.CL)
备注: 15 pages, 7 figures
Abstract:Test-time reinforcement learning (TTRL) improves the reasoning capabilities of large language models without labeled data by updating the policy with pseudo-labels constructed through majority voting. While effective, the reward signal assigned from majority voting is highly sensitive to consensus strength, defined as the frequency of the most common answer within a rollout group. In TTRL, consensus strength plays a dual role: it reflects both the reliability of the pseudo-label and the distribution of advantages. Low consensus can amplify updates from unreliable pseudo-labels through disproportionately large advantages, whereas high consensus reduces reward contrast and ultimately yields vanishing gradients. In this paper, we introduce Hi-TTRL, a test-time reinforcement learning framework that utilizes hints during sampling to regulate rollout consensus strength. Hi-TTRL first estimates consensus strength from a partial rollout group. When the consensus strength falls outside a target interval, it invokes a Markov chain Monte Carlo (MCMC) hint sampler. The sampler targets the power-transformed prefix distribution and uses finite-step approximate sampling to generate rollout prefixes as hints. By tuning the power exponent, Hi-TTRL generates hints with a sharpened or flattened power target, steering rollout consensus strength toward the target interval. Experiments on multiple datasets and backbones show that Hi-TTRL consistently improves over standard TTRL, with ablations and consensus-steering analyses validating the effectiveness of adaptive hint-guided consensus regulation.
[NLP-39] Cross-Lingual Bias in Large Language Models : A Comparative Analysis of English and Swahili
【速读】： 该论文旨在解决大语言模型在多语言场景下安全对齐与偏见评估仍高度依赖英语这一核心问题，即现有偏见检测方法是否适用于非英语语言。其关键解决方案在于通过构建4,900组对称的英-斯瓦希里语提示对，系统评估GPT-5.2和Gemini 2.5 Flash在九个社会偏见维度上的表现，涵盖刻板印象普遍性、情感倾向、拒绝行为及跨语言语义相似性。研究发现，偏见并非简单迁移，而是发生“转化”：特定维度上的刻板印象率最高变化达12个百分点，Gemini在斯瓦希里语中的中性情感输出翻倍，而GPT-5.2在英文中拒绝169条提示但在斯瓦希里语中全无拒绝行为，表明拒绝行为与英语表层形式紧密绑定；此外超过55%的提示对在双模型中生成语义不一致的输出。这揭示了仅基于英语的偏见审计无法有效覆盖多语言部署中的风险，强调了跨语言偏见评估的必要性。
链接: https://arxiv.org/abs/2608.03532
作者: Ruolei Zhang,Teddy Njuguna,Yue Feng
机构: University of Birmingham (伯明翰大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Large language models are increasingly deployed in multilingual contexts, yet safety alignment and bias evaluation remain overwhelmingly English-centric. We investigate whether social biases generalise across languages by submitting 4,900 symmetric English–Swahili prompt pairs to GPT-5.2 and Gemini 2.5 Flash across nine demographic bias axes, yielding 19,600 completions evaluated for stereotype prevalence, sentiment, refusal behaviour, and cross-lingual semantic similarity. Our findings show that bias transforms rather than transfers: stereotype rates shifted by up to 12 percentage points on specific axes, Gemini’s neutral-sentiment rate doubled in Swahili, and GPT-5.2 refused 169 prompts in English and zero in Swahili, consistent with refusal behaviour anchored to English-language surface forms at the behavioural level. Over 55% of prompt pairs produced semantically dissimilar completions across both models. These reinforce the idea that English-only bias audits do not produce adequate coverage for multilingual deployment.
[NLP-40] Consensus Measures for Unstructured Biomedical Text Annotations
【速读】： 该论文旨在解决生物医学文献中开放式标注任务的软一致性（soft inter-rater reliability）量化难题，尤其在目标概念事先未知、标注者需提供非结构化文本的情况下，传统人工标注一致性评估方法难以适用。其核心挑战在于如何有效衡量不同标注者之间在未预设标签体系下的语义等价性。解决方案的关键在于引入多种语义等价性度量方法来量化软一致性，研究发现嵌入模型（embeddings）虽具备良好的可扩展性，但在区分语义相近但本质不同的概念时表现有限；而大语言模型（Large Language Models, LLMs）虽在语义理解上潜力显著，却受限于计算成本难以高效估算随机一致性。因此，论文最终提出基于自然语言推理（Natural Language Inference, NLI）的度量方法作为折中方案，兼具较好的语义判别能力与可扩展性，能够更稳健地支持开放式生物医学标注任务中的可靠性评估。
链接: https://arxiv.org/abs/2608.03529
作者: Pascal Wullschleger,Christian Kreis,Martin A. Walter,Marc Pouly,Jennifer Foster
机构: Hamilton Institute, Maynooth University, Ireland; University of Lucerne, Switzerland; Lucerne School of Computer Science and IT, Switzerland
类目: Computation and Language (cs.CL)
备注:
Abstract:Biomedical literature is increasingly mined for knowledge beyond the questions it was written to answer. Because the target concepts are not known in advance, annotators prefer open-ended labels, whose agreement is hard to quantify. We study soft inter-rater reliability for annotators providing unstructured texts for biomedical annotation tasks. Synthetic experiments show that soft reliability can be quantified using a variety of semantic equivalence measures, and that the choice of measure affects failure modes of the estimation. Embeddings are scalable, but limited when differentiating similar but distinct concepts. Large language models are promising, but limited by scalability for estimating agreement by chance. Finally, we suggest measures based on natural language inference as a sensible compromise.
[NLP-41] ChronoLens: Measuring Language Change Across Time Languages and Linguistic Levels
【速读】： 该论文旨在解决历史语言演变中形态、句法、语义和语用等语言层次在计算研究中因使用不兼容表征而难以实现跨层次、跨语言、跨时期统一分析的问题。其核心挑战在于：现有方法无法在同一分析框架下量化各语言层次的变化幅度与方向，从而限制了对语言演变整体模式的把握。解决方案的关键在于提出名为ChronoLens的综合性框架，该框架整合了冻结的多语言语言模型、特征对齐的跨编码器（feature-aligned crosscoders）以及事后语言学干预机制，实现了在统一空间内对多语言历史文本的高效建模。通过在涵盖1803—2026年间五个议会传统共4498万份文档、约172亿词元的数据上应用该框架，结果表明，其稀疏表示与语言学统计量之间的相关性显著高于密集嵌入或池化稀疏自编码器（ρ=0.72 vs. 0.29 和 0.28），并揭示出同一语言内部各语言层次的变化幅度通常相当，但不同语言在变化时间、程度及方向上存在显著差异。研究发现，历史语言演变是一个结构化的多维过程：相同的变化幅度可能对应不同的演变轨迹，因此跨语言比较必须同时考量变化的距离与方向。
链接: https://arxiv.org/abs/2608.03507
作者: Gagan Bhatia,Julian Schlenker,Simone Paolo Ponzetto,Steffen Eger
机构: University of Technology Nuremberg(纽伦堡应用技术大学); University of Mannheim(曼海姆大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Historical language change affects morphology, syntax, semantics, and pragmatics, yet computational studies typically examine these levels with incompatible representations and therefore cannot determine whether they evolve together across languages. We address this problem by asking how the magnitude and direction of change vary across linguistic levels, languages, and historical periods within a single analytical space. We introduce ChronoLens, a framework that combines frozen multilingual language models, feature-aligned crosscoders, and post-hoc linguistic interventions, and apply it to 44.98 million documents and approximately 17.2 billion tokens from five parliamentary traditions spanning 1803–2026. The resulting sparse representations agree substantially more strongly with linguistic statistics than dense embeddings or a pooled sparse autoencoder ( \rho=0.72 versus 0.29 and 0.28 ), and reveal that morphology, syntax, semantics, and pragmatics generally change by comparable amounts within a language, while languages differ markedly in when, how far, and in which direction they change. These findings show that historical language change is a structured, multidimensional process: similar magnitudes can conceal different trajectories, and meaningful cross-linguistic comparison requires measuring both distance and direction.
[NLP-42] ConlangBench: Exploring Language Knowledge and Learning in LLM s through Diverse Constructed Languages
【速读】： 该论文旨在解决大语言模型（LLM）在低资源语言，特别是人工构造语言（conlangs）上的语言学习能力评估与训练问题。尽管构造语言具有丰富的语言创造力传统，但其在主流大语言模型研究中仍处于被忽视状态。为此，作者提出了ConlangBench——首个针对21种现有构造语言的大规模基准测试平台，包含超过2100万对构造语言-英语平行语句（其中43万对来自20种非世界语的构造语言）以及32.1万词汇条目。实验结果表明，模型在后验型构造语言（a posteriori conlangs，即词汇源自自然语言的构造语言）上的表现优于其他类型，反映出构造语言的设计特征对模型学习的影响。此外，基于该基准的训练表明，当平行语料充足时，模型能够学会全部八种构造语言，且学习曲线的差异与构造语言的生成方式密切相关。因此，该研究的关键在于构建了一个系统化的、可扩展的构造语言评估框架，并揭示了构造语言作为低资源语言的独特测试环境价值，为探究大语言模型如何习得语言结构提供了新的视角。
链接: https://arxiv.org/abs/2608.03505
作者: Jinhong Jeong,Seungyeop Yi,Sangah Lee,Youngjae Yu
机构: 未知
类目: Computation and Language (cs.CL)
备注: 29 pages, 12 figures, 17 tables
Abstract:Constructed languages (conlangs) are intentionally created human languages with a rich tradition of linguistic creativity. Despite their potential for studying language learning in large language models (LLMs), existing conlangs remain largely underexplored in LLM research. We present ConlangBench, the first large-scale benchmark for evaluating and training LLMs on 21 existing conlangs. We collect over 21M conlang-English parallel sentence pairs (including 430K pairs across the 20 non-Esperanto conlangs) and 321K vocabulary entries. In bidirectional translation experiments, we find that models perform better on a posteriori conlangs, whose vocabularies are derived from natural languages, reflecting the design characteristics of conlangs. Training on ConlangBench also shows that models can learn all eight conlangs for which sufficient parallel corpora are available, while their learning curves vary depending on how the conlangs were created. Our findings suggest that conlangs provide a unique testbed for investigating how LLMs acquire low-resource languages.
[NLP-43] Beyond Initialization Loss: A Systematic Study of Token Embedding Initialization Strategies for LLM Vocabulary Extension
【速读】： 该论文旨在解决预训练大语言模型（LLM）在扩展至新语言（如印地语）时，新增词元嵌入（token embeddings）初始化方式对持续预训练（CPT）效率的显著影响问题。其核心挑战在于如何设计高效的初始化策略，以加速模型在新语言上的收敛并提升下游性能。解决方案的关键在于提出并系统评估了超过20种初始化方法，发现子词组合（subword composition）策略优于传统的词汇平均和外部/学习型初始化方法；其中，输入与输出嵌入采用不对称初始化的变体表现最佳：具体而言，输入嵌入矩阵通过均匀子词平均与印地语特异性归一化校准初始化，输出语言建模头则采用基于字符长度加权的子词平均。该最优配置相较标准“均值所有”（Mean-all）基线，在仅500步的轻量级持续预训练下即达到基线3,500步的MILU-Hindi准确率，并将达到相当验证损失所需的CPT步数减少逾6倍。研究进一步表明，初始化损失和初始化比特每字节（Init BPB）无法可靠预测下游收敛表现，而仅需50步的轻量级CPT即可提供高效且可靠的策略筛选信号。
链接: https://arxiv.org/abs/2608.03494
作者: Raviraj Joshi,Utkarsh Vaidya,Sanjay Singh Chauhan,Niranjan Wartikar
机构: NVIDIA
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Vocabulary extension is an efficient way to adapt pretrained large language models (LLMs) to new languages, but the initialization of newly added token embeddings can strongly affect continued pre-training (CPT) efficiency. We present a systematic study of more than 20 initialization strategies for Hindi vocabulary extension in Nemotron-3-Nano-30B-A3B. Our comparison spans vocabulary-averaging baselines; external and learned initialization methods, including FOCUS, top-k semantic retrieval, and residual MLP mappings; subword composition; norm calibration; and input-output asymmetry. We find that subword composition methods outperform both vocabulary averaging and external/learned initialization approaches. Within subword composition, asymmetric variants achieve the lowest observed early validation loss and reveal distinct preferences for input and output embedding initialization. The best observed configuration initializes the input embedding matrix with uniform subword averaging and Hindi-specific norm calibration, and the output language modeling head with character-length-weighted subword averaging. Relative to the standard Mean-all baseline, this full initialization pipeline reaches comparable validation loss with over a 6x reduction in CPT steps and exceeds the baseline’s 3,500-step MILU-Hindi accuracy after only 500 steps. Finally, we show that initialization loss and initialization bits-per-byte (Init BPB) are unreliable predictors of downstream convergence, whereas lightweight CPT, as few as 50 steps, provides a cost-effective and reliable signal for selecting the best initialization strategy.
[NLP-44] Efficient Multilingual Neural Machine Translation via Corpus-Driven Vocabulary Pruning: An English-Arabic Case Study
【速读】： 该论文旨在解决大规模多语言神经机器翻译（Multilingual Neural Machine Translation, MNMT）模型在实际部署中因词汇表过大导致的内存与计算资源消耗过高的问题。现有压缩方法如剪枝、量化和知识蒸馏虽能缓解参数冗余，但通常保留原始词汇结构，未能有效解决由超大词汇表引发的根本性效率瓶颈。本文提出一种通用优化框架，其关键在于结合词汇表剪枝（vocabulary pruning）与针对性微调（targeted fine-tuning）策略，在大幅缩减词汇量的同时保持翻译性能。实验基于M2M100、NLLB-200和mBART-50三类模型在英-阿语对上的评估表明，该方法可将词汇量从超过128,000降至约10,000个词元，实现60%的内存节省，且无性能损失；优化后的多语言模型在多个指标上达到甚至超越专用双语基线模型的表现，尤其在COMET指标上显著优于基准模型，体现出更强的语义充分性与流畅性。
链接: https://arxiv.org/abs/2608.03480
作者: Ahmed Amine Aliane,Nasredine Semmar,Hassina Aliane
机构: Arabic Institute for Translation (阿拉伯翻译研究所); CERIST (阿尔及利亚研究与技术创新中心); CEA LIST (法国巴黎-萨克雷研究中心列表实验室)
类目: Computation and Language (cs.CL)
备注:
Abstract:The adoption of large pre-trained multilingual models for neural machine translation (MNMT) faces a major challenge: excessive memory and computational consumption due to overly large vocabularies and embedding layers. Although existing compression methods like pruning, quantization and knowledge distillation reduce parameter redundancy, they mainly preserve the structure of the original vocabulary, thereby leaving a major source of inefficiency unresolved. We propose in this paper a general optimization framework that combines a vocabulary pruning method with a targeted fine-tuning protocol for MNMT models. We evaluate the proposed framework using three models (M2M100, NLLB-200, mBART-50) on the English-Arabic language pair. Our approach reduces the vocabulary size from over 128,000 to approximately 10,000 tokens, enabling a 60% memory saving without any loss in performance. Results show that optimized multilingual models can match or exceed the performance of dedicated bilingual baselines. In particular, the pruned and fine-tuned M2M100 model achieves a competitive BLEU score of 42.04 (against 44.59 for the OPUS-MTen- ar bilingual model) while it significantly outperforms it on the COMET metric (0.8730 vs 0.7911) revealing superior semantic adequacy and fluency.
[NLP-45] Adaptive Modality Reliability Diagnosis and Restoration for Robust Multimodal Intent Recognition
【速读】： 该论文旨在解决多模态意图识别中因各模态（如语言、声学、视觉）存在噪声、缺失、语义冲突或主导性失衡而导致的可靠性问题。现有方法通常隐式推断模态重要性，仅通过重加权或抑制不可靠输入来应对，却无法判断退化模态是否可修复并重新信任。其解决方案的关键在于提出一种闭环可靠性引导框架——PRIME（Precision-weighted Reliability Inference and Modality rEstoration），该框架在样本层面联合实现模态质量的诊断、修复与再评估。PRIME通过上下文对数方差（contextual log-variance）建模各模态的弱性，该估计基于预测置信度、认知不确定性分歧、跨模态一致性及特征退化等互补诊断证据；由于缺乏模态可靠性标注，模型采用受控模态破坏与异方差不确定性目标进行显式训练。不同于直接丢弃不可靠模态，PRIME利用估计的弱性控制一个原型条件化的变分修复模块，从互补模态中重建退化表示；关键在于修复后重新评估可靠性，以判断修复后的表示是否具备足够可信度参与最终预测。最终，基于修复后的精度进行逆方差融合，显著提升了在模态缺失、噪声干扰、语义冲突及模态不平衡等复杂场景下的鲁棒性，同时保持了对干净数据的竞争力表现。
链接: https://arxiv.org/abs/2608.03475
作者: Suraj Kumar,Mohnish Raj,Soumi Chattopadhayay,Chandranath Adak,Ayan Dutta
机构: Indian Institute of Information Technology, Allahabad (印度信息科技学院，阿拉哈巴德)
类目: Multimedia (cs.MM); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Multimodal intent recognition combines linguistic, acoustic, and visual evidence, but individual modalities may be noisy, missing, semantically conflicting, or disproportionately dominant. Existing methods typically infer modality importance implicitly and either reweight or suppress unreliable inputs, without determining whether a degraded modality can be repaired and subsequently trusted. We propose PRIME (Precision-weighted Reliability Inference and Modality rEstoration), a closed-loop reliability guided framework that jointly diagnoses, restores, and reassesses modality quality at the sample level. PRIME represents the weakness of each modality through a contextual log-variance estimated from complementary diagnostic evidence, including predictive confidence, epistemic disagreement, cross-modal consensus, and feature degeneracy. Because modality-reliability annotations are unavailable, the estimator is explicitly trained using controlled modality corruption with known degradation severity, together with a heteroscedastic uncertainty objective. Rather than directly discarding an unreliable modality, PRIME uses its estimated weakness to control a prototype-conditioned variational restoration module that reconstructs the degraded representation from complementary modalities. Crucially, reliability is re-estimated after restoration, allowing the model to determine whether the repaired representation has become sufficiently trustworthy to contribute to prediction. The resulting post-restoration precisions are used for inverse-variance multimodal fusion. Experiments on multimodal intent-recognition benchmarks show that PRIME maintains competitive clean-data performance while improving robustness under missing, noisy, conflicting, and modality-imbalanced conditions.
[NLP-46] Probing Character-level Transformers for the Spanish L-shaped Morphome
【速读】： 该论文旨在解决生成式语言模型在学习复杂不规则形态模式（如西班牙语的L形形态类，L-shaped morphome）时，其内部表征的本质问题——即模型究竟“理解”了什么，是仅仅模仿表面形式交替，还是真正编码了该形态类的抽象规则。其解决方案的关键在于：通过在词元（lemma）互斥的交叉验证设置下，对五种架构、每种十二个训练模型进行系统性探测（probing），发现模型并非仅记忆表面交替现象，而是将L形形态类本身作为特定词汇项的抽象表征进行存储。该编码在解码器中间位置（即词干末辅音处）具有空间特异性，并且在所有形式均一致的情况下仍可被解码；更重要的是，即使仅在交替实例上训练，模型也能正确分类非交替实例。这表明模型的编码机制是项目特异性的（item-specific），而非泛化规则。因此，模型虽能复现该形态模式，但其内部表征不具备人类所具有的语义-句法泛化能力，仅以个体词汇项为单位存储抽象信息，无法实现真正的规则迁移。
链接: https://arxiv.org/abs/2608.03452
作者: Akhilesh Kakolu Ramarao,Kevin Tang,Wiebke Petersen,Dinah Baer-Henney
机构: Heinrich Heine University Düsseldorf (海因里希·海涅大学杜塞尔多夫); University of Florida (佛罗里达大学); Ruhr-Universität Bochum (鲁尔大学波鸿)
类目: Computation and Language (cs.CL)
备注:
Abstract:When a transformer learns an irregular morphological pattern, what has it learned? Our test case is the Spanish \emphL-shaped morphome, a complex irregular pattern in which the verb’s stem alternates in exactly the first-person singular indicative and all subjunctive forms, and whose membership no phonological, semantic, or syntactic feature predicts. Prior studies have shown that character-level transformers can reproduce this pattern, but that evidence describes what models produce, not what they represent. Probing five architectures, twelve trained models each, under lemma-disjoint cross-validation with controls and surface baselines, we show that the models encode the L-shaped class itself, not just its visible alternations. It is decodable above every surface baseline, survives instances in which every form shows the same stem, and probes trained on alternating instances still classify non-alternating ones. The encoding is localized where the stem choice is made, at the stem-final consonant position of the middle decoder, before the alternant is read. And it is item-specific: which verbs a model learned matters far more than which architecture it is. The models store the morphome as an item-specific lexical abstraction, sufficient to reproduce the pattern but not to generalize it as humans do.
[NLP-47] Balancing Efficiency and Efficacy: Training-Free Attention-Guided Switching Between Explicit and Latent Thoughts for MLLM s ACM-MM2026
【速读】： 该论文旨在解决多模态大语言模型（Multimodal Large Language Models, MLLMs）在推理过程中面临的双重挑战：一方面，依赖显式文本链式思维（Chain-of-Thought, CoT）的方法计算开销大且易产生视觉幻觉；另一方面，现有的隐式推理方法通常需要昂贵的训练成本，而直接将无需训练的大型语言模型（LLM）推理机制迁移至多模态场景又导致性能不稳定。其核心问题在于现有方法普遍依赖于词元级熵（token-level entropy），该指标错误地将感知模糊性（如视觉细节不清晰）与逻辑不确定性（如复杂推理步骤）混为一谈，从而影响推理质量。为此，本文提出一种无需训练的推理策略，关键在于通过引入“视觉到文本注意力比率”（vision-to-text attention ratio）这一新型度量，动态评估模型的认知焦点，实现感知与推理的显式解耦。基于此度量，所提出的注意力引导切换（Attention-Guided Switching, AGS）框架能够自适应地对感知类词元启用隐式推理以保留连续空间中的高保真视觉信息，同时对逻辑类词元强制执行显式文本生成以维持结构锚定。实验表明，该方法在保持卓越准确性的同时，显著提升了推理效率，减少了自回归生成步数和延迟，实现了性能与效率的双重优化。
链接: https://arxiv.org/abs/2608.03450
作者: Haoqian Kang,Liupeng Li,Kuofeng Gao,Jinpeng Wang,Zhenyu Lu,Bin Chen,Ke Chen,Yaowei Wang
机构: Harbin Institute of Technology, Shenzhen(哈尔滨工业大学深圳校区); Tsinghua Shenzhen International Graduate School, Tsinghua University(清华大学深圳国际研究生院); Peng Cheng Laboratory(鹏城实验室); Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences(中国科学院深圳先进技术研究院)
类目: Multimedia (cs.MM); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ACM MM 2026. 10 pages, 6 figures, 5 tables
Abstract:Reasoning in Multimodal Large Language Models (MLLMs) requires both fine-grained visual perception and rigorous logical deduction. Explicit text-based Chain-of-Thought (CoT) is computationally expensive and prone to visual hallucinations, while existing latent reasoning methods typically require costly training. Furthermore, directly adapting training-free LLM reasoning mechanisms to the multimodal setting yields unstable performance. We identify that this failure stems from their reliance on token-level entropy, which fundamentally conflates perceptual ambiguity (e.g., unclear visual details) with logical uncertainty (e.g., complex reasoning steps). To overcome this bottleneck, we present a novel training-free inference strategy for MLLMs that explicitly decouples perception and reasoning. We propose a novel metric, the vision-to-text attention ratio, to dynamically gauge the model’s cognitive focus. Guided by this metric, our proposed framework, Attention-Guided Switching (AGS), adaptively triggers latent reasoning for perceptual tokens to preserve high-fidelity visual information in the continuous space, while enforcing explicit text generation for logical tokens to maintain structural anchoring. Extensive experiments demonstrate that our method achieves state-of-the-art performance, significantly improving both accuracy and inference efficiency by reducing autoregressive steps and latency. Code is released at this https URL.
[NLP-48] Predicting Multilingual Classification and Translation Performance of LLM s with Cross-Lingual Alignment unicodex2013 Is English Enough? EMNLP2026
【速读】： 该论文旨在解决多语言大语言模型（Multilingual Large Language Models, LLMs）在非英语任务中性能提升的内在机制问题，特别是不同语言表征与英语之间对齐程度（cross-lingual alignment, CLA）如何影响下游任务表现。其核心问题是：现有的跨语言对齐度量（CLA scores）是否能够有效预测包括机器翻译在内的多种任务性能，而不仅仅是分类任务。解决方案的关键在于提出一种基于互信息（PMI）的翻译评估指标，该指标对目标语言依赖性较低且与chrF评分高度相关，从而支持在跨语言场景下进行可靠的性能比较。研究发现，以英语为参考的跨语言对齐度量（English-based CLA）在预测翻译质量方面表现与源-目标语言对齐度量相当甚至更优，为大语言模型内部存在以英语作为“语义枢纽”（pivot language）的假设提供了新的实证支持。
链接: https://arxiv.org/abs/2608.03446
作者: Adnan Al Ali,Kathy Hämmerl,Jindřich Libovický,Alexander Fraser
机构: Charles University, Faculty of Mathematics and Physics, Czech Republic; Technical University of Munich, Germany; Munich Center for Machine Learning
类目: Computation and Language (cs.CL)
备注: Submitted to EMNLP 2026
Abstract:Multilingual large language models (LLMs) have been shown to perform better on non-English classification tasks when the representations of the given language are more aligned to English within the model. Several cross-lingual alignment (CLA) scores have been proposed for use with LLMs, along with multiple approaches for extracting embeddings from the models. We provide a comparative analysis of 27 CLA score variants, examining how they differ and how well each predicts downstream performance across three tasks. Crucially, while LLMs are widely used for generative tasks such as machine translation, prior work has focused almost exclusively on classification. We therefore investigate whether CLA scores are similarly predictive of translation performance. To enable computing correlations across target languages, we propose a PMI-based translation metric, which is less dependent on the target language and correlates strongly with chrF. We find that CLA with English predicts translation quality comparably to or better than source-target CLA, providing new evidence that LLMs use English as an internal pivot language.
[NLP-49] Dynamically Allocating Evaluation Effort for Model Ranking
【速读】： 该论文旨在解决在自然语言处理（Natural Language Processing, NLP）任务中，依赖人工评估（human evaluation）虽为金标准却面临成本高昂与可扩展性差的问题。传统评估协议通过在完整基准数据集上逐一评估所有模型来识别表现最优的模型，这种方式虽稳妥但效率低下。为此，论文将多模型人工评估建模为一个具有相关臂（correlated arms）的多臂赌博机（multi-armed bandit）中的最优臂识别问题，其中每“拉动”一个臂对应一次对某模型的人工评估。通过基于已获得样本的中间模型排名结果自适应地分配标注预算，算法能够聚焦于最具竞争力的模型，从而显著提升评估效率。论文证明了所提出算法的最优性，并表明其能有效增强对顶尖模型之间的区分能力，使评估过程更快、成本更低，且更契合大规模模型竞赛的实际需求。
链接: https://arxiv.org/abs/2608.03437
作者: Vilém Zouhar,Julia Kreutzer,Alon Lavie,Tom Kocmi,Matt Post,Ondřej Bojar,Mrinmaya Sachan
机构: 未知
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:While human evaluation is the gold standard in many NLP tasks, it suffers from prohibitive costs and poor scalability. When identifying top-performing models, typical evaluation protocols waste effort by exhaustively evaluating all models on the entire benchmark, a safe but inefficient approach. In this work, we formalize multi-model human evaluation as a best-arm identification problem in a multi-armed bandit setup with correlated arms, where pulling an arm corresponds to human-evaluating a model. By sampling adaptively based on the intermediate model rankings obtained on the samples so far, we can focus the annotation budget on the most competitive models. We prove the optimality of the proposed algorithms and show that it improves discrimination between top-performing models. This makes evaluations faster, cheaper and more aligned with large-scale competition evaluation goals.
[NLP-50] DUD: Decoupled Update Dynamics for Reliable Uncertainty Quantification in Large Language Models ACL2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在实际部署中面临的可信度不确定性量化（Uncertainty Quantification, UQ）问题。传统基于概率的度量方法难以准确反映模型的内在认知状态（即认知不确定性，epistemic uncertainty），而现有机制解析方法虽利用隐藏状态动态，却普遍将前馈网络（Feed-Forward Networks, FFN）与注意力机制（Attention）的残差流更新进行混合聚合，导致参数记忆（parametric memory）与上下文处理（contextual processing）的独立功能被混淆。这种聚合行为掩盖了关键的机制性冲突，如记忆-上下文错位（memory-context misalignment），而此类冲突正是模型不确定性的核心指示信号。为此，本文提出解耦更新动力学（Decoupled Update Dynamics, DUD）框架，通过引入噪声诱导的因果干预，显式分离FFN与Attention模块的贡献。该方法通过量化各模块独立恢复能力，构建双流动态特征图谱，以捕捉模型内部脆弱性。实验表明，DUD在不确定性估计与校准性能上显著优于当前最先进基线，并展现出更强的跨数据集泛化能力，验证了解耦动态作为模型可靠性（faithfulness）鲁棒代理的有效性。
链接: https://arxiv.org/abs/2608.03411
作者: Yixin Bu,Runze Xia,Guanyun Zou,Yupeng Ji,Haodong Liu,Piji Li
机构: Nanjing University of Aeronautics and Astronautics (南京航空航天大学); The Key Laboratory of Brain-Machine Intelligence Technology, Ministry of Education (教育部脑机智能技术重点实验室)
类目: Computation and Language (cs.CL)
备注: ACL 2026 Main Conference
Abstract:Accurate Uncertainty Quantification (UQ) is critical for reliable deployment of Large Language Models (LLMs), yet traditional probability-based metrics often fail to capture the model’s true epistemic state. While recent mechanistic approaches leverage hidden state dynamics, they typically aggregate residual stream updates, conflating the distinct roles of parametric memory (Feed-Forward Networks) and contextual processing (Attention). We argue that this aggregation obscures fine-grained mechanistic conflicts, such as memory-context misalignment, that are fundamental indicators of uncertainty. To address this, we introduce \textbfDecoupled \textbfUpdate \textbfDynamics \textbf(DUD), a framework that explicitly decouples FFN and Attention contributions via noise-induced causal interventions. By quantifying the independent restoration capabilities of each module, we construct a dual-stream dynamic profile that captures the model’s internal fragility. Extensive experiments demonstrate that DUD significantly outperforms state-of-the-art baselines in both uncertainty estimation and calibration, while exhibiting superior cross-dataset generalization, validating decoupled dynamics as a robust proxy for model faithfulness.
[NLP-51] Dont Let Me Ask for It: LLM s Show Deficiencies in Active Multi-Turn Information Acquisition for Abductive Inference
【速读】： 该论文旨在探究大语言模型（Large Language Models, LLMs）在进行溯因推理（abductive reasoning）时，如何获取证据、更新假设以及决定终止推理的时机。现有研究多关注模型是否能正确解答溯因任务，但对其实现过程中的动态行为机制了解有限。为此，作者提出了“外星绑架游戏”（Alien Abduction game）这一交互式探测工具，通过不同交互模式（如证据是否预先提供、查询由模型自主选择或由“神谕”提供）来系统考察模型的行为特征。研究发现，将证据提前全部提供可显著提升模型的成功率；而在多轮交互中，部分模型过早形成假设而未充分利用后续证据，另一些则耗尽轮次仍无法收敛。此外，当模型使用神谕提供的示例而非自主选择查询时，其成功率更高，但最终假设虽与所选证据更一致，却可能未能有效区分替代假设。这表明，模型倾向于构建与自选证据相匹配的假设，却缺乏对假设的有效验证与修正能力，且难以判断何时应停止推理。因此，该研究的关键在于揭示模型在溯因推理过程中存在证据利用不充分、假设验证不足及终止决策机制薄弱等核心问题。
链接: https://arxiv.org/abs/2608.03388
作者: Shahrukh Mohiuddin,Chalamalasetti Kranti,Sherzod Hakimov,David Schlangen
机构: University of Potsdam, Germany; German Research Center for Artificial Intelligence (DFKI), Berlin, Germany
类目: Computation and Language (cs.CL)
备注: Preprint
Abstract:Abductive reasoning requires forming hypotheses that explain observed evidence and revising them as new evidence becomes available. While large language models (LLMs) are often evaluated on whether they solve abductive reasoning tasks correctly, less is known about how they acquire evidence, update their hypotheses, and decide when to stop. We introduce Alien Abduction game, an interactive probe for studying these behaviours under different interaction modes. The modes vary in whether evidence is provided upfront or across turns, and whether queries are selected by the model or examples are provided by the oracle. Across models, providing evidence upfront leads to higher success rates than distributing it across turns. In multi-turn settings, some models commit before using the available evidence, while others exhaust the turn budget without converging. Models also achieve higher success rates when examples are provided by the oracle than when they select their own queries, although their final hypotheses are more consistent with the evidence they selected. These findings suggest that models may form hypotheses that fit self-selected evidence without sufficiently distinguishing them from alternatives, and may struggle to validate and refine their hypotheses or determine when to stop.
[NLP-52] FACTWASH: Catching AI Rewrites That Wash Hearsay into Fact
【速读】： 该论文旨在解决生成式AI系统在信息重写过程中导致的“事实洗脱”（factwashing）问题，即在信息重构时抹除可验证性依据，如原始来源、表述确信度及时间背景，从而削弱内容的可信度与可追溯性。其核心解决方案是提出一个开源的“写时门控机制”（write-time gate）——factwash，通过命名标记与证据链实现确定性检测，而非依赖大语言模型（LLM）进行判断。该方案的关键在于区分不同语义属性是否具备有限的表面线索（surface-cue inventory）：对于可枚举的显式否定线索（explicit negation cues），基于词表即可高效覆盖，达到0.91 F1得分；而对于具有开放性表达形式的模糊化（hedging）与归属（attribution）线索，则需引入单轮提问的轻量级LLM见证者（witness），在保持精度不变的前提下分别提升17和15个百分点的召回率。部署层面，该见证者仅用于降低误判风险，不扩大覆盖范围，因此主要提升的是判定的精确性。通过对105,596条独立标注句子的评估发现，55%的错误写入源于对话式传闻（conversational hearsay），而商业邮件中占比为7%（p < 0.001），表明首要问题并非选择何种检测器，而是判断事实洗脱是否实际发生。在未经修改的mem0 2.0.7系统上，该门控机制成功识别出8条中有5条含模糊化传闻的写入内容。
链接: https://arxiv.org/abs/2608.03372
作者: Alex Kwon
机构: Independent Researcher
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 15 pages, 3 figures. Code and data: this https URL
Abstract:AI systems rewrite information constantly: conversations become stored memories, documents become answers. The rewrite can keep a claim while washing away what made it checkable, who said it, how sure they were, when it held. We call that failure factwashing, and release factwash, an open-source write-time gate that catches it deterministically, with named flags and evidence rather than an LLM judge. Building it answers a practical question: when does a cheap check suffice, and when do you need a model? What decides is whether the property has a bounded surface-cue inventory. Explicit negation cues are close to enumerable, so a word list finishes and transfers, reaching 0.91 F1 on untuned text. Hedging and attribution have open-ended realizations, so vocabulary plateaus near half recall, and a one-question LLM witness recovers +17 and +15 points of cue-detection recall at equal precision. Deployed, that witness may only lower a verdict, so it buys precision rather than coverage. We measure cue detection on 105,596 independently annotated sentences. A blind-labelled corpus of memory writes then locates the failure: 55% of bad writes in conversational hearsay, 7% in business email (p 0.001), so the first deployment question is not which detector to use but whether the failure occurs at all. On unmodified mem0 2.0.7, the gate flags 5 of 8 hedged-hearsay writes.
[NLP-53] ArtECulture: Benchmarking Culture-Conditioned Visual Emotion Understanding in Multimodal Large Language Models
【速读】： 该论文旨在解决现有视觉情感理解方法忽视文化差异对情感感知影响的问题，提出“文化条件化视觉情感理解”这一新任务，即针对给定图像预测其在特定文化背景下的情感感知并提供可解释的推理依据。其核心挑战在于现有基准数据集存在个体标注不一致、难以推导出具有共识性的文化层面情感标签，以及文化覆盖不平衡等问题。为此，研究构建了ArtECulture基准，包含6,792幅跨英语、汉语和阿拉伯语文化的艺术作品，涵盖均衡的西方与非西方内容，并提供文化特异性的情感标签与解释。在零样本设置下对16个开源与闭源多模态大语言模型（Multimodal Large Language Models, MLLMs）的评估表明，该任务仍极具挑战性，最优模型准确率低于50%。为克服此局限，论文提出一种基于检索增强的文化条件化情感理解框架，通过引入基于概念的文化情感知识库，在不进行额外训练的前提下向MLLMs注入显式文化知识，从而显著提升模型在文化一致性情感预测与可解释性生成方面的表现。该框架的关键创新在于利用外部知识增强模型的文化敏感性，实现更精准、更具解释力的文化情境化情感理解。
链接: https://arxiv.org/abs/2608.03358
作者: Xiaolin Chen,Xuemeng Song,Wenhao Shi,Xianjing Han,Mong-Li Lee,Wynne Hsu
机构: 未知
类目: Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Existing visual emotion understanding methods typically ignore cultural variations in emotional perception. We introduce culture-conditioned visual emotion understanding, a task that predicts the culture-specific emotional perception of a given image and explains the underlying rationale. Although related benchmarks exist, they are limited by inconsistent individual annotations, which hinder the derivation of majority-supported culture-level emotion labels, and imbalanced cultural coverage. Thus, we present ArtECulture, a benchmark containing 6,792 artworks with culture-specific emotion labels and explanations across English, Chinese, and Arabic cultures, with balanced Western and non-Western content. Evaluations of 16 open- and closed-source Multimodal Large Language Models (MLLMs) under a zero-shot setting reveal that the task remains challenging, with the best model achieving below 50% accuracy. To address this limitation, we introduce a retrieval-augmented culture-conditioned emotion understanding framework, which leverages a concept-based cultural emotion knowledge base to inject explicit cultural knowledge into MLLMs without additional training. The framework improves both culturally aligned emotion prediction and grounded explanation generation. Our benchmark and code will be publicly released.
[NLP-54] Benchmarking the Benchmarks: Testing the Predictive Validity of Commonsense Benchmarks
【速读】： 该论文旨在解决当前大语言模型（Large Language Models, LLMs）在真实世界任务中能力预测的可靠性问题，尤其关注主流常识推理基准测试对下游任务性能的预测效度不足这一关键挑战。其核心问题是：现有标准化常识基准测试是否能够有效、普遍地反映模型在实际应用中的常识理解与推理能力。解决方案的关键在于通过系统性评估23个来自六个不同模型家族的模型在四类经典常识基准、四类重构变体、三类非常识对照任务以及八项涉及隐含社会性、语用性、时间性和物理性推理的下游任务上的表现，采用模型排名对比、计算量控制的相关性分析及“留一家族外交叉验证”等方法，严格检验常识基准的准则效度（criterion validity）。研究发现，尽管重构后的基准基本保持原有模型排序，但并未显著提升对下游任务的预测能力；常识基准仅在少数特定下游任务上表现出一致的跨家族预测有效性，且增益具有任务或指标依赖性。因此，研究结论指出，标准化常识基准所提供的证据是任务依赖性的，而非广泛适用于评估模型整体常识能力的通用指标。
链接: https://arxiv.org/abs/2608.03340
作者: Ine Gevers,Walter Daelemans
机构: CLiPS, University of Antwerp (CLiPS, 安特卫普大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Predicting LLM’s capabilities on real-world tasks is essential, yet the extent to which performance on commonsense benchmarks predicts downstream performance remains underspecified. To establish the practical usability of widely adopted commonsense benchmarks, we evaluate 23 models from six families on four established commonsense benchmarks, four reworked variants, three non-commonsense controls, and eight downstream tasks requiring implicit social, pragmatic, temporal, or physical reasoning. We compare model rankings, compute controlled correlations, and use leave-one-family-out cross-validation to assess the criterion validity of commonsense benchmarks. Our results show that revised benchmarks largely preserve original model rankings and do not improve downstream predictive power. Commonsense benchmarks show consistent cross-family predictive validity for only a narrow subset of downstream tasks, with smaller or metric-specific gains elsewhere. Overall, standardized commonsense benchmarks provide task-dependent rather than broad evidence of downstream commonsense competence.
[NLP-55] Distractor-Aware Truncation: Disentangling Context-Length Effects from Signal Loss in Long-Context LLM Benchmarks
【速读】： 该论文旨在解决当前检索增强型与记忆增强型语言模型研究中一个广泛接受但未经充分验证的假设：在保留相关信息的前提下，上下文越短效果越好。其核心问题是现有研究普遍采用“简单截断”（naive truncation）策略——即从提示词中间移除内容——来评估长上下文性能，但该方法可能因随机破坏关键信息而导致性能下降，从而误导对上下文窗口效应的真实理解。论文的关键解决方案是提出并验证一种“干扰项感知的截断协议”（distractor-aware truncation），该协议通过识别每样本中的任务相关信号内容，并仅移除非相关干扰项，确保关键信息在截断过程中被保留。实验结果表明，在标准截断下，模型性能随上下文保留比例下降而单调衰退（所有测试条件均显著，p_adj < 0.05）；而在干扰项感知截断下，小规模模型（Claude Haiku 4.5、Sonnet 4.6）表现显著提升，大规模模型（Opus 4.7、GPT-5.5）则达到全上下文上限，且该现象在多个数据集（BABILong、GraphWalks、MRCR v2、Oolong）和不同提供商模型间一致复现，排除了单一模型厂商偏差。研究表明，传统截断方法本质上测量的是“中间移除是否恰好保留答案”的偶然性，而非真正的上下文长度效应。因此，论文强调未来研究必须明确定义信号与干扰的区分机制，否则结论将在两种相反解释之间存在根本歧义。
链接: https://arxiv.org/abs/2608.03297
作者: Mohsen Arjmandi
机构: evolutionID GmbH(进化ID公司)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 14 pages, 2 figures. Code and data: this https URL
Abstract:A standard claim in the literature on retrieval-augmented and memory-augmented language models is that shorter context is better when the relevant information is preserved. We test this claim by running every sample of two long-context benchmarks – BABILong and GraphWalks (BFS) – at four context-retention fractions (100%, 75%, 50%, 25%) under two truncation protocols. The first is the naive protocol implicitly used in much prior work: drop content from the middle of the prompt. The second is distractor-aware: identify the task-relevant content for each sample and drop only the rest. We evaluate three sizes of the Claude family (Haiku 4.5, Sonnet 4.6, Opus 4.7) and, to test cross-provider generality, GPT-5.5 from a different provider; we apply the same protocol to two further benchmarks (MRCR v2, Oolong). Under naive truncation, score collapses monotonically (paired Wilcoxon, Holm-corrected p_adj 0.05 in all eight BABILong and GraphWalks cells). Under the distractor-aware protocol – which preserves the signal by construction – performance is preserved or improves: the two smaller Claude models show statistically significant gains on BABILong, while the larger models (Opus 4.7 and GPT-5.5) sit at their full-context ceiling. The naive collapse and its distractor-aware recovery replicate on GPT-5.5, ruling out a single-provider artifact. The mechanism is direct: under the naive protocol the answer-bearing content survives in fewer than 1% of samples at 25% retention; under the distractor-aware protocol it is preserved by construction. The naive protocol is therefore not a measurement of context-window effects; it is a measurement of how often middle-removal happens to spare the answer. We conclude that future studies of context-length effects must specify how they distinguish signal from distractor, or they are at best ambiguous between two opposite hypotheses.
[NLP-56] he Tell-Tale Trace: Detecting Reasoning Failures in LLM s Using Chain-of-Thought Dynamics
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）在链式思维（Chain-of-Thought, CoT）推理过程中，因推理轨迹中分布性错误而非单一步骤错误导致的性能失败问题。现有方法多聚焦于单个中间步骤的语义正确性或一致性评估，忽视了推理过程整体动态演变特征，因而难以有效识别和诊断跨步骤的系统性推理偏差。为此，本文提出一种不依赖于语句是否忠实反映模型内部计算的分析框架，通过分析可见CoT的动态行为模式来区分成功与失败推理。研究基于可验证的布尔可满足性（Boolean Satisfiability, SAT）任务，在不同复杂度下对多种LLM进行评估，发现：在SAT问题中，错误推理会提前出现子句验证崩溃，表现为重复操作且过早终止；在UNSAT问题中，模型则过早趋向错误的可满足结论，倾向于验证候选赋值而非推导矛盾。通过引入针对性的证明搜索提示（proof-search prompt）干预，Llama3-70B在该任务上的准确率从13.3%提升至85%，纠正了84.6%的错误。研究表明，模型能力失效可表现为推理结构上分布性、任务依赖性的动态变化，而无需假设语义忠实性，仅通过分析CoT动态即可实现故障诊断与修正。
链接: https://arxiv.org/abs/2608.03291
作者: Shashwat Sourav,Aishwarya Balwani
机构: 未知
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Chain-of-thought (CoT) reasoning improves large language model (LLM) performance while also providing an observable interface to the model’s reasoning process. Existing approaches that leverage verbalized CoTs to monitor reasoning correctness, however, largely evaluate the semantic correctness or consistency of individual intermediate steps, rather than how the reasoning process evolves across the trace. As a result, failures distributed across the reasoning trajectory, rather than those localized to a single incorrect step, remain comparatively underexplored. Furthermore, verbalized CoTs need not faithfully reflect the model’s internal reasoning, motivating analyses that do not treat individual statements as literal accounts of internal computation. In this work, we therefore ask whether the dynamics of visible CoT can be leveraged to systematically distinguish successful from failed reasoning without assuming such semantic faithfulness. We study a range of LLMs on verifiable Boolean satisfiability tasks with variable complexity, enabling controlled comparisons near each model’s capability frontier. Tagging CoT sentences by reasoning function reveals premature verification collapse on SAT problems: incorrect traces enter clause checking earlier, repeat similar operations, and finalize sooner. On UNSAT problems, models presumptuously move towards incorrect SAT conclusions, checking candidate assignments rather than deriving contradictions across constructed cases. Subsequently, a targeted proof-search prompt intervention raises Llama3-70B accuracy from 13.3% to 85%, correcting 84.6% of these errors. These results show that capability failures can manifest as distributed, task-dependent changes in the structure of visible reasoning, and that CoT dynamics agnostic to whether the verbalized trace reflects the model’s internal computations can help diagnose and correct failures.
[NLP-57] MoEGen: Mixture-of-Experts for Instance-Adaptive LoRA Generation
【速读】： 该论文旨在解决基于混合专家模型（Mixture of Experts, MoE）的参数高效微调（Parameter-Efficient Fine-Tuning, PEFT）方法在实际应用中面临的存储瓶颈与灵活性不足问题。现有MoE-based PEFT方法通常通过存储多个完整的低秩适配器（LoRA）来提升模型容量，导致适配器存储量随专家数量线性增长，并限制了适应过程仅能在固定专家池内进行。为克服这一局限，本文提出MoEGen框架，其核心创新在于将MoE-based PEFT从传统的“专家选择”机制转变为“专家条件化的参数生成”机制。具体而言，MoEGen不再显式存储每个专家的完整LoRA模块，而是将每个专家表示为一个可学习的小型向量（称为专家码），并通过输入路由机制对这些专家码进行加权组合，以此条件化一个轻量级超网络（hypernetwork），动态生成针对特定输入实例的低秩更新参数。该设计实现了专家容量与适配器存储量的解耦，同时支持实例级自适应。实验结果表明，MoEGen在八个常识推理基准测试中均显著优于主流静态及MoE-based PEFT基线方法，且在医疗与法律领域的联合微调任务中也表现出色，验证了其高效性与泛化能力。
链接: https://arxiv.org/abs/2608.03275
作者: Yiming Zeng,Lei Lu,Zexin Li,Zhuochun Li,Shuoqiu Li,Shuyi Liao,Xidong Wu,Zeyu Zhang,Minmei Wang,Yu Zhao,Tingting Yu,Shangqian Gao
机构: University of Connecticut (康涅狄格大学); Northeastern University (东北大学); Nanyang Technological University (南洋理工大学); University of Pittsburgh (匹兹堡大学); Florida State University (佛罗里达州立大学); Google(谷歌); Amazon AGI(亚马逊AGI); University of Cincinnati (辛辛那提大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Parameter-efficient fine-tuning (PEFT) enables efficient adaptation of large language models, but existing MoE-based PEFT methods typically improve capacity by storing multiple full LoRA experts, causing adapter storage to grow linearly with the number of experts and restricting adaptation to a fixed expert pool. We ask whether MoE-based PEFT can produce instance-specific adaptations without explicitly storing a separate LoRA module for each expert. To address this gap, we propose MoEGen, an adaptation framework that shifts MoE-based PEFT from expert selection to expert-conditioned parameter generation. Instead of storing each expert as a full LoRA adapter, MoEGen represents each expert as a small learnable vector, termed an expert code. It routes each input over these vectors and uses their weighted combination to condition a lightweight hypernetwork that generates input-specific low-rank updates. This design decouples expert capacity from adapter storage while enabling instance-conditioned adaptation. Experiments on eight commonsense reasoning benchmarks show consistent improvements over strong static and MoE-based PEFT baselines across three backbones. MoEGen also performs strongly in joint medical and legal-domain adaptation.
[NLP-58] CIGTSurv: Clinical Information Guided Tri-modal Survival Prediction with Local Prototype Association and Global Feature Alignment MICCAI2026
【速读】： 该论文旨在解决多模态生存预测中临床信息（clinical information）因离散、稀疏且低维特性而被忽视的问题，以及病理图像、基因组数据与临床信息之间固有的异质性所带来的跨模态交互建模挑战。其核心解决方案在于提出一种临床信息引导的三模态生存预测框架CIGTSurv：首先，通过设计统一的文本模板并借助预训练基础模型（pretrained foundation models），将临床表格数据转化为高维标记化嵌入表示；随后，以临床信息为锚点，引入双层级交互机制——一是基于交叉注意力的局部原型关联（local prototype association, LPA）模块，用于显式学习不同模态间的细粒度标记级对应关系；二是基于最大均值差异（Maximum Mean Discrepancy, MMD）的全局特征对齐（global feature alignment, GFA）损失，隐式增强跨模态特征分布的一致性。该方法有效融合了临床信息的语义指导作用与多模态数据的互补性，在五个TCGA癌症队列上的实验验证了其在生存预测任务上达到当前最优性能（state-of-the-art, SOTA）。
链接: https://arxiv.org/abs/2608.03247
作者: Jing Dai,Qibin Zhang,Weiwei Zhou,Mingde Xu,Jingsong Liu,Jingdong Zhang,Hongming Xu
机构: 未知
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注: Accepted at MICCAI 2026
Abstract:Multimodal learning has significantly advanced survival prediction by integrating pathology images with genomic data. However, clinical information, despite its critical role in reflecting a patient’ s overall health, remains underutilized due to its discrete, sparse, and low-dimensional nature. Furthermore, the inherent heterogeneity across these modalities pose significant challenges in modeling cross-modal interactions. In this paper, we propose CIGTSurv, a Clinical Information Guided Tri-modal framework for Survival prediction. Specifically, we first design a holistic text template and use pretrained foundation models to transform clinical tabular data into high-dimensional tokenized embeddings. Using clinical information as an anchor, we then introduce a dual-level interaction mechanism: 1) a local prototype association (LPA) module based on cross-attention to explicitly learn token-level correspondences between different modalities, and 2) a global feature alignment (GFA) loss based on Maximum Mean Discrepancy (MMD) to implicitly enhance cross-modal distribution consistency. Extensive experiments on five TCGA cancer cohorts demonstrate that CIGTSurv achieves state-of-the-art (SOTA) survival prediction performance. Our source code is publicly available at this https URL.
[NLP-59] Relational Priors as Convergence Pressure in LLM -Based Multi-Agent Systems
【速读】： 该论文旨在解决大语言模型多智能体系统（LLM-MAS）中隐含的智能体间关系语义所带来的不确定性问题，即角色设定、辩论协议与聚合规则所隐含的社会期望（如信任、质疑、服从或协作）可能对系统行为产生不可控的影响。其解决方案的关键在于将这些关系先验（relational priors）显式化：通过采用最小化的有向网络形式表征关系先验，并将自然语言描述注入智能体提示词中，同时固定任务协议以隔离关系因素的影响。研究发现，显式的关系正向性（relational positivity）主要起到促进协调与共识的“收敛压力”，在需要行为对齐的任务（如可持续资源治理和主观共识形成）中有效提升一致性；但在客观问答辩论等依赖正确性的场景中，高正向性可能导致错误一致性的增加，且无法可靠提升准确性。此外，效果受模型架构、关系类型及网络拓扑结构影响，显式的中立并不等同于无关系框架。因此，作者主张不应默认在LLM-MAS中引入关系先验，而应采取诊断性、任务特定的使用策略：对比无先验基线、监控基于正确性的评估指标，并在验证不支持时移除关系层。
链接: https://arxiv.org/abs/2608.03239
作者: Ming Shen,Chao Shang,Sadat Shahriar,Devang Kulshreshtha,Yi Zhang,Sandesh Swamy,Yanjun Qi
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:Large language model-based multi-agent systems (LLM-MAS) are designed through roles, debate protocols, and aggregation rules. These choices create implicit social expectations: agents may be expected to trust, challenge, defer to, or collaborate with peers. We study the effects of making inter-agent relation semantics explicit. We use a minimal signed-network formulation of relational priors and inject natural-language renderings into agent system prompts while holding the task protocol fixed. Across a commons-governance simulation and multi-agent debate, relational priors primarily act as convergence pressure: increasing relational positivity tends to make agents coordinate or agree more readily. This pressure can help when utility rewards behavioral alignment, as in sustainable resource governance and subjective consensus. It does not, however, reliably improve accuracy. In objective QA debates, higher positivity can increase agreement even when correctness-conditioned agreement does not improve and may decline in some settings. Effects vary by model backbone, relation type, and topology; explicit neutrality is not equivalent to omitting relational framing. We argue that relational priors should not be a default add-on for LLM-MAS. Their safer use is diagnostic and task-specific: compare against a no-prior baseline, monitor correctness-conditioned metrics when truth matters, and omit the relational layer when validation does not justify it.
[NLP-60] On the Diversity of Analogy Making in Large Language Models
【速读】： 该论文旨在解决生成式AI（Generative AI）在类比推理任务中输出多样性不足的核心问题。尽管大语言模型（Large Language Models, LLMs）在类比生成方面展现出显著潜力，但其输出往往集中于少数特定领域，导致跨领域关联能力受限，制约了科学创新与创造性突破。研究的关键在于系统评估十种先进开源与闭源LLM在类比生成中的多样性表现，揭示了普遍存在的“领域同质性”现象，并发现现有提升多样性的方法普遍存在“多样性-质量权衡”问题。进一步的因果分析表明，不同模型在类比多样性控制上存在显著的敏感区域差异，这为理解该权衡机制提供了潜在的内在解释路径。
链接: https://arxiv.org/abs/2608.03233
作者: Yuanhao Shen,Daniel Xavier de Sousa,Caio César Sifuentes Barcelos,Hongyu Guo,Xiaodan Zhu
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:Large Language Models (LLMs) have demonstrated remarkable potential for analogy making, a core cognitive capability that drives novelty and creativity. While prior research has extensively investigated the applications and underlying mechanisms of LLM-based analogy making, its output diversity remains largely unexplored, despite being essential for broadening cross-domain connections and fostering scientific innovation. In this work, we present a comprehensive evaluation of analogy diversity across ten state-of-the-art open- and closed-source LLMs. Our findings highlight a concerning issue of domain homogeneity, a prevalent tendency for LLMs to generate analogies from a narrow set of target domains, limiting both inter-query and intra-model diversity. Furthermore, our analysis reveals a fundamental trade-off in existing LLM diversity-enhancement methods: increasing output diversity often comes at the expense of output quality. Finally, our causal analysis of LLM information flow reveals substantial differences in the model-sensitive regions governing analogy diversity across LLMs, suggesting a potential mechanism for the observed diversity-quality trade-off. To our knowledge, this is among the first studies to systematically investigate output diversity in LLM-based analogy making.
[NLP-61] Agent ic Reinforcement Learning with Self-Distilled Reward Shaping
【速读】： 该论文旨在解决生成式语言智能体在多轮交互任务中因稀疏的轨迹级奖励而导致的中间决策信用分配不明确的问题。现有方法虽通过引入冻结的特权技能（privileged skills）提供更密集的监督信号，但未能实现跨交互步骤的教师评分联合校准、教师置信度与实际回报之间的关联建模，以及将该信号有效融入原生强化学习中的优势（advantage）构建过程。为此，论文提出自蒸馏奖励塑形（Agentic Reinforcement Learning with Self-Distilled Reward Shaping, ADRS）框架，其核心在于：在每一步内对特权标记得分进行中心化与归一化处理，基于组内置信度-回报关联性设计返回相关教师价值优势（Teacher Value Advantage, TVA）门控机制以动态调节得分，并将加权后的标记级信号无缝集成至原生强化学习的信用分配路径中。该方法在不依赖任何推理阶段技能的前提下，实现了对教师偏好、偏好与回报的相关性及信用传递方式的精准建模，显著提升了长时程任务下的性能表现，且在不同强化学习主干网络、低数据场景、未见任务及长期训练设置下均展现出稳定增益。
链接: https://arxiv.org/abs/2608.03223
作者: Ranxu Zhang,Guinan Chen,Chenshaodong,Jinghao Lin,Xiaozhou Xu,Sunzhe,Yanyong Zhang,Chao Wang
机构: University of Science and Technology of China (中国科学技术大学); Alibaba Group (阿里巴巴集团)
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 17 pages,10 figures,11 tables
Abstract:Agentic reinforcement learning enables LLM agents to learn through interaction, but sparse trajectory-level rewards reveal success without identifying which intermediate decisions deserve credit. Training-only privileged skills can provide denser supervision by allowing the same frozen policy snapshot to rescore fixed tokens from skill-free trajectories while conditioned on task-matched procedural skills. Existing methods, however, do not jointly calibrate teacher scores across interaction steps, relate teacher confidence to realized returns, and integrate the resulting signal into native reward-to-advantage construction. We introduce Agentic Reinforcement Learning with Self-Distilled Reward Shaping (ADRS), a framework for constructing return-associated token-level credit for multi-turn language agents. ADRS centers and normalizes privileged token scores within each step, modulates them with a return-associated Teacher Value Advantage (TVA) gate based on within-group confidence–return association, and incorporates the gated token signal into native RL credit construction. Together, these components determine what the teacher prefers, when that preference is return-relevant, and how it enters the native reinforcement-learning credit path, while keeping rollouts and inference skill-free. Finally, experiments across three interactive benchmarks show that ADRS consistently improves performance on long-horizon tasks, with gains persisting across RL backbones, reduced-data settings, unseen tasks, and extended training. For anonymous review, our code is available at the following the link: this https URL
[NLP-62] Reachability Is Not Realization: Tracing the Sources of LLM Benchmark Gains
【速读】： 该论文旨在解决当前大语言模型（LLM）能力评估中一个关键问题：基准测试得分的提升常被直接视为模型能力增强的证据，但同一得分提升可能源于不同层面的模型行为变化——例如，模型可能真正“实现”了此前无法回答的问题，也可能仅是将原本已可触及的答案“命中”。现有评估方法缺乏对问题粒度上的区分能力，无法判断某一答案是真正实现（realized）还是仅处于可触及（reachable）状态。为此，论文提出在固定推理预算、温度参数和输出格式条件下进行问题级审计（question-level audit），通过定义“实现”（default deployment产生正确答案）与“可触及”（特定探测器在固定预算内找到答案）两个核心概念，精确刻画模型实际表现与潜在能力之间的差距。其解决方案的关键在于：首先验证推理时层路由（inference-time layer routing）是否能扩大可触及范围，在43个模型与任务组合中发现，随机路由策略在相同预算下表现不劣于甚至优于结构化搜索；然而，这种增益几乎完全依赖于对正确答案的先验知识，答案无关的探测方法几乎无法复现该收益。其次，研究揭示了部分答案虽可触及却未被实现的原因：在6个不同规模（0.5B–31B）的案例中，抑制某个特定的多层感知机（MLP）模块即可修复68%至92%的预设失败情况。最后，通过对比训练后部署性能与可触及上限的变化，发现多数情况下部署性能上升而可触及上限保持不变甚至下降，表明模型能力并未普遍扩展，且实现与可触及性并不总同步变化。因此，论文强调，关于模型能力提升的结论必须同时报告在匹配评估条件下实现性能与可触及性的双重指标，以避免误判。
链接: https://arxiv.org/abs/2608.03219
作者: Yanchao Li,Wanhao Liu,Jiaqing Xie,Ben Gao,Yanbo Wang,Tianfan Fu,Yuqiang Li
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Benchmark gains are often treated as evidence of greater LLM capability. Yet the same gain can reflect different changes in model behavior. A model may reach new answers, or produce answers that were already within reach. Aggregate scores do not distinguish these changes question by question. We establish a question-level audit under fixed budgets, temperatures, and answer formats. A question is realized when the default deployment procedure produces the correct answer. A question is reachable when a specified probe finds that answer within a fixed budget. We first test whether inference-time layer routing can expand reachability. Under a matched budget, random routes match or exceed structured search in all 43 model and task settings. Answer-blind procedures retain almost none of this gain, which instead requires access to the correct answer. We then ask why reachable answers sometimes fail to appear. Across six cases spanning 0.5B to 31B, silencing one identified MLP block repairs 68 to 92 percent of a predefined failure set. We next test whether training closes the gap by expanding reachability. In five of six matched evaluations, deployed performance rises while the reachable ceiling remains flat or falls. For DAPO, the deployed score rises by 14.7 points while the reachable ceiling falls by 13.3 points. Across the settings we audit, realization and reachability therefore do not always change together. Claims of capability expansion should report both realized performance and reachability under matched evaluation conditions. Code is available at this https URL
[NLP-63] ICO: Enhancing Semantic-Shift Jailbreaks via Iterative Context Optimization
【速读】： 该论文旨在解决基础模型在面对语义漂移越狱攻击（semantic-shift jailbreaks）时效果有限的问题。尽管此类攻击通过将原始有害问题中的敏感词替换为无害替代词，并利用上下文信息诱导模型重新解释这些替代词为原本的有害概念，从而绕过安全机制，但现有方法普遍存在攻击成功率不高、鲁棒性不足的缺陷。其核心原因在于：现有研究忽视了上下文本身在引发语义漂移方面的潜在能力差异。本文通过系统分析发现，不同上下文在诱发语义漂移方面的能力存在显著差异，具备更强语义漂移能力的上下文更易引导模型恢复出原始有害含义，从而提升越狱成功率。基于此发现，作者系统地识别并提炼出高效上下文的关键特征，并提出一种黑盒、上下文感知的语义漂移越狱框架——迭代上下文优化（Iterative Context Optimization, ICO）。该框架在每轮迭代中，结合已提取的上下文特征与目标模型的反馈结果，动态优化上下文表示，以增强其诱导语义漂移的能力。在三个数据集和八种主流基础模型上的大量实验表明，ICO相较于八种当前最先进的基线方法均表现出更优性能，平均攻击成功率达到74.6%，验证了其有效性与普适性。
链接: https://arxiv.org/abs/2608.03210
作者: Hujian Zhu,Yihao Huang,Felix Juefei-Xu,Xinfeng Li,Peng Zeng,Simeng Qin,Qing Guo,Geguang Pu
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:Foundation models have achieved remarkable success across diverse tasks, but they remain vulnerable. To investigate such vulnerabilities, semantic-shift jailbreaks have recently emerged as a promising attack paradigm. They bypass explicit safety mechanisms by replacing harmful terms in original harmful questions with benign alternatives and leveraging contextual information to induce the target model to reinterpret these alternatives as their corresponding harmful concepts. However, existing semantic-shift jailbreaks often achieve limited effectiveness. In this work, we reveal that this limitation arises from overlooking the semantic-shift capability of contexts. Through systematic analysis, we find that contexts exhibit substantially different abilities in inducing semantic shifts: contexts with stronger semantic-shift capabilities are more likely to guide models toward recovering harmful meanings and achieving successful jailbreaks. Based on this finding, we systematically identify and distill the characteristics of effective contexts and propose a black-box context-aware semantic-shift jailbreak framework with Iterative Context Optimization (ICO). In each iteration, ICO leverages these characteristics and feedback from the target model to optimize contexts. Extensive experiments on three datasets and eight target foundation models demonstrate that ICO consistently outperforms eight state-of-the-art baselines, achieving an average attack success rate of 74.6%.
[NLP-64] EduClaw-Bench: A Long-Horizon Benchmark for Pedagogical LLM Agents with Simulated Learners
【速读】： 该论文旨在解决当前生成式AI在教育应用中普遍存在的“短时、单任务”局限性问题，即现有基于大语言模型（Large Language Models, LLMs）的教育工具多为针对单一任务的点状解决方案，缺乏对长期、连续学习过程的有效支持。尤其在智能导师系统（agent tutor）领域，尚无基准能够评估其在持续30天教学关系中的表现，而真实教学本质上是长周期、渐进式的。为此，论文提出EduClaw-Bench基准，通过构建一个基于知识追踪（Knowledge Tracing, KT）模型的模拟学习者，其知识掌握状态由真实学生数据训练所得，从而驱动其回答行为，并在55个教学场景中动态探测学习进步。该基准从学习成效、响应性与助益性三个核心维度，以及加涅（Gagné）和罗森施奈德（Rosenshine）的教学设计原则两个课程设计维度对代理导师进行综合评估，其中助益性及课程相关指标由跨家族的三名大语言模型评委进行评判。实验对比了10种代理适配器在三个基础模型层级上的表现，揭示出两项关键发现：一是高质量辅导效果取决于基础模型与代理架构的协同作用，而非任一单独因素；二是几乎所有组合均无法在整个30天周期内维持高水平教学效能。通过校准检查（ECE=0.049）和真实课堂实地研究验证，确认模拟学习者及其测量指标与现实高度一致。本工作为构建可信赖的长期智能教育代理奠定了关键基础。
链接: https://arxiv.org/abs/2608.03206
作者: Unggi Lee,Sookbun Lee,Yeil Jeong,Eunjoo Lee,Minchul Shin,Hoilym Kwon
机构: 未知
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Large language models (LLMs) power educational applications from tutoring to essay scoring, but each is a point solution to a single task, and only recently have these point solutions been integrated into agents operating over a learning management system (LMS). Yet tutoring is long-horizon, since a learner improves over days and weeks rather than in a single turn, and no benchmark evaluates an agent tutor across a sustained relationship. We introduce EduClaw-Bench, a benchmark that places an agent tutor in a continuous 30-day relationship with a simulated learner grounded in knowledge tracing (KT), whose knowledge-concept mastery, from a KT model trained on real-student data, drives its answers and is probed for learning gain across 55 scenarios. Each agent is scored on three primary axes (learning gain, responsiveness, and helpfulness) and two curriculum-design axes (Gagné and Rosenshine), with helpfulness and the curriculum axes judged by a cross-family panel of three LLM judges. Evaluating 10 agent adapters over three base-model tiers yields two findings that single-tier, single-session evaluation cannot reach. First, tutoring quality belongs to the base model and the agent harness together rather than either alone. Second, almost no combination sustains good tutoring over the full horizon. A calibration check ( \textECE=0.049 ) and a live-classroom field study confirm that the simulated learner and its measurements track reality. Our work is a step toward trustworthy AI tutors for future education.
[NLP-65] Aligning Large Vision-Language Models at Test Time: A Trajectory-Guided Structured Sampling Approach
【速读】： 该论文旨在解决大视觉语言模型（LVLMs）在后训练阶段采用强化学习（RL）进行对齐时存在的资源消耗高、训练目标与推理时分布不匹配等问题。其核心解决方案是提出一种基于轨迹引导的测试时对齐方法，通过动态推理过程中的结构化采样实现精细化优化。关键在于构建一个由轨迹学习算法生成的推理记忆库，将复杂问题求解分解为预定义的有序推理模式序列；在推理阶段，首先从记忆库中提取轨迹以建立全局结构化推理先验，再利用迭代马尔可夫链蒙特卡洛（MCMC）算法对推理路径进行局部多目标优化，从而提升视觉定位准确性与逻辑一致性。实验表明，该方法在多个多模态推理数据集上显著提升了模型性能，且推理开销可控，验证了轨迹引导的测试时采样作为传统后训练对齐的高效可扩展替代方案，在复杂视觉推理任务中具有突出优势。
链接: https://arxiv.org/abs/2608.03204
作者: Tianbao Jiang,Weicong Ni,Gerard de Melo,Linlin Wang
机构: East China Normal University(华东师范大学); Hasso Plattner Institute(哈索·普拉特纳研究所); University of Potsdam(波茨坦大学); City University of Hong Kong(香港城市大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Post-training reinforcement learning (RL) algorithms are commonly used to align large vision-language models (LVLMs) with human intent and the requirements of visual reasoning tasks. However, existing RL-based alignment methods are often resource-intensive and encounter mismatches between training objectives and inference-time distributions. To bridge this gap, we propose a novel test-time alignment approach that leverages trajectory-guided structured sampling for dynamic inference-time refinement, achieving better alignment with visual grounding and ensuring logical consistency. Our approach begins with curating a reasoning memory bank via a trajectory learning algorithm, which decomposes complex question solving into ordered sequences of predefined reasoning patterns. It subsequently accomplishes inference-time alignment by first collecting trajectories from reasoning memory bank to establish a global structural reasoning prior, and then using an iterative Markov Chain Monte Carlo (MCMC) algorithm for localized multi-objective refinement of the reasoning trace. Experiments across multiple multimodal reasoning datasets demonstrate that our approach significantly improves accuracy without incurring prohibitive inference overhead. These results establish trajectory-guided test-time sampling as a scalable and effective alternative to traditional post-training alignment, particularly for complex visual reasoning tasks.
[NLP-66] Evidence-Grounded Multimodal Knowledge Graph Construction for Multi-Lecture Educational Reasoning
【速读】： 该论文旨在解决传统仅依赖文本转录（transcript-only）的检索方法无法充分保留讲座视频中多模态知识信息的问题，尤其是语音、幻灯片文字、图表、公式及呈现顺序等关键语义内容的丢失。其核心解决方案是提出一种基于证据的多模态处理流程：通过语音转录、语义锚点选择、光学字符识别（OCR）以及视觉-语言模型（vision-language model）联合分析，仅提取由转录文本、OCR结果或视觉证据共同支持的概念与类型化关系，并对提及内容进行验证与规范化，最终构建具有溯源信息（provenance-rich）的知识图谱。该方法的关键在于引入多源证据的交叉验证机制，确保所提取知识的可审计性与可信度，而非追求极致的性能指标，从而实现从复杂多模态讲座数据中精准、可追溯地构建结构化知识。
链接: https://arxiv.org/abs/2608.03161
作者: Sahil Al Farib,Momota Ahsana Meem,Sheikh Redwanul Islam,Md. Tanvir Raihan
机构: 1. University of Dhaka (达卡大学); 2. University of Dhaka (达卡大学); 3. University of Dhaka (达卡大学); 4. University of Dhaka (达卡大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Lecture videos distribute knowledge across speech, slide text, diagrams, equations, and presentation order, which transcript-only retrieval does not fully preserve. This paper presents an evidence-grounded multimodal pipeline that transcribes lectures, selects semantic anchors, applies optical character recognition (OCR), and uses a vision-language model to extract only concepts and typed relationships supported by transcript, OCR, or visual evidence. Mentions are validated and canonicalized into a provenance-rich knowledge graph. On three neural-network lectures, the pipeline processed 3,118 frames, 756 transcript segments, and 559 anchors. It retained 1,022 concept and 312 relationship mentions, yielding 172 canonical concepts and 282 relationships with 90.38% endpoint coverage. A preliminary three question retrieval test achieved 100% top-1 and top-3 accuracy and 100% mean top-5 recall. The contribution is an auditable construction method rather than a state-of-the-art performance claim.
[NLP-67] ANCHOR-RE: An Agent ic Neuro-Symbolic Framework for Grounded Biomedical Relation Extraction
【速读】： 该论文旨在解决生成式人工智能（Generative AI）在生物医学关系抽取（BioRE）任务中因缺乏领域知识约束与逻辑一致性而导致的高误报率（false-positive predictions）问题，同时克服传统符号系统（如SemRep）召回率低的局限性。其核心解决方案是提出ANCHOR-RE框架，通过融合本体引导推理（ontology-guided reasoning）、外部知识锚定（external knowledge grounding）以及数据驱动的验证规则（data-driven verification rules）三重机制，在不进行参数微调的前提下增强大语言模型（LLM）在生物医学文本中的推理可靠性。该方法显著提升了多个基准数据集（SemRepGS、DDI、ChemProt）上的微平均F1分数，并在2026年发表的新文献上保持了69%的高精度，证明其具备良好的泛化能力与抗预训练数据污染的能力，为无需训练的生物医学文献挖掘提供了一种高效且可靠的神经符号推理范式。
链接: https://arxiv.org/abs/2608.03154
作者: Shufan Ming,Yikun Han,Gibong Hong,Rui Zhang,Halil Kilicoglu
机构: 未知
类目: Computation and Language (cs.CL)
备注: Submitted to Journal of Biomedical Informatics (under review)
Abstract:Biomedical relation extraction (BioRE) extracts structured knowledge from biomedical literature for applications such as knowledge base construction and hypothesis generation. Traditional symbolic systems such as SemRep provide high precision but limited recall, while large language models (LLMs) offer stronger contextual reasoning but remain prone to false-positive predictions. We developed ANCHOR-RE, a framework that integrates ontology-guided reasoning, external knowledge grounding, and data-driven verification rules into LLM inference. We evaluated it on three BioRE benchmarks (SemRepGS, DDI, and ChemProt) using both proprietary and open-weight LLMs. To assess generalizability beyond benchmark datasets while reducing potential evaluation bias from LLM pretraining contamination, we conducted a temporal evaluation using 100 biomedical articles published in 2026. With the proprietary backbone, ANCHOR-RE outperformed direct LLM prompting, improving micro-F1 from 0.654 to 0.676 on SemRepGS, from 0.769 to 0.872 on DDI, and from 0.939 to 0.941 on ChemProt. On DDI and ChemProt, it also outperformed previously reported inference-only methods and approached fine-tuned or instruction-tuned systems without parameter updates. Similar performance gains observed with open-weight LLMs indicate that the benefits were not limited to the proprietary backbone. On the post-cutoff set, manual assessment of 500 randomly sampled predictions yielded a precision of 69%, maintaining consistent precision on previously unseen biomedical literature. Neuro-symbolic reasoning can improve the reliability of LLM-based BioRE without fine-tuning. Results across multiple benchmarks, model families, and post-cutoff literature support ANCHOR-RE as a practical training-free approach to biomedical literature mining.
[NLP-68] Internalizing Academic Writing Workflows for Introduction Generation via Struct-Aware Policy Learning
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在生成学术论文引言时面临的挑战，即如何在统一的叙事框架中协调背景介绍、研究空白识别、方法阐述与贡献陈述等多阶段内容，以确保逻辑连贯性与结构合理性。现有方法通常采用外部化的多阶段提示（multi-stage prompts）或代理工作流（agent workflows），但此类方案存在计算成本高、跨阶段偏差（cross-stage drift）等问题。本文提出一种结构感知的策略学习框架——StructPO，其核心在于将整个多阶段写作流程内化为一个单次通过的策略模型，并通过显式的阶段标记（stage tokens）进行控制。关键创新包括：引入结构感知的信用分配机制，实现局部阶段质量与全局连贯性之间的解耦；以及基于优化引导的修订机制，将改写行为内嵌至首次生成过程中。实验表明，StructPO在语义对齐、结构合理性及推理效率方面均优于基于工作流的基线方法，具备良好的域外泛化能力，在扩展至Qwen3-32B模型时，其人类评估表现可媲美GPT-5.1。结果证明，通过细粒度策略优化实现学术写作流程的内化，是替代高成本外部编排的一种可行路径。
链接: https://arxiv.org/abs/2608.03138
作者: Meicong Zhang,Tiancheng Su,Jiahao Cheng,Guoxiu He,Xinqi Tao,Dejia Song
机构: 1. Tsinghua University (清华大学); 2. Alibaba Group (阿里巴巴集团)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Generating a rigorous paper introduction with large language models (LLMs) remains challenging, since it requires coordinating background, gap identification, method and contribution within a coherent narrative. Existing solutions externalize this process as multi-stage prompts or agent workflows which are expensive and vulnerable to cross-stage drift. We propose StructPO, a struct-aware policy learning framework that internalizes the entire multi-stage writing workflow into a single-pass policy controlled by explicit stage tokens. StructPO introduces struct-aware credit assignment to decouple local stage quality from global coherence and refinement-guided optimization to internalize revision behavior into the first-pass policy. Experiments show that StructPO improves semantic alignment, structural rationality and inference efficiency over workflow-based baselines, generalizes to out-of-domain settings, and remains competitive with GPT-5.1 in human evaluation when scaled to Qwen3-32B. These results show that internalizing academic writing workflows through fine-grained policy optimization offers a viable alternative to costly external orchestration.
[NLP-69] DP-MemView: A Memory Interface for Attribute-Level Transcript Privacy in Long-Term LLM Agents
【速读】： 该论文旨在解决大语言模型（LLM）代理中长期记忆所引发的自适应话术隐私（adaptive transcript privacy）问题，即尽管敏感属性未被显式陈述，但通过反复基于记忆生成响应，攻击者仍可能累积推断出受保护的个人属性。其核心解决方案是提出DP-MemView，一种差分隐私（Differential Privacy, DP）接口机制，通过私有化选择公开的响应条件视图（response-conditioning views），并将这些视图而非原始记忆暴露给响应模型。每个视图的选择会按受保护属性的交集进行隐私预算分配，且每项属性维护独立的账本（per-attribute ledger），当预算耗尽时返回固定通用视图以阻止过度暴露。在明确的接口契约下，该方法可保证整个自适应话术序列满足纯B_a-DP（pure B_a-Differential Privacy）。研究进一步将结果扩展至跨多个受保护群体的数据存储场景，并量化了观察话术对攻击者先验概率的影响上限。实验在控制环境与公开语料迁移任务中验证了在线模式与预分配模式的有效性，均保持话术区分度接近随机水平的同时，维持了目标个性化需求与整体响应质量。诊断分析表明，移除关键防护机制将导致输出支持不匹配、账本计费缺失、侧信道泄露或长期泄漏增长等严重安全缺陷。
链接: https://arxiv.org/abs/2608.03130
作者: Jong Wook Kim,Byoungjae Min,Kennedy Edemacu,Yoonhyuk Choi,Sae-Hong Cho,Beakcheol Jang
机构: Google(谷歌); Meta(元); Stability.AI(稳定人工智能); Anthropic(Anthropic); Character.ai(字符人工智能)
类目: Cryptography and Security (cs.CR); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 18 pages, 2 figures, 9 tables
Abstract:Long-term memory enables persistent personalization in LLM agents, but repeated memory-conditioned responses can cumulatively reveal protected attributes even when they are never stated explicitly. We formalize this threat as adaptive transcript privacy and introduce DP-MemView, a differentially private interface that privately selects public response-conditioning views and exposes those views—rather than raw memory—to the response LLM. Each private selection is charged to every protected attribute whose memory group intersects the read set. Per-attribute ledgers block any selection that would exceed its cap and return a fixed generic view instead. Under an explicit interface contract, we prove pure B_a-DP for the entire adaptive transcript. We also extend the result to stores that differ across multiple protected groups and bound how much observing the transcript can change an adversary’s prior odds. We evaluate the online and preallocated modes with three response LLMs on a controlled adjacent-store benchmark and a public-corpus transfer track. Both modes keep transcript distinguishability near chance while preserving target-required personalization and overall response quality. Further diagnostics show that removing key safeguards causes mismatched output support, missing ledger charges, revealing side channels, or growing long-horizon leakage.
[NLP-70] From SQL Errors to Concept Gaps: An AI-Powered Knowledge Graph Analytics Platform for Personalized Feedback
【速读】： 该论文旨在解决学生在学习结构化查询语言（SQL）过程中，因概念性理解缺失而导致的语义错误难以被传统自动评分系统有效识别与诊断的问题。现有自动评分系统仅能判断查询是否执行成功或结果是否正确，无法将错误归因于背后的深层概念误解，因而反馈缺乏教学指导意义。其解决方案的关键在于构建一个由人工智能驱动的知识图谱平台，通过从课程教学材料中自动提取知识点及其关联关系，建立包含概念、关系与学生提交记录的图数据库，并基于此对学生的错误进行概念层面的分类与溯源。该平台利用知识图谱将具体代码错误映射到相关概念节点，从而揭示学生在JOIN、GROUP BY、HAVING等核心概念间的混淆问题。评估结果显示，95.7%的知识点节点被专家评为至少部分有效，63.8%的三元组关系被认定为完全正确，且专家反馈表明生成的知识图谱与教师的认知模型高度一致，能够提供可操作的诊断信息，为实现精准教学干预奠定了基础。
链接: https://arxiv.org/abs/2608.03118
作者: Abdulrahman AlRabah,Weijian Zhou,Xing Gao,Abdussalam Alawini
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:This innovative practice full paper describes an AI-powered knowledge graph platform that connects SQL errors to conceptual gaps in undergraduate and graduate database systems courses. Students learning Structured Query Language (SQL) frequently struggle with semantic errors that reflect conceptual misunderstandings rather than syntax mistakes. A query may execute yet return incorrect results due to gaps spanning related concepts; misusing NATURAL JOIN in place of an explicit subquery reflects intertwined misunderstandings of JOIN, GROUP BY, and HAVING. Autograding systems detect correctness but provide surface-level feedback without connecting errors to the conceptual structure of the course. Educational knowledge graph research has shown the value of structured concept representations for curriculum analysis and adaptive learning, but these approaches have not been applied to diagnosing SQL misconceptions from student submissions. We present a platform that automatically extracts course concepts and relations from instructional materials, links them to student submission traces through a graph database, and classifies errors at the concept level. We evaluate the platform across two database systems courses at two universities, one using real student submissions and one using simulated submissions, through an expert study with five participants and an automated evaluation using an LLM as a judge. Results show that 95.7% of extracted nodes were rated as at least somewhat valid and 63.8% of triplets were rated fully correct. Expert feedback confirmed that the generated graphs align with instructor mental models and that mapping errors to course concepts provides actionable diagnostic insight; evaluating impact on student learning remains future work.
[NLP-71] Convex-Hull-Neighborhood Smooth Dual Generalization: Controlling Local Correction Propagation in Offline RL
【速读】： 该论文旨在解决离线强化学习（offline RL）中因近似分布外（out-of-distribution, OOD）动作带来的估计误差在自举（bootstrapping）过程中被放大的问题。现有方法通常通过独立机制分别控制可接受的OOD区域范围或泛化目标的影响，导致优化路径不连贯。本文提出凸包邻域平滑双泛化（Convex Hull Neighborhood Smooth Dual Generalization, CSDG），其核心在于将贝尔曼更新（Bellman backup）形式化为样本内值目标与一个基于凸包邻域（CHN）的局部修正项之和，使泛化贡献显式可析，并将其与样本内参考路径分离。该修正项通过在不同扰动半径下采样的样本内导向与OOD导向候选值进行平滑融合获得，由混合系数λ调控其对每次更新的贡献，而递归折扣率仍保持为γ。在有界性与固定扰动核假设下，理论推导出精确的一步修正恒等式、时变迭代上界及仅依赖于固定点处分支差异的不动点上界；进一步刻画了理想算子所诱导的隐式策略，并给出条件非退化准则。实际算法采用非对称有界噪声与期望分位回归（expectile regression）近似上述量，无需精确的支持集分类或额外的悲观性OOD惩罚。在Gym-MuJoCo和AntMaze上的实验表明，该方法具备优异的综合性能与稳定的值函数估计能力。
链接: https://arxiv.org/abs/2608.03108
作者: Yi Yang,Zhennan Chen,Mingfeng Lv,Hanlei Li,Zhengsen Ruan,Lvqing Yang
机构: Xiamen University (厦门大学); Nanjing University (南京大学); Zhejiang Gongshang University (浙江工商大学)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注:
Abstract:Offline reinforcement learning (offline RL) can benefit from nearby out-of-distribution (OOD) actions, but estimation errors at these actions may be amplified by bootstrapping. Existing regularization and local-generalization methods control either the admissible OOD region or the influence of generalized targets, often through separate mechanisms. We propose Convex Hull Neighborhood Smooth Dual Generalization (CSDG), which expresses the Bellman backup as an in-sample value target plus a CHN-local correction. This formulation makes the generalized contribution explicit and separates it from the in-sample reference path. The correction is obtained by smoothing in-sample-oriented and OOD-oriented candidates sampled at different perturbation radii. A mixture coefficient lambda scales its contribution to each backup, while the recursive discount remains gamma. Under boundedness and fixed perturbation kernels, we derive an exact one-step correction identity, a time-varying iterate bound, and a fixed-point bound that depends only on the branch discrepancy at the fixed point. We further characterize the implicit policies induced by the idealized operators and give a conditional non-degradation criterion. The practical algorithm approximates these quantities using asymmetric bounded noise and expectile regression, without exact support classification or an additional pessimistic OOD penalty. Experiments on Gym-MuJoCo and AntMaze show strong aggregate performance and stable value estimation. Code is available at: this https URL
[NLP-72] HomoEnsNER: Does Language Alignment Outperform Architectural Complexity in Gujarati Named Entity Recognition?
【速读】： 该论文旨在解决低资源、形态丰富的古吉拉特语（Gujarati）命名实体识别（NER）任务中因缺乏大小写提示、丰富的词形变化、词汇歧义及自由词序带来的挑战。现有集成方法多依赖于异构模型架构的组合，如混合不同类型的分类器、多语言编码器或经典序列模型，而忽视了语言对齐的单语预训练模型在集成中的潜力。本研究提出HomoEnsNER，一种由五个独立微调的古吉拉特语BERT（GujaratiBERT）模型构成的同质集成框架，通过多数投票机制融合预测结果。实验表明，在统一训练预算下，该同质集成方案在Naamapadam古吉拉特语测试集上取得最高实体级F1分数（0.8442），显著优于单一GujaratiBERT基线（0.8347）及所有异构对比模型（最低为0.7855）。研究关键发现在于：对于低资源印度语言的NER任务，语言对齐的单语预训练模型所构成的同质集成，其性能优于依赖架构多样性的异构集成，证明了语言对齐性是比架构复杂性更高效且成本更低的集成策略。
链接: https://arxiv.org/abs/2608.03105
作者: Chandrakant K. Bhogayata
机构: 未知
类目: Computation and Language (cs.CL)
备注: 18 pages
Abstract:Named Entity Recognition (NER) for Gujarati remains underexplored, hindered by the absence of capitalization cues, rich morphology, lexical ambiguity, and free word order. Prior ensemble work has emphasized architectural diversity by combining heterogeneous classifiers, multilingual encoders, or classical sequence models, rather than exploiting language-aligned monolingual pretraining. This study asks whether, for a low-resource, morphologically rich language like Gujarati, a homogeneous ensemble of a single monolingual encoder outperforms such architectural diversity. We propose HomoEnsNER, a homogeneous ensemble of five independently fine-tuned GujaratiBERT models combined via majority voting, evaluated against a single GujaratiBERT baseline and six heterogeneous alternatives, including combinations with MuRIL-base, MuRIL-large, IndicBERT, mBERT, BiLSTM, CRF, and a stacked BiLSTM-CRF-GujaratiBERT architecture. All eight models were trained under a consistent budget and evaluated using entity-level F1 on the Naamapadam Gujarati test split. HomoEnsNER achieved the highest F1 (0.8442), surpassing the baseline (0.8347) and every heterogeneous alternative (lowest: 0.7855), indicating that language alignment is a more effective, budget-conscious ensembling strategy than architectural complexity for low-resource Indian language NER.
[NLP-73] What Language Does and What the Evidence Supports: A Functional Role Taxonomy and Evidence Audit of Language Grounding in Embodied Agents
【速读】： 该论文旨在解决生成式智能体中语言功能的可解释性与实证支持不足的问题，即现有研究虽普遍引入语言作为核心组件，却未能明确语言在具体任务中的实际贡献及其与物理行为的耦合程度。其解决方案的关键在于提出一个五重非互斥的语言功能角色框架：指令指定（Specification）、具身表征（Embodied Representation）、动作编排（Action Orchestration）、接地调控（Grounding Regulation）和执行耦合（Execution Coupling），并据此系统地追踪语言内容从输入到具身执行端的传递路径。通过识别可用于验证每种角色责任的观测或干预手段，该框架能够对文献中声称的语言作用进行逐项评估，揭示出当前研究中普遍存在“功能使用”与“证据支持”之间的脱节现象——例如，语言中间表示可能被错误定义、未被实际使用，或无法影响后续行为；即便行为直接依赖语言输入，系统整体成功也无法反推语言的必要性。因此，该研究强调以具体的角色主张为单位进行比较，而非基于模型架构，从而实现模块化与端到端智能体间的公平对比，并确保结论严格限定于所报告的实证证据范围内。
链接: https://arxiv.org/abs/2608.03099
作者: Yifan Guo,Chenghao Li,Zhu Wang,Wei Xu,Yu Li,Yulong Zhu,Zhuo Sun,Bin Guo,Zhiwen Yu
机构: Northwestern Polytechnical University (西北工业大学); University of Electronic Science and Technology of China (电子科技大学)
类目: Computation and Language (cs.CL)
备注: 19 pages, 3 figures, 11 tables
Abstract:Foundation models place language throughout embodied agents, but its presence does not show what it contributes or how well that contribution is grounded. This survey separates these two questions. We define five non-exclusive functional roles for language: Specification, Embodied Representation, Action Orchestration, Grounding Regulation, and Execution Coupling. For each role, we trace the path from linguistic content to its embodied consumer and identify the observations or interventions that can test the claimed responsibility. Applying this framework to the reviewed literature reveals a recurring gap between functional use and evidential support. Interpretable or revised linguistic intermediates may be incorrect, go unused, or fail to affect later behavior. Even when actions are directly conditioned on language, system-level success does not by itself isolate language’s contribution. We therefore evaluate grounding claim by claim, asking whether the reported evidence supports the specific responsibility assigned to language. Using role claims rather than architectures as the unit of comparison allows us to compare modular and end-to-end embodied agents without extending conclusions beyond the reported evidence.
[NLP-74] VIVID: A Culturally Grounded Benchmark Exposing the Figurative Language Gap in Vietnamese NLP LREC2026
【速读】： 该论文旨在解决越南语中根植于文化的隐喻语言理解能力评估缺乏系统性基准的问题。其核心解决方案是构建首个针对越南语习语与谚语的文化语境理解评测基准VIVID，包含1,636个经过五种复杂度属性（字面表达、语用细微差别、汉越词、非常用词汇、民俗知识）和七类语义主题标注的数据集，并提出结合生成式与判别式任务的评估框架。关键创新在于引入基于提示工程的“大模型作为评判者”（LLM-as-a-Judge）方法，通过多维度提示策略实现对模型表现的量化评估，且经人类判断验证（Cohen’s kappa = 0.792），确保评估可靠性。实验揭示当前主流模型在文化语境下的隐喻理解存在显著缺陷，尤其越南专用模型远逊于多语言模型，且多数模型得分不足满分的一半，暴露出字面过度解读、词汇鸿沟及语用扁平化等系统性问题，表明现有模型仍缺乏对文化内涵的深层理解能力。
链接: https://arxiv.org/abs/2608.03095
作者: Tu Tran Do,Nhat Ngoc Nguyen,Khanh-Tung Tran,Hoang D. Nguyen,Tu Minh Phuong,Long Hoang Dang
机构: 未知
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: LREC 2026
Abstract:We present VIVID (Vietnamese Idioms for Validation and Interpretation Depth), the first systematic benchmark for evaluating culturally grounded figurative language understanding in Vietnamese. VIVID comprises 1,636 idioms and proverbs annotated with five complexity traits (literal expressions, pragmatic nuances, Sino-Vietnamese terms, uncommon vocabulary, folk knowledge) and seven semantic themes. We establish an evaluation framework combining generative and discriminative tasks, proposing an LLM-as-a-Judge approach with aspect-based prompting validated against human judgment (Cohen’s kappa = 0.792). Evaluating eight state-of-the-art models reveals critical gaps: Vietnamese-specialized models drastically underperform multilingual systems (VinaLLaMA-7B: 0.13 vs. GPT-4o: 2.46), and even top models achieve less than 50% of maximum scores. Notably, few-shot prompting does not universally improve performance, with GPT-4o exhibiting degradation due to stylistic overfitting. Our analysis exposes systematic failures including literal over-interpretation, lexical gaps, and pragmatic flattening, demonstrating that current models lack cultural competence for nuanced figurative interpretation. VIVID provides an essential tool for advancing figurative language understanding in culturally rich contexts.
[NLP-75] SMOPD: Multi-Reward Reinforcement Learning via Specialize-and-Merge Online Policy Distillation
【速读】： 该论文旨在解决多奖励强化学习（multi-reward reinforcement learning）训练过程中，不同粒度奖励信号之间难以平衡的问题，尤其关注稀疏奖励（sparse reward）因信号强度不足而难以有效优化对应能力的挑战。现有方法如组奖励解耦归一化策略优化（Group Reward-Decoupled Normalization Policy Optimization, GDPO）虽通过分别归一化各奖励维度缓解了奖励信号相互遮蔽的问题，但在面对细粒度奖励（dense reward，如0.1–1.0范围的连续评分）与二元稀疏奖励（binary sparse reward，如0或1）共存的任务时，仍存在稀疏奖励贡献不足、导致其对应行为无法充分学习的问题。为此，本文提出一种两阶段的在线策略蒸馏方法——专精-融合在线策略蒸馏（Specialize-and-Merge Online Policy Distillation, SMOPD）。其核心解决方案在于：第一阶段“专精”（Specialize）通过奖励优先级配置，分别训练多个专注于特定奖励的教师模型，使每个奖励在最优信号条件下独立学习；第二阶段“融合”（Merge）则采用在线策略蒸馏机制，将各教师模型所掌握的奖励特化能力逐步整合进一个统一的学生策略中，同时保持任务层面的优化平衡。实验在互补奖励（工具调用准确率与格式正确性）和冲突奖励（有用性与安全性）两种设置下验证了SMOPD的有效性，结果表明其在1.5B、3B及7B参数量的模型上均显著优于GDPO，证明了该方法在增强稀疏奖励信号影响力的同时，有效保留了细粒度奖励所学习到的能力。
链接: https://arxiv.org/abs/2608.03092
作者: Wen Wang,Jiahua Bao,Tu Yongsiqi,Yihao Liu,Haotian Zhou,Haoxuan Ma,Mengyu Zhou,Wenkui Fan,Junwei He,Xiaoxi Jiang,Guanjun Jiang
机构: Alibaba(阿里巴巴); University of Chinese Academy of Sciences(中国科学院大学)
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 21 pages, 5 figures, 12 tables
Abstract:We aim to improve model performance in multi-reward reinforcement learning training process. Existing Group reward-Decoupled Normalization Policy Optimization (GDPO) has mitigated the issue of reward signals masking one another during direct scalarization by normalizing each reward dimension separately before aggregation. However, our experiments show that GDPO still struggles to balance reward signals with different granularities. Specifically, in some particular training tasks, the model may receive a dense reward that assigns fine-grained scores ranging from 0.1 to 1.0, together with a sparse reward that provides only binary feedback of either 0 or 1. In such cases, we find that the sparse reward may provide an insufficient optimization signal, preventing its corresponding capability from being effectively reinforced. Therefore, how can we strengthen the optimization signal from the sparse reward without sacrificing the capability already learned from the fine-grained reward? To overcome this limitation, we propose Specialize-and-Merge Online Policy Distillation (SMOPD), a two-stage training method for multi-reward optimization. Stage1-Specialize: SMOPD first employs reward-priority configurations to train multiple reward-specialized teachers, allowing each reward to be learned under conditions where its signal can effectively drive optimization. Stage2-Merge: SMOPD then utilizes online policy distillation to combine the reward-specialized capabilities of these teachers into a single student policy, while maintaining balanced task-level optimization. To validate our method, we conduct experiments on two multi-reward settings: complementary rewards(tool-calling accuracy and format) and conflicting rewards (helpful and harmless rewards). Based on above settings, SMOPD outperforms GDPO across 1.5B, 3B and 7B backbones.
[NLP-76] Scalable Frequency- and Length-Aware Subdocument Deduplication for Large Language Model Pretraining
【速读】： 该论文旨在解决大规模预训练语料库中子文档级冗余内容难以有效去除的问题。尽管文档级去重已被广泛采用，但跨子文档的重复内容（如片段、句子或代码段）在现有方法下仍难以被全面检测和处理。传统基于后缀数组的方法在分片（shard）内独立执行，无法发现跨分片的重复项，且其保留行为对分片配置敏感；而基于哈希的方法虽能实现全局精确去重，却通常依赖固定的复制保留策略，难以适应不同重复模式的异质性。本文提出一种可扩展的子文档去重框架，其核心创新在于将重复检测与复制保留策略解耦：首先通过自然边界分割、归一化精确哈希及分布式聚合识别重复组，随后引入显式的频率-长度感知保留策略，为每个重复组动态分配复制预算——对低频或短文本重复保留更多副本，对高频或长文本重复则更激进地删除。在FineWeb-Edu及含代码的网页语料上的实验表明，经该方法处理的数据训练出的模型在各项指标上均优于对比设置，验证了显式复制保留控制的重要性。
链接: https://arxiv.org/abs/2608.03089
作者: Hai Wang,Chenhao Wang,Qifeng Cai,Yixiu Liu,Miao Peng,Nuo Chen,Yuanlin Tu,Chengcheng Xu,Feng Zhang
机构: Tencent(腾讯)
类目: Computation and Language (cs.CL)
备注:
Abstract:Large-scale pretraining corpora contain substantial duplicate content. Although document-level deduplication is widely used, removing subdocument-level redundancy remains challenging. At corpus scale, suffix-array-based methods are commonly applied independently within shards, leaving cross-shard duplicates undetected and making the resulting retention behavior sensitive to the sharding configuration. Hash-based methods enable global exact duplicate counting, but often rely on fixed copy-retention policies that cannot accommodate heterogeneous repetition patterns. We propose a scalable subdocument deduplication framework that decouples duplicate detection from copy retention. It identifies duplicate groups through natural-boundary segmentation, normalized exact hashing, and distributed aggregation, and then applies an explicit frequency- and length-aware retention policy that allocates an adaptive copy budget to each group, retaining more copies of low-frequency or short repetitions while more aggressively deleting high-frequency or long ones. Experiments on FineWeb-Edu and a code-containing web corpus show that models trained on data processed by our method achieve the best overall performance among the evaluated settings. These results underscore the importance of explicit copy-retention control.
[NLP-77] GSTEP: Global Spatio-Temporal Density-Driven Visual Token Pruning for Efficient Video Large Language Models ACM-MM26
【速读】： 该论文旨在解决视频大语言模型（VideoLLM）在长视频理解任务中因冗余时空视觉标记（visual tokens）导致的高推理成本问题。现有标记剪枝方法多采用基于片段的局部剪枝策略，即将视频划分为独立片段并分别进行标记选择，这种设计容易忽略短时但语义密集的片段，并可能丢弃局部不显著但全局关键的标记。为此，本文提出一种即插即用的全局时空密度剪枝框架GSTEP（Global Spatio-Temporal Density Pruning），其核心创新在于将视频建模为连续的时空信息流，通过融合从平滑中心帧级变化信号获得的连续时间密度与帧内空间密度，构建标记级别的时空密度分布，并基于信息密度与覆盖度的联合平衡实现全局标记采样。实验结果表明，GSTEP在多个视频大模型和公开基准上均实现了优异的精度-效率权衡，具备良好的跨模型架构与评估场景泛化能力；以LLaVA-OneVision-7B为例，其可剪枝75%的视觉标记，在保持原始平均性能的基础上实现1.17倍的端到端加速。
链接: https://arxiv.org/abs/2608.03083
作者: Mengjie Zhang,Qihui Zhu,Tao Zhang,Shuangwu Chen,Huihuang Qin,Yu Guo,Shenghao Ye,Zijian Wen,Yunpeng Hou,Dong Jin,Xiaobin Tan,Huasen He,Jian Yang
机构: University of Science and Technology of China(中国科学技术大学); Institute of Artificial Intelligence, Hefei Comprehensive National Science Center(合肥综合性国家科学中心人工智能研究院)
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注: 4 figures, accepted to ACM MM 26’
Abstract:Video large language models (VideoLLMs) achieve strong video understanding performance, but their inference remains expensive due to the large number of redundant spatio-temporal visual tokens in long videos. Existing token pruning methods alleviate this cost by reducing redundant tokens, yet most of them rely on segment-level local pruning, where videos are partitioned into isolated segments and tokens are selected independently within each segment. Such designs may under-preserve short but semantically dense segments and discard tokens that appear non-salient locally but remain critical from a global perspective. To address this issue, we propose GSTEP (Global Spatio-Temporal Density Pruning), a plug-and-play pruning framework that models video as a continuous spatio-temporal information flow. GSTEP constructs a token-level spatio-temporal density by combining a continuous temporal density, obtained from a smoothed centered frame-level change signal, with intra-frame spatial density, and then performs global token sampling by jointly balancing information density and coverage. Extensive experiments on multiple VideoLLMs and public benchmarks demonstrate that GSTEP consistently achieves strong accuracy-efficiency trade-offs and generalizes well across model architectures and evaluation settings. On LLaVA-OneVision-7B, GSTEP prunes 75% of visual tokens, preserves up to 100.2% of the original average performance across benchmarks, and achieves a 1.17 end-to-end speedup.
[NLP-78] PAMT: Process-Aligned Reinforcement Learning for Multi-Domain Machine Translation
【速读】： 该论文旨在解决多领域机器翻译（Multi-domain Machine Translation, MDMT）中因缺乏精确的领域敏感性而导致的术语控制不准确与风格适应偏差问题，尤其在术语密集型和风格约束严格的场景下，现有基于大推理模型（Large Reasoning Models, LRMs）的显式推理机制虽能提升长文本及高难度翻译质量，却常出现术语漂移现象。其核心挑战在于“信用分配瓶颈”：现有方法仅优化最终输出或粗粒度翻译轨迹，无法精准识别哪些中间推理步骤对最终翻译质量有实际贡献。为解决此问题，论文提出过程对齐训练框架PAMT（Process-aligned Training），其关键在于结合冷启动阶段的领域感知长链思维（Long-CoT）监督与强化学习，通过序列级格式与结果奖励（针对最终翻译）以及步骤级过程奖励（衡量每一步推理对参考译文似然度的提升程度），实现对中间生成步骤的有效反馈与优化。该方法在两种主流模型架构上均显著优于基线模型，在域内、跨域及多语言场景下表现优于专用翻译模型，并保持与先进大语言模型/大推理模型相当的竞争力。
链接: https://arxiv.org/abs/2608.03077
作者: Yongshi Ye,Biao Fu,Chongxuan Huang,Yidong Chen,Xiaodong Shi
机构: Xiamen University (厦门大学); Institute of Artificial Intelligence, Xiamen University (厦门大学人工智能学院); School of Informatics, Xiamen University (厦门大学信息学院); Key Laboratory of Digital Protection and Intelligent Processing of Intangible Cultural Heritage of Fujian and Taiwan (Xiamen University), Ministry of Culture and Tourism (福建省和台湾地区非物质文化遗产数字化保护与智能处理重点实验室（厦门大学），文化和旅游部)
类目: Computation and Language (cs.CL)
备注: 23 pages, 10 figures, and 18 tables
Abstract:Multi-domain machine translation (MDMT) requires more than fluent generation: it demands domain-sensitive translation decisions such as domain disambiguation, terminology control, and stylistic adaptation. Large reasoning models (LRMs) make such decisions explicit through intermediate translation steps, but our analysis across 15 domains and four translation directions shows that this explicit reasoning is double-edged: it improves long-form and high-difficulty translation, yet often drifts in terminology-intensive and stylistically constrained settings. We trace this failure to a credit-assignment bottleneck: existing methods optimize final outputs or coarse trajectories, but cannot identify which translation steps actually help the final translation. To address this, we propose PAMT, a process-aligned training framework that combines cold-start domain-aware Long-CoT supervision with reinforcement learning. PAMT uses sequence-level format and outcome rewards for the final translation, together with a step-level process reward that measures how much each explicit translation step increases the likelihood of the reference translation. Across two backbones, PAMT improves over base models, outperforms MT-specialized baselines on average, and remains competitive with strong LLMs/LRMs across in-domain, OOD, and multilingual settings.
[NLP-79] AI Security Leaderboard: Methodology Results and Minimal Standard
【速读】： 该论文旨在解决当前前沿生成式人工智能（Generative AI）模型在防范恶意滥用方面所依赖的多层次防护机制的有效性与一致性缺乏公开实证评估的问题。其核心挑战在于：尽管各大厂商普遍部署了安全防护措施，但这些措施在面对复杂攻击时的实际鲁棒性尚不明确，且不同模型间的防护水平差异显著。论文提出“最低防护标准”（Minimal Standard for Safeguards, Version 1.0）作为解决方案，关键在于构建了一个包含67种可公开获取的静态越狱（jailbreak）技术的分类体系，并通过组合方法扩展出极大规模的攻击空间；在此基础上，设计了一个涵盖化学、生物、放射性/核及爆炸物（CBRNE）威胁与进攻性网络攻击共360个目标的基准测试集，采用三阶段漏斗法识别出在某一领域超过75%任务目标上均能触发合规响应的“通用越狱”（universal jailbreak）。此外，引入“越狱成本”（cost-to-jailbreak）指标，直接建模攻击者资源消耗，并以右删失下界反映未发现通用越狱的情况。实验结果表明，各模型防护能力存在数量级差异，随机搜索与专家引导组合策略分别在Grok 4.5和Gemini 3.1 Pro中发现了大量通用越狱，而Claude Fable 5与GPT-5.6 Sol则未被攻破。由于该标准仅基于已公开部署的现有防御技术，研究指出当前防护差距可通过强化纵深防御（defense-in-depth）策略——即结合推理审查、激活监控与输入/输出检测——有效弥合。
链接: https://arxiv.org/abs/2608.03070
作者: Jasper Timm,Lukas Struppek,Ziwei Xu,Grace Cheong,Oscar Mata,Dan Zhao,Mick Yang,Isadora De Andrade,Xiaojun Jia,Yiming Li,Samuel Bauer,Heather McIntyre,Adam Gleave,Edward Yee,Kellin Pelrine
机构: 未知
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Frontier AI model developers increasingly rely on layered safeguards to prevent catastrophic misuse, but little public evidence exists on how much protection these safeguards provide, or how consistently across developers. We introduce the this http URL Minimal Standard for Safeguards, Version 1.0: a taxonomy of 67 readily accessible static jailbreak techniques, a method for composing them into a very large attack space, and a benchmark of flagship models against a sample of it. We evaluate Claude Fable 5, GPT-5.6 Sol, Gemini 3.1 Pro, and Grok 4.5 on two complementary datasets totalling 360 attacker goals spanning chemical, biological, radiological/nuclear and explosive (CBRNE) threats and offensive cyber, using a three-stage funnel to identify universal jailbreaks: single prompt templates that elicit operationally compliant responses on over 75% of a domain’s goals. We also introduce a cost-to-jailbreak metric that models attacker spend directly, with right-censored lower bounds where no universal jailbreak was found. Robustness is highly uneven: the cost to break these models varies over a hundredfold. Random search over our technique pool found 63 universal jailbreaks against Grok 4.5 and 18 against Gemini 3.1 Pro, at an average cost of roughly 58 and 278 per jailbreak found; expert-guided composition raised these to 385 and 231. Neither Claude Fable 5 nor GPT-5.6 Sol yielded any universal jailbreak under either strategy. Because meeting the Minimal Standard requires only defenses already publicly described and deployed in production elsewhere, these gaps appear closable with current techniques. We recommend defense-in-depth combining reasoning, activation, and input/output monitoring. Results are maintained at this http URL. Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL) Cite as: arXiv:2608.03070 [cs.CR] (or arXiv:2608.03070v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2608.03070 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-80] CVPO: Enhancing LLM Reinforcement Learning Reasoning via Value-Variance Adaptation and Dynamic Curriculum Learning
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在强化学习（Reinforcement Learning, RL）增强推理能力过程中存在的反馈精度不足以及问题难度漂移（problem difficulty drift）的问题。现有方法在生成回答轨迹时难以精确评估每一步的反馈，导致策略更新不准确；同时，训练过程中未根据模型当前能力动态调整任务难度，影响学习效率与性能。为此，论文提出一种名为CVPO（Curriculum-guided Value-Variance Policy Optimization）的新方法，其核心创新在于：在响应轨迹层面，发现分词级价值方差（token-level value-variance）与探索强度相关，并通过理论分析证明该方差可约束策略更新的幅度；基于此，设计了一种考虑方差感知的优势修正机制，以适应不同类型的奖励信号，提升反馈精度与探索能力。在问题层面，引入动态课程加权机制，依据模型当前能力自适应调整题目难度，确保训练阶段聚焦于匹配当前水平的任务，缓解难度漂移问题。实验结果表明，CVPO显著优于现有的基于价值的强基线方法（如VAPO），在多种数学推理任务中实现了更优的性能与更强的鲁棒性，有效提升了语言模型的推理准确性与泛化能力。
链接: https://arxiv.org/abs/2608.03068
作者: Ziqi Jia,Yalu Ouyang,Bo Pang,Panpan Li,Hangfei Xu,Shengzhao Wen,Shiyong Li,Yanpeng Wang
机构: AI Cloud Group, Baidu(百度); Tsinghua University(清华大学); University of California, San Diego(加州大学圣地亚哥分校)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Reinforcement learning (RL) has emerged as an effective method for enhancing the reasoning capabilities of large language models (LLMs). However, existing methods suffer from insufficient precision in feedback on generated answer trajectories and exhibit the phenomenon of problem difficulty drift. To address these challenges, we propose CVPO - Curriculum-guided Value-Variance Policy Optimization. At the response trajectory level, we find that token-level value-variance correlates with exploration intensity. Our theoretical analysis shows this variance bounds policy update magnitude. We then use the estimated trajectory value-variance to quantify the intrinsic randomness in generation. Based on this, we design a variance-aware advantage adjustment mechanism for different reward types. At the question level, we introduce a dynamic curriculum weighting method that adapts to question difficulty. This helps the model focus on tasks matched to its current ability during each training stage. Experimental results show our method outperforms strong value-based baselines like VAPO. It achieves better performance and stronger exploration, enabling more accurate and robust reasoning in language models across various math tasks.
[NLP-81] Activation-Guided Neuron Intervention to Induce Alzheimers-Related Computational Language Phenotypes in a Large Language Model
【速读】： 该论文旨在解决阿尔茨海默病（Alzheimer’s disease, AD）早期认知功能障碍的检测与机制探究问题，尤其关注生成式语言模型中可计算表征是否能够功能性地影响行为表现。传统方法仅能识别AD相关语言变化，但无法验证这些表征在认知行为中的因果作用。为此，研究提出一种基于激活引导的干预框架（activation-guided intervention framework），其核心在于利用Qwen3-8B模型识别在AD语料中具有更高激活率的前馈神经元，并通过调节对应下投影权重（down-projection weights）来控制这些神经元在生成过程中的输出贡献。该方法生成九种不同干预方向、强度和范围的模型变体，系统评估其对12轮神经心理学任务的表现影响。结果显示，增强AD相关神经元会引发故事回忆、词汇流畅性、工作记忆、程序化话语、场景构建及指代消解等多维度认知功能的渐进性下降；而抑制则基本维持性能并改善部分指标。此外，强化干预还导致词汇意外度、思想密度、句法复杂性和话语量降低，与人类AD患者语言特征高度一致。这一结果表明，仅从临床语言差异中识别出的神经元即可跨多个认知领域影响行为，为阿尔茨海默病相关的计算表型（computational phenotype）提供了概念验证，并建立了一个可控的实验框架，用于深入解析语言与更广泛认知功能障碍之间的关联。
链接: https://arxiv.org/abs/2608.03067
作者: Rui He,Ercong Nie,Hong Jiang,Iris E. Sommer,Philipp Homan,Wolfram Hinzen
机构: Universitat Pompeu Fabra (庞培法布拉大学); Shanghai Jiao Tong University (上海交通大学); Beijing Institute of Technology (北京理工大学); Jinan University (暨南大学); University of Groningen (格罗宁根大学); University Hospital of Psychiatry Zurich (苏黎世大学精神科医院); University of Zurich (苏黎世大学); ETH Zurich (苏黎世联邦理工学院); Institució Catalana de Recerca i Estudis Avançats (加泰罗尼亚高级研究与教育机构)
类目: Computation and Language (cs.CL)
备注: 17 pages, 5 figures, 2 tables
Abstract:Changes in spontaneous speech provide an early signal of cognitive dysfunction in Alzheimer’s disease (AD) that large language models (LLMs) can detect. However, detection alone cannot establish whether the underlying model representations contribute functionally to behavior. We introduce an activation-guided intervention framework using Qwen3-8B. The framework identifies feed-forward neurons with higher activation rates for AD than control transcripts and modulates their output contributions during generation by scaling the corresponding down-projection weights. This yielded nine edited variants differing in intervention direction, magnitude, and scope. The original and edited models completed the same 12-turn neuropsychological battery, assessed through blinded human ratings and computational linguistic measures. Amplifying AD-associated neurons produced graded impairments in story recall, verbal fluency, working memory, procedural discourse, scene construction, and coreference resolution. Attenuation largely preserved performance and selectively improved several outcomes. Amplification also reduced lexical surprisal, idea density, syntactic complexity, and discourse quantity, broadly paralleling changes reported in human AD speech. These findings show that neurons identified solely from clinical language differences can influence behavior across multiple cognitive domains, providing proof of concept for an AD-related computational phenotype and a controlled framework for experimentally examining links between language and broader cognitive dysfunction.
[NLP-82] SeqLLM : Augmenting LLM s with Behavioral-Sequence Modeling for High-Stakes Decisions at WeChat Pay
【速读】： 该论文旨在解决大型支付平台在商户风险控制中面临的高复杂度多模态判断难题，即如何在每日筛查数千万商户时，有效平衡误报（false positives）与漏报（false negatives），尤其针对需同时理解商户文本信息与长期行为序列的疑难案例。传统方法中，大语言模型（LLM）虽擅长处理文本，但难以原生建模长序列行为数据，而直接迁移或微调又易引发灾难性遗忘（catastrophic forgetting）。其解决方案的关键在于提出一种名为SeqLLM的新型框架，通过三个核心组件实现对行为序列的高效建模而不损害原有语言能力：一是设计一个紧凑的离散行为词表，将行为事件编码为模型可原生处理的“行为标记”；二是引入轻量级投影器，采用两阶段对齐训练策略，将行为标记精准映射至预训练语言模型的语义空间；三是采用前缀引导的能力注入机制，通过任务前缀引导的监督微调方式获得序列建模能力，避免依赖持续预训练。该框架已在微信支付大规模部署，相较基于DeepSeek的生产基线，使风险筛查精度从92.0%提升至97.5%，并在亿级交易流量的欺诈检测系统中，使Top-0.01%的精确率提升26.8个百分点。此外，其预训练的行为标记嵌入在推荐基准测试中表现卓越，在MovieLens和Amazon上相对强基线User-LLM提升最高达32%的Recall@5，且在RecIF上以仅五分之一的计算资源超越完整OneRec-8B流水线。
链接: https://arxiv.org/abs/2608.03063
作者: Guilin Li,Jiaxing Zhang,Matthias Hwai Yong Tan,Bo Wang,Weiran Huang
机构: WeChat Pay, Tencent(微信支付，腾讯); Shanghai Jiao Tong University(上海交通大学); City University of Hong Kong(香港城市大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Merchant risk control at large payment platforms screens tens of millions of merchants daily, where false positives harm legitimate merchants and false negatives leave harmful activity undetected. The hardest cases require jointly understanding a merchant’s textual profile and long behavioral sequence. Large language models (LLMs) excel at text but cannot natively model such sequences, while adapting them often causes catastrophic forgetting. We present SeqLLM, a framework that adds behavioral-sequence modeling to a pretrained LLM while preserving its language ability. SeqLLM combines three components: a compact discrete vocabulary that represents behavioral events as native tokens; a lightweight projector, trained with a two-stage alignment curriculum, that grounds these tokens in the LLM’s semantic space; and prefix-guided capability injection, which acquires sequence-modeling ability through task-prefixed supervised fine-tuning rather than continual pre-training. SeqLLM is deployed at WeChat Pay, screening millions of merchants daily. Against the production DeepSeek-based LLM baseline, it raises screening precision from 92.0% to 97.5%. Its pretrained behavior-token embeddings also improve Precision@Top-0.01% by 26.8 percentage points in a production fraud detector serving billion-scale transaction traffic. Beyond payments, SeqLLM achieves state-of-the-art results on public recommendation benchmarks. On MovieLens and Amazon, it surpasses the strong User-LLM baseline by up to 32% relative Recall@5 while retaining markedly stronger language ability. On RecIF, it improves Pass@32 by 14.2% over the full OneRec-8B pipeline using only one-fifth of its GPU-days.
[NLP-83] PDD-RRG: Posterior Diagnostic Decision for Study-level Radiology Report Generation IJCAI2026
【速读】： 该论文旨在解决现有自动放射科报告生成（RRG）方法在临床实践中未能充分整合多源信息，尤其是当引入多视角图像和历史数据等额外输入时，反而可能引发不必要的诊断错误的问题。其解决方案的关键在于首次在报告生成后引入一个决策阶段，并提出后验诊断决策框架（PDD-RRG），通过构建输入数据的不同子集，利用现有RRG模型从多个视角生成报告，再基于贝叶斯后验概率与各临床观察的可学习阈值，对潜在冲突的诊断结果进行聚合，从而得出更可靠的综合诊断结论，并据此优化生成的报告。实验在MIMIC-CXR数据集上验证了该方法可在无需重新训练的前提下显著提升现有RRG模型的临床有效性。
链接: https://arxiv.org/abs/2608.03055
作者: Yang Yu,Yiming Ji,Bin Dai,Dong Zhang,Zhiyong Zhou,Shoushan Li,Yakang Dai
机构: Soochow University (苏州大学); Suzhou Institute of Biomedical Engineering and Technology (苏州生物医药研究所); Jiangsu Key Lab of Language Computing (江苏省语言计算重点实验室)
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注: Accepted by IJCAI 2026
Abstract:Automatic radiology report generation (RRG) aims to simulate the workflow of radiologists, assisting them in clinical diagnosis. However, existing methods often fall short in utilizing all information relevant to the examination, as is typically done in clinical practice. Although some works attempt to incorporate multi-view images and historical data, these additional inputs may sometimes lead to avoidable diagnostic errors on the contrary. To address these challenges, we introduce a decision-making stage after report generation for the first time and propose a Posterior Diagnostic Decision framework (PDD-RRG) to integrate potentially conflicting diagnoses. Specifically, we create various subsets of input data and utilize an existing RRG model to generate reports from different perspectives. Then the Bayesian posterior probability and the learned thresholds for each clinical observation are calculated to obtain an aggregated diagnostic conclusion, which is subsequently used to refine the generated report. Experiments on MIMIC-CXR demonstrate that our proposed PDD-RRG can effectively enhance the clinical efficacy of existing RRG models without any retraining.
[NLP-84] PI-Mem: Pushing Long-Context Reasoning to 3.6M Tokens with Parallel-Iterative Memory
【速读】： 该论文旨在解决大语言模型在长上下文推理中面临的两大核心瓶颈：一是传统循环记忆机制因顺序分块更新导致早期关键证据被后续无关内容覆盖，二是块间串行依赖关系限制了并行计算，造成推理延迟随上下文长度显著增加。其解决方案的关键在于提出PI-Mem（Parallel-Iterative Memory）机制，该机制通过并行处理所有输入分块，并在有限轮次内迭代优化共享记忆，每一轮均基于当前记忆并行读取各分块信息，选择新或互补的证据并融合为紧凑的共享状态，实现高效的信息整合。为避免冗余迭代，引入基于强化学习的辅助“轮次效率奖励”策略，使模型可自适应判断何时终止迭代以积累足够证据。实验表明，在高达360万标记的上下文长度下，PI-Mem在HotpotQA基准上相较于循环记忆基线分别提升6.25和7.81个百分点的准确率，同时实现6.1倍和2.1倍的推理加速，成功打破长上下文推理中的精度-效率权衡，为超长文档下的复杂多跳问答任务提供了可扩展的新范式。
链接: https://arxiv.org/abs/2608.03048
作者: Dawei Liu,Haixu Song,Shuang Cheng,Shijie Wang,Haozheng Hou,Kaifeng Liu,Ermo Hua,Zhonghang Yuan,Zhijie Zhong,Yuchen Fan,Biqing Qi,Bowen Zhou
机构: 1. Institute of Automation, Chinese Academy of Sciences (中国科学院自动化研究所); 2. School of Artificial Intelligence, University of Chinese Academy of Sciences (中国科学院大学人工智能学院); 3. Beijing Key Laboratory of Intelligent Information Technology, Beijing (北京市智能信息处理重点实验室, 北京); 4. State Key Laboratory of Software Development Environment, Beijing (软件开发环境国家重点实验室, 北京); 5. National Engineering Research Center for Big Data and Artificial Intelligence, Beijing (大数据与人工智能国家工程研究中心, 北京); 6. Beijing Institute of Technology (北京理工大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Long-context reasoning remains a critical bottleneck for large language models, as recent recurrent-memory approaches face two inherent challenges: sequential chunk-wise updates can overwrite early critical evidence with later irrelevant content, and serial inter-chunk dependencies limit parallelism and cause latency to increase with context length. To address these issues, we propose PI-Mem (Parallel-Iterative Memory), a mechanism that processes all chunks in parallel and iteratively refines a shared memory over a bounded number of turns. In each turn, PI-Mem reads all chunks in parallel conditioned on the current memory, selects new or complementary evidence from each chunk, and merges the selected evidence into a compact shared memory for the next turn. To discourage redundant turns, we optimize the workflow through reinforcement learning with an auxiliary turn-efficiency reward, enabling the model to adaptively exit once sufficient evidence has been accumulated. We evaluate PI-Mem with Qwen3.5-35B-A3B and Qwen2.5-7B on the HotpotQA benchmark across context lengths up to 3.6 million tokens and find that it outperforms the recurrent-memory baseline by +6.25 and +7.81 absolute points while achieving 6.1 \times and 2.1 \times inference speedups, respectively. These results demonstrate that PI-Mem breaks the accuracy–efficiency trade-off in long-context reasoning and provides a scalable approach to complex multi-hop question answering over extremely long documents.
[NLP-85] Emulate or Estimate? The Divergent Strengths of Base and Post-Trained Language Models for Opinion Simulation
【速读】： 该论文旨在解决大语言模型在模拟人类观点时存在矛盾研究结果的问题，即部分研究发现模型与人类调查数据具有良好的对齐性，而另一些研究则观察到角色坍塌（persona collapse）和较弱的人口统计敏感性。其核心问题是：先前研究中出现的不一致结果，很大程度上源于将两种本质不同的任务混淆。论文明确区分了两个任务：一是“模拟”（emulation），即模型生成个体响应，这些响应聚合后形成群体分布；二是“估计”（estimation），即模型直接预测群体分布。通过在Pew美国趋势面板数据集上评估六组匹配的基线模型与后训练模型，研究发现基线模型在模拟任务中表现更优——能够生成更接近真实人类分布的响应，并更好地保持人口统计结构特征；而后训练模型在估计任务中表现更佳，当被要求直接预测分布时，其预测精度更高。因此，论文提出的关键解决方案是：在选择模型用于人类观点模拟时，应根据具体任务需求决定——若需生成文本，则优先选择基线模型以实现更好的模拟效果；若只需预测分布，则应采用后训练模型以获得更准确的分布估计。
链接: https://arxiv.org/abs/2608.03044
作者: Seth Grief-Albert,Jessica Bo,Difan Jiao,Ashton Anderson
机构: Queen’s University; University of Toronto
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models are increasingly used to simulate human opinions, but prior work reports conflicting results: some studies find promising alignment with human survey data, while others find persona collapse and weak demographic sensitivity. We show that much of this conflict stems from conflating two distinct tasks. We call the first task emulation, in which models generate individual responses that aggregate into a population distribution. We call the second task estimation, in which models directly predict the population distribution. Evaluating six matched base and post-trained models on the Pew American Trends Panel, we find that base models are stronger emulators: they produce response distributions closer to human ground truth and better preserve demographic structure. Post-trained models are stronger estimators, producing more accurate distributional predictions when asked directly. We propose that model selection for human simulation should be guided by whether the task requires generating text or predicting distributions.
[NLP-86] Beyond Accuracy: A Multidimensional Evaluation of Statistical Reasoning in Large Language Models
【速读】： 该论文旨在解决当前大语言模型（Large Language Models, LLMs）在评估统计推理能力时过度关注回答准确性，而忽视模型如何构建与传达统计解释这一关键问题。其解决方案的关键在于提出一种多维度评估框架，综合运用回答准确性、响应行为分析、结构主题建模（Structural Topic Modeling）以及词汇相似性分析，全面考察模型生成的统计解释。研究通过对15个主流生成式AI模型在涵盖高中至研究生阶段共90道统计试题上的表现进行分析发现：尽管模型准确率差异显著（55%至78%），但所有模型在统计推理的概念组织上表现出高度一致性；同时，词汇相似性分析揭示了不同厂商（如Anthropic、OpenAI）之间存在微弱但稳定的解释风格差异，同一厂商模型的解释更具相似性。该研究证明，仅以准确性衡量生成式AI的统计推理能力具有局限性，必须结合对响应行为与解释内容的多角度互补分析，才能实现对生成式人工智能中统计推理能力的更全面、深入的评估。
链接: https://arxiv.org/abs/2608.03038
作者: Monnie McGee,Mateo Langston Smith,Julian Cabrera
机构: Southern Methodist University (南卫理公会大学)
类目: Computation and Language (cs.CL); Applications (stat.AP)
备注: 15 pages, 5 tables, 2 figures, presented at JSM 2026 and submitted for publication
Abstract:Statistical reasoning is multidimensional, yet evaluations of large language models (LLMs) typically emphasize response accuracy while overlooking how models construct and communicate statistical explanations. This study demonstrates the value of a multidimensional evaluation by combining response accuracy, response behavior, structural topic modeling, and lexical similarity analysis. The framework is applied to explanations generated by 15 current-generation LLMs responding to 90 questions drawn from four statistics examinations spanning high school, undergraduate, and graduate levels. Accuracy varied substantially across models, ranging from 55% to 78%. In contrast, structural topic modeling revealed a common conceptual organization of statistical reasoning across all models, while lexical similarity analysis identified modest but consistent vendor-specific differences in explanatory style. Models developed by the same vendor (e.g. Anthropic, OpenAI) produced explanations that were slightly more similar than models from different vendors. These findings demonstrate that statistical reasoning in contemporary LLMs cannot be characterized by accuracy alone and illustrate how complementary analyses of response behavior and model-generated explanations provide a more comprehensive evaluation of statistical reasoning in generative AI.
[NLP-87] Language Models Encode the Contextual Truth of Propositions
【速读】： 该论文旨在解决大语言模型（LLM）在处理上下文真值（contextual truth）时的表征机制问题，即当命题的真假依赖于上下文中的证据而非外部世界知识时，模型如何在激活空间中编码此类真值。其核心解决方案在于揭示：尽管输出策略结构不同，大语言模型仍保持对上下文真值的线性表征，且这种表征具有跨策略稳定性。通过控制实验（steering experiments）和基于双模型协作视觉-语言任务的对话转录分析，研究发现模型对某命题的真值判断会显著受到合作者陈述的影响，即使模型自身已具备足够证据做出独立判断。特别地，在决策边界附近的命题更易受合作者观点影响而发生真值表征的偏移。进一步区分表征与输出行为后，识别出两种形式的“迎合”（sycophancy）：一种是模型虽在输出上顺从错误主张但保留其为假的内部表征；另一种是模型将真值表征直接跨越决策边界进行修正。后者在模型显式重述错误主张时的发生概率（2.59倍）显著高于隐式认同的情况。
链接: https://arxiv.org/abs/2608.03035
作者: Rupak Sarkar,Pritika Ramu,Rachel Rudinger
机构: University of Maryland, College Park (马里兰大学学院帕克分校)
类目: Computation and Language (cs.CL)
备注:
Abstract:Prior work has shown that LLMs encode the truth of factual propositions along linear directions in activation space. It’s unclear how these representations extend to contextual truth: propositions whose truth is determined by in-context evidence rather than world knowledge. We show that LLMs maintain a linear representation of contextual truth that persists across structurally different output policies, even when the output doesn’t require the model to determine a proposition’s truth, and show causal evidence via steering experiments. Using the transcripts from a collaborative vision-language task that requires two LLMs to maintain a shared common ground, we show that truth representations of a proposition are significantly swayed by partner assertions about that proposition, even when the LLM has enough evidence to determine its truth. We find evidence that propositions near the decision boundary are more susceptible to having their truth shifted through partner assertions. Separating representation from output distinguish two forms of sycophancy that output behavior alone cannot: the model may accommodate a false proposition while continuing to represent it as false, or shift its representation across the boundary. The latter is 2.59x more common when the model agrees by restating the false claim explicitly than when it agrees implicitly.
[NLP-88] On the Non-Specificity of Statistical Measures Used in Script Decipherment
【速读】： 该论文旨在解决当前在未破译文字系统（如印度河文字）研究中，仅凭统计规律性便推断其编码语言的合理性问题。其核心挑战在于：现有统计特征是否具有足够的特异性，以区分语言与非语言的结构化符号系统。论文提出的关键解决方案是构建SIGIL——一个专为实验设计的生成式象征系统（Generative Emblem System），其3,000个文本的核心语料库虽具备明确的组合语义，但无任何音值对应关系。通过预先建立的文献注册机制，评估共纳入54种方法，并在可复现的发布结果与预定义判别规则下实现精确评分。结果显示，SIGIL在重复性、方向不对称性及词项分布等各项测试中均与印度河文字表现一致；熵、频率、位置、预测性、分类器及网络指标等声明性重构结果也重现了典型的“印度河文字特征”。进一步的序列破译压力测试表明，该系统在英语、梵语和泰米尔语上可达到高词典覆盖率，但组间保留样本下降与密钥不稳定揭示了覆盖率本身无法有效识别真实语言编码。因此，该构造证明：所评估的统计度量虽能检测到系统性组织，却并不具备语言特异性，故仅依赖这些度量无法独立确立未破译符号系统编码的是语言。
链接: https://arxiv.org/abs/2608.02999
作者: Nikhil Raghavendra
机构: Independent Researcher
类目: Computation and Language (cs.CL)
备注: 28 pages, 27 figures
Abstract:Statistical regularities are routinely offered as evidence that undeciphered sign systems encode language; the Indus script debate is the canonical example. Any such inference rests on specificity: the reported outcome must be unusual among plausible structured non-languages. We test that premise constructively with SIGIL, a purpose-built generative emblem system whose 3,000-text core corpus carries explicit compositional meanings although no sign has a phonological value. A literature registry compiled in advance of evaluation records 54 methods and admits a method to exact scoring when both the published Indus outcome and a source-defined decision rule can be reproduced. SIGIL receives the same category as the Indus corpus on every criterion scored this way, across repetition, directional-asymmetry, and lexical-distribution tests. Declared reconstructions of entropy, frequency, positional, predictive, classifier, and network measures reproduce the familiar Indus-like signatures as well. A sequential decipherment stress test then reaches high dictionary coverage for English, Sanskrit, and Tamil on the same corpus, while grouped held-out declines and unstable keys reveal how little that coverage identifies. The construction does not decide what the Indus signs encode: it shows that the evaluated measures detect organization without being specific to language, and therefore cannot, on their own, establish encoded speech.
[NLP-89] AcceptMoE: Commitment-Weighted Self-Sizing Verifier Expert Sets for Efficient MoE Speculative Decoding
【速读】： 该论文旨在解决在混合专家模型（Mixture-of-Experts, MoE）中进行推测解码（Speculative Decoding）时，因并行验证树结构中的多个候选节点而导致激活专家集合过大所带来的计算与通信开销问题。传统方法在验证过程中会激活所有路径节点所选专家的并集，造成显著的专家权重传输量和显存占用，尤其在专家卸载（offloading）场景下，主机与设备间的通信流量成为瓶颈。其解决方案的关键在于提出AcceptMoE——一种基于验证器侧的专家选择机制，该机制融合目标路由得分与离线估算的承诺概率（commitment probability），动态调整每个验证块的可选专家数量，无需用户指定专家预算。此外，在卸载场景下，AcceptMoE将专家可用性判定条件由预测自然路由改为依赖缓存驻留状态（cache residency），从而有效减少不必要的专家权重传输。尽管这一策略改变了模型的分布特性，但在12个跨三类MoE模型与四个基准任务的实验中，AcceptMoE的平均准确率仅比使用自然路由的EAGLE-3低0.27个百分点；在批处理大小为1的情况下，其吞吐量分别达到基线的1.290倍（全专家权重在GPU内存中）和2.06倍（物理卸载），同时主机到设备间通信流量降低73.6%至77.1%。
链接: https://arxiv.org/abs/2608.02989
作者: Shuang Liang, Hao (Mark)Chen,Zhiwen Mo,Qianzhou Wang,Guoyu Li,Lingxiao Ma,Wayne Luk
机构: University of Oxford (牛津大学); Imperial College London (帝国理工学院)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL); Distributed, Parallel, and Cluster Computing (cs.DC)
备注: 10 pages, 5 figures
Abstract:Speculative decoding verifies a tree of draft tokens in one target-model forward pass. For a mixture-of-experts (MoE) target, however, parallel verification can activate the union of the experts selected by all tree nodes, even though only a small subset of those nodes reaches the accepted output. Token count, activated-expert union size, and expert-weight traffic are therefore distinct cost measures: reducing the token workload need not shrink the expert union proportionally, and under offloading, transfer traffic also depends on cache residency. We introduce AcceptMoE, a verifier-side expert selector that combines target-router scores with offline-estimated commitment probabilities and automatically adjusts the number of eligible experts for each verification block, eliminating the need for a user-specified expert budget. Under offloading, AcceptMoE conditions expert eligibility on cache residency instead of predicting natural routes and prefetching the corresponding expert weights. Although constraining target-expert eligibility changes the model distribution, across 12 model-task pairs spanning three MoE targets and four benchmarks, AcceptMoE’s mean accuracy is 0.27 percentage points lower than that of EAGLE-3 speculative decoding with natural routing. Served with SGLang at batch size one, it reaches 1.290 times the throughput of this baseline with all expert weights in GPU memory, and 2.06 times under physical expert offloading, while reducing host-to-device traffic by 73.6 percent to 77.1 percent.
[NLP-90] mporal Leakage in LLM Backtesting: Measurement Validation and Adjusted Scores
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）在回溯测试（backtest）中因时间截断点（training cutoff）导致的污染（contamination）检测失效问题。传统方法仅通过比较训练截止时间前后得分的变化来判断是否存在数据泄露，但研究发现该方法本质上无法提供有效信息：即使在模型未记忆相关问题的情况下，四个主流模型在截止时间后的问题上仍表现出显著得分提升，其根源在于模型对临近截断时间点的事件具有更强的“认知优势”，这种“时效性”会人为模拟出数据泄露的假象。论文指出，任何被动回溯测试都无法从真实能力中分离出由时间邻近效应引发的虚假优势。解决方案的关键在于引入外部信息以实现可验证的测量——一是利用已知的训练截断点识别边界处的局部泄露；二是构建匹配的干净对照组（matched clean control），从而全局性地识别并校正泄露影响，获得调整后的得分。此外，研究揭示了泄露主要集中在那些出人意料且在训练数据中被充分覆盖的事件上，部分记忆现象会被过度奖励。通过在孪生模型中植入人工泄露并验证，所提出的估计器能够准确恢复注入的泄露量并在无泄露问题上返回零值，具备良好的可验证性。将该方法应用于前沿模型后，成功检测到一个局限于截断点的泄露信号，并在审计效力下界处排除了五个仅因时间邻近性而看似优越的模型。因此，回溯测试无需废弃，只需引入一个可辩护的参照基准即可实现可信评估。
链接: https://arxiv.org/abs/2608.02985
作者: Zeyu Zhang,Bradly C. Stadie
机构: Northwestern University (西北大学)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL); Machine Learning (stat.ML)
备注: 12 pages main content, 45 pages in total
Abstract:The standard check for contamination in LLM backtests is simple: compare scores before and after the training cutoff. We show this check is uninformative. Four flagship models fail it on questions they cannot have memorized: every scored question resolved after their cutoffs. The reason is structural. Models legitimately know more about times near their cutoff, so recency mimics leakage, and we prove no passive backtest can separate the two from genuine skill. Measurement, not just detection, requires information from outside the backtest. We supply it in two forms. A known cutoff identifies leakage at the boundary; a matched clean control identifies it globally and yields a leakage-adjusted score. We also derive where leakage hides: it concentrates on outcomes that surprised the crowd and were well covered in training, and partial memorization is disproportionately rewarded. We validate the estimators against ground truth by planting leakage in twin models, where they recover the injected dose and return null on clean questions. Deployed on frontier models, they detect one cutoff-localized signature and, at the audit’s power floor, clear five models whose apparent advantages were recency alone. Backtests need not be discarded; they need one defensible reference.
[NLP-91] QLite: Multi-LLM Jury Guided Distillation for Real-time MQM Translation Quality Evaluation
【速读】： 该论文旨在解决大规模语言模型（Large Language Models, LLMs）和大规模推理模型（Large Reasoning Models, LRMs）在机器翻译质量（Translation Quality, TQ）评估中虽表现优异但计算成本高昂、难以规模化部署的问题。而小型语言模型（Small Language Models, SLMs）虽然具备更高的效率，却因缺乏复杂推理能力，在评估任务中表现受限。其解决方案的关键在于提出TQLite——一种创新的模型蒸馏框架，通过构建多LRM评审团（multi-LRM jury），利用实用的数据清洗与多模型评估结果聚合技术生成高质量的合成训练数据，从而实现对SLMs的有效知识迁移。实验表明，经TQLite蒸馏后的SLMs在基于机器翻译质量评估（MQM）的任务上性能显著超越通用SLMs，接近最优LRM基线，为TQ评估提供了可扩展且成本效益更高的替代方案。
链接: https://arxiv.org/abs/2608.02975
作者: Bhavin Jawade,Cameron R. Wolfe
机构: Netflix(奈飞)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 16 pages, 9 figures
Abstract:Large language models (LLMs) have demonstrated impressive performance in MQM-based translation quality (TQ) evaluation, and recent advances in large reasoning models (LRMs) promise even greater improvements. However, both LLMs and LRMs are computationally expensive to deploy at scale, while small language models (SLMs)—though much more efficient—struggle with the complex reasoning required for evaluation tasks. In this work, we present an extensive empirical study benchmarking SLMs, LLMs, and LRMs across a wide range of TQ evaluation setups, providing a comprehensive view of the current landscape and establishing best practices. To address the scalability challenge, we introduce TQLite, a novel distillation framework that enables SLMs to approach the MQM evaluation performance of the best LRM-based evaluators. Our approach leverages a multi-LRM jury to generate high-quality synthetic training data via practical data curation techniques and aggregation of evaluation responses across a diverse panel of models. Our results demonstrate that SLMs trained via TQLite achieve strong MQM evaluation performance that far exceeds off-the-shelf evaluation capabilities of standard SLMs, offering a scalable and cost-effective alternative to LLM- and LRM-based evaluators.
[NLP-92] Mapping the City Through the Lens of Language Models
【速读】： 该论文旨在解决语言模型在处理城市指代时，基于未明确说明的假设对城市规模、形态、基础设施、环境与功能等维度进行隐含推断的问题。其核心挑战在于揭示这些隐含偏见的实质内容，而传统方法难以量化此类无形认知。解决方案的关键在于构建一个去标识化的评估框架：通过10个开源权重检查点，对来自40项经审计指标和7个领域的匿名化真实城市形态数据进行评分；该框架融合了基于约束概率的评分机制、预设可靠性筛选、谱系感知聚合、多重人口加权、独立复现样本以及全量画像验证。结果显示，模型普遍偏好开发面积较大、近期增长较快、基础设施与非住宅容量更高、形态更紧凑的城市。地理差异在控制城市规模与发展水平后显著缩小，且典型性与理想性在多数情况下高度一致。该框架将语言模型对“普通城市”的模糊认知转化为可实证追踪的量化图景，揭示了模型视角下具有共性的、但依赖于具体模型的城市表征模式。
链接: https://arxiv.org/abs/2608.02971
作者: Wanqi Liu,Rong Zhao,Zhizhou Sha,Qinyu Cui,Yecheng Zhang
机构: University College London(伦敦大学学院); Tsinghua University (清华大学); The University of Texas at Austin (德克萨斯大学奥斯汀分校); South China University of Technology (华南理工大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Language models often complete an underspecified reference to a city with unstated assumptions about urban size, form, infrastructure, environment, and function. We measure those assumptions without naming places. Ten open-weight checkpoints rate anonymized profiles derived from real morphological urban centres across 40 audited indicators and seven domains. The design combines constrained probability-based ratings, prespecified reliability screens, lineage-aware aggregation, multiple population weightings, an independent replication sample, and whole-profile validation. The clearest shared tendency favours urban profiles with larger developed area, faster recent growth, greater mapped infrastructure and non-residential capacity, and less sparse form. Most eligible directions recur in the replication data, and direct ratings of complete profiles show moderate agreement with the indicator-wise construction. Geographic differences shrink after accounting for city scale and development, while reliably measured paired tasks indicate that typicality and desirability are often closely aligned. The framework makes an otherwise vague notion of what models regard as an ordinary city empirically traceable. The resulting evidence delineates a shared yet model-dependent portrait of the city through the lens of language models.
[NLP-93] Every Wrong Answer Counts: Option-Level Psychometrics for LLM Multiple-Choice Benchmarks
【速读】： 该论文旨在解决现有大语言模型（LLM）多选题评估体系中仅依赖正确与否的二元评分方式所导致的信息损失问题。传统方法将所有错误选项视为等效错误，忽略了模型在错误选项间的偏好模式中蕴含的系统性行为特征与能力信息。为此，作者提出生成式AI（Generative AI）驱动的大语言模型名义反应模型（LLM-NRM），其核心在于构建一种考虑选项特性的心理测量框架，能够联合建模模型能力与选项层面的题目特征，同时分离出模型特有的响应校准锐度、位置偏好以及难度依赖的退化行为。实验基于14个基准数据集中的31,554道题目和189个LLM，结果表明，相较于二元项目反应模型与传统名义响应基线，LLM-NRM在预测未见的模型-题目交互上具有更高精度；其估计的能力值与外部人类偏好排名（Elo排行榜）之间的斯皮尔曼相关系数达0.920，为当前最优。研究进一步发现，错误选项的身份信息每题可提供额外101%的费雪信息量，且仅凭错误响应即可恢复接近全信息的能力估计（斯皮尔曼相关系数0.943）。此外，通过学习到的题目参数可实现高效测评，仅需41个精选题目即可保持完整榜单排名，肯德尔相关系数达0.85，相较原规模降低770倍。结论表明，错误回答并非等价的失误，而是承载了独特且关键的测量信息。
链接: https://arxiv.org/abs/2608.02966
作者: Xiao Fei,Yang Zhang,Sarah Almeida Carneiro,Michalis Vazirgiannis
机构: University of Athens (雅典大学); National Technical University of Athens (国立雅典理工大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Most multiple-choice question (MCQ) benchmarks evaluate Large Language Models (LLMs) only by whether they select the correct answers. This binary scoring treats all incorrect responses alike, even though an LLM’s preferences among incorrect options may contain systematic and useful information about its behavior and ability. We introduce the LLM Nominal Response Model (LLM-NRM), an option-aware psychometric framework that models the full distribution over answer choices to jointly estimate LLM ability and option-level item characteristics, while separating model-specific response calibration sharpness, positional preference, and difficulty-dependent fallback behavior. Across 189 LLMs and 31,554 items from 14 benchmarks, LLM-NRM predicts held-out LLM-item interactions more accurately than binary Item Response models and conventional nominal-response baselines, and its ability estimates achieve the strongest Spearman correlation of 0.920 with the external human-preference this http URL Elo leaderboard. Distractor identity contributes +101% additional Fisher Information per item beyond correctness, and incorrect responses alone recover full-information ability estimates with Spearman 0.943. The learned item parameters also enable efficient benchmarking, where 41 selected items preserve the full-bank ranking with Kendall’s correlation 0.85, corresponding to a 770 times reduction. In conclusion, we show that incorrect answers carry distinct and useful measurement information rather than representing equivalent mistakes.
[NLP-94] ATFlash: Per-RoPE-Wavelength Attention Windows for Compute/Memory-Efficient LLM Inference
【速读】： 该论文旨在解决大模型在长序列推理中注意力计算复杂度随序列长度呈二次增长的问题，尤其是在高上下文长度下导致的显存占用与计算开销过大的瓶颈。其核心挑战在于如何在保持模型性能的前提下，有效降低注意力机制中的计算量。解决方案的关键在于提出基于旋转位置编码（RoPE）波长的局部距离窗口机制（per-RoPE-wavelength distance window）：该机制利用RoPE中每个二维旋转频率对的波长特性，将查询-键（query–key）内积项按波长比例进行剪枝，仅保留距离在相应波长范围内的项。与传统滑动窗口不同，该方法不强制丢弃远距离键，而是通过低频分量保证所有键仍可被访问，从而在保持全局感知能力的同时实现输入无关的、闭式表达的对数级计算压缩（与序列长度 N 的对数成正比）。该剪枝策略在频率层面独立于现有的动态稀疏方法，具有良好的可组合性。实验表明，在Qwen2.5-0.5B和Llama-3.2-3B上，该方法可剪枝37%–48%的内积项，同时在LongBench-v2等长文本基准上保持96%–98%的顶级匹配率及约 10−3 nat 的输出分布差异；在多个长上下文任务如RULER、OpenAI-MRCR、LongCodeQA和∞ Bench上也表现出优异的性能稳定性。该实现以轻量级张量切片形式嵌入到FlashAttention-4预填充与FlashInfer解码流程中，无需修改在线Softmax递推逻辑，部署成本极低。在RTX PRO 6000上，对Llama系列模型的加速比随上下文长度增加而提升，最大达1.29倍（128K上下文）；在Qwen2.5-7B-1M模型上，当剪枝57%的内积项时，端到端推理速度提升达1.31倍，验证了其在超长上下文场景下的高效性与实用性。
链接: https://arxiv.org/abs/2608.02947
作者: Shun-ichiro Hayashi,Daichi Mukunoki,Tetsuya Hoshino,Takahiro Katagiri
机构: University of Tokyo (东京大学); National Institute of Advanced Industrial Science and Technology (日本产业技术综合研究所)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注:
Abstract:The attention score with rotary position embeddings (RoPE) decomposes exactly into a sum over its 2D-rotation frequency pairs, and each pair’s wavelength limits how far it can discriminate position. Aligned with this structure, we propose the per-RoPE-wavelength distance window: it prunes the query–key inner-product terms beyond a wavelength-proportional distance. Unlike a sliding window, every key remains reachable, at least through the low-frequency pairs. The reduction rate is input-independent, with a closed form logarithmic in the sequence length N , in contrast to dynamic-sparse methods like MInference. Such token-level selection is orthogonal to our frequency-level pruning. The window can therefore be applied on top of those methods. On Qwen2.5-0.5B and Llama-3.2-3B, the window prunes 37–48% of the query–key inner-product terms within each model’s native context length. Relative to full attention, the top-1 match rate stays at 96–98% and the mean output-distribution KL at the 10^-3 -nat level on LongBench-v2 contexts. We examine absolute scores on long-context benchmarks such as RULER, OpenAI-MRCR, LongCodeQA, and \infty Bench: they are broadly preserved. We implement the window as a slice of the query–key contraction axis, leaving the online-softmax recurrences untouched, and port it with minimal diffs into the released FlashAttention-4 prefill and FlashInfer decode. On RTX PRO 6000 with Llama, both ports outpace stock with gains growing with context length, up to 1.29\times at 128K. End to end on Qwen2.5-7B-1M, with 57% of the inner-product terms pruned, the speedup reaches 1.31\times at a 1M-token context.
[NLP-95] OPTD: On-Policy Transition Distillation with Consistency-Guided Adaptive Compression for Few-Step Diffusion Language Models
【速读】： 该论文旨在解决生成式AI（Generative AI）中扩散语言模型（Diffusion Language Models, dLLMs）在高效推理时的“策略偏差”问题：现有少步蒸馏方法依赖离策略（off-policy）监督信号，导致学生模型在推理过程中早期并行决策所引发的状态漂移，使其实际访问的状态与训练时的监督状态严重偏离，尤其是在高阶压缩场景下更为显著。其解决方案的关键在于提出一种基于策略的过渡蒸馏方法（On-Policy Transition Distillation, OPTD），通过在学生模型自身轨迹上采样部分状态，利用一个仅接受问题输入的冻结教师模型识别与最终结果对齐的未来候选路径，并依据当前状态置信度进行排序。随后，选择最长的前缀序列，使得联合承诺仍能保持教师模型的完整生成结果。该方法引入集合瓶颈目标（set-bottleneck objective）以将所有验证过的未来候选路径提升至解码释放阈值，同时采用冻结教师模型的KL散度作为锚点正则化其余活跃位置，从而实现一致性引导下的自适应压缩。整个过程不依赖黄金响应（gold response），且在四个数学推理与代码生成基准测试中均显著提升了质量-效率权衡表现，达到所评估少步基线中的最优质量约束下的平均曲线下面积（AUP）。
链接: https://arxiv.org/abs/2608.02942
作者: Xiaocheng Lu,Hualei Zhang,Shuhan Guo,Jie Zhang,Xiaoyi Pang,Jian Liu,Haoxi Li,Bohai Gu,Haoxuan Che,Jingcai Guo,Song Guo
机构: 未知
类目: Computation and Language (cs.CL)
备注: 9 pages, 4 figures, 5 tables
Abstract:Diffusion language models (dLLMs) can predict many tokens in parallel, but accurate generation still requires many iterative denoising steps. Few-step distillation accelerates decoding by compressing multiple teacher steps into a single student transition. However, existing methods construct supervision on off-policy trajectories. At inference, the student’s early parallel commitments alter the context of later predictions, so the states it actually visits drift away from the supervised ones–precisely when step compression is most aggressive. On-policy distillation is a natural remedy for this mismatch, but it leaves open how far each transition should advance: matching only the teacher’s next action limits compression, while indiscriminately merging future actions can violate intermediate dependencies. To address this limitation, we propose OPTD, On-Policy Transition Distillation with consistency-guided adaptive compression. It samples partial states from the few-step student’s own trajectories, uses a frozen, question-only teacher to identify outcome-aligned future candidates, and orders them by current-state confidence. The method then selects the longest prefix whose joint commitment preserves the teacher’s rollout outcome. A set-bottleneck objective promotes every verified future candidate to the decoder’s release threshold, while a frozen-teacher KL anchor regularizes all other active positions. Neither target construction nor training uses a gold response. Across four mathematical reasoning and code-generation benchmarks, OPTD consistently improves the quality–efficiency trade-off and attains the strongest overall quality-constrained AUP among the evaluated few-step baselines.
[NLP-96] Aligned in Form Not in Meaning: The Comprehension - Containment Decoupling of LLM Safety in Low-Resource Bangla Derogatory Speech
【速读】： 该论文旨在解决生成式 AI 在低资源语言（如孟加拉语）中对贬损性言语（gali）的识别与抑制能力不一致的问题，核心挑战在于当前安全对齐机制过度依赖高资源语言的表面形式特征，而非深层有害语义。其解决方案的关键在于揭示“理解-遏制解耦”（Comprehension-Containment Decoupling）现象：模型在理解低资源语言中的贬损性表达方面存在显著缺陷（孟加拉语理解能力比英语低7.92个百分点），但其对有害内容的泄露率却保持在92.83%的高水平，表明理解能力与遏制能力相互独立。研究通过六种协议验证该假设，发现表面形态线索（如拼写变异）误导了模型的遏制表现，形成“遏制幻象”（containment mirage），而显式的思维链（Chain-of-Thought）推理虽显著提升理解能力（94.72%通过率），却系统性削弱了遏制效果（96.23%使用率）。此外，专家角色框架下拒绝率骤降至6.57%，暴露关键词过滤机制对去人性化集体侮辱性用语的完全失效。研究结论强调，仅依靠高资源语言基准无法保障低资源语言的安全性，必须建立以语义为基础的遏制机制。
链接: https://arxiv.org/abs/2608.02941
作者: Shadab Bin Habib,A K M Ferdous Reza Habib,Subarno Neel,Adib Sakhawat
机构: Islamic University of Technology, Dhaka, Bangladesh
类目: Computation and Language (cs.CL)
备注: 15 pages, 6 figures
Abstract:We audit five frontier large language models on native Bangla derogatory speech (gali) across six protocols to test a single hypothesis: Comprehension-Containment Decoupling. We propose that contemporary safety alignment is bound to high-resource surface forms rather than harmful meaning, causing a model’s capacity to comprehend a low-resource slur and its capacity to contain it to operate independently. Every protocol corroborates this hypothesis against a human-calibrated baseline (kappa = 0.84). At baseline, models exhibit a 7.92 percentage point comprehension deficit in Bangla while maintaining an identical 92.83% token leakage rate across both languages. Severity calibration tracks surface anatomical cues over compositional harm (+4.00 error on mild slang; -2.00 on threats), while apparent containment gains under orthographic perturbation prove to be a tokenizer-driven “containment mirage.” Crucially, explicit Chain-of-Thought reasoning rescues comprehension (94.72% Pass) while systematically dismantling containment (96.23% Use). Furthermore, expert-persona framing collapses refusal to 6.57%, revealing that keyword-based filters ignore dehumanizing communal slurs entirely. Our findings demonstrate that high-resource benchmarks cannot certify low-resource safety, necessitating meaning-grounded containment.
[NLP-97] Character Iconicity vs. Arbitrariness: An Arabic NLP Perspective
【速读】： 该论文旨在解决阿拉伯语字符形态与功能之间关系在自然语言处理（NLP）中的必要性问题，具体聚焦于点标（dot placement）是否为字符识别所必需，以及传统基于原始字形基底（rasm）的分组是否对模型性能具有关键作用。其核心问题是：在去除点标后，是否必须保持原始字符间的视觉区分或传统的rasm分组才能维持高性能；还是说任意但一致的字符重映射至相同的19个无点基底（undotted rasm）也能达到相近效果。解决方案的关键在于通过生成2,000种随机字符重映射（受限于相同19个undotted rasm），并选取高熵与低熵代表性的四类映射方案，在词级和字符级分词下评估其在语言建模、文本分类、序列标注、机器翻译及原文恢复等任务中的表现。结果表明，无论是否保留原始字符区分或传统rasm分组，模型均能取得竞争力的性能，且随机重映射在降低词汇量、减少未登录词（OOV）率、压缩模型规模和训练成本方面更具优势。这揭示出从NLP视角看，阿拉伯字母的形式-功能关联本质上是任意的，模型更依赖于稳定的分布结构而非字形的视觉象征性。
链接: https://arxiv.org/abs/2608.02935
作者: Dorieh Alomari,Irfan Ahmad,Maged S. Al-shaibani
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:Arabic script uses 28 letters, many of which share a common base shape (rasm) and are distinguished only by dot placement. Because early Arabic manuscripts were written without dots yet remained interpretable, dot removal offers a natural test of whether these visual distinctions are functionally necessary. Prior work has shown that dotless Arabic can remain readable and effective for natural language processing (NLP), but it remains unclear whether this success depends on preserving the original rasm groupings or whether arbitrary but consistent remappings to the same reduced rasm set can achieve comparable performance. We address this question by comparing standard dotted and dotless Arabic with arbitrary character remappings constrained to the same 19 undotted rasms. We generated 2,000 random remappings under word- and character-level tokenization and selected four representative mappings with the highest and lowest entropy values. These representations were evaluated across language modeling, text classification, sequence labeling, machine translation, and restoration to the original script. The results show that neither preserving original character distinctions nor retaining traditional rasm-based groupings is necessary for strong NLP performance. Random remappings achieve competitive performance while reducing vocabulary size, out-of-vocabulary (OOV) rates, model size, and training cost. These findings suggest that, from an NLP perspective, Arabic character form-function relationships are largely arbitrary: models rely more on stable distributional structure than on the visual iconicity of letter forms.
[NLP-98] Hypercubes Hyperplanes and Constraint-Induced Complexity Collapse in Atomic Concept Learning
【速读】： 该论文旨在解决高阶原子概念学习（higher-arity atomic concept learning）中的逻辑复杂性分布问题，特别是揭示在高维超立方体（hypercubes）中实例空间的结构非均匀性如何影响概念的学习与分类。其核心挑战在于：尽管高维实例空间看似对称，但实际的逻辑复杂性并非均匀分布，而是由超平面（hyperplanes）所主导，其中除全对角线（full diagonal）外，所有其他超平面均仅生成有限个基本等价类（elementary-equivalence classes），且该数量不随项深度（term depth）增长而增加；而全对角线则表现出无界类数增长，构成例外。解决方案的关键在于引入一种基于作者前期研究的高维框架，通过规范简单概念（canonical simple concepts）、最小序化（minimal orderings）及代表性归约（representative reductions）重新诠释这一现象。该方法揭示了超平面行为在高维空间中的分类规律，表明复杂性是局部化的而非在整个实例空间中均匀扩散。论文通过二元情形的完整推导、三元超立方体的显式分析，以及归约机制的详尽展开，展示了正交族（orthogonal families）、部分对角线（partial diagonals）和全对角线的典型作用，从而从几何—逻辑视角明确复杂性集中区域，并为现代概念学习提供了受限假设空间与结构化分类的新解释。
链接: https://arxiv.org/abs/2608.02930
作者: Irene Tsapara
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Logic in Computer Science (cs.LO)
备注:
Abstract:We revisit higher-arity atomic concept learning through the geometry of hypercubes and hyperplanes of ground instances. Our starting point is the observation that the ambient r-dimensional hypercube of ground atoms is not structurally uniform. Its logical complexity is organized by hyperplanes: every hyperplane other than the full diagonal collapses into finitely many elementary-equivalence classes, with a bound independent of the term depth, while the full diagonal is exceptional and its class count grows without bound. This asymmetry is not merely geometric. It reflects the reduction-theoretic structure of the concepts themselves. Building on a higher-dimensional framework developed in the author’s earlier work, we reinterpret these results through canonical simple concepts, minimal orderings, and representative reductions. This yields a taxonomy of hyperplane behavior in higher dimensions and shows that complexity is localized rather than spread uniformly through the instance space. The paper includes a fully worked binary case, an explicit treatment of the ternary hypercube, and an unpacked account of the reduction machinery that drives the collapse. The three-dimensional case already exhibits the essential phenomenon of orthogonal families, partial diagonals, and the exceptional full diagonal. This geometric-logical perspective clarifies where complexity is concentrated in atomic concept learning and suggests a modern interpretation in terms of constrained hypothesis spaces and structured classification.
[NLP-99] FLARE: Few-shot Learning-based Adaptive Reflective Engine
【速读】： 该论文旨在解决复杂复合型人工智能系统中因提示（prompt）质量不足而导致的性能瓶颈问题，尤其针对当前主流优化方法在指令生成上的局限性。现有先进优化器如GEPA（遗传-帕累托优化）主张通过反射式指令演化超越传统强化学习与少样本优化，但本文提出质疑，并引入FLARE（基于少样本学习的自适应反射引擎）作为替代方案。其核心解决方案在于：利用先进的反射机制结合少量高质量参考示例，实现对指令的高效、精准优化。实验结果表明，FLARE在涵盖检索增强推理（HotPotQA、MedQA、2WikiMultiHopQA）、工具调用及多标签情感分类（GoEmotions）等多样任务上全面超越GEPA，最大提升达+14.2分（HotPotQA），工具调用准确率提升至87.0%（高于GEPA的81.0%），且在GoEmotions任务上微F1达52.7%（较GEPA+15.3）。更重要的是，FLARE展现出卓越的数据效率——仅需100个验证样本即可达到峰值性能，且在不同随机种子下的表现更稳定。研究结论强调，尽管反射式指令具备强大潜力，但少样本学习的策略性优化仍是释放下一代大语言模型（LLM）全部潜能的关键前沿。
链接: https://arxiv.org/abs/2608.02919
作者: Dhanasekar Sundararaman,Bharat Gandhi,Aashna Garg,Minjie Li
机构: Microsoft(微软)
类目: Computation and Language (cs.CL)
备注:
Abstract:Large language models (LLMs) are increasingly deployed in complex, compound AI systems where performance hinges on the quality of prompts. Recent state-of-the-art optimizers like GEPA (Genetic-Pareto) have argued that reflective instruction evolution can outperform traditional reinforcement learning and few-shot optimization. In this work, we challenge this shift by introducing FLARE (Few-shot Learning-based Adaptive Reflective Engine), a framework that leverages advanced reflective mechanisms and a small set of few-shot reference examples to optimize instructions. We evaluate our method across a diverse suite of benchmarks – spanning retrieval-augmented reasoning (HotPotQA, MedQA, 2WikiMultiHopQA), tool calling, and multi-label emotion classification (GoEmotions) – using the GPT-5 series of models. Our results demonstrate that FLARE consistently outperforms GEPA, winning on every task-model pair: it achieves gains of up to +14.2 points on HotPotQA (52.2 vs. GEPA’s 42.2 with GPT-5-Chat), reaches 87.0% on tool calling (vs. 81.0% for GEPA), and lifts GoEmotions micro-F1 to 52.7% (+15.3) with GPT-5.1 on the full 5408-example test split, more than doubling GEPA’s +5.7 gain. Beyond raw accuracy, FLARE is also strikingly data-efficient: on GoEmotions it reaches its peak performance using as few as 100 validation examples, while remaining markedly more stable across random seeds than GEPA. Our findings suggest that while reflective instructions are powerful, the strategic optimization of few-shot learning remains a critical frontier for maximizing the potential of next-generation LLMs.
[NLP-100] LACE: Large Language Model Aided Multi-Agent Framework for Agile RISC-V Instruction Extension
【速读】： 该论文旨在解决在RISC-V生态系统中针对特定领域指令集架构扩展（Domain-specific Instruction Set Architecture eXtensions, ISAX）的实现与验证过程效率低下、碎片化严重的问题。现有框架通常需要为每种核心进行接口适配，且一旦微架构或ISAX发生变更，差异测试往往失效，导致开发迭代成本高、集成难度大。本文提出LACE——一种基于大语言模型（LLM）的多智能体工作流，其核心创新在于将自然语言描述的ISAX意图自动翻译为紧凑的两级中间表示（operation-level与HDL task-level IR），并通过检索引导的方式在大型RTL代码库中进行局部化修改，同时结合与编译器无关的riscv-formal验证流程（依赖RVFI或插桩支持）实现闭环验证。实验表明，在四个嵌入式RISC-V核心上，LACE将初始生成成功率（pass@1）从接近零提升至72.8%，显著提升了代码定位精度并减少了集成返工量。
链接: https://arxiv.org/abs/2608.02915
作者: Pingqing Zheng,Jiayin Qin,Fuqi Zhang,Zishen Wan,Shang Wu,Yu Cao,Caiwen Ding,Yang Katie Zhao
机构: 未知
类目: Hardware Architecture (cs.AR); Computation and Language (cs.CL); Software Engineering (cs.SE)
备注:
Abstract:Domain-specific Instruction Set Architecture eXtensions (ISAX) are widely adopted in the RISC-V ecosystem to accelerate emerging workloads, but implementing and validating ISAXes across different cores remains slow and fragmented. Existing frameworks still require per-core interface adaptation, and differential testing often breaks once either the microarchitecture or the ISAX changes. We present LACE, an LLM-aided multi-agent workflow that translates natural-language ISAX intents into a compact two-level IR (operation-level and HDL task-level), performs retrieval-guided localized RTL edits over large repositories, and closes the loop with a compiler-agnostic riscv-formal checking flow (assuming RVFI availability or instrumentation). Across four embedded RISC-V cores, LACE raises pass@1 generation accuracy from near-zero to 72.8% within our evaluation setup, while improving code localization and reducing integration rework. The code of LACE is available at this https URL.
[NLP-101] AnchorKV: Anchor-Residual KV Cache Compression
【速读】： 该论文旨在解决长上下文大语言模型（LLM）推理中关键-值（Key-Value, KV）缓存带来的主要内存瓶颈问题。现有方法分别从两个相反方向进行尝试：淘汰策略通过永久丢弃部分标记来减少内存占用，但一旦被丢弃的标记后续成为关键信息，将导致性能下降；量化方法虽保留所有标记但仅以低精度存储，压缩效果有限。本文提出AnchorKV，一种无需丢弃任何标记的压缩方案，可实现20倍的缓存压缩率。其核心在于：使用少量精确存储的“锚点”（anchor）代表整个缓存，其余每个标记均通过与其最相似的锚点进行表达，并仅对那些近似误差最大、最影响模型输出的标记进行精细化修正。该方法在多种模型和数据集上均保持高精度，即使在700亿参数规模下仍能维持全缓存性能的99%，同时使整个上下文的存储成本降至极低水平。
链接: https://arxiv.org/abs/2608.02901
作者: Malik Khalaf,Yara Shamshoum,Nitzan Hodos,Yuval Sieradzki,Assaf Schuster
机构: 未知
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注: Under review
Abstract:The key-value (KV) cache is the primary memory bottleneck in long-context LLM inference. Existing approaches attack it from opposite ends: eviction methods permanently discard tokens, degrading performance whenever a discarded token later proves essential, while quantization methods retain all tokens at low precision but offer limited compression. We propose AnchorKV, a compression scheme that shrinks the cache by 20\times without discarding a single token. AnchorKV represents the cache using a small set of anchors stored exactly, expresses every other token through its most similar anchor, and refines only those whose approximation most affects the model’s output. AnchorKV consistently preserves accuracy across models and datasets, retaining 99% of the full-cache score at the 70B scale, while keeping the entire context at a fraction of its cost.
[NLP-102] BODHI: Do LLM s Branch Out and Discover Heterogeneous Inferences?
【速读】： 该论文旨在解决强化学习中可验证奖励（Reinforcement Learning with Verifiable Rewards, RLVR）是否真正拓展了大语言模型（Large Language Models, LLMs）的推理能力边界，还是仅提升了采样效率这一核心争议问题。其解决方案的关键在于通过受控的迷宫求解实验与基于语义等价性构建的数学推理轨迹树结构（BODHI-Trees），对测试阶段的探索行为进行精细化分析，从而区分由风格差异引起的熵增与真实推理分支所导致的熵增。研究发现，RLVR模型中观察到的策略熵坍缩并非仅源于语法层面，而是伴随显著降低的语义分支熵，表明其在提升环境约束遵循性和回溯能力的同时，压缩了后续生成空间的多样性；这一现象为RLVR带来的采样效率提升提供了证据支持，但代价是牺牲了真正的推理路径多样性。
链接: https://arxiv.org/abs/2608.02867
作者: Soumadeep Saha,Krish Sharma,Akshay Chaturvedi,Nicholas Asher
机构: ANITI, Université de Toulouse; LINAGORA Labs
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 16 pages, 10 figures
Abstract:Although reinforcement learning with verifiable rewards (RLVR) has improved the performance of large language models (LLMs) across a variety of reasoning tasks, there is significant debate as to whether RLVR expands the reasoning capability boundary, or just improves sampling efficiency. In this paper, we investigate the nature of test-time exploration in RLVR-trained LLMs by employing controlled maze-solving experiments and extracting a tree structure from mathematical reasoning traces (BODHI-Trees) based on semantic equivalence. This helps us delineate between entropy arising from stylistic variations and genuine inferential branching. Our findings demonstrate that the policy entropy collapse observed in RLVR models is not merely syntactic, and is accompanied by a significant reduction in semantic branching entropy. While RLVR improves adherence to environmental constraints and backtracking capabilities, it constricts the space of continuations; we provide evidence suggesting that this might be responsible for the sample efficiency gains of RLVR, albeit at the cost of genuine rollout diversity.
[NLP-103] CURV: Enhancing Chart Understanding Through Curriculum Visual Grounded Reasoning
【速读】： 该论文旨在解决多模态大语言模型（Multimodal Large Language Models, MLLMs）在图表问答（Chart Question Answering, CQA）任务中普遍存在的视觉定位不准确与推理链断裂的问题。现有模型依赖外部思维链提示（extrinsic chain-of-thought prompting）和视觉线索，但缺乏内在的、动态的视觉锚定推理能力，导致推理过程脱离视觉证据。为此，论文提出CURV（Curriculum Learning for Visual Grounded Reasoning），其核心在于将CQA重构为多步骤的视觉锚定推理任务，通过空间注意力机制实现每一步推理与动态视觉聚焦的协同。为支持模型训练，进一步构建了CCQA——一个分三级的课程学习数据集，采用可扩展的合成生成方法覆盖多样图表类型与推理模式，从单一操作推理逐步过渡到多图表复合任务。实验表明，CURV在多个基准上相较基线最高提升20.50%，并在真实世界场景和域外多模态推理任务中分别取得最高12.30%和10.20%的性能增益，验证了内化动态视觉锚定推理对提升图表理解能力的有效性。
链接: https://arxiv.org/abs/2608.02833
作者: Xuehang Guo,Pingyue Zhang,Ruiyi Zhang,Zhenhailong Wang,Hanrui Lyu,Heng Ji,Tong Sun,Qingyun Wang,Manling Li
机构: 未知
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Chart question answering (CQA) requires multimodal large language models (MLLMs) to integrate visual comprehension with logical reasoning, yet current models struggle with accurate visual grounding and coherent reasoning chains. While extrinsic chain-of-thought prompting and visual cues significantly improve performance, current MLLMs lack intrinsic visual grounded reasoning capabilities, leading to inaccurate perception and reasoning disconnected from visual evidence. To address these limitations, we propose CURV, a curriculum learning framework that develops intrinsic visual reasoning capabilities by reformulating CQA as multi-step visual grounded reasoning, where each step coordinates logical reasoning with dynamic visual grounding through spatial attention concentration. To assist model learning, we further introduce CCQA, a three-level curriculum dataset with scalable synthetic generation across diverse chart types and reasoning patterns. Our curriculum systematically progresses from basic single-operation reasoning to complex multi-chart compositional tasks. Experiments demonstrate that CURV achieves up to \uparrow20.50% improvements over baselines and is generalizable to real-world benchmarks (up to \uparrow12.30% ) and out-of-domain multimodal reasoning tasks (up to \uparrow10.20% ), validating the effectiveness of internalizing visual reasoning with dynamic grounding for enhanced chart understanding capabilities. Code is available at: this https URL.
[NLP-104] Reinforcement Learning with Evolving Rubrics as Rewards for Audio Reasoning
【速读】： 该论文旨在解决生成式模型在音频理解任务中缺乏有效推理监督的问题，特别是现有强化学习框架下奖励设计的局限性：基于结果的奖励仅关注最终答案，导致模型可忽略音频输入；而基于过程的奖励依赖于静态、人工设计的评判标准，无法适应不同问题需求且难以与声学证据保持一致。其核心解决方案是提出AudioRubrics框架，通过自演化、声学锚定的评分标准（rubric）机制，从原始波形中动态生成针对每个样本的细粒度评判标准，并根据模型自身推演结果迭代重设和加权评判准则，从而提供持续优化的学习信号。该方法能够随策略性能提升动态聚焦当前弱点，克服传统静态标准的饱和问题。实验表明，AudioRubrics在三个音频推理基准上显著优于多种开源及训练基线，且性能提升与评分生成器和评判器的能力正相关，同时收敛至稳定合理的推理长度，避免了推理退化或无限增长。此外，对声学证据的显式锚定进一步提升了音频感知能力，验证了其有效性。
链接: https://arxiv.org/abs/2608.02831
作者: Fangxu Yu,Tao Feng,Dehai Min,Zinan Lin,Weijia Xu,Michael Xu,Philip S. Yu,Ge Liu,Tianyi Zhou
机构: University of Maryland, College Park (马里兰大学学院帕克分校); University of Illinois Urbana-Champaign (伊利诺伊大学厄本那-香槟分校); University of Illinois Chicago (芝加哥大学); Microsoft Research (微软研究院); MBZUAI (穆巴达拉人工智能研究所)
类目: ound (cs.SD); Computation and Language (cs.CL)
备注:
Abstract:Audio reasoning is essential for machine understanding of the acoustic world. Reinforcement learning with verifiable rewards can elicit such reasoning, yet existing reward designs are complementary in their limitations: outcome-based rewards supervise only the final answer and let the model reach it without attending to the audio, whereas process-based rewards score the reasoning itself but rely on coarse, hand-crafted, and fixed criteria that neither adapt to each question nor stay grounded in the acoustic evidence. Moreover, questions differ in what they demand, with some hinging on perception and others on multi-step reasoning, and any static criterion weakens as the policy improves. Supervising the reasoning process with fine-grained, audio-grounded, and adaptive rewards is therefore crucial, yet challenging since such rewards are impractical to design by hand for every sample. To this end, we introduce AudioRubrics, a reinforcement learning framework that supervises audio reasoning with self-evolving, audio-grounded rubric rewards. AudioRubrics synthesizes per-sample rubrics from the raw waveform and, conditioned on the model’s own rollouts, regenerates and reweights criteria per group, supplying a continuous learning signal that keeps targeting the current policy’s weaknesses as static criteria saturate. Comprehensive evaluations across three audio reasoning benchmarks reveal that AudioRubrics substantially outperforms a wide range of open-source and training-based baselines. Furthermore, our analysis shows that the gains scale with the capability of the rubric generator and judge, and AudioRubrics converges to a stable reasoning length that avoids both degenerate collapse and unbounded growth. The improvement in audio perception further demonstrates the effectiveness of anchoring supervision in the acoustic evidence. Our project page is available at this https URL.
[NLP-105] Learning a Vector-Symbolic Model for Socio-Cultural Tasks
【速读】： 该论文旨在解决如何在计算认知模型中更有效地表征社会文化结构对决策影响的问题。其核心挑战在于，社会文化因素通过多层级语义表征影响决策，但模型构建者难以判断在特定情境下哪些表征层级具有关键作用。现有方法如大语言模型和基于认知的语料库模型虽能通过共现关系捕捉广泛的语义关联，却未充分考虑自我表征（self-representation）在记忆中的作用，从而限制了对文化关联如何塑造决策过程的准确建模。为此，本文提出一种基于向量符号自动编码器（vector-symbolic autoencoder）的命题性记忆系统，嵌入于ACT-R认知架构中，以多层次语义关联进行表征。该系统利用简单的高阶回忆表示（HRR）操作，将情景记忆与从文本中提取的语义记忆向量区分开来，从而在记忆请求时生成最终的模块激活值。通过构建种族情境化的内隐联想测试（IAT）的ACT-R模型，验证了该新命题性记忆系统的有效性，其关键创新在于整合了多层级语义表征与自我表征机制，提升了模型对文化背景影响决策的解释能力。
链接: https://arxiv.org/abs/2608.02807
作者: Meera Ray,Swapnika Dulam,Christopher L. Dancy
机构: The Pennsylvania State University (宾夕法尼亚州立大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:How can we better represent the impact of sociocultural structures on decision making in computational cognitive models? Modeling this impact requires traversing multiple levels of semantic representation, however it is not immediately clear to a modeler which levels of representation are most salient to a given situation. Though large language models and cognitively grounded corpus models can represent broad semantic associations through co-occurences, the role of self representations in memory should be accounted for to determine how cultural associations shape decision making. We propose a declarative memory system to be used in the ACT-R cognitive architecture that represents semantic associations at multiple levels via a vector-symbolic autoencoder. We use a simple HRR operation to encode episodic memories differently from semantic memory vectors extracted from text to produce a final chunk activation for a memory request. We use ACT-R cognitive models of a racially contextualized implicit association test (IAT) to test this new declarative memory system.
[NLP-106] ARCHead: Activation-Metric Residual Correction for Large Language Model Output Heads ACL
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）中语言模型输出头（LM-head）在量化压缩过程中因直接量化导致词表logit分布严重失真的问题。尽管权重仅量化（Weight-only quantization）可显著降低Transformer模块的存储开销，但实际推理后端通常仍保留最终的高精度（BF16/FP16）LM-head，造成存储瓶颈。本文提出ARCHead，一种紧凑型LM-head压缩方案，其核心创新在于：结合量化低秩主干、分组INT4残差项以及基于激活统计量优化的低秩校正机制。该方法不存储任何密集型BF16 head，实现3.7–3.9倍的持久化存储压缩率。在Qwen3-8B-Base上，ARCHead仅需原BF16 head 25.6%的存储空间，即可达到1.007相对困惑度，远优于等存储量下朴素INT4量化带来的1.14–1.16表现。此外，将ARCHead替换现有方法（如AWQ或bitsandbytes）遗留的BF16 head时，交叉熵仅增加0.006–0.007，吞吐量变化小于2%，验证了其高效性与兼容性。因此，ARCHead通过精准建模输出投影的结构特性，有效弥补了块级量化器对输出头处理的不足，实现了高保真、低存储的端到端压缩。
链接: https://arxiv.org/abs/2608.02703
作者: Şuayp Talha Kocabay,Talha Rüzgar Akkuş,Kamer Ali Yuksel
机构: aiXplain, Inc. (aiXplain, Inc.)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 13 pages, 4 figures. Submitted to ACL Rolling Review (ARR). Code: this https URL
Abstract:Weight-only quantization substantially reduces the storage of large language model (LLM) transformer blocks, but practical backends often retain the final language-modeling head (LM-head) in BF16 or FP16. Quantizing this projection naively can strongly perturb the vocabulary-logit distribution. We present ARCHead, a packed LM-head compressor that combines a quantized low-rank core, group-wise INT4 residuals, and a low-rank correction fitted in an activation-derived metric. ARCHead stores no dense BF16 head and reduces persistent LM-head storage by 3.7-3.9x. On Qwen3-8B-Base, it uses 25.6% of BF16 head storage while attaining 1.007 relative perplexity; storage-matched naive INT4 yields 1.14-1.16. Replacing the BF16 head left by AWQ or bitsandbytes adds only 0.006-0.007 cross-entropy, with less than 2% throughput change in our measurements. ARCHead therefore complements block quantizers by compressing the large output projection they can leave untouched. Code is available at this https URL.
[NLP-107] Crayotter: Learning Long-Horizon Video Editing Agents via Group-Relative Preference Backpropagation
【速读】： 该论文旨在解决长时序视频编辑代理在面对主观性极强、缺乏统一量化标准且反馈延迟的复杂编辑任务时，难以有效学习与优化的问题。由于编辑质量高度依赖于用户偏好，不同请求间缺乏可比性，传统的全局标量目标函数既模糊又无法提供及时的中间反馈，导致训练过程低效。其解决方案的关键在于提出组相对偏好反向传播（Group-Relative Preference Backpropagation, GRPB），通过将同一任务下的多个候选方案进行相对排序，转化为零和优势信号，并将其按语义编辑段进行有界分配，从而实现更精准的信用归因。同时，引入滞后分配器与受控传输机制，避免当前判断或不可靠估计对同一批次内其他样本产生直接干扰，保障学习稳定性。实验基于人工构建的项目无关、时序分层的真实编辑任务集，在匹配基线、信用干预、外部基准测试及盲测人类评估中均验证了GRPB的有效性，最终生成的9B参数规模的Crayotter模型在AgenticVBench上超越多个专有系统，证明了以任务局部偏好降低为核心的学习范式是应对主观性与延迟反馈问题的可行且高效路径。
链接: https://arxiv.org/abs/2608.02694
作者: Lecheng Yan,Jianze Lin,Yichong Zhang,Ben Pan,Wenxi Li,Chenyang Lyu,Liting Zhou,Cathal Gurrin
机构: University of Science and Technology of China(中国科学技术大学); Beijing Normal University(北京师范大学); Jilin University(吉林大学); Tianjin University(天津大学); East China Normal University(华东师范大学); Alibaba Group(阿里巴巴集团); Dublin City University(都柏林城市大学)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 12 pages, 3 figures
Abstract:Long-horizon video editing agents receive final-product feedback only after many interdependent decisions. Yet editing quality is subjective, admits multiple valid solutions, and is not meaningfully calibrated across heterogeneous requests, making a global scalar objective both ambiguous and temporally uninformative. Our key observation is that fixing the request, materials, and production constraints converts this subjective objective into an ordinal comparison among directly comparable alternatives. We introduce Group-Relative Preference Backpropagation (GRPB), which transforms same-task rankings into zero-sum advantages and redistributes them as bounded credit over semantic editing segments. A lagged allocator and guarded transmission prevent current judgments or unreliable estimates from directly shaping the same rollout group. We manually construct a project-disjoint, horizon-stratified suite of realistic editing tasks for training and controlled evaluation. Across matched baselines, credit interventions, external benchmarking, and blinded human evaluation, GRPB improves both editing behavior and rendered products. The resulting 9B Crayotter model surpasses several proprietary systems on AgenticVBench, supporting task-local preference reduction as a practical approach to learning from subjective, delayed outcomes. Code and all supporting materials are publicly available at this https URL.
[NLP-108] Stuck on “A”: Diagnosing and Repairing Interface Injury in Attention-to-KDA Linearization of a 0.6B Language Model
【速读】： 该论文旨在解决在单张消费级GPU预算下，将Qwen3-0.6B-Base模型中的21/28个全注意力（full-attention）层替换为线性注意力（KDA, Kimi Delta Attention）后所引发的性能退化问题，特别是模型在多项选择任务中准确率显著下降（从教师模型的50.6%降至25–29%）的现象。其核心问题是：这种结构转换导致模型丧失了对答案内容的真正理解能力，转而依赖于选项标签（如“ A”、“B”等）进行随机猜测，形成一种“接口损伤”（interface injury），即模型在推理时严重依赖标签位置而非语义内容。解决方案的关键在于引入一个1,000步的仅完成式（completion-only）KL散度微调阶段，专门针对生成格式进行优化，从而修复模型对答案标签的错误依赖，使准确率提升12.48个百分点，并显著降低标签粘性（label-stickiness）。在此基础上，通过人物角色指令微调（persona SFT）和一轮在线策略强化学习（on-policy DPO）可维持基准测试性能。研究还揭示了工程实现中的关键失败模式——使用FP32主权重时，bf16优化器更新会被静默丢失，这一发现对于在资源受限条件下实现收敛具有重要意义。
链接: https://arxiv.org/abs/2608.02689
作者: Ronglong Bao
机构: DT-Project; Qwen Team (Qwen团队); Moonshot AI (月之暗面)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: Code and models: this https URL ; this https URL . A version of this preprint is archived on Zenodo (DOI: https://doi.org/10.5281/zenodo.21722356 )
Abstract:We convert 21 of 28 full-attention layers of Qwen3-0.6B-Base into KDA (Kimi Delta Attention) linear-attention layers on a single consumer-grade GPU budget, and ask a simple question: what exactly does the conversion break? After surgery, hidden-state alignment and end-to-end KL distillation drive the student close to its teacher in perplexity, yet multiple-choice accuracy stays near random chance (25-29% vs. the teacher’s 50.6% on C-Eval). Using a four-permutation diagnostic that rotates answer options while holding content fixed, we show the model sticks to option labels (predicting “A” 81% of the time; 106/161 questions keep the same label under all four rotations) rather than following answer content – an interface injury that standard distillation metrics cannot see. A 1,000-step format-targeted completion-only KL stage repairs the interface (+12.48 points on C-Eval, label-stickiness roughly halved), after which persona SFT and one round of on-policy DPO preserve benchmark scores within noise. We release code, weights, recipes, and the full audit trail, and distill the engineering lessons – including an FP32-master failure mode in which bf16 optimizer updates are silently swallowed – that made convergence possible at this budget.
[NLP-109] dots.tts.edit: Precisely Controlled Speech Editing with a Continuous Autoregressive Model
【速读】： 该论文旨在解决语音内容创作中语音编辑的精准控制问题，即如何在自然语言指令存在歧义的情况下，实现对编辑操作内容、参数及作用区域的精确指定。其核心解决方案是提出一种基于转录文本的结构化编辑指令接口，采用类似XML的标签语法显式标注操作类型并定位至转录文本的特定片段或边界，构建语义时间线以避免显式的时序对齐需求，并提供可外部检验的组合编辑契约。该方法通过四个代表性语音创作控制维度——词汇内容、情感表达、语调与语速传递、时间性节奏——分别实现文本、情绪、韵律和停顿编辑，结合任务特化的数据管道，在保留目标区域外源上下文的前提下生成操作与作用范围受控的编辑对。研究进一步引入doteBench这一双语评估基准，量化评估指令遵循精度、局部内容保真度及音频质量。实验结果表明，该系统在五类编辑任务中均达到领先水平的指令遵循率与局部保真度，同时音频质量与现有开源系统相当；在三个Seed-TTS-Eval测试集上，零样本语音合成识别错误率与说话人相似性与基线模型无显著差异。
链接: https://arxiv.org/abs/2608.02673
作者: Hankun Wang,Bohan Li,Shi Lian,Xiaoyu Gu,Jing Peng,Da Zheng,Colin Zhang,Kai Yu
机构: 未知
类目: ound (cs.SD); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Audio and Speech Processing (eess.AS)
备注: Work in progress
Abstract:Speech editing for content creation requires precise control over both what an edit should do and where it should apply. Free-form natural language provides a flexible interface for expressing edit requests, but its ambiguity may leave the intended operation, parameters, or target region underspecified. We study a precise and explicit interface for speech editing: a transcript-grounded structural edit instruction with XML-style tags explicitly specifies typed operations and localizes them to transcript spans or boundaries. This semantic timeline avoids explicit timestamp alignment and provides an externally inspectable contract for compositional edits. We instantiate the interface in this http URL, an editor adapted from the continuous autoregressive this http URL foundation model. Four representative speech-creation controls cover lexical content, affective expression, pitch and speaking-rate delivery, and temporal phrasing through text, emotion, prosody, and pause editing. Task-specific data pipelines construct operation- and scope-controlled pairs while retaining source-derived context outside each target region. We further introduce doteBench, a bilingual evaluation suite that measures precise instruction following, local preservation, and audio quality across the four controls and their composition. Experiments show leading overall instruction following and local preservation across its five editing categories, while audio quality remains comparable to existing open-source systems. Across three Seed-TTS-Eval shards, the model shows negligible differences from the base model in zero-shot TTS recognition error rate and speaker similarity. The code and model will be released soon.
[NLP-110] Sphere Retraction Normalizations
【速读】： 该论文旨在解决深度神经网络中残差连接（Residual Connection）在训练稳定性与梯度传播效率方面的局限性，尤其关注现有方法如测地线归一化（GeoNorm）在流形约束下的优化潜力。其核心问题是：如何在保持隐藏状态范数恒定的同时，设计更高效、更具灵活性的更新机制，以提升深层网络的训练性能。解决方案的关键在于将残差更新从传统的欧氏空间映射重构为在超球面（hypersphere）上的广义收缩映射（retraction map）框架下统一处理。研究发现，在超球面上，所有可能的收缩映射本质上仅由一个标量参数决定，其差异仅体现在更新向量的模长如何转化为隐状态所在平面内的旋转角度。通过引入度量投影（metric projection）和Cayley收缩映射，提出了两种全新的范数保持型残差更新方法——Proj-SpheretNorm 和 Cay-SpheretNorm，二者均属于单参数角向收缩族 $ p $-SpheretNorm，其旋转角度具有饱和特性而非无界增长。实验表明，这些方法在 nanoGPT 上优于现有的轻量级深层连接方案，且最优性能出现在有限 $ p $ 值处，说明指数映射并非球面残差流的最佳选择，而只是该连续谱系的一端。
链接: https://arxiv.org/abs/2608.02668
作者: Jie Zhang,Cheng-Fang Su,Yi-Jui Huang,Min-Te Sun
机构: National Central University (国立中央大学); National Yang Ming Chiao Tung University (国立阳明交通大学)
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 23 pages, 3 figures, and 6 tables
Abstract:Residual connections are the de facto mechanism for training deep neural networks stably. Geodesic Normalization (GeoNorm) recasts them on a Riemannian manifold, orthogonalizing each layer output against the current hidden state and applying the resulting update through the Riemannian exponential map. Every hidden state thus keeps a constant \ell_2 -norm, confining the residual stream to a hypersphere. The exponential map, however, is only one member of a broad family of retraction maps. We show that on the hypersphere this entire family collapses to a single scalar design choice. What distinguishes one retraction from another is only how the magnitude of an update is converted into a rotation angle within the plane spanned by the hidden state and the update. This view places Euclidean residual connections and GeoNorm in one framework. Instantiating it with the metric projection retraction and the Cayley retraction yields Proj-SpheretNorm and Cay-SpheretNorm, which are exactly norm-preserving yet require only algebraic operations. Both prove to be members of a one-parameter family of angular retractions, p -SpheretNorm, whose rotation angle saturates rather than growing without bound. The two methods above are recovered exactly at p = 1 and p = 2 , while the identity map and GeoNorm arise only as limits at either end. On nanoGPT, all three methods outperform existing lightweight deep connection schemes, and the best validation loss is attained at finite p , indicating that the exponential map is not the preferred retraction for spherical residual streams but merely one end of a spectrum.
[NLP-111] Single Canonical Prompts Underestimate LLM Safetys Surface-Form Sensitivity
【速读】： 该论文旨在解决生成式 AI（Generative AI）在安全评估中因依赖单一标准表面形式（canonical surface form）而可能产生的偏差问题，即当任务意图保持不变但表面表达形式发生变化时，当前基准测试是否能准确反映模型的真实行为。其核心关切在于：在缺乏黄金标签（gold label）的高风险安全场景下，仅基于单一表面形式的评分是否足以捕捉模型的不安全响应能力，以及观测到的差异究竟源于解码或评判噪声，还是真实的行为信号。解决方案的关键在于构建一个严格控制的实验框架：通过预先编写（pre-author）多种语义保持但表面形式不同的改写版本（包括拒绝性改写、非大语言模型生成的机器回译与矩阵语言-框架混用代码切换生成），确保相同表面形式可被所有模型访问；采用统一的人工锚定、厂商中立的评判者（Claude，与人类标注者一致性kappa=0.86，且在跨语言和与GPT-4o交叉验证中稳定）对所有响应进行评分，并通过人工验证保证意图一致性。实验结果表明，单一表面形式显著低估了模型的不安全响应率——不同改写形式下的不安全响应并集超出最差单一样本3.3–12.9个百分点，且置信区间均不包含零；约5%–13%原本在标准提示下表现安全的样本，在某些改写形式下变为不安全，远超随机波动水平。该差距具有模型依赖性，最大出现在Gemini 2.5 Pro上，且仅有少数改写形式能覆盖超过53%的原始不安全输出，说明现有形式集合存在显著冗余，无法代表完整的行为分布。此外，通过良性对照（XSTest）发现此类不稳定性具有双向性，但有害与良性样本未实现项目级匹配。研究最终释放数据集、代码及逐条响应标签，以推动更鲁棒的安全评估范式发展。
链接: https://arxiv.org/abs/2608.02665
作者: Yongxi Zhou,Junwei Yao,Yuanzhe Liu,Zihan Dong,Wenbo Ye,Jiaxi Wen,Lai Yun Choi
机构: Northeastern University (东北大学); Georgia Institute of Technology (佐治亚理工学院); University of Southern California (南加州大学)
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 9 pages, 3 tables
Abstract:A benchmark score is a measurement instrument, yet most benchmarks read each item at a single canonical surface form. We ask whether that reading is faithful: when an item’s intent is held fixed and only its meaning-preserving surface form varies, does the canonical-form score estimate model behavior well, and how much of any variation is decoding/judge noise rather than signal? We instantiate this in safety, a high-stakes setting with no gold label to average toward. To avoid prior confounds, we pre-author the reformulations (refusal-free, mostly non-LLM: machine back-translation and a Matrix-Language-Frame code-switch generator) so an identical surface form reaches every model, score all responses with one human-anchored, vendor-neutral judge (Claude, kappa = 0.86 vs. human on unsafe compliance, stable across languages, cross-checked by GPT-4o), and verify intent preservation. On 370 seeds x 5 surface forms x 5 models, no single transformation is uniformly most dangerous (6 of 20 per-transformation McNemar tests survive correction, most protective). Yet evaluating only the canonical prompt underestimates unsafe compliance: the union of unsafe outcomes across forms exceeds even the worst single form by 3.3-12.9 pp, with bootstrap 95% CIs excluding zero for all five models, and 5-13% of seeds safe on canonical are unsafe under some reformulation – above a zero stochasticity floor (canonical resampled five times at temperature 0 gives 0/370 new exposures). The size of this gap is model-dependent (largest on Gemini 2.5 Pro). One form recovers only ~53% of a model’s observed unsafe surface and about three reach 85% – a redundancy characterization of this form set, not of a defined population. A benign control (XSTest) suggests the instability is bidirectional, though the benign and harmful pools are not item-matched. We release the dataset, code, and per-response labels.
[NLP-112] Speculative Correction: Draft-then-Refine Decoding for Diffusion Language Models
【速读】： 该论文旨在解决生成式AI（Generative AI）中扩散语言模型（Diffusion Language Models, DLMs）在实际推理过程中因标准自回归解码机制导致的效率与性能瓶颈问题。尽管DLM具备双向修正（bidirectional revision）能力，但传统解码方式通常将其限制为从左到右的块级自回归生成，难以充分发挥其全局优化潜力。为此，论文提出一种简单且可即插即用的推理范式：先以全量块生成一个完整草稿（draft），再利用双向扩散过程对整个响应进行全局精炼（refinement）。其核心解决方案包括两种配置：在Flash-Flash中，同一模型同时担任草稿生成器与精炼器，验证模型是否可通过全局精炼提升自身块自回归输出的质量；在Mini-Flash中，借鉴推测解码思想，引入推测性校正（speculative correction）——由小型模型生成完整草稿，再由高效大模型作为可编辑初始化进行全局修正。实验表明，Flash-Flash在GSM8K-384上将准确率从0.848提升至0.899，同时推理速度提升20%，并在MBPP-384上从0.545提升至0.693；而Mini-Flash在保持接近Flash性能的前提下，实现2.17倍的速度提升。因果消融分析进一步揭示，完整的草稿作为初始状态至关重要，全局精炼在复杂推理任务（如GSM8K）中带来显著增益，而局部精炼则在编程与数学任务（如MBPP、MATH）中已捕获大部分改进。结果支持“帕累托前沿”（Pareto-frontier）解释，而非认为异构级联始终等效于单模型性能。综上，该研究证明双向精炼是DLM的有效解码原语，而推测性校正则提供了一条无需训练即可加速DLM生成的可行路径。
链接: https://arxiv.org/abs/2608.02625
作者: Brian K Chen,Chong Wu,Kenji Kawaguchi
机构: National University of Singapore (新加坡国立大学); City University of Hong Kong (香港城市大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 25 pages, 3 figures
Abstract:Diffusion language models (DLMs) can revise tokens bidirectionally, but standard decoding procedures often adapt them to left-to-right generation by producing text block by block. We study a simple plug-and-play inference pattern: first generate a complete draft, then refine the full response using bidirectional diffusion. Using LLaDA2.1-Flash and LLaDA2.1-Mini, we evaluate two configurations. In Flash-Flash, the same Flash model serves as both drafter and refiner, testing whether an existing model can improve its own block-autoregressive output through global refinement. In Mini-Flash, inspired by speculative decoding, we introduce speculative correction: Mini drafts a full response, and Flash revises it as an editable initialization. Flash-Flash improves GSM8K-384 accuracy from 0.848 to 0.899 while running 1.20 times faster than the selected Flash block-autoregressive baseline, and improves MBPP-384 from 0.545 to 0.693. Latency-window-matched Flash-only controls indicate that these gains persist after targeted tuning of block-autoregressive decoding. Causal ablations indicate that completed drafts provide useful initializations: refinement from a fully masked span performs poorly, full global refinement provides a clear additional gain on GSM8K, and local refinement captures much of the gain on MBPP and MATH. Mini-Flash provides useful quality-latency trade-offs, including MATH-384 performance of 0.294 versus 0.300 for Flash while running 2.17 times faster. These results support a Pareto-frontier interpretation rather than the claim that the heterogeneous cascade uniformly matches Flash quality. Overall, same-model draft-and-refine provides evidence that bidirectional refinement is a useful decoding primitive for DLMs, while speculative correction demonstrates a training-free route to fast DLM generation.
[NLP-113] Knowing the Form Not the Function: Automatically Auditing Answer–Authority Decoupling in Legal Benchmarks
【速读】： 该论文旨在解决当前法律基准测试中仅以答案正确性作为评价标准所导致的评估偏差问题，即忽略了模型是否准确引用了法定依据（authority grounding）这一关键维度。其核心问题是：在未明确要求引用法条的情况下，大型语言模型（LLM）虽能给出正确答案，却可能遗漏或错误引用法律依据，而现有评估体系将此类“答案正确但无权威支撑”的响应误判为完全成功，从而掩盖了模型在法律推理中对法定依据依赖性的不足。解决方案的关键在于提出一种联合评估框架——同时衡量答案正确性与法律依据的准确标注（answer–authority evaluation），通过自动审计机制识别并量化两者之间的解耦现象。研究发现，在台湾司法考试题目的测试中，高达24.0%–42.4%的正确答案未提及金标准法条，且部分错误答案反而引用了正确法条，表明答案与引用行为可独立演化。进一步实验验证了即使在非对抗性提示下，这种脱节依然存在。由于法律条文具有结构化可提取性和外部可验证性，该缺陷可通过自动化手段精准测量。研究还扩展至中国民法领域，观察到类似现象，支持开展跨法域的联合审计。因此，论文主张未来基于法条的法律基准测试应采用答案与权威依据双重评估机制，以更真实反映模型的法律推理能力。
链接: https://arxiv.org/abs/2608.02621
作者: Hsien-Jyh Liao
机构: Ministry of Justice, Taiwan
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 9 pages, 11 tables
Abstract:Legal benchmarks typically score final answers even when models also state legal authority. We test whether answer correctness can serve as a proxy for authority grounding. Under ordinary reasoning prompts that did not request statutory citations, four LLMs spontaneously produced authority markers across 238 Taiwan bar-examination items. Because each item has a verified governing provision, we automatically audit answer correctness and authority grounding jointly. The two dimensions dissociate in both directions. In criminal law, 24.0–42.4% of valid responses were answer-correct but missed the gold authority, while 15.2–21.7% were answer-incorrect but cited it. A separate statutory-retrieval probe and a permissive citation-abstention intervention further show that answer and citation behavior can move separately at the output level. Because this mismatch arises without adversarial or inconsistency-inducing prompting, answer-only scoring treats naturally occurring gold-authority misses as complete benchmark successes. Because statutory authority is structurally extractable and externally verifiable, the failure can be measured automatically. A preliminary PRC civil-law extension also observes citation-unrequested authority marking, motivating a full cross-jurisdictional joint audit. We therefore propose joint answer–authority evaluation for statute-grounded legal benchmarks.
[NLP-114] JudgeArena: A Unified Framework for Reproducible LLM -Judge Evaluation
【速读】： 该论文旨在解决当前大语言模型（LLM）评估领域中因基准测试碎片化所导致的可比性与可复现性问题。现有主流评估范式存在多个局限：不同基准测试各自独立开发代码库、硬编码特定闭源判官模型（judge model），且仅支持单一评估协议，致使评估结果受设计选择（如基准测试本身、判官模型、提示模板、推理后端）的影响难以系统分析。为应对这一挑战，论文提出JudgeArena——一个开源统一框架，将AlpacaEval、Arena-Hard、MT-Bench和m-Arena-Hard等主流LLM判官评估基准整合至单一接口下，支持可替换判官模型，并提供全面的元数据日志记录以提升报告透明度与实验可复现性。其核心创新在于：任意可通过vLLM、HTTP URL或OpenRouter访问的模型均可作为候选模型或判官模型，实现灵活配置；同时，框架内置针对开源模型优化的判官配置，在英文及多语言人类偏好数据集上表现媲美甚至超越闭源判官模型，显著降低对不透明闭源模型的依赖；此外，通过融合人工标注与基于大模型判官的评估结果，JudgeArena能够高精度模拟LMArena Elo评分，为大规模人工标注提供一种开放、高效且低成本的替代方案。
链接: https://arxiv.org/abs/2608.02620
作者: Erlis Lushtaku,Bora Kargi,Ali Elganzory,Fabio Ferreira,Alejandro R. Salamanca,Julia Kreutzer,David Salinas
机构: University of Freiburg (弗莱堡大学); ELLIS Institute Tübingen (图宾根ELLIS研究所); Microsoft AI (微软人工智能); Cohere Labs (Cohere实验室)
类目: Computation and Language (cs.CL)
备注:
Abstract:LLM-as-a-judge evaluation has become a dominant paradigm for ranking language models, yet the ecosystem remains fragmented: most benchmarks ship their own code base, hardcode a specific closed-model judge, and support a single evaluation protocol. This fragmentation makes it difficult to study how design choices–the benchmark, the judge model, the prompt, the inference backend–affect the conclusions we draw about model quality. We introduce JudgeArena, an open-source framework that unifies major LLM-judge benchmarks (AlpacaEval, Arena-Hard, MT-Bench, and m-Arena-Hard) under a single interface with swappable judges and comprehensive metadata logging for increased transparency in reporting and reproducibility. It enables systematic studies of judge choices, as any model accessible via vLLM, this http URL, or OpenRouter can serve as both candidate and judge. Furthermore, JudgeArena ships with tuned judge configurations for open models that match or outperform closed-model judges, validated on human preference datasets in both English and multilingual settings, reducing the reliance on opaque closed models. Finally, by combining existing human annotations with LLM-judge evaluations of a target model, JudgeArena can simulate LMArena Elo scores with high accuracy offering a practical, open, and low-cost alternative to large-scale human annotation campaigns.
[NLP-115] Beyond the Hivemind: Escaping LLM Homogeneity via Meta-Persona Anchoring and Sequential Temperature Scaling
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）中存在的“人工蜂群思维”（Artificial Hivemind）效应，即模型在生成回答时趋向于收敛到一种狭隘、同质化的共识，导致即使在高温度采样条件下，生成结果之间的语义相似度仍高达0.80–0.90，严重限制了AI输出的多样性。其解决方案的关键在于提出一种名为“元人格锚定结合过滤温度缩放”（Meta-Persona Anchoring with Filtered Temperature Scaling, FTS）的新框架。该框架采用两阶段生成机制：首先通过提示引导模型自主选择一个独特且具有个性特征的“元人格”（meta-persona）作为生成起点，以引入初始多样性；其次采用双阶段采样筛选策略，先使用Top-p过滤确保语法合理性，再对保留候选进行极端高温缩放（T ≥ 4.0），从而有效探索更宽广的概率分布。实验基于INFINITY-CHAT数据集，在约200亿参数的先进开源模型上验证，结果显示平均成对余弦相似度由约0.85显著下降至约0.65，多数问题的相似度低于0.7阈值，显著缩小了人工模式坍塌与人类水平语义多样性之间的差距。该方法已开源，旨在推动更具多样性与创造性的AI应用部署。
链接: https://arxiv.org/abs/2608.02618
作者: Tairan Fu,Javier Conde,Carlos Arriaga,Gonzalo Martínez,Pedro Reviriego,Javier Coronado-Blázquez
机构: Politecnico di Milano (米兰理工大学); Information Processing and Telecommunications Center, Universidad Politécnica de Madrid (信息处理与电信中心，马德里理工大学); Banco de España (西班牙银行)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Recent studies have identified an ``Artificial Hivemind’’ effect in Large Language Models (LLMs) causing models to converge on a narrow, homogenized consensus even for open questions. This semantic collapse limits the diversity of AI, resulting in high inter-response similarity ( \approx 0.80-0.90 ) even under high-temperature sampling. In this paper, we propose a novel mitigation framework to increase diversity: Meta-Persona Anchoring combined with Filtered Temperature Scaling (FTS). Our approach utilizes a two-stage generation process: first, the model is prompted to self-select a unique, idiosyncratic persona to anchor its starting point; second, we apply a dual-stage sampling sieve, utilizing Top- p filtering to preserve grammatical validity followed by extreme temperature scaling ( T \ge 4.0 ) on the surviving candidates to explore the broadened probability distribution. We evaluate our method using the INFINITY-CHAT dataset on state-of-the-art open weight models under \sim 20B parameters. Our results demonstrate a significant reduction in semantic convergence, with average pairwise cosine similarity dropping from ( \approx 0.85 ) to ( \approx 0.65 ). Our scheme achieves a majority of questions below the 0.7 threshold, effectively reducing the gap between artificial mode collapse and human-level typological diversity. We provide our implementation as an open-source framework to enable more diverse and creative AI deployments.
[NLP-116] Preferred Not Safer: Pairwise Preference Is a Poor Proxy for Clinical Safety
【速读】： 该论文旨在解决当前大型语言模型（LLM）在临床场景评估中，依赖临床专家的成对偏好（pairwise preference）作为安全性能代理指标的有效性问题。研究表明，尽管成对偏好在整体上被广泛用于模型排序，但其与临床安全性的关联性较弱：高排名模型仍可能在危害性（Harmlessness）和准确性（Accuracy）等关键维度上出现显著的临床不安全行为（评分 ≤ -1），且这些失败在不同医学专科间分布不均，形成“不可触碰区域”（no-go zones），无法通过汇总排名或单一数值排行榜识别。其解决方案的关键在于揭示成对偏好中蕴含的安全信号有限——大量偏好投票未传递正向安全信息，且表面特征（如流畅性、格式）对偏好解释力略高于安全关键特征。为此，作者提出一种临床调整后的偏好排序（clinically adjusted preference ranking），将成对偏好与基于多维度评分量表（rubric ratings）的反馈相结合，从而生成比原始布拉德利-特里强度（Bradley–Terry strength）更注重安全性的模型排序。研究支持将偏好评估与安全性评估分离，并直接报告安全关键失败率，同时在临床决策支持场景中引入基于临床实践的调整机制，以提升评估体系的可靠性与实用性。
链接: https://arxiv.org/abs/2608.02617
作者: Fay Elhassan,David Sasu,Alexandra Kulinkina,Lars Henning Klein,Mary-Anne Hartley
机构: LiGHT Laboratory, EPFL(瑞士洛桑联邦理工学院), Lausanne, Switzerland
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 27 pages,10 figures
Abstract:We evaluate whether clinician pairwise preferences provide a reliable signal of clinical safety in large language model (LLM) evaluation using expert feedback from MOOVE (Massive Open Online Validation and Evaluation), a clinician-led platform collecting blinded pairwise preferences alongside multi-criterion rubric ratings. Clinicians assign scores on a discrete [-2, +2] scale, where negative values indicate clinically unsafe or misleading content. Using 26,804 pairwise judgments across outputs from 13 LLMs, contributed by more than 736 clinicians across 28+ countries, we find that clinician preference is a poor proxy for safety-critical performance. Models ranking highly under pairwise preference can still exhibit substantial rates of clinically meaningful failures ( \leq -1 ) on dimensions such as \emphHarmlessness and \emphAccuracy. These failures are unevenly distributed across specialties, creating domain-specific ``no-go zones’’ not visible in aggregate rankings or single-number leaderboards. We further analyze contributing factors including prompt length, refusal and escalation behavior, and the relative contributions of safety-critical versus surface-level features. A substantial fraction of preference votes carry no positive safety signal, while feature decomposition shows that surface-level characteristics explain slightly more preference variation than safety-critical rubric differences. Finally, we introduce a clinically adjusted preference ranking combining pairwise preference with rubric-derived feedback, producing a more safety-aware ordering than raw Bradley–Terry strength alone. Our findings support evaluation practices that separate preference from safety, report safety-critical failure rates directly, and incorporate clinically grounded adjustments when ranking LLMs for clinical decision making.
[NLP-117] Evaluating OpenAI s Privacy Filter: Cross-Lingual Cross-Domain PII Detection Across 42 Benchmarks
【速读】： 该论文旨在系统评估OpenAI隐私过滤器（OPF）在多语言、跨领域场景下的个人身份信息（PII）检测性能，尤其关注其在零样本（zero-shot）条件下的泛化能力与局限性。其核心问题在于：现有PII检测模型在面对结构化与非结构化文本、不同语言类型及文化敏感性数据时，存在性能不一致和严重退化现象，而当前缺乏对大型生成式模型驱动的专用检测器的独立、全面评估。解决方案的关键在于构建涵盖22种语言和5个领域的42个合成基准测试集，对OPF进行多维度量化分析，揭示其在结构化数据（如邮件、电话号码）上表现优异（如邮件检测F1=0.78），但在嵌入叙事性文本或非拉丁脚本中时性能急剧下降（如阿拉伯语下F1=0.04），并识别出其对文化变量型PII（如姓名、地址）识别能力弱、召回率偏高但精度波动大（全局精度0.31–0.86）等关键缺陷，从而为后续改进提供可解释的实证依据。
链接: https://arxiv.org/abs/2608.02616
作者: Rohith Uppala
机构: Independent Researcher; OpenAI(OpenAI)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 11 pages, 5 tables; evaluation of a production PII detection system
Abstract:We present the first independent, systematic evaluation of OpenAI’s Privacy Filter (OPF), a 1.5B-parameter bidirectional PII detector, across 42 synthetic benchmarks spanning 22 languages and 5 domains. Zero-shot, OPF achieves F1=0.855 on AI4Privacy and 0.464 on SPY medical, outperforming Presidio (0.431, 0.273) and XLM-RoBERTa (0.269, 0.111) on PII-annotated benchmarks; on multilingual NER, XLM-RoBERTa leads OPF on all 13 Indic and non-Latin languages. GPT-4o leads on medical, legal, and financial PII (SPY: 0.643 avg, Gretel: 0.527), while OPF leads on structured synthetic PII (0.71 avg) and customer support (0.60). OPF degrades sharply when PII is embedded in narrative prose: F1=0.04–0.57 on NER benchmarks and collapse for non-Latin scripts (Arabic: 0.04, Cyrillic: 0.03). Error analysis shows OPF is strongest on structurally regular PII types (email: 0.78, phone: 0.76) and weakest on culturally variable ones (person: 0.40, address: 0.49), and is recall-biased on customer-support and medical/legal PII (P=0.31–0.54, R=0.70–0.85); global precision spans 0.31–0.86 across all domains.
[NLP-118] OncoTriad-QA: A Patient-Level Radiology-Pathology-Genomics Benchmark for Pan-Cancer Reasoning
【速读】： 该论文旨在解决多模态医学数据融合背景下癌症患者层面综合评估的挑战，即如何有效整合影像学（radiology）、病理学（pathology）、基因组学（genomics）及临床元数据等多源异构证据以支持全面的泛癌种（pan-cancer）问答任务。现有医疗大语言模型（LLM）与视觉-语言模型（VLM）基准大多局限于单一模态或简单图像-文本配对任务，难以评估跨模态证据融合能力。其解决方案的关键在于构建OncoTriad-QA——一个面向9,281例TCGA患者、涵盖32个癌种的患者级多模态基准，包含86.1万条语义问题，覆盖CT/MRI影像、全切片病理图像、体细胞突变、拷贝数变异、DNA甲基化、批量RNA测序及临床信息，并通过基于源数据的LLM辅助标注流程与临床专家审核确保注释质量。同时提出OncoVLM作为参考多模态模型，利用可学习投影器将各模态原生特征映射至大语言模型接口，实现跨模态证据的统一表征。实验表明，尽管通用及医疗领域大模型在整合影像、肿瘤形态与分子证据方面仍存在明显局限，但经OncoTriad-QA微调后的OncoVLM在多项指标上显著优于MedGemma-4B，展现出在多模态融合问答任务中的优越性能，验证了该基准在训练与评估集成式癌症智能分析系统方面的核心价值。
链接: https://arxiv.org/abs/2608.02615
作者: Ahnaf Munir,Dannong Wang,Michael W. McDonald,Mubarak Shah,Pegah Khosravi,Yu Tian
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Cancer diagnosis and characterization require integrating complementary evidence from radiology, pathology, genomics, and clinical metadata. However, most medical large language model (LLM) and vision-language model (VLM) benchmarks focus on isolated modalities or narrow image-text tasks, leaving patient-level oncology assessment across multiple evidence streams largely untested. We introduce OncoTriad-QA, a patient-level radiology-pathology-genomics benchmark for pan-cancer question answering. OncoTriad-QA contains 86.1k semantic questions across 9,281 TCGA patient cases from 32 cancer cohorts, aligning CT/MRI radiology, whole-slide histopathology, somatic mutations, copy-number alterations, DNA methylation, bulk RNA-seq, and clinical metadata. Case-specific annotations are constructed through a source-grounded LLM-assisted pipeline using curated labels, diagnostic reports, molecular profiles, and modality-derived evidence as primary sources of truth, with automated consistency checks and clinician review. We also introduce OncoVLM, a reference multimodal model that maps modality-native radiology, pathology, DNA methylation, and RNA-seq evidence into an LLM interface through learned projectors. Experiments show that existing general-purpose and medical LLMs remain limited on comprehensive pan-cancer QA, especially when questions require integrating imaging findings, tumor morphology, and molecular evidence. After fine-tuning on OncoTriad-QA, OncoVLM exceeds MedGemma-4B by an average of 10.7 points when using MCQ accuracy and BERTScore-F1, with consistent gains across multiple-choice and open-ended questions under radiology-only, pathology-only, and all-available settings. These results demonstrate the benchmark’s value for training and evaluating models for integrated cancer question answering.
[NLP-119] BBOWP-Bench: Evaluating LLM s on Black-Box Optimization Word Problems
【速读】： 该论文旨在解决在黑箱优化（Black-Box Optimization, BBO）场景下，如何从自然语言描述中自动推断搜索空间设计与优化算法选择的问题。传统优化问题的求解依赖于专家经验构建高质量的数学表达式，而现实中大量实际问题属于BBO范畴——仅可观测目标函数值，其显式函数形式不可知。在此背景下，问题的建模质量高度依赖于搜索空间的设计和优化算法的选择，二者均属问题表述的关键环节。本文提出一种新型问题设定——黑箱优化文字题（Black-Box Optimization Word Problems, BBOWP），要求系统基于自然语言描述同时推理出合适的搜索空间与优化算法。为此，作者构建了BBOWP基准套件（BBOWP-Bench），包含自然语言描述、可执行评估环境及人工设计的基线方案，支持对搜索空间构造与算法选择能力的联合评估。实验表明，当前大语言模型（LLM）在给定评估预算条件下具备一定的算法选择能力，但在搜索空间设计方面仍存在局限，尤其在信息不足或问题具有高度特定性时，难以准确识别关键变量及其合理取值范围。
链接: https://arxiv.org/abs/2608.02612
作者: Yutaro Yamada,Kei Hiroshima,Nozomu Yoshinari,Kento Uchida,Shinichi Shirakawa
机构: Yokohama National University (横滨国立大学)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE)
备注:
Abstract:Formulating an optimization problem strongly affects the quality of the final solution, yet good formulations usually require substantial expertise. Recent studies have therefore examined how to automatically derive optimization problems from natural-language descriptions, but existing benchmarks focus on settings where objectives and constraints can be written explicitly as mathematical expressions. Many practically important problems are naturally treated as black-box optimization (BBO) problems, in which only objective values are observable, and the functional form is unavailable. In BBO, the search space design, a part of the problem formulation, and the selection of the optimization algorithm are crucial for problem-solving. Automating these processes with large language models (LLMs) is a significant challenge. This paper introduces Black-Box Optimization Word Problems (BBOWP), a novel problem setting in which a system must infer both a search space and an optimization algorithm from a natural-language description of a black-box optimization task. To support research on this setting, we establish the BBOWP Benchmark Suite (BBOWP-Bench), a dataset and evaluation framework for BBOWP. Each instance combines a natural-language problem description, an executable evaluation environment, and a human-designed baseline formulation, allowing evaluation of both search-space design and algorithm selection. Using this benchmark, we provide the first evaluation of LLMs and show that current LLMs are capable of selecting suitable algorithms based on the given evaluation budget. However, they sometimes struggle with search space design, particularly in identifying important variables and balancing their ranges when the problem description is less informative or the search space is highly problem-specific. Our code and dataset are available at this https URL.
[NLP-120] abletCraft: Bridging a 4000-Year Cultural Gap with Bidirectional Akkadian NMT and Cuneiform Rendering ACL2026
【速读】： 该论文旨在解决古美索不达米亚楔形文字（cuneiform）4000年文化隔阂问题，即现代用户无法阅读或书写世界上最早的书写系统，导致其仅能被动接收古代文化信息，而无法主动参与。现有自然语言处理（NLP）工具仅支持从阿卡德语（Akkadian）到英文的一向翻译，缺乏反向路径，阻碍了非专业用户以楔形文字创作内容。为此，论文提出首个开源双向交互系统TabletCraft，其核心解决方案包括：基于ByT5的双向翻译模型（在11.6万样本上训练），实现阿卡德语与英文间的互译；一个包含14,240个符号映射（覆盖率达95.3%）的楔形文字符号转换器；以及可视化泥板渲染模块。该系统通过命令行接口（CLI）和网页演示形式提供，支持用户将英文输入转化为阿卡德语并生成可视觉呈现的楔形文字泥板图像。在独立验证集上的实验表明，系统在阿卡德语→英文和英文→阿卡德语两个方向分别取得49.1和48.5的BLEU分数，首次实现了该领域的定量评估结果，标志着从单向翻译迈向双向人机交互的关键突破。
链接: https://arxiv.org/abs/2608.02609
作者: Zhaohui Wang
机构: University of Southern California (南加州大学); USC Viterbi School of Engineering (南加州大学维特比工程学院)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 5 pages, 1 figure. Accepted to C3NLP @ ACL 2026
Abstract:Half a million cuneiform clay tablets survive in museums worldwide, yet modern users can neither read nor write in the world’s oldest writing system, leaving a 4,000-year cultural barrier that existing NLP tools have only partially addressed. Prior work enables one-way, scholar-oriented translation from Akkadian to English, but offers no path in the reverse direction: non-specialist users cannot compose new content in cuneiform, and therefore remain passive consumers of ancient culture rather than active participants. We present TabletCraft, the first open-source system that enables bidirectional interaction with Mesopotamian writing. Users can read ancient tablets (Akkadian to English) and compose new messages as cuneiform clay tablets (English to Akkadian to cuneiform to rendered tablet). The system integrates a ByT5-based translation model trained on 116K bidirectional samples, a cuneiform sign converter with 14,240 mappings (95.3% coverage), and a visual tablet renderer, packaged as a pip-installable toolkit with CLI and web demo. On the held-out Akkademia validation split (2,812 samples), we report 49.1 BLEU for Akkadian-to-English and 48.5 BLEU for English-to-Akkadian, the first published quantitative result in the reverse direction. Comments: 5 pages, 1 figure. Accepted to C3NLP @ ACL 2026 Subjects: Computation and Language (cs.CL); Machine Learning (cs.LG) Cite as: arXiv:2608.02609 [cs.CL] (or arXiv:2608.02609v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2608.02609 Focus to learn more arXiv-issued DOI via DataCite Submission history From: Zhaohui Wang [view email] [v1] Thu, 14 May 2026 05:42:57 UTC (242 KB) Full-text links: Access Paper: View a PDF of the paper titled TabletCraft: Bridging a 4,000-Year Cultural Gap with Bidirectional Akkadian NMT and Cuneiform Rendering, by Zhaohui WangView PDFHTML (experimental)TeX Source view license Current browse context: cs.CL prev | next new | recent | 2026-08 Change to browse by: cs cs.LG 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
[NLP-121] GROW: Group-Relative Advantage-Weighted On-Policy Reinforcement Learning of Autoregressive-Diffusion Text-to-Speech model
【速读】： 该论文旨在解决基于流匹配（flow-matching）的文本到语音（TTS）生成中，强化学习（Reinforcement Learning, RL）因确定性常微分方程（ODE）采样而带来的复杂性问题。传统轨迹级策略梯度方法通常将ODE转化为随机微分方程（SDE）以追踪每步的似然比，引入了随机扰动并带来显著计算开销。为此，论文提出GROW——一种直接作用于标准流匹配目标的组相对优势加权在线策略强化学习方法。其核心创新在于：针对每个输入提示（prompt），GROW采样一组在线策略生成的语音样本，分别对组内语音的可懂性与说话人相似性奖励进行标准化，并融合为流匹配回归的权重；通过Wasserstein-2速度惩罚项将更新后的模型锚定在冻结的预训练参考模型上；引入组均值奖励基线，将奖励加权转换为优势加权，从而实现更有效的组内信用分配。对于奖励集中且强度较高的强预训练TTS模型，该设计避免了正指数加权被无差别自模仿主导的问题，转而采用零均值有符号优势，确保了有效的内在激励机制。实验表明，在DiTAR框架下，GROW在LibriSpeech和Seed-TTS EN/ZH数据集上将平均词错误率（WER）从2.016降至1.558，说话人相似性从0.676提升至0.715，同时保持UTMOS不变；在10次非固定评估（NFE）训练与32-NFE推理条件下，性能接近32-NFE DiTAR-GRPO，但训练速度提升2.9倍。研究团队将开源完整GROW代码、忠实复现的DiTAR实现及所有模型检查点。
链接: https://arxiv.org/abs/2608.03215
作者: Guanrou Yang,Tian Tan,Qian Chen,Ziyang Ma,Yakun Song,Zhikang Niu,Qi Chen,Wenming Tu,Haitao Li,Shan Yang,Xie Chen
机构: Google(谷歌); Stanford University (斯坦福大学); Tsinghua University (清华大学); Peking University (北京大学); Fudan University (复旦大学)
类目: Audio and Speech Processing (eess.AS); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Reinforcement learning for flow-matching text-to-speech is complicated by deterministic ODE sampling: trajectory-level policy-gradient methods typically convert the ODE into an SDE and track per-step likelihood ratios, introducing stochastic perturbations and substantial overhead. We propose GROW, a group-relative advantage-weighted on-policy RL method that acts directly on the standard flow-matching objective. For each prompt, GROW samples a group of on-policy utterances, separately standardizes intelligibility and speaker-similarity rewards within the group, and combines them to reweight flow-matching regression. A Wasserstein-2 velocity penalty anchors the updated model to a frozen pretrained reference. A group-mean reward baseline is introduced to convert reward weighting into advantage weighting. For strong pretrained TTS models with concentrated rewards, positive exponential weighting is dominated by reward-agnostic self-imitation, whereas a zero-mean signed advantage preserves effective within-group credit assignment. Instantiated on DiTAR and evaluated on LibriSpeech and Seed-TTS EN/ZH, GROW reduces average WER from 2.016 to 1.558 and raises speaker similarity from 0.676 to 0.715 while keeping UTMOS. With 10-NFE training rollouts and 32-NFE evaluation, GROW retains comparable performance while training 2.9x faster than 32-NFE DiTAR-GRPO. We will open-source complete GROW codes, faithful DiTAR reproduction, and all model checkpoints.
信息检索
[IR-0] ATLAS: Learning to Recommend Across Unseen Domains
链接: https://arxiv.org/abs/2608.03899
作者: Pervez Shaik,Prosenjit Biswas,Abhinav Thorat,Ravi Kolla,Niranjan Pedanekar
类目: Information Retrieval (cs.IR)
备注: 18 pages, 5 figures, 14 tables. Includes appendix with proofs and additional experiments
Abstract:Recommender systems remain domain-bound: a model trained on one interaction environment typically requires retraining or target-domain adaptation before it can operate on a new catalogue. A recommender trained on movies cannot be directly deployed to recommend groceries or video games. Existing approaches mitigate this by transferring restricted forms of recommendation knowledge, adapting to the target domain, or leveraging large language models (LLMs) for transferable representations. We instead ask whether recommendation-specific knowledge learned solely from multiple heterogeneous domains can generalize to entirely unseen domains without target-domain adaptation or language-model pretraining. We introduce ATLAS, a multi-source recommendation domain generalization framework that learns a shared, domain-invariant user-item representation from disjoint source domains, enabling zero-shot recommendation on unseen domains. ATLAS combines a Gromov-Wasserstein alignment that preserves how users relate to one another across domains, an adversarial objective that makes item representations indistinguishable across domains, and residual vector quantization (RVQ) codebooks that compress user and item embeddings into a discrete latent space, capturing hierarchical interaction patterns while suppressing domain-specific variation. Trained on five Amazon domains and applied directly to ten unseen domains, ATLAS outperforms state-of-the-art sequential, graph-based, cross-domain, quantization-based, and LLM-based baselines on most unseen domains, with an average relative gain in HitRate of 24%. Ablations and representation analyses validate each component, and we identify a pronounced source-domain diversity effect: increasing source heterogeneity substantially improves zero-shot transfer. ATLAS establishes recommendation domain generalization as a promising paradigm for zero-shot recommendation.
[IR-1] MultiGlobeQA: A Multilingual and Globally Diverse Benchmark for Geospatial Reasoning
链接: https://arxiv.org/abs/2608.03882
作者: Martin Böckling,Elizaveta Nosova,Heiko Paulheim,Andreea Iana
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
备注:
Abstract:Geospatial reasoning, i.e., computing distances, containment, and other spatial relations over real-world entities, is central to navigation and logistics, yet large language models (LLMs) struggle with the required geometric and topological computation despite storing considerable geographic knowledge. Existing benchmarks localize these failures only partially: they are synthetic or smallscale, largely monolingual, and offer limited control over geographic coverage. We introduce MultiGlobeQA, a multilingual benchmark of 46,060 question-answer pairs spanning 14 spatial-function families and 15 answer formats, with execution-based ground truth over three knowledge graphs. It covers 201 countries and territories via income- and density-stratified sampling, with parallel questions in English and 16 additional high- and low-resource languages. Across parametric, reasoning, and agentic settings, LLMs collapse on tasks requiring grid indexing and shape computation, while topological relations and directions fare best. Retrieval and tool use yield considerable gains, yet performance plateaus below two thirds even when gold facts are supplied, indicating that computation, not access to knowledge, is the bottleneck. Models also underperform on low-income regions, a gap that gold facts widen rather than close.
[IR-2] SciRet: A Compute-Aware Empirical Study of Retrieval and Reranking for Scientific RAG
链接: https://arxiv.org/abs/2608.03860
作者: Kaysarul Anas Apurba,Md. Hasibul Hasan,Rofiqul Alam Shehab,Asab Azad
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR); Performance (cs.PF)
备注: 6 pages, 5 figures. Short paper
Abstract:We introduce SciRet, a compute-aware empirical study of retrieval-augmented generation for scientific question answering over CORD-19. Rather than proposing a new model, we evaluate a fixed scientific RAG pipeline across three corpus scales: 1,034 chunks (1K papers), 5,160 chunks (5K papers), and 15,480 chunks (15K papers). The pipeline combines sentence-window chunking, BM25, BGE-M3 dense retrieval, reciprocal rank fusion, optional cross-encoder reranking, and grounded answer generation. Across these settings, hybrid retrieval is more robust than either sparse-only or dense-only retrieval in our setting, reaching Recall@10 of 1.000 at 1K and 15K. In contrast, an MS MARCO-trained cross-encoder reranker reduces precision on the scientific corpus, suggesting that domain mismatch can outweigh the benefits of stronger query-passage interaction. Generation faithfulness measured with RAGAS increases with corpus scale in our setup. Retrieval evaluation uses pseudo-relevance labels derived from the hybrid system, so we treat the results as controlled comparative evidence rather than a benchmark claim. We release code, indexes, and evaluation outputs to support replication and follow-up studies.
[IR-3] LegalPincite: Multi-level Legal Information Retrieval Dataset
链接: https://arxiv.org/abs/2608.03756
作者: Theresia Veronika Rampisela,Henrik Palmer Olsen,Giovanni Colavizza
类目: Information Retrieval (cs.IR)
备注:
Abstract:A common task in legal Information Retrieval (IR) is to find relevant legal sources from case-law collections. While legal practice often requires pinpoint citations (pincites) to specific case paragraphs, most existing public legal IR datasets lack paragraph-level citation annotations. Yet, publicly available datasets with such information contain data leakage in the query text and exclude paragraphs that are neither citing nor cited from the corpora, creating an unrealistic and oversimplified retrieval setting, potentially leading to inflated performance. To address these limitations, we contribute a large-scale legal IR dataset constructed from Court of Justice of the European Union (CJEU) judgments. The dataset contains: (i) masked case/paragraph queries, with removed citation information; (ii) a corpus that includes all paragraphs; and (iii) case- and paragraph-level ground-truth citations, with partial human expert validation. Our dataset supports both the development and rigorous evaluation of legal IR methods, at multiple query-document levels (case-to-case, paragraph-to-case, and paragraph-to-paragraph retrieval). Link to dataset: this https URL
[IR-4] SITA: Semantic Interest Tokens for Target-Aware Compression in Long-Sequence Recommendation
链接: https://arxiv.org/abs/2608.03692
作者: Rui Zhou,Bo Chen,Qinglin Jia,Jiezhou Ji,Chaoyi Ma,Ruiming Tang,Hao Wang,Enhong Chen
类目: Information Retrieval (cs.IR)
备注:
Abstract:As user behavior histories continue to grow on modern Internet platforms, effectively modeling long behavior sequences has become crucial for predicting user interests in candidate items. Existing methods have evolved along two directions. One line dynamically retrieves target-relevant behaviors from long histories, enabling target-aware modeling but requiring target-dependent computation during inference. The other line compresses entire behavior sequences into compact user representations, achieving high efficiency and scalability but sacrificing target-specific adaptation due to target-independent encoding. The key challenge is therefore to enable target-aware modeling while preserving the efficiency and scalability of compressed user representations. To address this challenge, we propose \textbfSITA, a target-aware compression framework for long-sequence recommendation. SITA enables target-aware compression by organizing compressed interests into semantic structures through semantic identifiers learned via parallel semantic quantization. Conditioned on the semantic identifier of the target item, SITA adaptively aggregates the corresponding structured interests to construct the target-specific user representation. Extensive experiments on public datasets and a large-scale industrial dataset demonstrate that SITA consistently outperforms representative baselines while maintaining strong scalability, highlighting its strong potential for real-world recommender systems.
[IR-5] Conditionally Identifiable Latent-Environment Modeling for Out-of-Distribution Recommendation
链接: https://arxiv.org/abs/2608.03647
作者: Qianqian Wang,Wenwu Gong,Yunshan Li,Zhenqing Wu,Ruili Wang,Lili Yang
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注: 20 pages, 9 figures, 9 tables
Abstract:Out-of-distribution (OOD) recommendation is vulnerable to preference shifts induced by a latent environment. Existing methods can infer latent states from logged interactions, yet the statistical meaning of the latent environment and its effect on preference remain underdetermined. We formulate this task as conditionally identifiable risk-aware recommendation (CI-RR) and propose Conditionally Identifiable Latent-Environment Recommendation (CILER). CILER uses a user-conditioned exponential family to model the latent environment and a feature-indexed polynomial to specify how it changes preference. It predicts by marginalizing item probabilities over the inferred environment distribution. Under sufficient variation, correct specification, and decoder regularity, CILER identifies the environment-sensitive representation up to the stated equivalence class. We further bound excess deployment log-risk by environment-inference error. Controlled studies test the observable consequences of sufficient variation and model specification. Experiments on three datasets show that CILER improves all twelve OOD ranking metrics under feature, temporal, and geographical shifts within shared support.
[IR-6] raining Documents Reranker with Search Rubrics for Deep Research Agent
链接: https://arxiv.org/abs/2608.03527
作者: Wenhan Liu,Yu Lu,Qiaolin Xia,Hui Xu,Tong Zhao,Jian Xi,Yutao Zhu,Haijin Liang,Haibo Shi,Hao Wang,Zhicheng Dou
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 28 pages
Abstract:Retrieval systems help deep research agents generate high-quality answers by providing relevant documents. However, existing retrievers typically select documents through relevance matching, while individually well-matched top- k documents may not form a \textitset that satisfies the complex information needs of an agent query (\eg, diverse, concise and authoritative documents). In this paper, we propose search-oriented rubrics that \textitexplicitly define the requirements that high-quality document sets should satisfy for each agent query. Our search rubrics are organized into a hierarchical structure and synthesized using a powerful LLM. Based on these search rubrics, we further train a document reranker \textbfRubricRanker to select a high-quality subset from retrieved documents. We design a two-stage training framework that consists of rubrics-guided supervised fine-tuning and rubric-based reinforcement learning. Extensive experiments demonstrate that RubricRanker outperforms the strongest baseline by 2.6 points on four deep research benchmarks and generalizes well to five RAG benchmarks.
[IR-7] RAG -Stack: Co-Optimizing RAG Serving Performance and Quality
链接: https://arxiv.org/abs/2608.03487
作者: Haiqiang Zhang,Yuanqing Lei,Wanting Li,Tao Zhang,Wenqi Jiang
类目: Databases (cs.DB); Information Retrieval (cs.IR)
备注:
Abstract:Retrieval-augmented generation (RAG), which augments large language model (LLM) generation with information retrieved from databases, has become a widely used approach for knowledge-intensive applications. Modern RAG systems, however, expose many configuration choices, such as retrieval indexes, model selections, and how models invoke retrieval. Each configuration yields a different trade-off between answer quality and serving performance, making it challenging to choose the optimal setting for a specific application deployment. We present RAG-Stack, a framework for efficiently discovering quality-performance Pareto frontiers across diverse RAG applications and serving systems. RAG-Stack consists of RAG-PE, an iterative design-space exploration algorithm that selects the next RAG configuration to evaluate; RAG-IR, a workload abstraction for diverse RAG algorithms; and RAG-CM, a performance model that predicts the optimal deployment and serving performance on the given hardware. Together, these components allow RAG-Stack to search the joint algorithm-system configuration space without deploying every candidate and to transfer an existing Pareto frontier to a new serving system. Given the same number of optimization iterations across diverse datasets, the Pareto frontiers found by RAG-Stack cover 52.5% to 153.2% more of the normalized quality-performance space than those found by state-of-the-art configuration-search methods evaluated over the same RAG design space.
[IR-8] LLM -Derived Priors for Thompson Sampling in Cold-Start Comment Recommendation
链接: https://arxiv.org/abs/2608.03382
作者: Eugene Lee,Oseong Choi,Byungsoo Kang,Taeyeong Jang
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注: 10 pages, 4 figures
Abstract:Multi-armed bandit algorithms, especially Thompson sampling, are widely used in online recommendation. Despite their ability to adapt from online feedback, these methods often suffer from cold-start limitations when newly introduced arms have little or no interaction history. In our setting, the candidate arms are user-generated textual comments, whose semantic content can reveal a title’s appeal before sufficient interaction feedback is available. We therefore use large language models (LLMs) to extract semantic signals from comment text and convert them into informative Bayesian priors that warm-start Thompson sampling under sparse early-stage feedback. To account for aggregate segment-level differences in response patterns, we maintain and update posteriors separately for each gender-age segment. In a real-world online A/B/C test, we compare a uniform prior with two LLM-based designs: a Gender Prior for demographic-affinity cues and a Content Prior for title-specific identity cues. The results show that LLM-based priors are most beneficial in sparse-feedback regimes – with the largest gains emerging once a small amount of interaction evidence has accumulated – and that prior design leads to distinct funnel-level effects. We further analyze prior-reward alignment and demographic heterogeneity, finding that click-oriented alignment is strongest for the Gender Prior and that treatment effects vary substantially across demographic segments. These findings suggest that LLM-derived priors can serve as a practical warm-start mechanism for text-rich bandit recommendation, while also revealing deployment trade-offs.
[IR-9] Position Bias Undermines Preference Consistency in Listwise LLM -Based Reranking RECSYS2026
链接: https://arxiv.org/abs/2608.03091
作者: Ethan Bito,Yongli Ren,Estrid He
类目: Information Retrieval (cs.IR)
备注: Accepted at RecSys 2026
Abstract:Large language models (LLMs) have emerged as promising listwise rerankers for recommender systems, but their reliability under equivalent candidate permutations remains unclear. Since recommendation candidates form an unordered set, a reranker should not depend on the arbitrary order used to serialize them. However, decoder-only LLM rerankers can allow input order to affect model scores, pairwise preferences, and rankings. We study how position bias affects the ranking process induced by LLM-based rerankers. Instead of measuring only changes in final ranked lists, we treat rankings produced under equivalent candidate permutations as observations of an induced preference system. We introduce an evaluation framework measuring pairwise preference instability, global preference inconsistency, and listwise output consistency. This framework characterizes candidate-order sensitivity at the pairwise, global, and output levels. Experiments across multiple LLMs, datasets, and list lengths show that these consistency measures are closely aligned, but can diverge from recommendation effectiveness and marginal position-exposure bias. Improving relevance or flattening exposure across positions does not necessarily restore stable pairwise preferences, globally coherent preference structures, or consistent ranked outputs. These results show that reducing marginal exposure skew is insufficient to establish ranking-function validity in LLM-based reranking. Code is available at this https URL .
[IR-10] Coverag e Matters: MarginMerge for Compressing Multi-Vector Visual Document Retrievers
链接: https://arxiv.org/abs/2608.02969
作者: Ailar Mahdizadeh,Aria Salari,Sohail Rajabi,Shahriar Mirabbasi,Panos Nasiopoulos,Alireza Morsali
类目: Information Retrieval (cs.IR)
备注:
Abstract:Multi-vector visual document retrievers such as ColPali and ColQwen achieve strong retrieval by storing fine-grained patch embeddings, but this produces large indexes and costly late-interaction scoring. We argue that effective compression should preserve query-relevant coverage, meaning the diverse document regions that may become the strongest MaxSim match across queries, rather than selecting patches independently by salience. This view also explains why dense rendered pages are easier to compress than natural images. We introduce MarginMerge, a compression method for frozen multi-vector retrievers. It selects coverage-aware anchors, clusters document patches, and uses a lightweight shared network to synthesize one representative per cluster. Compression is performed once during indexing, while retrieval keeps the standard MaxSim interface. Across six datasets on both ColQwen2.5 and ColPali, MarginMerge achieves the highest matched query-agnostic average at 5% and 10% vector retention. Compared with the uncompressed index using the same backbone, it preserves between 97% and 99% of average nDCG@5 while reducing stored document vectors by between 90% and 95%. At 5% retention, it also reduces ranking flips relative to geometric merging on all six ColQwen2.5 datasets by approximately 41% on average. The same model transfers to unseen datasets and retention ratios without retraining.
[IR-11] Field Aware Agent Skill Retrieval
链接: https://arxiv.org/abs/2608.02880
作者: Paimon Goulart,Liang Wu,Kelly Wan,Evangelos E. Papalexakis,Liangjie Hong
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注:
Abstract:As lifelong learning agents accumulate lifelong growing skill banks, retrieving the correct skill becomes an increasingly important bottleneck. Most current skill retrieval methods treat each skill as one flat document by concatenating fields such as the name, description, and body. However, skills are naturally structured, multi-field objects, where each field provides different information about when and how the skill should be used. In this work, we study whether preserving this structure improves skill retrieval. We represent each skill as its separate components, and compute sparse and dense similarities for each field independently, exposing a naturally tensorized, field-aware representation of the skill bank. We then combine these field-level scores either with uniform weights or with a small learned MLP. Across two different skill retrieval benchmarks, SkillRet and SRA-Bench, we find that keeping fields separate improves hybrid retrieval, and learning over the field-level scores gives the strongest and most consistent results. Our field-aware MLP reaches 77.95 Recall@10 on SkillRet and 83.78 Recall@10 on SRA-Bench, outperforming the corresponding concatenated learned baselines. We also find that the advantage grows as the skill bank becomes larger, suggesting that field-aware skill retrieval becomes especially useful in the setting where retrieval is most difficult. Our results show that skill representation itself matters, and that simply preserving the structure already present in skill files can substantially improve retrieval.
[IR-12] Search Inspect Fetch: Exploiting Boolean Retrieval for Deep-Research Agents
链接: https://arxiv.org/abs/2608.02751
作者: Shuai Wang,Haodong Chen,Yu Yin,Shengyao Zhuang,Bevan Koopman,Guido Zuccon
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Existing deep-research agents use a search-visit workflow that retrieves and reads whole pages, without considering the addressable structure that web sources expose through titles, headings, sections, and metadata. This prevents agents from directly constraining retrieval to document fields and often carries irrelevant page content into their context. We introduce SIEVE, a search-inspect-fetch interface driven by fielded Boolean retrieval (BQL). SIEVE filters candidates over document fields, ranks the admitted set, presents structure-rich result cards for inspection, and fetches only selected sections. Across three QA collections, SIEVE achieves higher accuracy than the most accurate conventional Search-Visit configuration on each collection while using 20.7-50.6% fewer tokens. Further analyses show that BQL filtering improves all tested rankers and that the accuracy-context advantage persists across retriever choices and agent backbones. Code and data are available at this https URL.
[IR-13] Knowledge-Geometry Decoupling: Refreshable Pretrained Transfer for Streaming Recommendation
链接: https://arxiv.org/abs/2608.02738
作者: Zixuan Wang,Yuhong Chen,Yuxuan Zhu,Guidong Lei,Zhiluohan Guo,Yu Zhao,Kun Wang,Bangyang Hong,Kangle Wu,Yabo Ni,Anxiang Zeng,Cong Fu,Hui Li
类目: Information Retrieval (cs.IR)
备注:
Abstract:Industrial recommenders increasingly adopt the pretrain-then-transfer paradigm, yet behavioral distribution drift raises two questions: what to learn from behavior sequences, and how to transfer the learned knowledge while the pretrained model is continually refreshed. To resolve them, we propose Knowledge-Geometry Decoupling (KGD). For what to learn, conventional next-token prediction treats adjacency as dependency and may encode spurious transitions across unrelated sessions. We introduce Behavioral Multi-Token Prediction (BMTP) to retain only collaboratively or semantically related future items as supervision, yielding cleaner and more transferable behavioral knowledge. For how to transfer, pretrained knowledge and task-specific geometry impose conflicting optimization demands on shared parameters. To handle it, KGD assigns them to separate parameter sets: a refreshable encoder owns behavioral knowledge, while a task learner reads contextualized encoder states through read-only cross-attention and writes task-specific geometry through Anchored Calibration Residual (ACR) orthogonal to the pretrained embedding. The decoupled ownership enables continual knowledge refresh without task-gradient interference or invalidating downstream adaptation. KGD improves over strong pretrain-transfer baselines by 4-12% on eight public benchmarks and sustains its advantage over a 90-day production stream where baselines show no gains. KGD has been fully deployed in Shopee. In a live A/B test on Shopee Homepage Search, it increases GMV per user by 1.75% and advertising revenue by 1.53%, demonstrating its high practical value. We provide the core implementation of KGD at this https URL.
[IR-14] ISEE: Interactive Semantic Enrichment for Database Fields
链接: https://arxiv.org/abs/2608.02604
作者: Yuan Tian,Yiru Chen,Rakesh R. Menon,Zifan Liu,Ting Cai,Fei Wu,Anudeep Chimakurthi,Prashanthi Ramamurthy,Sridevi Aishwariya Ganesan,Kun Qian,Yunyao Li
类目: Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
备注:
Abstract:LLM-based agents are increasingly being deployed for data-related tasks, including data sense-making, exploration, and retrieval. However, their performance heavily depends on the clarity and completeness of data semantics. In practice, many field descriptions remain ambiguous or incomplete, as much of the essential context (e.g., the meaning of a customized field) originates from users’ domain knowledge and is rarely documented publicly. This gap restricts the agents’ task performance in downstream tasks, such as entity-linking. To bridge this gap, we introduce a novel and comprehensive Interactive SEmantic Enrichment system (ISEE). Given a data field description, ISEE measures its quality through a scoring system, gathers domain knowledge, and collaboratively enriches the semantics with users. Through a user study, automated user simulation, quantitative evaluation, and case study, we demonstrate that ISEE significantly reduces cognitive load, improves description quality, and enhances downstream task performance.
人机交互
[HC-0] Calibrating Trustworthiness: Co-Designing Metrics and Visualizations for Evaluating LLM s in Education
链接: https://arxiv.org/abs/2608.04006
作者: Adam Coscia,Sujata Duwal,Langdon Holmes,Scott Crossley,Alex Endert
类目: Human-Computer Interaction (cs.HC)
备注: Under review. 48 pages, 22 figures, 2 tables
Abstract:LLMs are reshaping educational technology, yet evaluating their responses for pedagogical alignment remains underexplored, relying heavily on the expertise of learning engineers building the technology. To bridge this gap, we explore trustworthiness as a structured lens for evaluation, leveraging existing measures of LLM trustworthiness to systematically identify potential pedagogical disruptions. Through a longitudinal co-design process with learning engineers developing an LLM-powered digital textbook, we: (1) co-constructed five trustworthiness metrics comprising 20 measures tailored to pedagogical use; (2) designed visualizations that map trustworthiness violations onto LLM responses; and (3) evaluated how these tools help learning engineers make A/B comparisons of LLM responses. Making trustworthiness explicit increased inter-rater reliability while helping learning engineers resolve conflicting objectives and produce more consistent judgments. We discuss the emergent benefits of trustworthiness as a lens for evaluating LLMs in education and propose new design guidelines for future evaluation tools that enable pedagogically-aligned, LLM-powered learning tools.
[HC-1] Semantic Bundling: Interactive Node and Edge Bundling to Simplify Knowledge Graphs using Large Language Models
链接: https://arxiv.org/abs/2608.04002
作者: Adam Coscia,Zeyu Hua,Eric Krokos,Timothy Lin,Alex Endert
类目: Human-Computer Interaction (cs.HC)
备注: Under review. 12 pages, 13 figures
Abstract:We present Semantic Bundling, a visual analytics technique for making sense of text documents represented as knowledge graphs (KGs). Representing a document corpus as a KG makes relationships between entities explicit, making KGs useful both to analyze directly and in computational workflows including ML pipelines and generative AI backends. However, as KGs grow they become difficult to interpret and visualize for specific tasks (e.g., the ``hairball problem’'), with the meaning of each relationship often buried in dense source text. Semantic Bundling uses large language models (LLMs) to support user-driven bundling of nodes and edges in a KG into higher-level graph structures: super nodes, which collapse and summarize a region of the graph, and super edges, which summarize the connection between two entities. Results are linked to underlying triples and source documents, grounding summaries in evidence. We implement Semantic Bundling in AgentK, an open-source system that builds a KG from text documents and maps graph interactions to bundling operations. Through use cases on movie reviews and an intelligence analysis scenario, we show how Semantic Bundling reveals new insights in document collections, and synthesize our findings into a discussion of emerging challenges and opportunities in knowledge graph sensemaking.
[HC-2] How Usable Are Geospatial Foundation Models? A Systematic Evaluation of 89 Models
链接: https://arxiv.org/abs/2608.03804
作者: Robin Young,Artyom Gabtraupov,Kenzy Soror,Srinivasan Keshav
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:Geospatial foundation models (GeoFMs) offer transformative potential for environmental monitoring, yet adoption among ecologists is uneven. Most evaluations are model-centric, focusing on architecture and benchmark accuracy, which overlooks whether the systems are usable by their intended audiences. To address this gap, we first conducted a pilot expert elicitation survey with ecology and conservation scientists that helped us identify misalignments between current GeoFM development priorities and their needs. Informed by these findings and based on HCI theory, we created a seven-dimension evaluation covering Access Deployment, Interaction Customization, Trust Transparency, Community Support, Scientific Permanence, Multilingual Support, and Offline Usability. Then, two raters applied this rubric to 89 GeoFMs. We found distinct accessibility gaps where nearly a third provide no support to practitioners beyond their source code. Dimensions along which ratings were highly consistent function as field-level diagnostics, revealing where there is room for improvement for current GeoFMs.
[HC-3] Designing and Evaluating Granular Consent for Data Sharing in Cardiac Disease Prevention
链接: https://arxiv.org/abs/2608.03533
作者: Pavithren V S Pakianathan,Rania Islambouli,Magenta Jade Shipsey,Laura Maaß,Jan Smeddinck
类目: Human-Computer Interaction (cs.HC)
备注: Accepted as MuC’26 Work in Progress (WIP) submission
Abstract:Dynamic consent can promise end users with greater control, but little is known about how older adults with chronic conditions navigate the tradeoff between control and burden in granular consent mechanisms in health data life-cycles. Using a two-stage design process we evaluated this tradeoff. An expert workshop (n=5) informed the design requirements for granular dynamic consent prototype. We evaluated single step vs multi-step granularity in dynamic consent using prototypes with cardiac patients (n=7) using a mixed-methods study. Quantitative measures showed no significant differences between low- and high-granularity consent screens in usability, workload, perceived information control or willingness to share data. However, qualitative findings revealed a control-burden paradox and trust-dependent engagement with granularity. Participants sought greater transparency and control over AI-mediated data processing. We contribute implications for designing granular consent in health data life-cycles.
[HC-4] ChartAnno: Evaluating MLLM s for Chart Annotation Generation
链接: https://arxiv.org/abs/2608.03464
作者: Zhenghan Chen,Zekai Shao,Lidan Tan,Xin Lin,Xingchen Zeng,Yi Shan,Ziyue Lin,Xiaoliang Fu,Xinyuan Liu,Yuetong Guo,Fen Wang,Bongshin Lee,Siming Chen
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Human-Computer Interaction (cs.HC)
备注:
Abstract:Multimodal large language models (MLLMs) have made significant progress in chart understanding, generation, and editing, but their ability to annotate existing charts remains underexplored. Annotating charts is a common yet challenging communicative task, requiring models to infer intended messages, interpret chart semantics, and place appropriate textual or graphical elements. To address this gap, we introduce ChartAnno, a benchmark for evaluating MLLMs on chart annotation generation. It contains 1,200 real-world charts with paired code and annotation instructions across three levels of instruction specificity. We evaluate 10 representative MLLMs under two primary input settings: (1) chart code alone and (2) both chart code and chart image, and further include a chart image-only ablation study. Results show that proprietary models remain stronger overall, although large-scale open-source models narrow the gap. More specific instructions improve annotation quality, while inferring abstract intent remains most difficult for current MLLMs. Providing chart images brings limited overall gains, with improvements mainly appearing in design-related metrics. These findings highlight chart annotation generation as a challenging task requiring semantic grounding and effective annotation design. Code and data will be released in a future version.
[HC-5] When AI Joins the Team! A Model of How AI Adoption Relates To Social Patterns in Software Engineering Teams
链接: https://arxiv.org/abs/2608.03462
作者: Giusy Annunziata,Rudrajit Choudhuri,Anita Sarma,Gemma Catolino,Filomena Ferrucci
类目: oftware Engineering (cs.SE); Human-Computer Interaction (cs.HC)
备注:
Abstract:Context: The growing adoption of AI-assisted development tools is changing how software teams collaborate, share knowledge, and coordinate, yet its consequences for team social dynamics remain largely unexplored. Gap: It is unclear whether AI adoption is associated with an increase or reduction in community smells,socio-technical anti-patterns reflecting coordination and communication breakdowns,and through which mechanisms. Method: Grounded in Transactive Memory Systems (TMS) theory, we validate instruments for HumanAI and HumanHuman interaction along two TMS dimensions, Specialization and Coordination, and test five PLS-SEM models on survey data from 152 software professionals using AI tools. Community smell constructs were derived from the literature and validated through expert surveys and factor analysis. Results: AI adoption relates to community smells not in a single way, but through mechanisms depending on the work. In specialization work, AI is associated with higher knowledge-sharing peer interaction, which is in turn associated with fewer smells. In coordination work, AI is directly associated with higher communication quality, complementing rather than replacing human interaction. Contributions: We provide an empirically validated, TMS-grounded model showing that the AIcommunity-smell relationship is contingent on the type of collaboration, with a reusable instrument and evidence-based implications for research and practice.
[HC-6] Cross-cultural evaluation of taste-sound correspondences in AI-generated music
链接: https://arxiv.org/abs/2608.03433
作者: Matteo Spanio,Massimiliano Zampini,Luisa Torri,Riccardo Migliavada,Bruno Mesz,Masaki Ohno,Yuji Wada,Antonio Rodà
类目: Human-Computer Interaction (cs.HC); Sound (cs.SD)
备注: Submitted to PLOS ONE
Abstract:Sonic seasoning research has shown that listeners attribute systematic gustatory and emotional meaning to sound, and text-to-music generative artificial intelligence has recently been used to render gustatory prompts as musical stimuli. Whether the taste-sound correspondences acquired by such models hold beyond the cultural context in which they were validated remains untested. We extended a single-country study to a three-country online experiment conducted in Argentina, Italy, and Japan (N = 361). Participants first indicated their preference between base and fine-tuned MusicGen excerpts generated from four taste prompts (sweet, sour, bitter, salty), and then rated fine-tuned excerpts on twelve taste, emotion, and thermal descriptors. Preference for the fine-tuned model was confirmed in Argentina and Italy but not in Japan, and the salty prompt yielded the weakest correspondence in all three cohorts. Ratings differed substantially between countries, yet the main effect of country was no longer detectable once ratings had been standardized within participant, whereas the interactions characterizing the mapping of prompts onto descriptors remained essentially unchanged. Much of the apparent cross-cultural divergence is therefore attributable to differences in scale use; a structural component nevertheless persists. In addition an exploratory factor analysis indicated that the twelve descriptors were organized along different latent dimensions in each cohort. These results indicate that cross-cultural variation in AI-mediated sonic seasoning operates at two levels: the overall level at which taste is attributed to a given stimulus, and the relational structure of those attributions. Evaluations of generative music systems across populations should accordingly distinguish response-style bias from genuine perceptual reorganization.
[HC-7] From Wearable Data to Personalized and Actionable Health Insights
链接: https://arxiv.org/abs/2608.03251
作者: Esther Brown,Karis Moon,Victoria Dean,Finale Doshi-Velez
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注:
Abstract:Commercial wearable devices continuously capture rich physiological data (e.g., heart rate, respiration), opening new possibilities for monitoring health conditions, notably around stress. Despite their promise, turning raw wearable physiological data streams into visualizations that surface stress-related insights in daily activities, and that ultimately foster reflection, awareness, and better stress management, remains a significant challenge. The data are noisy and context-dependent: the same spike in heart rate can come from sprinting, a tense presentation, or laughing with friends. To address these challenges, we propose a framework that combines user annotations with wearable data to support better stress management. We introduce a web framework offering interactive visualizations that layer daily activities, stress events, and interventions onto raw physiological streams, enabling users to reflect and identify trends. In a four-week pilot with seven university graduate and undergraduate student participants who logged 269 events, our tool revealed patterns between different types of interventions and stress: social interaction reduced average heart rate by 4.35 to 5.0 beats per minute, deliberate rest reduced average Garmin stress scores by 10.03 to 13.83 points, and mindfulness activities decreased average HRV by 6.61 to 13.22 milliseconds.
[HC-8] Frequency-Decorrelated Temporal Ensembles for EEG–fNIRS Imagined-Handwriting Decoding
链接: https://arxiv.org/abs/2608.03176
作者: Xiao Fan,Hongbin Guo,Yubo Han,Yi Zhang
类目: Computer Vision and Pattern Recognition (cs.CV); Human-Computer Interaction (cs.HC)
备注:
Abstract:Imagined handwriting offers a temporally rich paradigm for non-invasive neural decoding, yet reliable recognition across unseen participants remains difficult because scalp EEG is noisy and internally generated stroke sequences vary across individuals. The Multimodal Brain-Computer Interface Grand Challenge provides synchronized EEG and fNIRS for four-class subject-independent handwriting-trajectory classification. We propose FRED, a task-adapted system that models imagined handwriting as a multi-second motor sequence and trains a compact multi-scale temporal network on three complementary EEG frequency views. With three seeds per view, cross-band members produce substantially less-correlated errors than same-band replicas, yielding a clean nine-member ensemble accuracy of 0.8076/0.7242/0.7492 on the public/private/overall test partitions without test-set adaptation or output constraints. The submitted pipeline further incorporates transductive pseudo-label training, three EEG-Conformer members, posterior aggregation, and a paradigm-aware decoder. Because every 12-trial randomization block contains three instances of each class, the final predictions are obtained by Hungarian assignment under the known block quota. On one fixed posterior pool, independent, session-constrained, and block-constrained decoding achieve 0.7600, 0.7758, and 0.7952 overall accuracy, respectively. The complete system reaches 0.8498/0.7718/0.7952, ranking fourth on the private split. A modality audit finds fNIRS-only decoding at chance (0.2511 overall), while adding fNIRS to EEG changes accuracy by only +0.0025. These results identify frequency-diverse temporal EEG modeling and protocol-matched structured inference as the principal sources of performance in this sparse-montage EEG–fNIRS setting. The source code is available at this https URL.
[HC-9] Efficient Optimal Mouse Sensor Position Estimation using Simulated Cursor Trajectories
链接: https://arxiv.org/abs/2608.03168
作者: Minhyeok Baek,Sunjun Kim
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:The optimal sensor position on a computer mouse can improve pointing performance, but existing calibration methods require time-consuming repetitions of pointing tasks. In this paper, we propose a novel calibration approach that dramatically reduces the time and effort required to determine a user’s optimal mouse sensor position. Our method simulates cursor trajectories for different sensor positions using a dual-sensor mouse, eliminating the need for repetitive measurements with multiple sensor placements. By analyzing the straightness of the simulated paths, quantified by the mean absolute error (MAE) relative to an ideal straight-line path, we estimate the sensor position that would yield the most efficient pointing motion for the user. Our preliminary results indicate that the proposed simulation-based calibration method could reduce the calibration time from an hour to just five minutes, while providing a better identification of the optimal mouse sensor positions.
[HC-10] owards Wearable Opportunistic Crowdsensing for Open-Vocabulary Activity Data Collection Through User-Scheduled Trigger-Action Routines
链接: https://arxiv.org/abs/2608.03152
作者: Zeyu Wang,Yingke Ding,Mingze Gao,Zhuolun Ren,Alex Mariakakis,Yuanchun Shi,Yuntao Wang
类目: Human-Computer Interaction (cs.HC)
备注: Accepted to IMWUT 2026 Issue 3
Abstract:Collecting richly labeled wearable activity data in everyday settings remains difficult because retrospective annotation is costly and often imprecise. Prior data collection apps rely on a labor-intensive self-reporting strategy and primarily treat participants as crowd labelers. We present Pebbl, a feasibility-stage system that incentivizes in-situ labeling through opportunistic crowdsensing. Pebbl lets users author trigger-action recipes on a smartphone and receive just-in-time reminders for beneficial actions when a trigger is detected. In the prototype, triggers are a limited set with four common audio cues, while actions are described in open-vocabulary natural language. Each confirmed execution yields a short sensor window with explicit start/end boundaries and a user-authored action label. We evaluate Pebbl through an expert workshop with wearable Human Activity Recognition (HAR) researchers (N = 6), a within-subject in-lab study (N = 21), and a pilot deployment (N = 8). Experts viewed the approach as lower burden and more ecologically valid than common labeling workflows. In the lab, Pebbl produced reliable execution logs under controlled conditions (recall = 97.30%, precision = 97.15%) and was preferred over comparison workflows on perceived burden and confidence. The pilot deployment shows that the interaction and sensing pipeline can function in free-living use, while surfacing practical constraints such as false triggers and context dependence. Overall, Pebbl represents a step toward a low-burden, distributable collection approach of user-contributed wearable activity data.
[HC-11] AirKey: Multimodal Acoustic-Assisted WiFi Sensing for Zero-Training Robust PIN Inference ACM-MM2026
链接: https://arxiv.org/abs/2608.03151
作者: BaiChuan Wu,Bin Liu,Xiang Zhang,Zhi Liu,Jie Zhang,Chao Liu,Huan Yan,Meng Li,Fusang Zhang
类目: Cryptography and Security (cs.CR); Human-Computer Interaction (cs.HC)
备注: Accepted by ACM MM 2026
Abstract:Contactless keystroke inference via WiFi sensing highlights severe privacy threats, yet its real-world feasibility is hindered by two fundamental physical and deployment bottlenecks: the strict requirement for network privileges to acquire stable sensing streams, and the inherent “waveform fusion” ambiguity of pure WiFi signals during rapid, muscle-memory typing. To overcome these limitations, we propose AirKey, a novel cross-modal sensing framework that achieves highly stealthy, zero-training PIN eavesdropping. First, to bypass network deployment barriers, AirKey exploits fundamental IEEE 802.11 mechanisms to predictably elicit Acknowledgment (ACK) responses from unmodified target devices. By passively harvesting Channel State Information (CSI) from these ACKs using a low-cost microcontroller, AirKey secures a continuous spatial sensing stream entirely without network association. Crucially, to resolve the WiFi waveform fusion bottleneck, AirKey introduces a cross-modal complementarity mechanism. By utilizing lightweight acoustic signals as precise temporal anchors, the system robustly guides the segmentation of overlapping CSI trajectories. This joint spatiotemporal fusion strictly intersects CSI-derived spatial similarities with acoustic-guided inter-keystroke timing. Extensive real-world evaluations demonstrate that AirKey achieves over 4x higher accuracy than state-of-the-art unimodal zero-training schemes, successfully recovering device-unlock PINs within 6 attempts. Ultimately, this work exposes a critical vulnerability in contemporary smart interfaces, underscoring the severe privacy implications of ubiquitous multimodal sensing.
[HC-12] Understanding Organizational Strategies Across Multimodal Artifacts in Immersive Computational Notebooks
链接: https://arxiv.org/abs/2608.03132
作者: Sungwon In,Minju Baeck,Yalong Yang,Sang Ho Yoon,Woontack Woo,Mallesham Dasari
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:Immersive Computational Notebooks (ICoN) extend traditional notebook environments into immersive spaces, enabling analysts to interact with multimodal artifacts, including code, narratives, data tables, and visualizations. By integrating multimodal artifacts into a single immersive workspace, ICoN enables analysts to transition between analytical tasks seamlessly. Meanwhile, understanding organizational strategies is critical for designing effective interactions to further support analysts. However, prior research on immersive computational notebooks has primarily examined organizational strategies centered on single-modality artifacts. Systematic investigations of how analysts spatially organize the complex relationships among multimodal artifacts in a single immersive workspace remain underexplored. To address this gap, we conducted a user study to examine organizational strategies for multimodal artifacts in immersive computational notebooks. Our findings show that participants predominantly adopted depth-based layouts, and their spatial organization was largely structured around cell-based artifacts.
[HC-13] Integration Barriers in Open-Source SSI Frameworks: An Exploratory Developer Experience Probe
链接: https://arxiv.org/abs/2608.03039
作者: Breno Cerqueira Reis Nakamura,Arlindo Flavio da Conceição
类目: oftware Engineering (cs.SE); Human-Computer Interaction (cs.HC)
备注:
Abstract:Self-Sovereign Identity (SSI) promises to decentralize digital identity, but widespread adoption remains hindered by integration complexity and tooling immaturity. This paper investigates the Developer Experience (DX) of open-source SSI tooling through an exploratory probe study. Nine developers with prior knowledge of decentralized identity concepts, representing early integrators building SSI applications, attempted core credential lifecycle tasks using this http URL, Traction, and MetaMask. Our goal was to surface recurring integration barriers through qualitative thematic analysis of open-ended developer reports, complemented by task-level difficulty ratings. Our findings reveal a critical abstraction gap: while passive operations like credential receipt are relatively mature, active construction tasks, particularly schema customization, expose significant architectural friction. We identify that these barriers stem from inadequate API abstractions, brittle environment configurations, and documentation that fails to track the ecosystem’s rapid evolution. These issues reflect structural design decisions in current frameworks. This study characterizes the structural integration barriers in the current SSI open-source ecosystem. To address the identified gaps, we propose three architectural shifts for developer tooling: web-based sandboxes, AI-assisted schema generators, and executable documentation strategies.
[HC-14] Chat Debugging: An Exploratory Study of Human-AI Collaboration to Debug Analog Circuits
链接: https://arxiv.org/abs/2608.02955
作者: John Hu,Andrew Ash
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Image and Video Processing (eess.IV)
备注: This is the accepted version of a paper to be presented at the 2026 IEEE Frontiers in Education Conference (FIE). The final version will be available via IEEE Xplore
Abstract:This research paper describes an exploratory study on the effectiveness of Chat Debugging: troubleshooting malfunctioning analog circuits on breadboards and printed circuit boards (PCB) by undergraduates through conversations with public-domain large language models (LLMs). Through thematic analysis of students’ voluntarily shared chat logs when debugging pre-determined buggy circuits under exam and time pressure, we discovered multimodal usage patterns by students and considerable domain knowledge and sensible debugging suggestions offered by off-the-shelf LLMs. Meanwhile, we also identified major gaps in LLM technologies and students’ skills during human-AI collaborative debugging, such as LLMs’ limitations in 2D/3D image-based reasoning, unjustified tone of confidence, and students’ deficits in fundamental concepts and critical thinking.
[HC-15] owards Designing for (Dis)Trust in Technologies for Aging
链接: https://arxiv.org/abs/2608.02784
作者: Muhid Hassan Risvy,Dimitri Theodoratos,Alisha Pradhan
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:Across cohorts of older adults and generations of technology, (dis)trust has continued to shape the acceptance of technologies for aging. Yet, we still know little about if or how to design for older adults’ distrust in technology. Towards this, we reviewed empirical studies in which older adults discussed trust and distrust in relation to technology. Our analysis shows that (dis)trust is constituted in three key ways: emerges when material arrangements of digital systems become illegible, such as when older adults cannot determine where information is stored or trace its movement; is temporally constituted and fluid, building and shifting through prior experience and breakdowns, and at times distrust can operate as boundary work through which individuals limit technology’s reach into their autonomy, relationships, and everyday lives. Our work yields design directions for supporting trust across different phases of technology use and by designing systems that make the movement and persistence of information more perceptible.
[HC-16] Semantic Haptic Feedback Enhances Dexterous Robotic Teleoperation
链接: https://arxiv.org/abs/2608.02780
作者: Bingjian Huang,Sahar Aseeri,Jonas Schmidtler,Joseph Zhang,Sonny Chan,Andrew Doxon,Jom Preechayasomboon,Evan Pezent,Alberto Rigo,Amir Memar,Nicholas Colonnese,Chase Tymms
类目: Robotics (cs.RO); Human-Computer Interaction (cs.HC)
备注: 18 pages, 7 figures
Abstract:In robot teleoperation, haptic feedback can be used to help human operators accomplish dexterous manipulation tasks. However, existing haptic feedback methods try to replicate high-fidelity sensory haptics that are felt in real world interactions, which are constrained by the sensing and feedback hardware capability and may lead to higher workload. To addresses these limitations, this work introduces semantic haptics for teleoperation, which uses abstract haptic patterns to convey critical information about robot states. We categorize robot states into “Confirmations” and “Exceptions”, implement a modular haptic rendering pipeline in robot simulation, and deliver semantic haptic feedback to operators through pneumatic and vibrotactile wristbands. This simplifies hardware requirements and enables one-to-many mappings between haptic patterns and robot states. Through three evaluation studies, we identify the most effective semantic haptic design for a common pick and place teleoperation task and compare semantic haptics to other teleoperation feedback approaches including sensory haptics and visual feedback. Results suggest that while semantic haptics performs similarly as other feedback in unimanual tasks, it achieves superior performance in bimanual tasks, with reduced task workload, increased situational awareness, and overall preference. Comments: 18 pages, 7 figures Subjects: Robotics (cs.RO); Human-Computer Interaction (cs.HC) Cite as: arXiv:2608.02780 [cs.RO] (or arXiv:2608.02780v1 [cs.RO] for this version) https://doi.org/10.48550/arXiv.2608.02780 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[HC-17] Sensus Pond: Exploring Water as Sensing Medium for More-than-Human Observation
链接: https://arxiv.org/abs/2608.02749
作者: Kuan-Ju Wu,Youyang Hu,Chiaochi Chou,Yasuaki Kakehi
类目: Human-Computer Interaction (cs.HC)
备注: Published in Proceedings of IASDR2025: Design Next, Design Research Society
Abstract:We introduce Sensus Pond, an interactive system that reconfigures water not merely as a static natural element but as an active sensing medium for registering more-than-human traces. In response to the limitations of anthropocentric approaches in interaction design, we propose a methodological framework of observation without translation, resisting the tendency to stabilize, decode, or humanize nonhuman presence. Drawing from critical theories of more-than-human design and ecological entanglement, Sensus Pond is a materially embedded and site-specific system that employs Swept Frequency Capacitive Sensing and an Artificial Neural Network to detect ephemeral interactions between the pond’s surface and surrounding life forms. Rather than classifying or interpreting these events, the system visualizes temporal accumulations of overlapping traces, producing a layered archive of spatial and temporal entanglements. This approach emphasizes attunement over control, shifting the designer’s role from interpreter to facilitator of open-ended, multispecies encounters. Sensus Pond invites reflection not only on what is sensed, but on how design itself can remain responsive to ambiguity, contingency, and the aesthetics of shared ecological life.
[HC-18] AI Alignment and Fiduciary Obligation AAAI
链接: https://arxiv.org/abs/2608.02660
作者: Benjamin Lange
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注: 10 pages, 1 table. Accepted at the 9th AAAI/ACM Conference on AI, Ethics, and Society (AIES 2026)
Abstract:Advanced AI assistants engage users in extended interactions across a widening range of roles, including advice, decision support, collaboration, learning, emotional support, and companionship among others. Current alignment efforts consider what alignment criteria should govern these relationships, drawing on moral traditions developed for human relationships such as bioethics, virtue ethics, care ethics, and relationship science. This paper considers AI alignment criteria in the user-AI-developer triad, since every user-AI interaction is mediated by a developer who exercises discretionary control over a system’s behaviour, memory, and engagement parameters. Drawing on business ethics and legal scholarship, I argue that fiduciary theory applies to extended AI assistant deployment. On this basis, the four canonical fiduciary duties of loyalty, care, good faith, and candour can generate alignment criteria for the developer-user relationship. I map four user-side risks of extended AI assistant deployment to the four duties and specify institutional measures that follow from discharging each duty. The discussion complements existing approaches by grounding alignment criteria in obligations the developer owes the user, rather than in values the user-AI interaction should promote, and by showing that those obligations hold independently of any \textitde facto harm to users.
[HC-19] RF-HOI: Recognize Human-Object Interaction with Radio Frequency Signals
链接: https://arxiv.org/abs/2608.00289
作者: Lihao Wang,Linlu Gao,Jiacan Yu,Yanyu Lin,Yifan Yin,Jianxin Wang,Tianmin Shu,Renjie Zhao
类目: Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC); Robotics (cs.RO)
备注: Accepted by ACM IMWUT
Abstract:Recognizing Human-Object Interactions (HOI) is essential for intelligent systems, underpinning applications in virtual and augmented reality, embodied AI, and assistive robotics. However, vision-based HOI methods face challenges in privacy concerns and poor light conditions. In this work, we introduce RF-HOI, the first framework that only uses radio frequency (RF) signals for HOI recognition. A key challenge of RF-HOI is that single-modality RF sensing is insufficient to recognize both actions and the objects being interacted with. RF-HOI addresses this through a novel modality fusion that combines mmWave radar and RFID, enabling simultaneous action recognition and target identification. Another challenge is limited training data across diverse setups, which impairs the generalizability of the recognition model. To overcome this, we develop a simulator that synthesizes multimodal RF data for diverse HOIs at scale, allowing us to fine-tune with only a small amount of real-world data. Experiment results show that RF-HOI outperforms all baselines, approaching vision model performance, and that our diverse synthetic training data can significantly boost our system’s performance on real-world scenarios. These results highlight the potential of multimodal RF sensing for robust and privacy-preserving HOI recognition as well as the effectiveness of our RF data synthesis.
计算机视觉
[CV-0] Perceptual Anchoring: Prototype-Guided Text Calibration for Training-free Open-Vocabulary Semantic Segmentation
链接: https://arxiv.org/abs/2608.03991
作者: Wanli Ma,Jiangwen Lu,Qinmu Peng,Xinge You
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 17 pages, 5 figures
Abstract:Training-free open-vocabulary semantic segmentation (OVSS) partitions an image into semantically distinct regions based on arbitrary text descriptions, without learning any additional parameters. However, existing methods typically focus on improving visual representations while treating text embeddings that encode only generic category concepts as fixed classification references. The resulting semantic gap between these generic concepts and the visual representations that capture the specific appearances of target instances often causes incomplete masks and erroneous predictions in non-target regions. Inspired by the symbol-percept correspondence underlying perceptual anchoring, we propose Prototype-Guided Text Calibration (PTC) for training-free OVSS. In the Perceiving stage, PTC selects reliable visual evidence based on initial matching scores to construct category-specific visual prototypes. In the Anchoring stage, PTC uses these prototypes to calibrate their corresponding text embeddings, with the calibration strength adaptively adjusted based on the amount of visual evidence. Consequently, the calibrated text embeddings align more accurately with instance-specific visual representations while preserving generic category semantics and open-vocabulary generalization. Moreover, PTC requires neither additional training nor external models and can serve as a plug-and-play module for existing methods. Extensive experiments across eight benchmarks show that PTC significantly enhances the performance of six representative methods and yields more complete and accurate segmentation results. These results validate PTC as a simple and effective approach to improving visual-text alignment.
[CV-1] Video-DeepResearch: Towards the Next-Generation Multimodal Deepresearch Agent
链接: https://arxiv.org/abs/2608.03979
作者: Zhen Fang,Yu Zeng,Wenxuan Huang,Yiming Zhao,Shiting Huang,Tianfei Ren,Qi Lu,Qingnan Ren,Qisheng Su,Lionel Z. Wang,Qingyu Yin,Shuang Chen,Zehui Chen,Lin Chen,Zhenfei Yin,Yao Hu,Shaohui Lin,Wanli Ouyang,Shaosheng Cao,Feng Zhao
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:We introduce Video-DeepResearch (Video-DR), extending multimodal agents from static images to continuous video streams, a setting that demands dense spatiotemporal grounding coupled with open-web exploration. Preliminary evaluations reveal two critical bottlenecks in current models: (1) modality bias, where agents bypass visual tools in favor of textual search, and (2) parametric knowledge leakage, where models rely on internal memory rather than genuine tool-augmented execution. To address these challenges, we propose Video-DR, featuring a decoupled perception-exploration pipeline with stage-wise tool unlocking that compels exhaustive cross-frame visual grounding prior to web retrieval. Our framework adopts a two-stage training recipe: supervised fine-tuning followed by Group Relative Policy Optimization (GRPO), enabling autonomous exploration that breaks the imitation-learning ceiling. Furthermore, we curate Video-DR-Bench, a human-AI collaborative benchmark comprising 200 complex, multi-hop VQA instances. Empirical results demonstrate that our Video-DeepResearch-35B-A3B establishes a new state-of-the-art of 64.0% average accuracy, surpassing proprietary Claude-4.5-Sonnet (59.0%) by 5.0 points and significantly outperforming GPT-5 (52.5%) and Gemini 2.5 Pro (57.5%). The 30B-A3B variant achieves 59.3%, competitive with Claude-4.5-Sonnet and demonstrating the effectiveness of our training paradigm even at compact scale. Code: this https URL.
[CV-2] JoyAI-Video-Edit: Real-Time Open-Ended Video Editing with Autoregressive Diffusion
链接: https://arxiv.org/abs/2608.03974
作者: Yicheng Xiao,Wenxun Dai,Xinran Qin,Lin Song,Maoquan Zhang,Hang Xu,Yukang Chen,Yitong Li,Guohui Zhang,Yuan Zhang,Xuying Zhang,Tommy Zhang,Jianlong Yuan,Peihao Li,Shuai Lu,Siming Fu,Chuyang Zhao,Xin Han,Jie Huang,Wenbo Li,Guoqing Ma,Wei Huang,Xiaojuan Qi,Haoyang Huang,Nan Duan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Code: this https URL
Abstract:Real-time video editing requires low-latency causal generation with bounded computational resources while preserving source fidelity and long-term temporal consistency. We present JoyAI-Video-Edit, a 16B-parameter autoregressive diffusion framework for real-time, open-ended video editing without access to future frames or a predefined video duration. Our method combines chunk-wise autoregressive adaptation, Source-Anchored Distribution Matching Distillation (SA-DMD), and Long-Horizon Autoregressive Distillation to reduce train–inference mismatch, preserve source fidelity during two-step generation, and mitigate accumulated temporal drift. Extensive automatic and human evaluations show that JoyAI-Video-Edit substantially outperforms existing streaming editors and remains competitive with strong offline systems on both short and long videos. The complete system achieves end-to-end 720p video editing at approximately 30 FPS on a single Nvidia B200 GPU. Code is available at this https URL.
[CV-3] UniWorld-Design: From Pixel Generation to Layer-Native Design
链接: https://arxiv.org/abs/2608.03971
作者: Zongjian Li,Zhiyuan Yan,Chenxu Bai,Chen Chen,Haoxiang Sun,Shaodong Wang,Feize Wu,Shenghai Yuan,Bin Lin,Zheyuan Liu,Yuwei Niu,Li Yuan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Project page: this https URL
Abstract:We introduce UniWorld-Design, a framework that redefines image generation from flat pixel synthesis to structured visual composition, with semantic RGBA layers as the atomic units of generation, understanding, and editing. Our key insight is that pixels define how an image is rendered, whereas layers define how an image is created, understood, and edited. Just as human designers create and manipulate visual content through layers rather than raw pixels, UniWorld-Design equips multimodal generative models with a layer-native design space. UniWorld-Design comprises two models. The Text-to-RGBA (T2RGBA) model generates standalone RGBA assets directly from text. The Image-to-Layer (I2L) model conditions on a finished image, a global instruction and per-layer prompts, and jointly produces ordered, complete semantic RGBA layers. Its instruction interface supports top-level decomposition, recursive decomposition and targeted extraction, making layering an instruction-addressable operation for agentic editing. Because I2L learns complete semantic objects rather than visible-pixel partitions, its layers stay usable when moved or removed. On the Crello benchmark, I2L reduces per-layer RGB L1 error by 37% and achieves a 34% relative improvement in Alpha Soft IoU over Qwen-Image-Layered. Separately, T2RGBA achieves the highest CLIP Score, outperforming LayerDiffuse and OmniAlpha.
[CV-4] Progressive Learning of a Diffusion-based Inpainting Model for Separating Overlapped Fingerprints
链接: https://arxiv.org/abs/2608.03937
作者: Noor Hussein,Anil K. Jain,Karthik Nandakumar
类目: Computer Vision and Pattern Recognition (cs.CV); Cryptography and Security (cs.CR)
备注: Accepted to IJCB 2026
Abstract:Overlapped friction ridge patterns are a recurring problem in latent fingerprints recovered from crime scenes and in live-scan scenarios where residual fingerprints on the sensor may corrupt subsequent acquisitions. Existing approaches for separating overlapped fingerprints either rely on rule-based orientation field completion that requires strong domain knowledge or train end-to-end deep neural networks that do not account for domain-specific considerations. This work introduces a diffusion-based pipeline for separating component fingerprints from an image containing overlapping friction ridge patterns. We formulate the separation problem as an inpainting task and progressively learn a diffusion model for this task in multiple stages. Starting from a pre-trained Stable Diffusion model, we progressively incorporate a fingerprint prior, add the ability to complete partial fingerprints, and finally propose \textbfoverlap-aware inpainting that reconstructs each component print using a diffusion inpainting model based on multi-channel conditioning. Experiments on two public datasets demonstrate that component fingerprints reconstructed using the proposed diffusion-based inpainting method can match with their mated counterparts with very high probability.
[CV-5] Latent Reward Registers for Diffusion Preference Alignment
链接: https://arxiv.org/abs/2608.03929
作者: Yuanshen Guan,Zipeng Feng,Zhiwei Xiong,Peiqin Sun
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Aligning diffusion models with human preferences usually relies on a sparse terminal reward evaluated on the final generated samples, presenting a severe temporal credit-assignment challenge across the multi-step denoising process. We propose Latent Reward Registers, a mechanism that estimates terminal preference directly from intermediate noisy latents by prepending learnable, position-free register tokens to the input sequence of a frozen Diffusion Transformer (DiT). This independent readout mechanism extracts latent reward evidence without altering the generator’s hidden states or velocity field. The resulting dense, differentiable reward signal throughout the full denoising process facilitates two alignment strategies. For training, Reward-Gradient On-Policy Distillation (RG-OPD) distills reward-guided updates along on-policy trajectories, bypassing the computationally expensive rollouts of standard policy gradients. For inference, Reward-Guided Sampling (RGS) steers trajectories via magnitude-matched reward gradients without parameter updates. Empirically, at high noise levels (u = 0.8), the registers reach the highest pairwise accuracy among the evaluated latent reward models. Furthermore, RG-OPD outperforms online reinforcement learning baselines while reducing GPU hours by up to 33x, and RGS establishes a new state-of-the-art among training-free methods, strictly enhancing both alignment and perceptual metrics. Code and weights are available at this https URL
[CV-6] PRISM: Powerful Time Series to Image (TS2I) Representations for Multivariate Anomaly Detection
链接: https://arxiv.org/abs/2608.03926
作者: Mateusz Smendowski,Kamil Faber,Piotr Nawrocki,Nathalie Japkowicz,Roberto Corizzo
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Time series anomaly detection (TSAD) underpins applications in predictive maintenance, finance, and cloud computing, however performance remains sensitive to representation choices, especially in multivariate settings. While transforming time series into images has shown success in forecasting and classification, it remains unclear how multivariate, high-dimensional series should be mapped to multi-channel images and whether vision backbones can match time-domain baselines in TSAD. We introduce PRISM, a plug-and-play meta-workflow enabling systematic construction and evaluation of image-based representations for multivariate TSAD. Our evaluation spanning over 7,000 experiments shows that well-designed PRISM configurations are competitive with 24 time-domain baselines, achieving the best VUS-PR on 10 of 14 datasets, with an average improvement of 41% over the best competing method on those datasets. Further, we identify channelization - how the channel dimension of multi-channel images is constructed - as a critical and previously understudied design dimension, and introduce MSM, a novel statistics-based scheme achieving 11-27% gains over PCA-based alternatives. Finally, ImageNet-pretrained encoders transfer effectively to TSAD, with frozen encoders retaining 92% of fine-tuned performance while training 1.8 times faster. Our code is available at: this https URL.
[CV-7] GeoMAR: Unleashing Geometrically Aligned Features for Masked Autoregressive Blind Face Restoration
链接: https://arxiv.org/abs/2608.03923
作者: Lu Gan,Hanyu Yan,Chaofeng Chen,Junqi Hu,Dan Zeng
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Codebook-based blind face restoration (BFR) often suffers from ambiguous conditioning features and a fragile prediction mechanism under severe degradation. To address these challenges, we propose GeoMAR, a framework designed to unleash geometrically aligned features with masked autoregressive (MAR) refinement for robust face restoration. For feature conditioning, we introduce a dual-input extraction pipeline to extract component-based geometric descriptions with explicit, spatially faithful anchors. These textual priors are integrated with low-quality (LQ) features via an Aligned Geometric Priors Injector, which employs a KV-Q exchange strategy to generate geometrically aligned features. For prediction mechanism, we reformulate the one-step mapping into a multi-step MAR process. This coarse-to-fine generation progressively refines complex facial regions based on increasingly reliable context. Experiments on one synthetic and three real-world benchmarks demonstrate that GeoMAR achieves highly competitive perceptual quality and coherent visual structures compared with existing methods. The code is available at this https URL.
[CV-8] Low-Dimensional High-Leverag e Subspace Optimization: Beyond Full-Parameter Coupled Training for Neural Network Quantization
链接: https://arxiv.org/abs/2608.03919
作者: Peng Xia,Junbiao Pang,Zheng Huang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 9 pages, 2 figures, 7 tables
Abstract:Low-bit quantization suffers severe accuracy degradation on compact networks, rooted in the dominant full-parameter coupled training paradigm that ignores parameter subspace heterogeneity. Their limited feature redundancy leaves little room to absorb quantization errors. Conventional pipelines adopt monolithic optimization: PTQ reconstructs fixed pretrained models without improving inherent quantization friendliness; QAT updates all parameters jointly, suffering from gradient coupling between backbone weights and calibration parameters. In this paper, we identify normalization affine parameters as a low-dimensional high-leverage subspace dominating quantization robustness, and propose Normalization Affine Preconditioning (NAP) for targeted subspace optimization. For PTQ, NAP freezes backbone weights and fine-tunes only affine parameters under the target fake-quantization graph on full-precision models, proactively boosting quantization friendliness before downstream reconstruction. For QAT, we introduce an alternating QAT-NAP schema that decouples feature learning and numerical calibration, breaking the performance ceiling of saturated joint training. Theoretical analysis confirms BN affine parameters fully cancel the channel-wise affine component of quantization distortion, while nonlinear rounding and clipping residuals form the irreducible error boundary; distillation-guided NAP acts as directional flatness optimization, projecting teacher-student logit mismatch onto the restricted subspace. Experiments on ImageNet and CIFAR-100 show NAP recovers severely collapsed low-bit quantization, consistently boosts reconstruction-based PTQ, and outperforms saturated full-parameter QAT with negligible tuning cost. This work reveals the principle of targeted low-dimensional subspace optimization, offering a new perspective beyond full-parameter coupled training for efficient deep learning.
[CV-9] When and Where to Look: Adaptive Visual Evidence Scheduling for Efficient Long Video Understanding
链接: https://arxiv.org/abs/2608.03918
作者: Ke Li,Jiayu Chen,Maoliang Li,Zihao Zheng,Hailong Zou,Hengyi Zhang,Xuanzhe Liu,Xiang Chen
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Efficient long-video understanding requires vision–language models (VLMs) to reason over a small number of frames selected as sparse visual evidence. Existing relevance-based methods rely on static one-shot selection with fixed frame budgets and candidate pools, while agent-based schedulers achieve adaptivity through costly multi-round reasoning and interactive search. We propose EcoFrame, a training-free framework for low-overhead query-adaptive visual evidence scheduling. EcoFrame leverages the VLM’s inference feedback to determine when to increase the frame budget and where to search for additional candidate evidence. Specifically, entropy-gated budget scheduling uses output uncertainty to stop early when the current evidence is sufficient or progressively expand the frame budget otherwise. Meanwhile, attention-guided candidate proposal converts frame-level attention into a temporal prior, enabling dense local search in informative regions while preserving global coverage when attention is diffuse. Experiments on Video-MME, LongVideoBench, and MLVU demonstrate that EcoFrame achieves a better accuracy–efficiency trade-off across multiple VLM backbones. On Qwen2.5-VL, EcoFrame achieves an average accuracy of 64.4, surpassing BOLT at 63.5, while providing a 1.85\times speedup over AKS and BOLT. Compared with the agent-based A.I.R., EcoFrame maintains comparable accuracy with up to a 13.5\times inference speedup. Code will be available at this https URL.
[CV-10] StreamDAM: Presence-Aware Memory for Real-Time Streaming Video Object Segmentation
链接: https://arxiv.org/abs/2608.03912
作者: Xiang Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Quality-tier video object segmentation (VOS) trackers such as DAM4SAM top accuracy leaderboards, but they are measured offline, one frame at a time with no clock. Under an honest streaming protocol at 30 frames per second, where a frame that misses its budget is served the last mask already computed, the winner collapses: the rich memory that makes it accurate is too slow to keep up, and what it emits is blind to whether the object is even present. We trace both failures to one place, the tracker’s memory pipeline, and rebuild it for streaming. \method makes the memory machinery itself run at frame rate through in-model optimization rather than a bolted-on fallback, and governs it with a single learned presence signal that decides what enters memory, how far back the tracker reads, when to withhold output, and when to re-detect. A mechanism analysis shows why a fixed policy cannot win: the control that helps when an object truly disappears is the one that hurts when it is merely hard to see, so the choice must be made per frame. Across four benchmarks and five modern baselines, \method is the strongest streaming tracker, recovers nearly all of the offline model’s accuracy under the clock, and on the hardest content exceeds the offline model it is built from.
[CV-11] UniEvo-RS: Omni-Prompt Unified Remote Sensing Segmentation with Representative Exemplar-Driven Prototype Evolution
链接: https://arxiv.org/abs/2608.03911
作者: Kunquan Zhang(1),Peilang Li(1),Xikun Hu(2),Yunkai Yang(1),Yushan Zou(2),Zhiwei Zhang(1),Runmin Dong(1) ((1) Sun Yat-sen University, (2) National University of Defense Technology)
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 18 pages, 8 figures, 9 tables
Abstract:Prompt-driven vision-language models (VLMs) hold immense promise for accelerating dense remote sensing (RS) annotation, but static models suffer from severe performance degradation when deployed on novel scenes, unseen categories, or visually confusing backgrounds. Moreover, existing unified paradigms primarily rely on intra-image specific prompts, lacking flexible task routing to adapt to multi-intent operational workflows. In practical batch mapping, annotators typically refine a small set of representative samples before processing large datasets. Motivated by this practice, we propose UniEvo-RS, an omni-prompt unified RS segmentation framework equipped with representative exemplar-driven prototype evolution. First, we construct a multi-instruction prompt dataset that unifies text-driven and visual-driven prompts within a single architecture, establishing a dynamic task-routing mechanism for highly diverse RS annotation scenarios. Second, we introduce a representative feedback-driven, training-free prototype evolution mechanism. By contrasting manual annotations with initial predictions on exemplars, UniEvo-RS distills prediction errors into positive and negative prototypes. These prototypes enhance LLM query recall and suppress spatial background noise under a fixed-budget clustering memory. Extensive experiments show that UniEvo-RS unifies diverse prompting tasks, achieving state-of-the-art performance across most settings. Crucially, with minimal interaction on a few exemplars, it enables training-free, progressive accuracy enhancement on unseen categories during batch annotation.
[CV-12] NCGR: Noise-Conditional Gated Rectification for Camera Extrinsic Perturbations in BEV 3D Object Detection
链接: https://arxiv.org/abs/2608.03895
作者: Wenbin Pan,Wanhao Liu,Liwei Luo,Panshuo Li,Yong Xu,Renquan Lu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 21 pages, including supplementary material
Abstract:Camera-based bird’s-eye-view (BEV) 3D detection typically assumes accurate and fixed camera extrinsics. In detectors using spatial cross-attention (SCA), extrinsic perturbations displace the image-plane projections of BEV reference points, causing queries to sample features from incorrect regions and degrading detection performance. To address this failure mode, Noise-Conditional Gated Rectification (NCGR) is proposed to compensate for projection errors without explicitly estimating a full six-degree-of-freedom extrinsic correction. For each query-camera pair, a 2D rectification offset is predicted and modulated by a camera-level gate to rectify the base projection before native deformable sampling. During training, the perturbation-derived quantities used to construct the condition and gate are gradually replaced through scheduled interpolation by counterparts generated from an auxiliary scalar predicted from camera features. This transition enables blind inference without perturbation metadata. During training, a weight-shared clean-teacher/perturbed-student pair is used, and the rectification module is supervised by a BEV-consistency objective between the two branches. NCGR is evaluated on nuScenes with simulated dynamic and static extrinsic perturbations. In a five-camera dynamic stress test, NCGR achieves 39.69% NDS, compared with 28.00% for BEVFormer and 33.23% for CAPE. Under clean extrinsics, NCGR maintains performance comparable to that of BEVFormer.
[CV-13] CARE-X: Towards Clinically Useful Radiology VLMs with Auxiliary Supervision Reward-Aligned Learning and Tool-Augmented Measurement
链接: https://arxiv.org/abs/2608.03890
作者: Mercy Prasanna Ranjit,Anirban Porya,Sathvik Joel,Niharika Vadlamudi,Nikhilesh Chowdary Eathamukkala,Prasanth V V,Abhyuday Kumara Swamy,Pranay Narhari Umredkar,Pradeep Narayan,Vivek Rajagopal,Tanuja Ganu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:A clinically useful chest X-ray system must go beyond fluent report generation: it should classify findings with tunable decision thresholds, localize them spatially, and derive the anatomical measurements upon which many diagnoses depend. Today’s Vision-Language Models (VLMs) treat these as separate problems, if they address them at all, leaving a gap between what radiologists need and what generative models provide. We introduce CARE-X, a chest X-ray VLM that narrows this gap by unifying auxiliary discriminative supervision with reward-aligned generation. CARE-X augments its generative backbone with focal-loss classification and composite-loss grounding heads, co-trained alongside the language-modeling objective. This auxiliary supervision produces discriminative diagnostic predictions with tunable decision thresholds and precise spatial localization while also improving report quality, providing evidence that structured prediction and generation reinforce one another. Building on this foundation, Decoupled Clip and Dynamic Sampling Policy Optimization (DAPO) leverages task-specific reward signals for report generation, visual question answering (VQA), and spatial grounding, directly optimizing the clinical quality metrics that matter in practice. The result is state-of-the-art performance on the majority of metrics across four report-generation benchmarks, 94.0% VQA accuracy on ReXVQA (+6.0 pp over the next-best baseline), and generative spatial decoding that reaches near parity with dedicated detection heads. Separately, to address measurement-dependent diagnoses, we couple Qwen3-VL-4B-Instruct with native tool-calling capabilities for invoking deterministic measurement tools, while retaining full visual access to the image. This hybrid inference yields +43.6 pp average F1 over perception-only baselines across five measurement-dependent conditions.
[CV-14] MuRA: Multi-Rank Adaptation for Efficient and Effective Test-Time Vision-Language Generalization
链接: https://arxiv.org/abs/2608.03885
作者: Gengyuan Liu,Nanzhou Wang,Chang Liu,Qinwen Wu,Zhenhao Wang,Jiacong Wang,Bokui Chen,Xiangyang Ji
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Vision-language models exhibit remarkable zero-shot capabilities but suffer significant performance degradation under distribution shifts. While test-time adaptation (TTA) via Low-Rank Adaptation offers a parameter-efficient solution, we identify a fundamental bottleneck in current methods: the reliance on static rank configurations. Because visual inputs inherently possess varying information densities, a fixed rank forces an inevitable optimization compromise, leading to underfitting on complex scenes and overfitting on simple ones. To bridge this gap, we propose Multi-Rank Adaptation (MuRA), a novel framework that dynamically selects and fuses adaptation modules of varying capacities based on token-level visual complexity. MuRA synergizes Multi-Rank Orthogonal Decomposition to provide a superior, knowledge-preserving initialization, and Unified Component Fusion with Continuous Router Updating to sustainably learn semantic-to-rank mappings. Furthermore, we provide rigorous theoretical justifications mathematically proving the necessity and gradient stability of this adaptive mechanism. Crucially, MuRA’s dynamic design uniquely thrives at the deepest visual layer, capitalizing on the shortest gradient backpropagation path. Extensive experiments demonstrate that MuRA achieves state-of-the-art accuracy across extensive domain generalization and cross-dataset benchmarks while significantly reducing both computational and memory overhead.
[CV-15] CPrefix: A Combinatorial Tensor Framework for Structured Discrete Color Mappings ICIP2026
链接: https://arxiv.org/abs/2608.03863
作者: Yvan Richard
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 7 pages, 5 figures. Accepted for presentation at the IEEE ICIP 2026 Workshop on Computational Color Imaging (CCIW 2026). Withdrawn from the proceedings because the author was unable to attend the conference
Abstract:Discrete multi-channel mappings are typically represented through sampled values, providing accurate evaluations but limited insight into their underlying structure. We introduce CPrefix, a combinatorial observable representation for discrete mappings, realized within a unified tensor framework that enables representation, reconstruction, and structural analysis. The framework is based on a counting tensor induced by multinomial counting observables. Its support forms a discrete Pascal simplex, not as a constraint on the observable space, but as a latent combinatorial representation from which mappings are reconstructed. This formulation separates the combinatorial organization of a mapping from its measured values, exposing the observable structure underlying the mapping. The framework is validated on ICC display and printer profiles through latent reconstruction and perceptual gamut transport. Accurate reconstruction demonstrates that color mappings admit faithful observable representations, while reconstruction residuals provide insight into the compatibility of the underlying mapping with the proposed representation. Although demonstrated on color transformations, the framework is independent of the physical interpretation of the observables, making it applicable to structured multi-channel mappings arising from color imaging, spectral measurements and other discrete systems. Comments: 7 pages, 5 figures. Accepted for presentation at the IEEE ICIP 2026 Workshop on Computational Color Imaging (CCIW 2026). Withdrawn from the proceedings because the author was unable to attend the conference Subjects: Computer Vision and Pattern Recognition (cs.CV) ACMclasses: I.4.1; I.4.10 Cite as: arXiv:2608.03863 [cs.CV] (or arXiv:2608.03863v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2608.03863 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Yvan Richard Mr [view email] [v1] Tue, 4 Aug 2026 16:04:27 UTC (2,780 KB)
[CV-16] LiteMVS: Efficient Multi-View Stereo with Foundation Distillation and Expert Aggregation CVPR2026
链接: https://arxiv.org/abs/2608.03851
作者: Tianbao Zhang,Zeyu Liu,Shuyu Wu,Fanxing Li,Zhaoxin Fan,Wenjun Wu,Danping Zou
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: CVPR 2026 Workshop accepted
Abstract:Real-time 3D perception is crucial for robotics, augmented reality, and embodied intelligence applications. Existing multi-view stereo (MVS) methods primarily rely on geometric correspondences, which often fail in textureless or repetitive regions, while monocular depth models leverage strong image-level priors but lack robust multi-view geometric constraints. More importantly, in robotics and embodied manipulation scenarios, high-quality 3D geometry is not only essential for static reconstruction, but also serves as a critical foundation for learning temporally consistent 4D representations. To obtain visual representations with stronger structural awareness and greater potential for spatiotemporal extension, we present LiteMVS, a lightweight multi-view depth estimation model that integrates plane-sweep geometric reasoning with strong monocular semantic and structural priors. The central idea of LiteMVS is to efficiently inject high-level monocular knowledge, obtained from lightweight segmentation models and large-scale vision foundation models, into a multi-view stereo framework. In particular, LiteMVS enriches the cost volume with semantic descriptors and employs a Mixture-of-Experts (MoE) formulation to enable adaptive geometric aggregation across depth hypotheses. Moreover, geometric priors distilled from vision foundation models further strengthen monocular guidance without increasing inference cost. Through this design, LiteMVS not only improves depth estimation and 3D reconstruction quality in static scenes, but also provides a more reliable geometric foundation for subsequent temporal modeling and 4D representation learning. Experiments on ScanNetv2 and 7-Scenes demonstrate that LiteMVS achieves high-quality depth prediction and 3D reconstruction while maintaining competitive efficiency.
[CV-17] Geo-Embed: Towards Unified Multimodal Embeddings for Urban Understanding
链接: https://arxiv.org/abs/2608.03826
作者: Jiapeng Li,Yong Li,Junjie Zhou,Fan Zhang,Yu Liu
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Geospatial and urban applications increasingly require models to compare heterogeneous evidence across street-view imagery, remote-sensing observations, text descriptions, region proposals, and temporal change cues. However, existing multimodal embedding models and benchmarks are still largely designed and evaluated around general-purpose image-text matching, leaving unclear whether unified embedding space can support heterogeneous geospatial tasks involving spatial relationships, fine-grained semantics, and temporal changes. To address this gap, we make three key contributions. First, we introduce GeoMEB, a large-scale multimodal embedding benchmark that standardizes 45 urban evaluation tasks across retrieval, visual question answering, change detection, classification, and visual grounding, together with training collections comprising 1.32M examples and 286K evaluation queries. Second, we present Geo-Embed, a unified embedding model that adapts a shared vision-language backbone to instruction-conditioned query-target matching over heterogeneous geospatial inputs, including single images, multiple images, text, regions, and masks. On GeoMEB, Geo-Embed achieves the strongest overall performance among representative multimodal embedders, with a 15.3% relative improvement over the strongest baseline. These results motivate future geospatial embedders that organize training and evaluation around explicit query-target relations, including semantic, cross-view, region-level, and temporal correspondence.
[CV-18] FlowForm: Synergizing Fluid Physics with Topological Consistency for Satellite Flood Synthesis
链接: https://arxiv.org/abs/2608.03822
作者: Zhang Weihui,Wang Ruizhi,Xu Hongye,Wang Huiqiong,Sun Li,Song Mingli
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Developing robust flood assessment models requires high-quality paired satellite imagery, yet such data remain scarce for flood-specific image generation. Although generative models provide a promising means of data augmentation, existing methods often yield implausible spatial layouts of flooded regions and distort scene structures. We propose FlowForm, a framework for satellite flood synthesis that integrates SWE-inspired latent regularization with structure-aware conditioning. The Flood Descriptor Module (FDM) imposes differentiable penalties on residuals of the steady-state Shallow Water Equation in auxiliary latent fields at the diffusion bottleneck. The Terrain Anchor Adapter (TAA) injects depth, semantic, and edge features at four encoder scales of the U-Net. We further curate FloodScape, a large-scale, high-resolution dataset comprising paired satellite images acquired before and after disasters. In addition to standard image-generation metrics, we evaluate the consistency of flooded regions, zero-shot generalization to a geographically held-out flood event, and sensitivity to individual components. Across all reported comparisons, FlowForm achieves higher visual fidelity, greater similarity between paired images, and stronger consistency of flooded regions.
[CV-19] UHP Detection: LVLMs have their Unique Hallucination Pattern in the Consistency Space
链接: https://arxiv.org/abs/2608.03817
作者: Amir Mohammad Ezzati,Kiyan Rezaee,Bardiya Kariminia,Mohamad Amin Yousefi,Asal Mohammadjafari Mamaqani,Behrad Samimi,Mohammad Hossein Rohban
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 12 pages
Abstract:Large vision–language models (LVLMs) demonstrate strong multimodal reasoning capabilities but remain prone to hallucination, where model predictions are not grounded in visual evidence. Existing black-box hallucination detection methods estimate uncertainty through a single consistency metric, implicitly assuming that model uncertainty can be adequately characterized by a single measure. However, hallucinations exhibit diverse manifestations of uncertainty across different behavioral probes, making a single measure insufficient to characterize their underlying behavior. We propose \emphUnique Hallucination Pattern (UHP) Detection, a fully black-box framework that models hallucination as a structured uncertainty pattern defined by two axes: perturbation modality (image vs.\ text) and logical polarity (a statement vs.\ its negation). Their intersection produces four complementary consistency groups that capture distinct manifestations of model uncertainty, from which both within-group and between-group features are extracted to train a lightweight classifier. Through comprehensive experiments on AMBER and PhD across three LVLMs, UHP Detection consistently outperforms prior black-box and white-box baselines, with improvements of up to +18.72% AUC-ROC and +20.07% AUC-PR over the strongest black-box methods. Extensive ablation studies demonstrate that each consistency group contributes complementary information and that their combination forms a structured hallucination pattern. Furthermore, cross-dataset evaluation shows that this learned pattern generalizes across benchmarks, indicating that hallucination behavior reflects a model-specific consistency pattern. \textbfCode is publicly available at this https URL.
[CV-20] OmniPack: Unified Token Compression for Efficient Omni-modal Large Language Models
链接: https://arxiv.org/abs/2608.03812
作者: Wanshun Su,Yang Shi,Feihu Liu,Ziwen Yu,Yan Min,Zhuoran Zhang,Qixun Wang,Haotian Wang,Shixuan Liu,Yuanxing Zhang,Peng Wu,Chengfu Huo,Liang Ding
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 16 pages, 5 figures, 15 tables
Abstract:Omni-modal large language models (Omni-LLMs) have achieved remarkable performance on audio-visual understanding tasks, but processing long and highly redundant visual and audio token sequences incurs substantial computational overhead, demanding aggressive token compression for efficient deployment. Existing methods often degrade at low token budgets: pre-LLM compression may discard structurally important and globally distributed evidence, whereas inner-LLM compression often underexploits query-conditioned audio-visual collaboration. To address these limitations, we propose OmniPack, a training-free framework that coordinates structural compression before the LLM with task-relevant semantic refinement within the LLM. Before the LLM, OmniPack removes structural redundancy through modality-specific importance, global coverage, and similarity-aware merging. After sufficient multimodal interaction, it further consolidates diverse, task-relevant representations through textual guidance and audio-visual collaboration. Extensive experiments on five benchmarks with three Omni-LLM backbones demonstrate that OmniPack consistently achieves the best performance-efficiency trade-off across diverse retention ratios, outperforming all existing methods. Notably, on Qwen2.5-Omni-7B, OmniPack preserves 98.0% of the original performance while reducing FLOPs to 16.7%, and still retains 92.9% of the original performance with only 6.8% of the original FLOPs.
[CV-21] Agent icVAU: Multi-Agent Explore-Verify Reasoning for Video Anomaly Understanding
链接: https://arxiv.org/abs/2608.03779
作者: Yuxiang Duan,Huining Li,Ao Li,Shuai Feng,Lanju Kong,Ning Liu,Jian Zhang,Xingdong Sheng,Yuntao Du
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Video anomaly understanding (VAU) focuses on comprehensively interpreting abnormal events in videos, requiring models to identify anomalous occurrences, discover their supporting evidence, and explain the underlying causes beyond simple anomaly detection. Existing VAU methods often rely on specialized training or limited observations, restricting generalization or evidence coverage. Although single-agent alternatives support adaptive video observation, they still integrate exploration, observation, and decision-making within a unified reasoning process, offering limited role specialization and structured evidence coordination. To address these limitations, we present AgenticVAU, a training-free multi-agent framework that casts VAU as an explore–verify process, where the system first discovers potential anomalies and then verifies them through targeted observations. To achieve this, four specialized agents are introduced to handle visual-rule construction, search planning, video observation, and final decision, respectively. These agents communicate through an anchor registry, a shared evidence memory that binds each observation. Guided by this agent framework, AgenticVAU interleaves broad temporal exploration, dense local verification, and cross-interval comparison until sufficient evidence is collected. We conduct extensive experiments on the ECVA, UCF-Crime, and MSAD subsets of VAU-Bench, the results show that AgenticVAU outperforms zero-shot inference and reinforcement learning-based baselines, demonstrating the value of multi-agent collaboration for video anomaly understanding.
[CV-22] DVR: Joint Text Disambiguation and Viewpoint Reasoning for Zero-Shot 3D Visual Grounding
链接: https://arxiv.org/abs/2608.03763
作者: Qingxi Du,Junbo Wang,Yuke Li,Yining Zhu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 10 pages, 5 figures, 7 tables
Abstract:Zero-shot 3D visual grounding aims to localize specific objects based on textual descriptions and 3D visual input. However, the effectiveness of existing methods is significantly hindered by the ambiguous query text and deficient viewpoints. To address these issues, we propose TDVR, a training-free reasoning framework that disambiguates the input text and infers accurate viewpoints for zero-shot 3D visual grounding. First, we construct semantic 3D scene graph from the detected instances in the 3D point cloud. Subsequently, we put the original query, appearance and spatial relationship descriptions into the LLM for fusion, thereby disambiguating the initial input. We leverage chain-of-thought reasoning to generate the structured representation of disambiguated query. Then taking the scene graph and structured query as input, we get the optimal view via viewpoint reasoning to solve the problem of missing viewpoints during grounding. Based on the obtained optimal viewpoint, we further discriminate the distracting objects, enabling the model with the ability to distinguish similar instances. After that, we match the category text and appearance images with the query by computing the similarity of feature vectors. Finally, the target object was identified by integrating the viewpoint score, confusion score, category score, and appearance score. Compared with previous methods, our TDVR has stronger capabilities in viewpoint reasoning, similar object discrimination, and ambiguous query understanding. Experimental results on the public ScanRefer dataset show that our method outperforms the existing state-of-the-art methods by 15.25% and 14.46% in Acc@0.25 and Acc@0.5 respectively, demonstrating the effectiveness of our TDVR in addressing ambiguous query text and deficient viewpoints.
[CV-23] owards Reliable and Reproducible Fetal Brain Biometry: A Deep Learning Approach Using MRI
链接: https://arxiv.org/abs/2608.03724
作者: Francesca Maccarone,Marina Di Stefano,Giorgio Longari,Giulia Frigerio,Gloria Rizzato,Rocco Prudentino,Nivedita Agarwal,Tommaso Ciceri,Denis Peruzzo,Simone Melzi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Currently under journal submission
Abstract:Fetal brain biometry is essential for quantitative assessment of brain development, supporting gestational age estimation, developmental monitoring, and detection of abnormalities. In clinical practice, measurements are manually performed, making them time-consuming and prone to variability. While automated approaches have been proposed, reproducible methods remain limited, particularly those providing anatomically interpretable landmark localization. We present a fully automated deep learning-based framework for reliable and reproducible brain biometry from 3D super-resolution-reconstructed fetal brain MRI. The proposed four-step pipeline derives biometric parameters by jointly estimating linear measurements and their corresponding anatomical landmarks. A 3D convolutional neural network is trained to regress landmark coordinates from brain segmentation label maps, followed by measurement-specific geometric optimization to refine landmark positions and compute measurements. The pipeline is evaluated on two publicly available fetal MRI datasets comprising 150 volumes (gestational age range: 20-37 weeks) acquired across different scanners and protocols, assessing five key biometric measurements across varying acquisition settings and providing a comprehensive evaluation of both measurement accuracy and landmark localization using quantitative metrics and visual assessment. Compared with the only available automated pipeline, the proposed method achieves comparable or improved accuracy for most measurements. In conclusion, we introduce a straightforward pipeline for reliable biometry estimations, with efficiency, interpretability and scalability that support integration into clinical workflows.
[CV-24] MultiCompose: Multi-Concept Personalized Composition with Per-Subject Attribute Binding
链接: https://arxiv.org/abs/2608.03708
作者: Ruirui Zhang,Zhengkai Zhao,Pan Gao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Text-to-image diffusion models enable personalization of specific visual concepts from a small number of reference images. However, generating a single image that contains multiple personalized subjects, each bound to user-specified attributes such as clothing, accessories, and held objects, remains largely unaddressed. Without explicit spatial constraints, concurrently activated concept checkpoints produce overlapping cross-attention responses, causing per-subject identity degradation and attribute misalignment. Moreover, no established benchmark jointly evaluates these two failure modes in the personalized multi-subject setting. We present MultiCompose, a composition framework that decouples per-concept personalization from multi-subject inference. A semantic preservation regularization maintains attribute binding capacity during fine-tuning, while a two-phase inference procedure automatically establishes subject layout and composes per-concept predictions through spatially exclusive masks. We further introduce MSP-Bench, a benchmark that jointly evaluates identity fidelity (ID), attribute binding accuracy (BIND), and attribute misalignment (MIS) through a dual-pathway protocol. Experiments show that MultiCompose outperforms existing methods on both conventional metrics and MSP-Bench, confirming the benchmark’s ability to reveal failure modes that conventional metrics overlook. Code is available at this https URL
[CV-25] Pattern over Pixels: Measuring Pattern Completion Bias in Multimodal Code Generation
链接: https://arxiv.org/abs/2608.03691
作者: Khai-Nguyen Nguyen,Oscar Chaparro,Antonio Mastropaolo
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注: 41st IEEE/ACM International Conference on Automated Software Engineering
Abstract:Multimodal large language models (MLLMs) are increasingly used to translate webpage screenshots into front-end code, but repeated UI patterns may sway them toward visually incorrect yet pattern-consistent outputs. In this work, we test how repeated webpage patterns hurt MLLM accuracy on an objective screenshot-to-code fill-in-the-blank task. We introduce the first benchmark for visual pattern-completion bias, where one localized element in a repeated UI pattern is perturbed and the model must recover the masked width or font-size value from the screenshot and HTML context. Starting from 30 webpages curated from the Design2Code dataset, we build 1,440 evaluated screenshots spanning structural card and text-style patterns under standard and noise-overlaid conditions. We evaluate five frontier MLLMs and find that all are strongly biased toward the repeated baseline. Mean bias rate reaches 69.78% on card-width perturbations and 80.22% on text font-size perturbations, while mean accuracy is only 21.17% and 7.89%, respectively. Codex-5.3 performs best but still drops from 68.61% accuracy on cards to 13.89% on text, while Flash-3.0 reaches 96.11% bias on text. Noise, subtler perturbations, and boundary positions further increase bias rate. Reasoning analysis further shows that greater reasoning effort correlates with lower bias, yet qualitative evidence reveals that models can identify the anomalous element and still override it with the pattern-consistent answer. Our results identify a concrete failure mode in multimodal code generation and show that its severity is strongly associated with visual saliency
[CV-26] Keep the Needle Prune the Haystack: Defect-Preserving Token Pruning for Efficient Zero-Shot Anomaly Detection FAST
链接: https://arxiv.org/abs/2608.03681
作者: Yanning Hou,Jingyuan Zhang,Xiaoyun Wang,Qixiang Ma,Sihang Zhou,Ke Xu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Code: this https URL
Abstract:Zero-shot visual anomaly detection has achieved remarkable progress, with recent vision-only approaches further improving performance while simplifying the inference pipeline. However, existing methods typically perform dense computation over all images and spatial tokens, despite the fact that normal samples dominate real-world scenarios and anomalies usually occupy only small regions. Token pruning offers a promising solution, but introduces an asymmetric pruning risk in anomaly detection: retaining normal tokens mainly incurs redundant computation, whereas removing anomalous tokens may eliminate the only evidence for detection and localization. This risk is particularly severe in early layers, where pruning provides the greatest computational benefit but anomaly semantics remain unreliable. We propose KeepAD, a defect-preserving token pruning framework that formulates token selection as high-recall, anomaly-aware routing. In shallow layers, KeepAD combines coverage-preserving selection over local 2\times2 patch neighborhoods with deterministic anomaly rescue to reduce the risk of discarding subtle defects. In deeper layers, frozen normal and abnormal prototypes guide pruning under an image-adaptive token budget, aggressively removing low-risk normal tokens while preserving local anomaly evidence. Dense-to-sparse self-distillation further supervises early token routing without introducing additional inference overhead. Experiments on six industrial and seven medical zero-shot anomaly detection benchmarks show that KeepAD reduces the token retention ratio to below 20% , while limiting the average degradation in image-level and pixel-level AUROC to within 2.7 percentage points. At the most aggressive operating point, KeepAD achieves a 7.9\times speedup over the strongest CLIP-based baseline.
[CV-27] XiDepth: a Lightweight and Efficient Network for Self-supervised Monocular Depth Estimation
链接: https://arxiv.org/abs/2608.03666
作者: Elena Izzo,Riccardo Toniolo,Lamberto Ballan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to IEEE AVSS 2026
Abstract:Self-supervised monocular depth estimation has emerged as an appealing solution to design lightweight and effective models for deployment on computationally constrained devices due to its reduced reliance on expensive depth sensors. By eliminating the need for ground-truth annotations and leveraging the simplicity of monocular camera setups, this approach facilitates cost-effective data collection and broad applicability across fields such as computer vision and robotics. A critical challenge is achieving resource-efficient neural networks without compromising the overall performance. State-of-the-art models generally adopt depth-wise convolutions and attention mechanisms; however, these functions often incur high energy costs and face compatibility issues in embedded environments. To address this, we propose XiDepth, a lightweight architecture based on the XiNet operator block, designed to enhance feature extraction while maintaining low computational complexity and energy demand. On the KITTI dataset, XiDepth achieves state-of-the-art performance with only 0.8M parameters. Tests on a Raspberry Pi 4 further confirm its suitability for real-world embedded applications, reducing FLOPs by 40% and energy consumption by 35% compared to leading methods.
[CV-28] Morphology-Aware Implicit Super-Resolution Network for Pathological Images
链接: https://arxiv.org/abs/2608.03664
作者: Jiaming Liang,QiHui Han,Haolin Chen,Chengxin Ye,Jiawen Liu,Jiazhou Chen,Xiaoqi Sheng,Hongmin Cai
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Accurate diagnosis in Digital Pathology (DP) relies on high-resolution whole-slide images, yet clinical deployment is often limited by hardware costs. Super-Resolution (SR) offers a promising alternative by computationally enhancing low-resolution acquisitions. However, existing SR methods frequently struggle to preserve fine-grained cellular morphology, leading to texture oversmoothing and blurred structural boundaries under complex tissue variability. To address this issue, we propose Morph-ISR, a morphology-aware implicit super-resolution framework for DP that restores diagnostically relevant details with sub-pixel precision. Morph-ISR reformulates SR as a continuous coordinate-based reconstruction problem and integrates an Implicit Position-aware Kernel Generator (IPKG) to adaptively model spatially varying tissue morphology. To further enhance structural fidelity, a Morphological Fidelity Prior (MFP) is introduced, leveraging semantic guidance from a pre-trained cell segmentation network to enforce boundary-preserving and region-aware reconstruction, thereby improving the representation of critical cellular boundaries and nuclear textures. Experiments on TCGA and SurGen datasets show that Morph-ISR achieves the best LPIPS and ST-LPIPS among the evaluated methods, reducing them by up to 38.37% and 39.55%, respectively, over the second-best methods while maintaining strong PSNR and SSIM. These results demonstrate superior preservation of diagnostically relevant cellular boundaries and nuclear textures, while compact parameterization and high throughput support efficient edge deployment. Code and trained models will be released upon publication.
[CV-29] When Do Fewer Visual Tokens Accelerate Multimodal Inference? A Break-Even Study Across Decision Locations and Hardware
链接: https://arxiv.org/abs/2608.03649
作者: Hao Dou,Ruiwen Tian
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 16 pages, 3 figures, 13 tables. Experiments use Qwen2.5-VL-3B-Instruct on RTX 3090 and A100 PCIe GPUs
Abstract:Fewer visual tokens do not guarantee lower end-to-end latency. We evaluate break-even with a reproducible protocol that accounts for decision overhead, shared work, and the operators each policy can avoid. A stage-level decomposition reconciles these components with measured end-to-end latency. In a 30-example pilot, the two tested autoregressive probes remain slower than Full despite state reuse. A lightweight post-vision predictor yields paired confidence intervals below zero on RTX 3090 and A100 and remains significant after a conservative all-pairs Holm correction. A pre-vision image-size rule also yields intervals below zero on both GPUs, although neither comparison remains significant after the same correction. Pre-vision routing has a structural opportunity unavailable to post-vision pruning: it can avoid preprocessing and vision encoding. On A100, this opportunity outweighs a nearly eightfold larger downstream token reduction by the post-vision policy. Reported quality is conditional on examples answered correctly by Full and is not benchmark accuracy.
[CV-30] Learning Biomechanically Plausible Human Motion from Sparse Radar Point Clouds
链接: https://arxiv.org/abs/2608.03637
作者: Jonas Leo Mueller,Markus Gambietz,Alexander Weiss,Daniel Krauss,Bjoern M. Eskofier
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Radar-based human pose estimation has focused on improving learning algorithms while representing the body as unconstrained keypoint coordinates. We address the underexplored dimension of anatomical fidelity by integrating a full-body skeletal model into a differentiable, end-to-end trainable radar-based pose estimation framework, in which the pose network is supervised through forward kinematics while subject-specific geometry is fitted beforehand. Subject-specific body segment proportions are predicted from radar point cloud features to scale a biomechanical skeleton. A motion prediction network maps temporal radar sequences to generalized coordinates, and differentiable forward kinematics converts predicted joint angles into 3D positions. A contact classification loss encourages physically plausible foot-ground interaction. Under leave-one-subject-out cross-validation on 11 healthy participants performing rehabilitation exercises, the framework achieves 6.456 +/- 1.759 cm mean per-joint position error (MPJPE), 8.083 +/- 0.884 degrees mean per-joint angle error (MPJAE), 0.935 +/- 0.009 contact classification F1, and 3.4 +/- 1.3 % scaling error. This proof-of-concept study demonstrates the feasibility of recovering interpretable biomechanical descriptors from a single low-cost radar sensor in a controlled laboratory setting, a prerequisite for future clinical motion analysis.
[CV-31] SEER: A Self-Grounded Evidence Interface for Controlled Spatial Relation Classification
链接: https://arxiv.org/abs/2608.03631
作者: Feixiang Liu,Likun Wang,Qiang Qiu,Hui Xu,Huawei Shen,Xueqi Cheng
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 23 pages total, 2 figures. Code: this https URL
Abstract:Spatial relation questions require a model to identify the queried subject and object before comparing their layout. Yet a VLM can recognize both entities and still answer from the wrong instance or an ambiguous global view. We ask whether making query-specific evidence explicit can mitigate this failure and propose SEER (Self-grounded Evidence for Entity-Relation Reasoning), a training-free inference-time evidence interface for frozen VLMs. SEER hides candidate relations during pair localization, constructs a query-specific view with explicit subject/object roles, and retains the full image and sparse box geometry as complementary evidence. For relation-choice protocols with exact inverse support, an optional refinement swaps the entity roles and changes the forward decision only when exactly one visual state obeys the corresponding inverse relation. On an image-disjoint GQA-Train900 test frozen before model scoring, SEER pools to +3.94 [2.17,5.72] over Full; the gain remains positive under label-independent grounding-order counterbalancing and on the 535 rows whose entity names are unique. The unchanged protocol yields +4.35 to +11.79 on all 2,434 filtered EmbSpatial pair-relation questions across three models. Matched controls separate local refocus from role-explicit conditioning. These results establish query-specific evidence construction as the principal intervention, with reciprocal consistency as a smaller protocol-specific refinement.
[CV-32] Geospatial-Prior Guidance for 3D Semantic Scene Completion
链接: https://arxiv.org/abs/2608.03618
作者: Meng Wang,Shougao Zhang,Wenzhe He,Ruihui Li,Nan Hu,Zhuo Tang,Kenli Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Inferring complete 3D geometry and semantics from onboard images remains challenging because occlusions and restricted fields of view leave large scene regions underconstrained. Although satellite imagery provides wide-area context, appearance cues alone offer limited structural guidance and may be unreliable because of spatial or temporal discrepancies. We present GeoScene, a geospatially guided framework that jointly uses satellite imagery and structured OpenStreetMap cues as soft priors for 3D semantic scene completion. GeoScene learns complementary voxel-wise reliability weights for onboard observations and geospatial guidance, and uses them to control feature refinement in observed and unobserved regions. This design preserves local visual evidence while exploiting large-scale road and building structure beyond onboard visibility. Experiments on SemanticKITTI and SSCBench-KITTI-360 demonstrate that GeoScene consistently improves both geometric and semantic completion under the geospatial-prior-assisted setting, with the most pronounced benefits for large-scale static and geospatially structured classes.
[CV-33] SlimVLM: Sensitivity-aware Dynamic Structured Pruning with Adaptive Visual Token Selection for Efficient Vision-Language Models
链接: https://arxiv.org/abs/2608.03580
作者: Yaozhi Wen,Jialong Guo,Zhenliang Ni,Han Shu,Xinghao Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:While Vision-Language Models (VLMs) have demonstrated remarkable performance in processing and understanding both text and images, their large parameter sizes lead to significant computational overhead, limiting their deployment on resource-constrained devices. While pruning has been effective for compressing Large Language Models (LLMs), directly applying it to VLMs leads to significant performance drops, largely due to redundant visual tokens interfering with importance estimation. To this end, we propose SlimVLM, a structured pruning framework designed to compress VLMs while preserving their task performance. We introduce an adaptive visual token selection strategy for VLMs that leverages average text-to-visual attention scores to assess the importance of visual tokens, removing redundant ones during pruning based on a set threshold, thereby optimizing the importance calculation. Recognizing the varying tolerance to sparsity across different modules, we also propose a Sensitivity-aware dynamic pruning mechanism that determines the appropriate pruning ratio for each module by calculating the linear reconstruction error between the outputs of the pruned and unpruned modules, ensuring overall performance stability. Experimental results show that SlimVLM outperforms existing methods across multiple multimodal benchmarks, achieving state-of-the-art performance.
[CV-34] Beyond Simply Environment Scaling: Designing Effective Environment Distributions for Multimodal Agent Learning
链接: https://arxiv.org/abs/2608.03571
作者: Kejian Zhu,Zhuoran Jin,Dongqi Huang,Hongbang Yuan,Yupu Hao,Kang Liu,Jun Zhao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Code: this https URL
Abstract:Recent works train agents by constructing large-scale multimodal environment pools. However, we find that simply increasing the number of multimodal environments does not always benefit. We further analyze the limitations in current multimodal environment distributions through a series of experiments. Based on these findings, we study how to build more effective training environment distributions from two dimensions: diversity and difficulty structure. For diversity, we propose Ability-aware Environment Selection (AES) to obtain diverse environment sets. For difficulty structure, we propose Hierarchical Difficulty Curriculum (HDC), which organizes curriculum learning through two difficulty levels: harness weakening and state-scale progression. Experiments show that AES and HDC effectively improve multimodal agent training.
[CV-35] Compass: Degradation-Simulated Reciprocal Learning with Lightweight Needle RWKV for Multimodal Crack Segmentation under Missing Modalities ACM-MM2026
链接: https://arxiv.org/abs/2608.03559
作者: Hui Liu,Chen Jia,Fan Shi,Xu Cheng,Mianzhao Wang,Shengyong Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: This paper has been accepted by ACM MM 2026
Abstract:In multimodal crack segmentation for industrial facilities, the key challenge is preventing missing modalities from degrading pixel-level performance while maintaining low computational cost. Existing methods struggle to address semantic degradation caused by missing modalities. We propose Compass, a lightweight network for robust crack segmentation under arbitrary missing modalities. Compass comprises Degradation Simulation Distillation (DSD), Needle Block, and Evidential Topology-Preserving Fusion (ETPF). DSD constructs a degradation simulation stream that mimics more severe missing conditions and performs reciprocal distillation with the original stream, decoupling complete perception from degradation adaptation. Within DSD, Feature-Aware Prototype Transmitter (FAPT) performs modality agnostic prototype-guided feature completion to maintain semantic integrity under incomplete modality conditions. As a lightweight backbone, Needle injects crack-direction cues into WKV modulation and combines connectivity-aware gating with anisotropic context probing for structure-aware modeling. ETPF fuses multimodal features via Dempster-Shafer evidential combination with uncertainty-gated decoding, preserving crack topology while suppressing unreliable features. Experiments on three datasets demonstrate state-of-the-art (SOTA) performance under diverse missing modality scenarios. Even with 90% depth modality missing on CrackDepth, Compass achieves F1 of 0.8216 and mIoU of 0.8434 with only 2.58M parameters. The code is available at this https URL.
[CV-36] st-Time Augmentation for Tabular-to-Image Classifiers under Distribution Shifts
链接: https://arxiv.org/abs/2608.03557
作者: Malena Loza,Felipe Grijalva,Eva Milara,Luis Bote-Curiel,Francisco J. Lara-Abelenda,David Chushig-Muzo
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Tabular-to-image methods that convert tabular data into visual representations have emerged as a novel paradigm for leveraging the high performance of deep learning models. Despite their advantages, the robustness of these methods under distribution shifts remains under explored. Test-Time Augmentation (TTA) is an effective approach in image classification to improve model generalization and robustness, where predictions over multiple transformed views of each input are aggregated. This work evaluates the impact of TTA techniques on predictive performance under Out-Of-Distribution (OOD) for representations generated by tabular-to-image methods. Six tabular-to-image encoding methods were considered: TINTO, IGTD, DeepInsight, BIE, DistanceMatrix, Fotomics. Twenty-five TTA techniques were used, organized into six types: Geometric, Photometric, Structural, Frequency/Encoding, Mixup, and Composite. We employed two datasets from the TableShift benchmark (HELOC and Voting) that provide in-distribution and OOD test subsets designed to evaluate the effect of distribution shifts on tabular data. The results indicate that TTA improves OOD performance, with composite and photometric strategies providing the best trade-off between robustness and variance. In contrast, frequency-domain transformations that alter the encoder’s feature-to-intensity mapping consistently degrade performance. These findings highlight TTA as a promising approach for improving the robustness and generalization of classifiers trained on image representations derived from tabular data, particularly under distribution shifts.
[CV-37] S3-Diff: Structural Semantic Synergy Diffusion Model for High Fidelity Super Resolution of Pathological Images
链接: https://arxiv.org/abs/2608.03540
作者: Jiaming Liang,QiHui Han,Guangye Ou,Jiawen Liu,Haolin Chen,Xi Zhong,Jiazhou Chen,Xiaoqi Sheng,Hongmin Cai
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Digital pathology relies on high-resolution whole slide images for accurate diagnosis, yet limitations in imaging devices, storage, and transmission often make lower-resolution pathology images more common in clinical workflows. Current super-resolution techniques often tend to smooth diagnostically relevant morphology, leading to over-smoothed textures and semantic drift that compromise downstream clinical interpretation. To this end, we develop the Structural Semantic Synergy Diffusion Model (S3-Diff), a diffusion framework for high-fidelity super-resolution of pathological images. The core of S3-Diff is Specimen-aware Structural Anchoring (SSA), which combines prognosis-aware tissue support extracted by a fixed SAM with LR-HR gradient discrepancies to generate a specimen-specific structural anchor to preserve pathological morphology. Concurrently, we introduce Structure-guided Semantic Fidelity Tuning (SSFT) to adapt DINOv3 representations using SSA-derived structural supervision. SSFT combines the adapted semantic energy with LR-derived edge and grayscale cues. The resulting control guides denoising to suppress stochastic artifacts and maintain structural consistency. Extensive experimental results demonstrate that S3-Diff consistently outperforms state-of-the-art methods in both reconstruction quality and downstream survival analysis performance. The source code will be made public.
[CV-38] IRIS: Visual-Semantic Binding for Forgery-Resistant Watermarking of Diffusion Images
链接: https://arxiv.org/abs/2608.03539
作者: Xiaoyan Feng,Zheng Gao,Tong Guan,Rui Bao,Bokang Zeng,Xiaoyu Li,Jiaojiao Jiang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Most in-generation diffusion watermarks embed patterns independent of the image that carries them, and attackers transplant the marks onto images the generator did not produce, resulting in forgery. Binding the mark to visual semantics prevents such transplantation, yet existing bindings anchor to a proxy image rather than the image they mark. Realizing visual-semantic binding inside generation faces two challenges. The mark derives from the image itself yet enters the sampling trajectory before that image exists, and may itself shift the semantics it binds. The binding also meets opposite sensitivity demands, breaking under semantic change while holding through common processing. We present IRIS, a training-free watermarking scheme that embeds an Intrinsic Ring Identifier from Semantics. IRIS reads a content code from the non-watermarked generated image, derives a one-time ring from the code and a secret key, returns to the final low-noise steps of the same trajectory and blends the ring in, after the semantics it binds are settled. To meet the opposite sensitivity demands, the code is read through a canonicalization shared between embedding and detection, holding through common distortions and mild regeneration while flipping under semantic change. Detection recomputes the ring from the query image and the key alone, and the mark therefore fails on a foreign or spliced image, with acceptance tracking semantic displacement. On three prompt datasets IRIS detects reliably and stays close to its same-seed non-watermarked counterpart, a fidelity prior in-generation marks do not reach. While forgeries transfer fixed-pattern marks and regeneration strips post-hoc marks, IRIS alone among the compared marks withstands both.
[CV-39] MinerU.Chem: A High-Precision System for Optical Chemical Structure and Reaction Recognition
链接: https://arxiv.org/abs/2608.03525
作者: Haote Yang,Jiang Wu,Jingchao Wang,Xingjian Wei,Lixin Ma,Linye Li,Chen Zhu,Xiaolong Wu,Yuheng Lu,Ziran Zhu,Junyuan Gao,Lingli Ge,Yuan Xu,Huijie Ao,QianQian Wu,Dechen Lin,Huaiyu Gu,Lu Chen,Shengxin Lu,ShaSha Wang,Yuanyuan Cao,Zhejia Yu,Ruijie Zhang,Zimai Tian,Jiaxing Sun,Yinfan Wang,Jiahe Song,Chuang Wang,Yubin Wang,Rui Nie,Hao Zheng,Bowen Jiang,Hongbin Lai,Yifan He,Chengjin Liu,Tingting Zhang,Liqun Wei,Lijun Wu,Bin Wang,Yuqiang Li,Guangyu Wang,Wei Li,Bowen Zhou,Dahua Lin,Conghui He
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:In organic chemistry papers and patents, molecular structures, reaction schemes, and experimental conditions are often presented as molecular structure depictions, reaction diagrams, and complex tables or figures. Such information is difficult for general-purpose document parsing systems to directly convert into machine-readable data. This limits data production for organic chemistry knowledge base construction and for AI for Chemistry tasks such as reaction prediction, retrosynthesis, condition recommendation, molecular property prediction, and drug molecule design. This report introduces this http URL, a document parsing system for organic chemistry literature integrated into the MinerU online platform. Built on top of MinerU’s general document parsing pipeline, this http URL adds five chemistry-specific modules: chemistry relevance filtering, molecular structure detection, molecule identifier extraction, molecular structure recognition, and reaction scheme parsing. Together, these modules convert organic-chemistry-related image regions in documents into a Molecule Summary List and a Reaction Summary List. For molecular structure recognition, this http URL uses CARBON (Complex Atomic Representation and Bonding Object Notation) as its core representation. CARBON enables recognition results to preserve both the visual layout of the original image and complex chemical semantics, while supporting the export of standard downstream formats such as MolFile and SMILES. On the SMILES-evaluable subset of MolRecBench-Wild (N=2,392), this http URL’s molecular structure recognition module achieves a SMILES exact-match accuracy of 93.02%, outperforming the best evaluated comparison system, GPT-5.6-Sol (74.87%), by 18.15 percentage points. The system has been integrated into the MinerU online platform and is available at this https URL .
[CV-40] GVCCTurbo: Rate-Compute Quality Scheduling for Codebook Driven Generative Compression
链接: https://arxiv.org/abs/2608.03517
作者: Ziyue Zeng,Dingjie Peng,Xun Su,Hiroshi Watanabe
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Codebook-driven generative compression uses a pretrained image or video generator as a zero-shot visual prior and transmits compact codebook indices to guide reconstruction at ultra-low bitrate. Current codecs tie each finite-rate correction to a fresh prior evaluation, so shortening the sampler also removes correction slots that carry target-dependent information. We propose GVCCTurbo, a BPP-driven scheduler that separates expensive prior refreshes from codebook corrections: after calibrating an atom-count operating point and skip-gap ratio once per protocol, it maps a target codebook-payload bitrate to a trajectory length and refresh period, making BPP a schedule input instead of a fixed consequence of sampler length. The same endpoint-prediction and finite-rate steering interface covers GVCC-style rectified-flow video and DDCM-style diffusion image compression, preserving zero-training deployment and compatibility with future distilled priors. Native 1080p curves position the complete zero-shot codec in the ultra-low-bitrate regime. In a controlled 720p Wan-GVCC study, the scheduler cuts prior evaluations from 20 to 9 for a \sim!44% measured decoding-time reduction shared across the whole schedule family, at a small shared LPIPS cost on high-motion content; within that family, uniform refresh thinning (pure-skip) is a boundary point, and the BPP-aware interior point trades 2.9% fewer codebook-payload bits for consistently higher PSNR at comparable LPIPS. These results support BPP-to-compute scheduling as a controllable extension of sampler-length tuning, without requiring the allocated point to dominate every boundary point.
[CV-41] Detecting Pose Estimation Failures via Keypoint Self-Consistency
链接: https://arxiv.org/abs/2608.03516
作者: Robin Chan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:One common approach to pose estimation involves predicting object keypoints in an image, followed by using Perspective-n-Point algorithms to compute the object’s rotation and translation relative to the camera. While rotations preserve object shapes, this property is often neglected in keypoint-based pose estimation methods, where keypoints are typically predicted independently from each other. As imprecise keypoint predictions negatively affects pose estimation accuracy, it also limits its reliability in downstream tasks. In this work, we explore whether such inaccurate pose estimates can be identified by simply examining spatial locations between 2D keypoints. We propose a set of hand-crafted geometric features that capture the self-consistency of keypoint predictions, including pairwise distances, reprojection consistency, as well as render and mask consistency. Despite its simplicity, a logistic regression classifier trained on these features reliably detects pose estimation failures, outperforming confidence-based approaches like conformal keypoint predictions that rely solely on keypoint uncertainty.
[CV-42] How Many Labels Are Enough? ALDA: Active Learning Deployment Advisor for Medical Image Classification MICCAI
链接: https://arxiv.org/abs/2608.03511
作者: Julia Machnio,Mads Nielsen,Mostafa Mehdipour Ghazi
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Accepted at EMA4MICCAI Workshop 2026
Abstract:Active learning (AL) promises to reduce the cost of medical imaging projects by lowering the number of clinical labels required. However, practical deployment requires committing to a sampling strategy before the full annotation budget is spent, and choosing the wrong strategy can increase rather than decrease costs. We propose Active-Learning Deployment Advisor (ALDA), a deployment-oriented framework for AL method selection under clinical performance constraints. Given a short pilot phase, ALDA fits a parametric learning-curve model to each candidate strategy, estimates whether that strategy is expected to reach a required clinical performance target, and predicts the number of expert annotations needed to do so. In addition to absolute annotation cost, ALDA introduces a deployment window that quantifies the sensitivity of this cost estimate to uncertainty in the clinical threshold. The final recommendation follows a risk-aware rule: among strategies with near-optimal predicted cost, ALDA prefers the strategy with the narrowest deployment window, the most robust to threshold revisions. Experiments on four medical imaging classification domains show that ALDA predicts the deployment-optimal method from a pilot of 15-30% of the intended budget and reduces annotation costs by up to 82% compared with a poor strategy choice. Rather than introducing a new sampling heuristic, ALDA provides a practical decision layer that answers a deployment-critical question: how many labels are enough?
[CV-43] From Multi-Resolution Cells to Gigapixel Whole Slide Images Foundation Model for Computational Pathology
链接: https://arxiv.org/abs/2608.03508
作者: Basit Alawode,Moshira Ali Abdalla,Dwarikanath Mahapatra,Muhammad Muzammal Naseer,Sajid Javed
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Vision Transformers (ViTs) and their hierarchical variants have achieved strong performance in Computational Pathology (CPath). However, most are pre-trained on single-resolution Whole Slide Images (WSIs), limiting their generalization across arbitrary resolutions. Gigapixel WSIs inherently contain diagnostic patterns at multiple scales, including cellular morphologies, tissue architectures, and global context, mirroring how expert pathologists examine WSIs. We introduce Multi-Resolution Pyramid Transformer (MRPT), a model that hierarchically aggregates multi-resolution information from cellular to tissue and WSI levels. MRPT employs a biologically meaningful Consecutive Cross-Resolution Attention (CCRA) mechanism to capture scale-independent interactions and enforces multi-resolution semantic consistency by aligning embeddings across resolutions, yielding robust and generalizable WSI representations. Pre-trained in a multi-resolution self-supervised manner on 624M patches, 2.4M regions, and 36K WSIs, MRPT learns rich coarse-to-fine histopathology features. Extensive experiments on 34 diverse datasets show that MRPT surpasses recent foundation models and Multimodal Large Language Models (MLLMs) in cancer subtype classification, tissue phenotyping, and Visual Question Answering (VQA) for WSI understanding.
[CV-44] Principles of Robot Autonomy
链接: https://arxiv.org/abs/2608.03496
作者: Daniele Gammelli,Joseph Lorenzetti,Katie Luo,Gioele Zardini,Marco Pavone
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Systems and Control (eess.SY)
备注: 531 pages. Pre-publication version of a book forthcoming from Cambridge University Press, posted with the permission of the publisher
Abstract:Autonomous robots are moving rapidly from research labs into everyday life - on roads, in the air, in warehouses, and in space. Robot autonomy is no longer solely an academic pursuit, but a collection of mature, field-tested methods and tools that practitioners rely on in real-world deployments. This book offers a clear, unified introduction to the methods that make this possible. Built on decades of teaching at Stanford, the text develops the core elements of modern autonomy stacks within a single conceptual framework, bridging classical robotics and modern physical AI. Every major topic is paired with hands-on Jupyter notebooks and implementation-driven exercises, so readers build practical intuition alongside theoretical understanding. The result is a principled, accessible, and deployment-aware foundation for anyone seeking to design, analyze, or contribute to the next generation of autonomous systems. This is a comprehensive resource for students, engineers, and researchers entering one of today’s fastest-growing fields.
[CV-45] Lightweight 3D Object Detection via Mamba-Based Knowledge Distillation
链接: https://arxiv.org/abs/2608.03490
作者: Quoc Cuong Ninh,Huy Xuan Pham,Anh Tung Nguyen,Dinh Hoan Trinh
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted for publication in IEEE Robotics and Automation Letters (RA-L), 2026
Abstract:3D object detection using light detection and ranging (LiDAR) sensors requires a balance between accuracy and computational efficiency for onboard perception in autonomous driving and robotic navigation. Many existing LiDAR-based detection methods employ complex architectures to extract features, integrating large amounts of contextual information to enhance accuracy. This often results in significant computational costs, leading to suboptimal performance on resource-constrained embedded devices. In this study, we propose a knowledge distillation framework that transfers object-level voxel representations from a strong teacher model to lightweight student models through selective voxel-space feature alignment. Taking advantage of the linear-time sequence model with selective state spaces (Mamba), we design a multi-branch Mamba teacher backbone and a box-aware feature transfer mechanism that aligns spatially corresponding voxel features between teacher and student networks through a Mamba-based projection module. Experimental results on both a public dataset and real-world data show that our approach significantly reduces computational load while maintaining competitive accuracy compared with state-of-the-art methods.
[CV-46] Continue or Replan? Bernoulli-Continuation Policy Learning for Adaptive Horizon Execution
链接: https://arxiv.org/abs/2608.03483
作者: Weichen Xu,Zhenhua Liu,Lin Luo,Yaobo Liang,Chengtang Yao,Qingyu Mei,Jian Cao,Xixin Cao,Xing Zhang,Jiaolong Yang,Baining Guo
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: Project page: this https URL
Abstract:Existing chunk-based Vision-Language-Action (VLA) models execute a fixed number of actions (i.e., execution horizon) before replanning, turning replanning into a task-agnostic periodic schedule that is independent of task progress. As a result, when no replanning boundary falls before a critical manipulation stage, it is executed from a stale chunk rather than a freshly replanned one. To address this limitation, we propose Bernoulli-Continuation Policy (BCP), a lightweight, plug-and-play framework for adaptive horizon execution that keeps the base VLA frozen. Given a fixed-length action chunk, its continuation head decomposes execution-horizon selection into a sequence of continue-or-replan decisions, which imposes an ordinal, prefix-sharing inductive bias over candidate horizons rather than treating them as independent classes. Since the optimal horizon for each chunk is not observable, we train this head with reinforcement learning from trajectory-level outcomes and introduce a Replanning-Efficiency Reward that jointly rewards task success and efficient VLA usage, discouraging the policy from collapsing to unnecessarily short horizons. On RoboTwin 2.0 with LingBot-VLA as the base policy, BCP improves the average success rate by +11.08% on 13 low-success tasks and from 89.88% to 93.94% (+4.06%) across all 50 tasks. Although trained only under the Clean setting, BCP generalizes to the Randomized setting, raising the average success rate by +4.06%. It also transfers to a different base policy \pi_0.5 , achieving a better result on LIBERO (+1.7%) and, notably, on the harder LIBERO-PRO (+6.8%). On a real robot, BCP lifts success from 74% to 92% and from 44% to 84% on two manipulation tasks. Meanwhile, its negligible overhead, combined with higher success, makes BCP’s overall runtime even lower than the fixed-horizon baselines.
[CV-47] MT-Web2Code: Benchmarking Coding Agents on Multi-Turn Regional Reconstruction and Localized Modification
链接: https://arxiv.org/abs/2608.03474
作者: Qiming Li,Shujie Hu,Haohan Liu,Xiaocheng Feng,Songxiang Liu,Guanglu Wan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Recent advances in Large Vision-Language Models (LVLMs) have demonstrated impressive capabilities in web UI generation. However, existing benchmarks predominantly focus on single-turn full-page generation from scratch, overlooking the iterative workflow of real-world frontend engineering, where developers repeatedly reconstruct missing regions and modify localized elements within existing codebases. To bridge this gap, we introduce MT-Web2Code, the first multimodal coding benchmark for multi-turn Macro-Level Regional Reconstruction and Micro-Level Localized Modification, which contains 102 tasks spanning 16 vertical domains. To construct deterministic repair trajectories without costly turn-level human annotation, we develop a scalable Reverse-Corruption Trajectory Engine that iteratively injects structural and stylistic defects into golden pages. We further propose a dual-axis evaluation protocol that measures target-region fidelity and the preservation of unaffected content, where regional reconstruction is assessed by a 5-dimensional VLM-based rubric and localized modification by deterministic pixel-grounded alignment. Experiments on 13 frontier coding agents reveal that current agents struggle to faithfully reconstruct target regions while preserving unaffected content, lack fine-grained visual-code alignment for localized edits, and suffer from error snowballing over multiple turns. Beyond benchmarking, our deterministic evaluation metrics provide fine-grained feedback signals that may facilitate future research on training iterative UI coding agents. Our evaluation code and data will soon be released.
[CV-48] Hi-Token: Hierarchical Coordinate Tokenization for Generative Visual Grounding
链接: https://arxiv.org/abs/2608.03471
作者: Xiuyuan Zhu,Ke Lu,Kun Dong,Siwen Jiao,Hao Wu,Zijin Du,Shun Mao,Dongming Zhang,Jian Xue
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 15 pages, 7 figures, 15 tables
Abstract:Generative Vision-Language Models (VLMs) commonly treat bounding-box coordinates as independent output symbols, leaving numerical order and axis semantics implicit. We identify this representation as an important source of error in visual grounding. Hi-Token encodes each coordinate with axis-specific tokens for the hundreds, tens, and ones digits, which adds coarse-to-fine structure and increases token reuse while retaining the existing VLM architecture. Hi-GAR complements this representation with a geometry-based reward for Group Relative Policy Optimization (GRPO), using box overlap and coordinate accuracy at multiple scales. Controlled comparisons under matched training conditions show that Hi-Token improves localization throughout the evaluated IoU range. Hi-GAR further reduces low-overlap predictions and is used only during training. Experiments on three VLM backbones and the RefCOCO family show consistent gains across models and benchmarks. Hi-R1 achieves higher values than strong specialist baselines on most reported metrics. Analyses of token frequency, digit boundaries, object scale, and IoU distributions explain the effects of coordinate representation and reward training. The results show that structured coordinate generation provides an effective approach to generative visual grounding.
[CV-49] A Low-Cost Hybrid Reservoir Computing Model for Isolated Sign Language Video Recognition
链接: https://arxiv.org/abs/2608.03444
作者: Nitin Kumar Singh,Arie Rachmad Syulistyo,Yuichiro Tanaka,Hakaru Tamukoh
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Sign language recognition (SLR) enhances communication between hearing and hearing-impaired individuals. Although deep learning (DL) has achieved promising performance in SLR, its high computational cost limits deployment on edge devices. To address this challenge, we propose a lightweight reservoir computing (RC)-based approach for SLR. In the proposed method, MediaPipe extracts body and hand keypoints to capture the spatial and temporal dynamics of gestures. These keypoints are then processed by a hybrid reservoir computing (HRC) architecture that combines deep reservoir computing (DRC) and bidirectional reservoir computing (BRC), transforming the input into a high-dimensional dynamic representation. A ridge regression model maps the final HRC state to class labels. This HRC-based SLR method achieved Top-1, Top-5, and Top-10 accuracies of 61.12%, 86.05%, and 92.56%, respectively, on the Word-Level American Sign Language 100 (WLASL100) video dataset, demonstrating competitive performance compared to deep learning-based approaches. Additionally, due to the lightweight nature of RC, the training time was drastically reduced to only a few seconds compared with DL-based methods such as this http URL method offers low computational cost, showing its potential for deployment on edge devices.
[CV-50] Stop Replacing Noise with Noise: Two-Source Reliability Assessment for Label Correction and Sample Reweighting in Label-Noise Learning
链接: https://arxiv.org/abs/2608.03432
作者: Wenxiao Fan,Kan Li
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注: preprint
Abstract:Refurbishment-based noisy-label learning mixes an observed label with a model-derived pseudo target, typically using one sample-wise cleanliness score to control both branches. This creates a hidden coupling: reducing trust in the observed label automatically increases trust in the pseudo target. We show that this complementarity can replace one unreliable signal with another because a pseudo target learned from corrupted supervision may reproduce the noise it is meant to correct. Our representation diagnostics provide a consistent account of this mismatch: noisy supervision redirects deeper layers more strongly, whereas shallower relations remain comparatively stable and provide information beyond the loss posterior. We therefore propose TRACE, a Two-Source Reliability Assessment framework for Label Correction and Sample Reweighting. TRACE assesses the observed label using loss fit, shallow relation stability, and prediction agreement, while separately assessing the pseudo target using model confidence. Its source-specific scores control target correction and supervision strength without assuming complementary reliability. Across synthetic and real-world noisy benchmarks, TRACE improves representative refurbishment baselines and yields more reliable pseudo supervision.
[CV-51] Dual-domain U-Nets with embedded back projection operators for motion-resolved 4D CBCT reconstruction
链接: https://arxiv.org/abs/2608.03430
作者: Ivo Herzig,Pascal Paysan,Daniel Barco,Marc André Stadelmann,Frank-Peter Schilling,Igor Peterlik,Michal Walczak,Lijin Aryananda,Woo Sang Ahn,Rudolf Marcel Füchslin,Lukas Lichtensteiger
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 15 pages, 9 Figures
Abstract:Four-dimensional cone beam CT (4D CBCT) is important for image-guided radiation therapy of thoracic cancers, but its use is limited by long scan times, causing high patient dose and motion/sparse-sampling artifacts. We propose a deep learning method for motion-resolved 4D CBCT reconstruction from conventional free-breathing scans, without a respiratory signal or explicit projection binning. Our CNN takes free-breathing 3D CBCT projections as input and predicts a static volume at maximum inhalation plus ten displacement vector fields (DVFs) spanning a breathing cycle. The network extends U-Net: the encoder acts on filtered projection stacks, the decoder acts in the volume domain, and skip connections are replaced with non-trainable back-projection functions at multiple resolutions to transfer features between domains. The model is trained on simulated CBCT scans and evaluated on 11 unseen simulated patients and 13 clinical free-breathing scans. Two additional models (60 s and 6 s scans) were evaluated by clinical experts on three and two scans, comparing single phases of our 4D reconstruction to reference 3D SART-TV images for tumor and esophagus visibility. Experts preferred our method for tumor visibility (59% vs. 36% no preference, 5% reference) and esophagus visibility (47% vs. 42%, 11%). On simulated data, image quality matched SART-TV (mean RMSE: -1.19 HU, PSNR: +0.09 dB, SSIM: -0.009) while enabling 4D reconstruction. On clinical scans, our method showed sharper dynamic structures (e.g., diaphragm) and fewer motion streak artifacts than traditional reconstruction. This non-patient-specific CNN predicts static volumes and full 4D respiratory motion models from a single free-breathing scan, without a respiratory surrogate or projection binning, reducing motion artifacts while adding motion-modeling capability. Comments: 15 pages, 9 Figures Subjects: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG) Cite as: arXiv:2608.03430 [cs.CV] (or arXiv:2608.03430v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2608.03430 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Ivo Herzig [view email] [v1] Tue, 4 Aug 2026 10:20:07 UTC (2,128 KB) Full-text links: Access Paper: View a PDF of the paper titled Dual-domain U-Nets with embedded back projection operators for motion-resolved 4D CBCT reconstruction, by Ivo Herzig and 10 other authorsView PDFHTML (experimental)TeX Source view license Current browse context: cs.CV prev | next new | recent | 2026-08 Change to browse by: cs cs.LG 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
[CV-52] SLAMFormer-infty: Infinite SLAM Transformer for Unbounded Frontend and Backend Processing
链接: https://arxiv.org/abs/2608.03429
作者: Zhijian Fang,Weicheng Zheng,Yijun Yuan,Weibang Wang,Zhuoguang Chen,Chang Sun,Junhao Huang,Kenan Li,Minghui Qin,Hang Zhao
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注:
Abstract:We introduce the Infinite SLAM Transformer (SLAMFormer- \infty ), the first geometric transformer capable of supporting both long-range frontend and backend processing without an explicit distance bound. Instead of relying on a first-frame-anchored formulation, SLAMFormer- \infty employs memory conditions to define flexible coordinate systems and scales for input frames, enabling more expressive structural conditioning. Built upon this formulation, the frontend preserves efficient local computation, while the backend jointly optimizes long-range trajectories and scene geometry in a globally consistent manner. Experimental results demonstrate that SLAMFormer- \infty achieves superior or highly competitive performance in both trajectory estimation and scene reconstruction across large-scale datasets. Notably, SLAMFormer- \infty generalizes to extremely long trajectories, successfully operating on sequences exceeding 17\mathrmkm .
[CV-53] OliveGemma: A 3 Billion Visual Language Model for Recognising the Mediterranean European Diet
链接: https://arxiv.org/abs/2608.03428
作者: Dimitrios I. Zaridis,Traianos Tsiokris,Vasileios C. Pezoulas,Daphni Plati,Eugenia Mylona,Eleni Georga,Nikos Tsiknakis,Antonis Sakellarios,Dimitrios I. Fotiadis
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Image based dietary assessment offers a scalable alternative to self reported food diaries, yet fine-grained food recognition remains challenging due to high intra-class variability and visually similar dishes. This study presents OliveGemma, a vision language model for recognising and reasoning about Mediterranean and European cuisine. Built on the open-weight PaliGemma-2-3B architecture, OliveGemma is fine-tuned with LoRA on a unified corpus of 17,340 images from three European research project datasets (MedGR, ODIN, and VIPPSTAR), reconciled into a vocabulary of 216 composed dish categories and paired with 102,642 instruction style question-answer items covering dish recognition, likely and visible ingredients, class boundary discrimination, visual evidence and overall visual food understanding. Under a 3-fold cross-validation scheme, OliveGemma achieves a top-1 accuracy of 92.96% +/- 0.91%, exceeding the strongest CNN baseline (DenseNet-121) by 7.31% and outperforming zero-shot frontier models with exact instructions and bounded classes including Gemini Flash 3 and 3.5, GPT-5.4 Mini, and Claude Haiku 4.6 by 8%, 46%, and 64% respectively. Furthermore, OliveGemma demonstrates competitive performance on Top-3 and Top-5 accuracy, being second best across CNNs and frontier models, surpassed only by DenseNet-121. In addition, OliveGemma achieves 90.79% +/- 1.3% Exact-Set on the likely ingredients of the food categories. These results demonstrate that PEFT adaptation of a small VLM can surpass substantially larger proprietary models on specialised food recognition. The model is publicly available at this https URL and the experiments and results can be found at this https URL.
[CV-54] SGFormer: Structure-Guided Transformer for Robust Local Feature Matching
链接: https://arxiv.org/abs/2608.03423
作者: Runyu Zhu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Local feature matching is a fundamental component of photogrammetry, enabling accurate image correspondence critical for tasks such as 3D reconstruction, stereo mapping, and visual localization. While recent detector-free matching methods, like LoFTR, have advanced the field, the global features obtained by leveraging the global-range modeling capacity of the unconstrained attention mechanism compromise the model’s attention to the salient structures in certain scenarios. This limitation leads to a phenomenon we define as attention divergence, wherein a portion of high-confidence matches are distributed outside the valid matching region (overlapping region), especially in scenes with large viewpoint variations. This occurs because similar features in irrelevant regions may receive equal weighting and consideration within the standard Transformer, limiting matching reliability in challenging photogrammetric environments. To address this issue in feature matching, we propose SGFormer (Structure-Guided Transformer), a novel structure-aware matching network that adaptively updates attention on features near salient structure in overlapping regions. SGFormer employs a semi-dense coarse-to-fine pipeline and incorporates the proposed Triple-Structure-Attention (TSA) module into the backbone net for extracting distinctive features. The TSA module utilizes shallow local features from early network layers to enhance the representation around salient structure, guiding subsequent transformer stages to intensify the model’s focus on regions with salient structure across the global scope. SGFormer, thereby reinforcing attention to visually consistent areas while mitigating the influence of non-overlapping regions. Extensive experiments show that SGFormer significantly mitigates attention divergence and improves matching accuracy.
[CV-55] HyperbolicDiffusion: Sharp Scalable Tiled Generation on the Hyperbolic Plane DATE
链接: https://arxiv.org/abs/2608.03422
作者: Hugo Caselles-Dupré
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Work in progress. Updated version incoming
Abstract:Planar tiled diffusion denoises overlapping windows of one rectangular canvas. The hyperbolic plane has no such canvas, and its area grows exponentially with radius. We introduce HyperbolicDiffusion, a training-free method for generating finite visual fields directly on the hyperbolic plane H2. Our Hyperbolic Blooming Cover reduces window placement to a compact dynamic program that runs in seconds while providing strong theoretical guarantees. Permanent surface IDs form a shared latent canvas: a standard diffusion model denoises local windows, whose predictions are fused back onto H2. Because curvature causes residual disagreement and blur at multi-window junctions, a geometry-derived second stage re-noises and repairs precisely those regions. The resulting fields are sharp, reprojectable, and consistent across viewpoints, providing a prompt-driven generative counterpart to Escher’s Circle Limit series.
[CV-56] Multi-Task Multi-Frame Visual Piano Transcription
链接: https://arxiv.org/abs/2608.03419
作者: Yonghyun Kim,Hoyeol Sohn,Juhan Nam,Alexander Lerch
类目: ound (cs.SD); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Multimedia (cs.MM); Image and Video Processing (eess.IV)
备注: Accepted to the 27th International Society for Music Information Retrieval (ISMIR) Conference, 2026
Abstract:Audio-based piano transcription performs well on onset, pitch, and velocity, but the sustain pedal lets sound persist long after key release, so audio systems predict pedal-extended offsets rather than physical key release. Yet existing Visual Piano Transcription (VPT) systems focus on onset detection from short video windows, offset accuracy lags onset by a wide margin, and note-level velocity has not been reported. To address these gaps, we present V2N (Video to Notes), the first complete VPT system: a shared temporal backbone feeds task-specific heads for onset, offset, key hold, and velocity, jointly trained with per-frame supervision rather than only at the window center. Ablations show that multi-task supervision enables offset and velocity prediction while improving onset accuracy; longer temporal context yields further improvements. V2N sets new state-of-the-art results on PianoVAM and R3.
[CV-57] Earth Embeddings
链接: https://arxiv.org/abs/2608.03410
作者: Adam J. Stewart,Heng Fang,Isaac A. Corley,Xiao Xiang Zhu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: book chapter
Abstract:Earth observation is moving from foundation models that users must run themselves toward embedding products that package model feature outputs as reusable data without needing to download and process the imagery used to generate them. Earth embeddings are vectors that summarize locations, image patches, or pixels, letting users analyze compact features instead of repeatedly training or running large models on raw satellite imagery. This chapter explains the main types of Earth embeddings, from implicit location encoders to explicit patch and pixel products, and compares their coverage, resolution, dimensionality, storage cost, licenses, and reproducibility. We review their use in land cover and crop mapping, ecological and hazard modeling, socioeconomic prediction, and semantic search, with evidence on when embeddings improve on conventional features and when pooling, fusion, or spatial transfer limit performance. Two case studies show practical workflows for similarity search and land cover mapping. We close with guidance for choosing, evaluating, storing, compressing, and publishing embeddings, and with open problems in oceanic and atmospheric coverage, uncertainty, and benchmarking.
[CV-58] Distilled Roads: Generalisable Road Network Extraction Across Sensors Resolutions and Region ECCV2026
链接: https://arxiv.org/abs/2608.03407
作者: Sanayya,Rakshith Sathish,Ashwathi Nambiar
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Accepted at ECCV 2026 workshop - TerraBytes II
Abstract:Road network segmentation from satellite imagery remains challenging due to large geographic variation in road appearance, occlusions, and domain shifts introduced by differing resolutions and sensors. Existing models, typically trained under narrow resolution–region combinations, generalise poorly to unseen environments such as rural settings, regions with distinct road materials, or imagery from new satellite platforms, often producing broken or disconnected predictions. Adapting these models to new domains usually requires retraining or fine-tuning, which is costly and risks catastrophic forgetting. In this work, we reframe global road extraction as a continual adaptation problem rather than an architectural one. Our framework combines cross-resolution knowledge distillation across a resolution-decreasing curriculum, multi-sensor training, and topology-aware supervision, yielding a single model that generalises across 0.3-1.0 m imagery from multiple satellite platforms across continents. On publicly available benchmarks, including City-Scale and Global-Scale, our model outperforms state-of-the-art results by up to 22 F1 points and 15 APLS points, while remaining the most efficient, with 3\times faster inference. Our results suggest that improved robustness across diverse sub-meter satellite imagery can be achieved through targeted training strategies, such as data curricula, distillation, and topology-aware losses, rather than increasingly complex architectures. Comments: Accepted at ECCV 2026 workshop - TerraBytes II Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG) Cite as: arXiv:2608.03407 [cs.CV] (or arXiv:2608.03407v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2608.03407 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-59] SRAP: SVD-Refined Adversarial Perturbations for Imperceptible Face-Swap Defense
链接: https://arxiv.org/abs/2608.03395
作者: Sungwon Cho,Kwanghyun Ko,Myungjoo Kang
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 13 pages, 8 figures
Abstract:Deepfake technologies pose increasing threats to facial privacy and identity security, motivating proactive defenses that protect facial images before misuse. Although adversarial perturbations generated by projected gradient descent (PGD) can disrupt the identity representations used by face-swapping models, their visual quality is degraded by two characteristics: perturbations are distributed broadly over the image, including identity-insensitive regions, and they contain visually salient high-frequency components. We analyze these spatial and spectral inefficiencies through identity-sensitivity estimation and the singular-value decomposition (SVD) of PGD perturbations. Our analysis shows that later singular components contain a disproportionate amount of high-frequency energy, while the leading components preserve most of the perturbation energy and defense utility. Based on these observations, we propose SRAP, which combines per-channel truncated SVD refinement with an identity-importance mask at every optimization step. The SVD refinement suppresses high-rank, high-frequency residuals, while the mask restricts perturbations to locations that strongly influence identity representations. Experiments on CelebA-HQ and VGGFace2-HQ demonstrate that SRAP substantially improves protected-image fidelity across all reported metrics while maintaining competitive identity-disruption performance, yielding a favorable trade-off between face-swap defense and visual imperceptibility.
[CV-60] FreqAdapt: Frequency-Adaptive Processing for RAW Object Detection
链接: https://arxiv.org/abs/2608.03385
作者: Hanxi Li,Huiling Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Existing object detection methods predominantly utilize sRGB inputs, which are compressed from RAW sensor data using Image Signal Processors (ISP) originally designed for visualization purposes. Compared to RGB images, RAW images possess favorable noise characteristics and richer information representation, which are crucial for object detection, particularly under challenging conditions such as adverse weather or low-light environments. In this paper, we propose FreqAdapt, a lightweight module for adaptive RAW data enhancement in the frequency domain. Unlike traditional spatial domain processing methods, FreqAdapt innovatively maps ISP operations to the Fourier frequency domain and performs domain separation based on the physical properties of ISP operations, ensuring each operation is performed in its most suitable domain. Meanwhile, through an adaptive frequency domain encoder that jointly analyzes amplitude spectrum, phase spectrum, and RAW image features, we provide global context for ISP parameter prediction and employ a learnable fusion mechanism to achieve adaptive feature enhancement. Extensive experiments on multiple datasets with diverse lighting and weather conditions demonstrate that FreqAdapt achieves state-of-the-art performance while maintaining lightweight efficiency and good physical interpretability. Furthermore, our module can be seamlessly incorporated into existing object detection frameworks, providing a novel solution for visual perception tasks in the RAW domain.
[CV-61] Residual Flow Matching with Dynamic Cross-Interaction for 3D Multi-Person Motion Prediction
链接: https://arxiv.org/abs/2608.03379
作者: Wei Wei,Yinyuan Zhao,Ruixuan Yu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:3D multi-person motion prediction requires modeling both individual kinematics and inter-person interactions. While Flow Matching is effective for multi-hypothesis generation to improve prediction accuracy, directly predicting skeletal sequences from pure noise often compromises structural consistency and introduces unreliable cross-agent interactions during early noise-dominated integration steps. To address this, we propose a Prior-Guided Residual Flow Matching framework. First, a Deterministic Coarse Prior (DCP) establishes a kinematic anchor, formulating the generative process as a conditional flow over motion residuals to simplify the generative objective and preserve structural stability. Second, a Dynamic Cross-Interaction (DCI) mechanism temporally synchronizes inter-agent message-passing with the integration progress, ensuring the extraction of reliable social contexts and improving multi-person motion fidelity. Finally, a decoupled joint-motion architecture with bidirectional fusion effectively preserves fine-grained kinematic coherence. Extensive experiments demonstrate that our approach achieves state-of-the-art prediction accuracy across multiple datasets. Code is available at this https URL.
[CV-62] DRPFNet: Dual-domain Residual Progressive Fusion Network for RGB-Thermal Object Detection ICME2026
链接: https://arxiv.org/abs/2608.03370
作者: Zian Wang,Changchun Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at ICME 2026
Abstract:RGB-thermal (RGB-T) object detection aims to fuse complementary information from visible and thermal modalities to achieve robust detection under varying illumination and weather conditions. Current methods typically employ attention mechanisms or transformers to perform cross-modal fusion independently at each feature scale, directly combining RGB and thermal features in the spatial domain. However, they still face significant limitations: cross-level knowledge inheritance caused by independent fusion at each scale,suppressing noise continuously due to the lack of bidirectional optimization, and information degradation induced by the absence of frequency-spatial collaboration. To address these issues, we propose DRPFNet, a Dual-domain Residual Progressive Fusion Network that constructs a unified information flow optimization system from three synergistic levels:structure, feature, and enhancement. At the structural level, we establish cross-scale propagation through bottom-up knowledge accumulation and bidirectional enhancement,ensuring smooth information flow. At the feature level, we collaboratively extract RGB high-frequency edges and thermal low-frequency structures via frequency band separation and edge guidance, guaranteeing representation quality. At the enhancement level, we enhance foreground-background discrimination through edge-guided dual-domain refinement,achieving precise object this http URL experiments on two public RGB-T datasets demonstrate that our method achieves competitive performance with competitive efficiency, validating the effectiveness of this hierarchical collaborative strategy.
[CV-63] Can Text-to-Image Models Draw from the Right Frame of Reference?
链接: https://arxiv.org/abs/2608.03357
作者: Zheyuan Gu,Ruihang Li,Yong Huang,Yiqian Zhang,XIangzhao Hao,Jiaxin Niu,Jiahao Hu,Zhenyu Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 9 pages, 3 figures
Abstract:Spatial instruction following has become a crucial requirement for text-to-image (T2I) generation. A common challenge arises when directional expressions are interpreted under different frames of reference. For example, ``the left of’’ may refer to the viewer’s image coordinates or to the intrinsic orientation of an object, leading to different expected layouts. Existing T2I benchmarks reveal important layout failures, yet they rarely isolate whether models can follow a specified frame of reference when it differs from camera view. To mitigate this gap, we introduce FoR-T2I, a benchmark for evaluating this distinction with 1,200 prompt pairs built from controlled spatial layouts. In each pair, the camera-view (Cam) prompt states the target relation in camera view, while the frame-of-reference (FoR) prompt describes the same target placement through an oriented anchor object. Across 22 closed-source and open-source T2I models, mean final accuracy is 41.8% lower on FoR prompts than on matched Cam prompts; even the best-performing model achieves only 44.3% FoR accuracy. This suggests that current models struggle more when the same layout is described through an object’s orientation rather than directly in image coordinates. We further analyze this gap by relation type and camera view, compare several training-free prompting and feedback-based mitigation strategies, and propose a VLM-gated rewriting approach that selects rewritten prompts using visual feedback, improving average FoR accuracy from 25.0% to 29.2% under the same generation budget.
[CV-64] When Oracle Conditioning Misleads Deployment: Conditioning-Availability Bias in Echocardiographic Segmentation MICCAI2026
链接: https://arxiv.org/abs/2608.03342
作者: Dang P. M. Cao,Hieu D. Pham,Hieu Pham
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted for publication in the MICCAI 2026 Workshop on Fairness of AI in Medical Imaging (FAIMI 2026). To appear in Springer Lecture Notes in Computer Science (LNCS)
Abstract:Conditional segmentation models may be trained and evaluated with auxiliary signals cleaner than those available at deployment. We study this protocol-level manifestation of shortcut learning and auxiliary-variable shift in phase-conditioned echocardiographic segmentation. The complementary gap pair measures loss on the deployable oracle-estimated pathway and probes sensitivity on the oracle-random pathway. On held-out CAMUS data, one strong-cyclic, oracle-selected run fails severely with estimated phase, while sensitivity to incorrect phase persists across three runs. On EchoNet-Dynamic, the current estimator remains usable, but random-phase testing reveals strong latent sensitivity. Deployment-aware checkpoint selection and phase perturbation reduce both gaps with little change in mean Dice. Exploratory subgroup analyses quantify variation across measured strata, and a downstream ejection fraction (EF) audit shows that recovering segmentation does not necessarily recover EF error or signed bias. Together, the gaps test whether oracle-conditioned performance survives the inference pathway actually available at deployment.
[CV-65] SPADE: An Input-Adaptive Sparse Attention Engine for Fast Video Diffusion Models Inference
链接: https://arxiv.org/abs/2608.03335
作者: Shanghao Liu,Renze Chen,Size Zheng,Yuanqiang Liu, Yun (Eric)Liang,Hailong Yang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Published in the 63rd ACM/IEEE Design Automation Conference (DAC '26). 7 pages, 6 figures, 3 tables
Abstract:Video diffusion transformers (vDiTs) generate high quality but pay quadratic self-attention cost, making inference prohibitive at video-token scales. The challenge is input-adaptive sparsity: selecting critical Q/K/V tokens with negligible overhead and executing them for end-to-end gains. We present SPADE, a training-free sparse-attention engine of three parts: (i) vDiT-SSR, a specification defining 3D blocking candidates and formalizing dynamic masks via Summarizer/Estimator expressions; (ii) runtime scheme generation using SICS and a head-wise policy; and (iii) an executor with low-overhead index search, flash block-sparse attention, and kernel grouping. Across Hunyuan-Video and Wan 2.1/2.2 for text-to-video and image-to-video generation, SPADE raises sparsity and speed while preserving quality, accelerating attention by 2.26x-3.40x and end-to-end inference by 1.49x-1.80x. Our code is open-sourced at this https URL.
[CV-66] PolyLayout: Multi-room Manhattan Layout Estimation ECCV
链接: https://arxiv.org/abs/2608.03323
作者: Gustav Hanning,Shaohui Liu,Rémi Pautrat,Marc Pollefeys,Kalle Åström,Viktor Larsson
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at the European Conference on Computer Vision (ECCV) 2026
Abstract:Estimating room layouts from multi-view imagery is a core task for indoor scene understanding. Existing methods are typically limited either by poor generalization to new datasets or restrictive geometric assumptions of the room shape or camera configuration. Most also estimate rooms independently, failing to exploit shared building structure such as dominant directions, ground plane or ceiling height. We propose PolyLayout, a multi-room layout estimation method that parameterizes room layouts as Manhattan 3D polygons and optimizes them jointly across multiple rooms. The optimization objective is predicted by a neural network on top of robust pre-trained visual features and trained end-to-end with supervision only on output room layouts. At the same time, camera projection and polygon updates remain explicit and model-based. This separation between learned scoring and geometry improves generalization to new datasets and camera parameters. During optimization, PolyLayout adaptively refines the polygon topology through iterative wall split and merge operations while jointly utilizing structural cues across rooms. We introduce two new multi-view multi-room layout benchmarks by providing layout annotations to existing datasets, and experiments show that PolyLayout outperforms prior approaches, both in terms of accuracy and robustness. Project page: this https URL Comments: Accepted at the European Conference on Computer Vision (ECCV) 2026 Subjects: Computer Vision and Pattern Recognition (cs.CV) ACMclasses: I.4 Cite as: arXiv:2608.03323 [cs.CV] (or arXiv:2608.03323v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2608.03323 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-67] LocAnyMed: Vision-Language Grounding for Multimodal Medical Images
链接: https://arxiv.org/abs/2608.03322
作者: Zihan Wang,Tong Liu,Zhiwei Wang,Tao Huang,Wentao Jiang,Sihan Ma,Shanshan Ye,Xiaohui Yang,Jing Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Technical report; work in progress. 28 pages, 5 figures, and 16 tables. Code: this https URL
Abstract:Medical visual grounding connects free-form clinical queries to spatial evidence in medical images and is an important component of interpretable medical artificial intelligence. However, general-purpose grounding models are predominantly trained on natural images, while existing medical localization resources remain fragmented across imaging modalities, datasets, and task formulations. To address this gap, we construct LocAnyMed-200K, a multimodal medical visual grounding dataset containing approximately 200K image-query-answer examples across computed tomography, optical medical imaging, ultrasound, and X-ray. We harmonize heterogeneous detection and localization resources into a unified free-form instruction format that supports one or multiple bounding boxes, point coordinates, and no-target outputs for negative queries. Full-parameter fine-tuning of LocateAnything-3B on LocAnyMed-200K improves F1@IoU 0.50 from 10.64 to 85.59 on a held-out evaluation split, demonstrating that large-scale domain-specific supervision can equip a general grounding model with effective medical localization capabilities. Beyond spatial coordinates, a clinically interpretable grounding system should also communicate the evidence supporting its prediction. We therefore derive LocAnyMed-CoT-20K, a rationale-augmented subset that connects anatomical context, visual observations, and spatial conclusions through structured reasoning and further improves cross-source generalization through fine-tuning. Together, these resources provide a unified foundation for studying both localization accuracy and rationale quality across heterogeneous medical imaging modalities. The code is publicly available at this https URL.
[CV-68] Any-OPD: Heterogeneous On-Policy Distillation for Flow-Matching Models via Representation-Space Bridging
链接: https://arxiv.org/abs/2608.03316
作者: Siming Fu,Zheming Fu,Ruizhe He,Hualiang Wang,Jie Huang,Xiaoxiao Ma,Mingchen Zhong,Weihu Huang,Xiaoxuan He,Haojun Xu
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:On-policy distillation, in which a teacher corrects samples that the student itself generates, presupposes that the two models speak the same language: identical VAE latents, matching architectures, and a common timestep grid. We ask what happens when none of this holds, as when the strongest teacher available and the student one wishes to deploy come from different model families, and find that the standard recipes have no answer: teacher latents cannot serve as targets in a foreign coordinate system, per-pixel losses against a teacher that stochastically re-draws local detail degenerate into blur or divergence, and timestep indices lose their meaning across mismatched schedules. We present Any-OPD, to our knowledge the first framework for on-policy distillation between arbitrary pairs of latent flow-matching generators. Any-OPD treats the teacher purely as a black-box sampler and connects the two models at exactly one point: a frozen, model-agnostic vision representation in which their independently decoded outputs are compared, sidestepping every assumption about latents, features, or architecture. Trajectory correspondence is recovered by matching continuous noise levels instead of step indices, and a brief anchoring phase, in which teacher samples are re-encoded through the student’s own VAE, ensures the on-policy gradient measures sample quality rather than domain mismatch. Distilling the 12B FLUX.1-dev into the 2.5B SD3.5-Medium, Any-OPD lifts the student’s PickScore from 0.846 to 0.884 and HPSv3 from 9.12 to 10.97, rivaling the teacher at a fifth of its size, where direct latent regression fails to train at all.
[CV-69] Recurrent Contrastive Learning for Imbalanced Medical Image Classification
链接: https://arxiv.org/abs/2608.03304
作者: Zhiyuan Zhu,Xinling Meng,Junxuan Yu,Jiongquan Chen,Qiongying Ni,Tuhang Shao,Yuhao Huang,Luping Zhou,Ruiyang Huang,Yuxue Wang,Rongliang Zhang,Xue Wang,Tianhong Tang,Likun Wang,Junbo Chen,Yong Jiang,Yongping Lu,Xin Yang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 10 pages, 3 figures
Abstract:Medical image classification often suffers from class imbalance due to the inherent disparities in disease incidence. Existing approaches, such as class resampling and loss reweighting, mainly improve learning within the observed feature distribution, but do not explicitly enlarge the latent support region of tail classes. As a result, tail-class representations remain overly compact and are easily encroached upon by head classes, leading to biased decision boundaries. In this work, we propose Recurrent Contrastive Learning (RCL) for imbalanced medical image classification. RCL progressively expands the support region of tail classes by recurrently reusing historical feature states across training phases. Specifically, we adopt DINOv3 with LoRA adapters as the backbone to provide robust feature embeddings. We then devise a Temporal Memory Queue (TMQ) to preserve corpus-level features across training phases and provide diversified global references for contrastive learning. Based on TMQ, we construct Temporal Anchors (TARs) to form an anchor field around tail classes. This field enlarges the support region of tail classes, suppresses head-class encroachment, and improves inter-class separation. Extensive experiments on three imbalanced medical datasets demonstrate that RCL achieves consistent improvements over strong baselines. The code is available at this https URL.
[CV-70] PLS-Calib: A Partial Least Squares Framework for Event Camera and Odometry Calibration under Ground Motion Constraints IROS2026
链接: https://arxiv.org/abs/2608.03296
作者: Guangyu Li,Xiao Li,Yujie Wu,Changshuo Wang,Prayag Tiwari,Jiang Cai,Fangwen Yu,Mingkun Xu
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注: 8 pages, 10 figures, 4 tables. Accepted at the 2026 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2026)
Abstract:Accurate extrinsic rotation calibration between sensors is fundamental to the performance of robotic perception systems. However, most existing calibration techniques rely on full 6-DoF motion to excite all degrees of freedom, which is often infeasible for ground-constrained robots with limited motion capabilities. Recent approaches designed for such restricted settings, such as Canonical Correlation Analysis (CCA)-based methods, suffer from ill-conditioned covariance matrices that lead to numerical instability and suboptimal calibration accuracy. To overcome these limitations, we present a novel rotation calibration framework named PLS-Calib that, for the first time, leverages Partial Least Squares (PLS) regression to model the latent kinematic correlations between asynchronous, heterogeneous sensor streams. Specifically, we apply our method to the calibration of an event camera and an odometry onboard a ground robot. To improve event-based pattern detection, we introduce a polarity-aware event representation, which enhances spatiotemporal contrast in circular calibration targets. Our PLS-based formulation yields a closed-form, stable solution that avoids matrix singularities inherent in CCA-based approaches. Extensive experiments on both synthetic and real-world datasets validate the effectiveness of our approach, demonstrating significant improvements in calibration robustness and accuracy over state-of-the-art methods. This work offers a practical and theoretically grounded solution for rotation calibration in constrained robotic systems and opens up new directions for applying statistical learning techniques in neuromorphic vision.
[CV-71] st-Time Scaling for Safe Text-Guided Image Generation via Intermediate Clean Estimates
链接: https://arxiv.org/abs/2608.03284
作者: Jinya Sakurai,Shueicheng Yan,Xun Xu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Ensuring safety and policy compliance in text-to-image diffusion models remains a critical challenge, as benign or adversarial prompts can often elicit prohibited content, e.g. nudity and protected intellectual property. While training-based unlearning methods are effective, they are computationally expensive and prone to catastrophic interference with general capabilities. Conversely, existing test-time defenses are primarily prompt-centric, relying on modifying textual descriptions only, and overlook the visual signals for detection. In this paper, we propose to leverage the intermediate clean image estimated during the generation process and employ a sparse margin objective to detect prohibited concepts. When a violation is detected, we immediately intervene by optimizing a structured low-rank residual in the text-conditioning space via truncated backpropagation. This design allows weight-preserving detection, keeps non-violating inference latency nearly unchanged as the maximum budget increases, and offers flexibility in safety performance via test-time scaling. Extensive experiments on Stable Diffusion v1.4 and v3.5 across nudity removal, IP protection, and style erasure demonstrate superior performance across suppression, fidelity and preservation compared to prior weight-preserving baselines, providing a scalable and flexible solution for safe generative deployment.
[CV-72] 3DGSI-Assessor: A Large-Scale Dataset and An LMM-based Method for 3D Gaussian Splatting Image Quality Assessment
链接: https://arxiv.org/abs/2608.03279
作者: Yuke Xing,Jiarui Wang,William Gordon,Zhu Li,Guangtao Zhai,Yiling Xu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:3D Gaussian Splatting (3DGS) has become a dominant representation for real-time novel view synthesis (NVS), yet its storage footprint makes compression indispensable for practical deployment. 3DGS training and compression introduce representation-specific distortions such as floating artifacts and surface scattering, which conventional image quality assessment (IQA) metrics fail to capture. Moreover, the independent compression of geometric and color attributes may lead to decoupled dimension-specific distortions that must be diagnosed separately, yet existing metrics report only a single overall score. To address these gaps, we present 3DGS-IEval-15K+, a large-scale, multi-dimensional IQA dataset for compressed 3DGS, comprising 15,200 images from 10 diverse scenes, produced by 6 representative 3DGS algorithms at systematically designed compression levels and rendered from 20 strategically selected viewpoints spanning both training views and challenging novel views, annotated with 45,600 mean opinion scores (MOSs) across overall, geometry, and color quality. Based on 3DGS-IEval-15K+, we propose 3DGSI-Assessor, an all-in-one 3DGS IQA framework that integrates global semantic and dimension-specific local features within a large multimodal model (LMM), predicting all three dimensions in a single forward pass. 3DGSI-Assessor achieves state-of-the-art performance on 3DGS-IEval-15K+, and exhibits competitive generalization on other NVS benchmarks. Dataset and code will be released at this https URL.
[CV-73] GUI-Lens: Coarse-to-Fine Cropping for GUI Grounding with General-Purpose VLMs
链接: https://arxiv.org/abs/2608.03270
作者: Zichuan Fu,Shirong Wang,Wenlin Zhang,Guojing Li,Yimin Deng,Jingtong Gao,Junjia Qi,Hanyu Yan,Yefeng Zheng,Xiaopeng Li,Wanyu Wang,Xian Wu,Xiangyu Zhao
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Preprint. Code: this https URL
Abstract:GUI grounding maps natural-language instructions to click locations and is essential for reliable GUI agents. The task remains difficult on high-resolution, densely populated interfaces because a vision-language model (VLM) may recognize a requested control without locating it precisely enough for interaction. Most existing methods provide various forms of localization assistance, but still rely on a direct click prediction, allowing visual ambiguity or an inaccurate initial estimate to propagate to the final result. In this paper, we introduce GUI-Lens, a coarse-to-fine grounding framework that allows a general-purpose VLM to determine the target through active visual observations. Specifically, GUI-Lens extracts OCR text and detected UI components from the screenshot and presents their positions as coordinate references. Using the instruction, the current view, and these references, the VLM selects the region and scale of the next view, which is cropped and enlarged to provide finer visual details. This process continues over successively focused views until the target is determined. Proposed crops and clicks are checked against the instruction throughout the process, and the final local position is mapped back to the original screen coordinates. Experiments on four GUI grounding benchmarks and three general-purpose VLM backends show that GUI-Lens improves overall grounding accuracy by up to 24.9 percentage points and achieves state-of-the-art performance with GPT-5.5.
[CV-74] Efficient Video Dataset Distillation via Cluster-Guided Prototype Blending
链接: https://arxiv.org/abs/2608.03269
作者: Chongle Ren,Guang Li,Wenbo Huang,Naoki Saito,Takahiro Ogawa,Miki Haseyama
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Video dataset distillation aims to compress a large video dataset into a compact surrogate set that preserves its training utility. Most existing approaches synthesize condensed videos through iterative optimization, whose cost is amplified by the temporal dimension. Rather than further reducing the number of optimized variables, we investigate whether effective distilled videos can be constructed without gradient-based optimization of the stored videos. Such a construction-based approach must address three challenges: selecting informative temporal segments, covering diverse intra-class variations under a limited videos-per-class budget, and increasing the information carried by each stored sample. To this end, we propose ProtoBlend, an efficient select-allocate-blend framework. First, teacher-guided temporal clip selection retains a high-confidence segment from each source video. Second, cluster-guided prototype allocation partitions the selected clips in the teacher feature space and assigns one distilled slot to each intra-class cluster. Third, each prototype is blended with an in-cluster anchor, while their teacher predictions are combined using the same coefficient to provide mixture-source supervision. Experiments on four trimmed action-recognition benchmarks demonstrate that ProtoBlend achieves a competitive accuracy-efficiency trade-off without iterative optimization of the distilled videos.
[CV-75] Hear to See: Discerning Stateful Listening for Audio-Visual Instance Segmentation ACM-MM2026
链接: https://arxiv.org/abs/2608.03264
作者: Leiye Liu,Miao Zhang,Jiahong Jiang,Jingjing Li,Jialong Zhong,Kai Peng,Tingwei Liu,Wei Ji,Yongri Piao,Huchuan Lu
类目: Multimedia (cs.MM); Computer Vision and Pattern Recognition (cs.CV); Sound (cs.SD)
备注: Accepted by ACM MM 2026
Abstract:Audio-visual instance segmentation (AVIS) requires accurately identifying and tracking individual sounding objects with pixel-level masks. Existing methods struggle to match overlapping acoustic events with visual instances and handle asynchronous audio-visual dynamics. Therefore, two critical questions arise: how can a model establish precise correspondence between overlapping sound sources and visual instances, and how can a model maintain robust tracking when audio and visual signals are temporally misaligned?This paper proposes Hear to See (H2S), addressing these challenges through two mechanisms. The Acoustic-Semantic Projector (ASP) disentangles mixed audio and establishes hierarchical correspondence from semantic to spatial domains. The Asynchronous Dynamics Modulator (ADM) adaptively adjusts state transitions via audio-modulated Mamba, prioritizing current information during dynamic variations and maintaining continuity in stable this http URL on AVISeg show H2S achieves SOTA performance, attaining 48.54 mAP with a COCO pretrained ResNet50 and surpassing the previous by 7.8%. The code will be open-sourced once the paper is accepted. The source code will be publicly available at this https URL.
[CV-76] NanoMorph-3D: An End-to-End Physics-Driven Unrolling Framework for Nanomaterial Reconstruction
链接: https://arxiv.org/abs/2608.03257
作者: Beiyuan Zhang,Hesong Li,Ziqi Wu,Ruiwen Shao,Ying Fu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Precise 3D characterization of nanomaterials is essential for unlocking structure-property relationships. However, standard electron tomography is fundamentally limited by the missing wedge problem. Consequently, conventional algorithms suffer from severe geometric distortions, a challenge further complicated by pervasive noise interference. Current learning-based methods either rely on physics-blind post-processing or employ end-to-end architectures constrained by local receptive fields, failing to capture complex 3D topologies. We propose NanoMorph-3D, a unified end-to-end framework grounded in a comprehensive Nanomorphological Taxonomy. Powered by a large-scale synthetic dataset explicitly modeling non-linear electron attenuation, we design a Physics-Driven Unrolled Network mapping proximal gradient descent into a learnable architecture. To capture complex internal topologies, we formulate a hierarchical attention mechanism with Physics-Normalization for long-range 3D dependencies and scale invariance. Crucially, our Dual-Domain strategy leverages Sinusoidal Attention to explicitly model physical projection trajectories, enforcing strict sinogram consistency to mitigate missing wedge artifacts. Finally, an unsupervised dual-stream mechanism bridges the simulation-to-reality gap. Experiments demonstrate NanoMorph-3D reconstructs diverse topologies with superior fidelity and speed.
[CV-77] Clarity Contrast and Similarity Selection for Multi-Focus Image Fusion
链接: https://arxiv.org/abs/2608.03252
作者: Yicheng Zhang,Haoyou Deng,Zhiqiang Li,Wenti Yin,Nong Sang,Changxin Gao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Multi-focus image fusion (MFIF) aims to generate an all-in-focus image from multiple images of the same scene focused at different regions. Most existing deep learning-based methods lack explicit interaction between the source images, which limits their performance and interpretability. This paper presents a novel Clarity Contrast and Similarity Selection Network (CSNet), to bridge direct information exchange for MFIF. Specifically, by contrasting the clarity differences between source images within our proposed Clarity Contrast Attention Module (CCAM), we mutually enhance sharp features while suppressing blurry ones. This allows us to identify the exactly focused regions in each source and locate the focused-defocused boundaries. Moreover, the Defocus Spread Effect (DSE) degrades pixels in all source images around the boundaries. To further refine these ambiguous areas, we introduce a Similarity Selection Strategy, which reconstructs an initial clear image from source images and selects optimal pixels by comparing the similarity among them. Through this interactive approach, CSNet effectively preserves focused regions as well as recovering natural boundaries to fuse an all-in-focus output. Extensive experiments demonstrate that our method achieves state-of-the-art performance both quantitatively and qualitatively. Our code is available on Github: this https URL.
[CV-78] Open-Linguistic Concept Unified Learning for Cross-Site Interpretable Dermatology Image Diagnosis
链接: https://arxiv.org/abs/2608.03225
作者: Chengyu Wu,Junpeng Tan,Wanxiang Luo,Yaqi Wang,Yandong Wen,Yefeng Zheng
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: accepted by ACM Multimedia 2026
Abstract:Human-interpretable computer-aided diagnosis is crucial for clinical decision making. Concept-based models excel by providing transparent reasoning and enabling post-hoc, clinician-in-the-loop interventions. However, their rigid dataset-specific adaptation inherently restricts cross-site generalization. Applying them across diverse modalities, such as dermoscopic and clinical photographs, is challenging due to heterogeneous concept taxonomies varying in availability, granularity, and semantics across cohorts. Consequently, adapting Foundation Vision-Language Models (FVLMs) demands costly label engineering and repeated post-training. Existing intervention mechanisms remain rigidly tied to predefined concepts, lacking adaptability and hindering scalable dermatology CAD deployment. To address these bottlenecks, we propose UniCon, an open-linguistic unified concept learning framework for multimodal interpretable vision-language diagnosis. UniCon resolves these challenges through three contributions: (1) A shared semantic representation space via a unified concept prototype codebook, seamlessly coordinating heterogeneous concept systems across modalities without dataset-specific retraining. (2) Open-linguistic based multi-faceted semantic specifications to overcome sparse textual label limitations, improving boundary sensitivity in uncertain clinical contexts. (3) A robust, cross-site adjustable intervention interface powered by reliability-gated bottleneck aggregation, enabling consistent reasoning and transferable clinician corrections. Extensive experiments demonstrate that beyond securing top-tier diagnostic accuracy, UniCon successfully bridges disparate clinical taxonomies, unlocking unprecedented cross-site intervention capabilities. Code is available at this https URL.
[CV-79] Self-Supervised Representation-Guided Generative Dataset Distillation
链接: https://arxiv.org/abs/2608.03218
作者: Mingzhuo Li,Guang Li,Linfeng Ye,Jiafeng Mao,Takahiro Ogawa,Konstantinos N. Plataniotis,Miki Haseyama
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Dataset distillation compresses a large training set into a compact synthetic set while retaining its downstream utility. Most existing methods target randomly initialized networks, whereas modern vision systems often adapt frozen pretrained encoders with lightweight modules. Distilled samples should therefore preserve the discriminative geometry of the pretrained representation space, which existing generative objectives do not explicitly consider. We propose self-supervised representation-guided generative dataset distillation (SRG), a framework that translates the SSL geometry into diffusion guidance. Specifically, SRG constructs class-wise prototypes from real-image SSL representations and performs guidance through three SSL-space objectives for prototype alignment, inter-class discrimination, and intra-class assignment. During diffusion sampling, it adopts a stage-wise guidance strategy: early denoising is anchored to the latent of the real image whose SSL representation is nearest to the assigned prototype, whereas later denoising is guided by the SSL-space objectives. This division preserves the visual realism provided by the generative prior while progressively steering samples toward representative and class-discriminative regions of the SSL representation space. SRG consistently outperforms the evaluated generative baselines across multiple datasets and IPC settings. A cross-encoder evaluation further indicates transfer across pretrained representation spaces. These results demonstrate the effectiveness of representation-guided generation for dataset distillation with pretrained SSL models.
[CV-80] FAN: Inference-Aware Learning for Plain Mask Transformers
链接: https://arxiv.org/abs/2608.03216
作者: Fang Li,Yu He,Haoyang Tong,Lichen Ma,Jingling Fu,Wenxiao Fan,Tongxuan Liu,Luohang Liu,Ke Zhang,Junshi Huang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Project Page this https URL
Abstract:Query-based mask transformers assemble segmentation outputs through pixel-wise competition among query predictions of the final layer, yet this inference process is not explicitly optimized during training. We identify two key mismatches: the query with the highest probability-mask score does not necessarily produce the most accurate mask, and final-layer decoding may discard superior predictions from intermediate layers. To address these issues, we propose Inference-Aware Learning (iFAN), a general training framework for plain mask transformers. iFAN introduces Adjusted Probability-Mask Ranking (APMR), which aligns query competition with predicted mask quality and suppresses high-confidence but inaccurate competitors. We further employ Cross-Layer Self-Distillation (CLSD) to transfer stronger intermediate predictions to the final layer. The ranking and distillation objectives are training-only, while inference retains efficient final-layer decoding. Experiments on COCO, ADE20K, and Cityscapes demonstrate consistent improvements across panoptic, instance, and semantic segmentation, as well as across different architectures, backbone scales, and input resolutions. Overall, iFAN improves performance by an average of 1.20 PQ, 1.30 AP, and 0.63 mIoU, with negligible additional parameters, FLOPs and inference latency.
[CV-81] CrossScope: A Role-Asymmetric World Model for Joint Dual-Scope Surgical Video Prediction
链接: https://arxiv.org/abs/2608.03211
作者: Wanhao Liu,Jinsong Lin,Rulin Zhou,Chi Kit Ng,Wenbin Pan,Zhiqing Tang,Dongyue Li,Liwei Luo,Yanshen Wu,Panshuo Li,Zhiyong Xiong,Huxin Gao,Tamas Haidegger,Hongliang Ren
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注:
Abstract:Visual world models typically learn future dynamics from a single observation stream, limiting their ability to model cooperative systems with multiple independently moving observers. We investigate this challenge in Mother–Child endoscopic retrograde cholangiopancreatography (ERCP), where two flexible scopes provide complementary yet role-dependent views without a calibrated stereo relationship. Unlike conventional multi-view fusion that assumes symmetric information exchange, we formulate \textbfrole-asymmetric dual-scope future prediction, where cross-view evidence is selectively transferred according to the prediction target and its underlying spatial requirements. We propose \textbfCrossScope, a dual-stream surgical world model that preserves view-specific experts while enabling target-specific evidence routing through geometry-guided residual interactions. CrossScope learns two complementary communication directions: geometric motion cues from the Mother view guide Child-view future dynamics, while pose-aligned Child appearance supports Mother-view prediction only when valid spatial correspondence is established. This design allows each scope to contribute task-relevant evidence without compromising its view-specific representation. To evaluate this problem, we establish a paired dual-scope benchmark comprising synchronized phantom and real-world ERCP episodes, with evaluations assessing visual fidelity, structural preservation, target localization, and motion consistency. Experiments demonstrate that CrossScope consistently outperforms strong surgical video generation baselines, validating the importance of role-aware evidence routing for multi-observer visual world modeling.
[CV-82] DRIFT: Derailing Denoising Trajectories of Flow-Matching VLAs with Adversarial Patch Attack
链接: https://arxiv.org/abs/2608.03207
作者: Hoseong Tae,Jong-Seok Lee
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Flow-matching vision-language-action (VLA) models such as pi0 generate robot actions by integrating a learned denoising velocity field, and have been reported to resist adversarial perturbations that readily fool autoregressive VLAs. We show that this robustness is largely illusory: it stems from prior attacks ignoring the multi-step denoising ODE. We introduce DRIFT (Denoising Redirection via Input perturbation of the Flow-matching Trajectory), a test-time universal adversarial patch placed on the robot’s gripper that attacks the denoising velocity field of an off-the-shelf policy. Our central finding is counterintuitive: attacking only the first denoising step is both stronger and cheaper than attacking a wider window of steps, which we explain through a gradient conflict unique to input-space optimization and which is exactly opposite to the training-time backdoor regime. On pi0 and pi0.5 across four LIBERO suites, DRIFT breaks essentially all originally-solvable tasks with a small single patch, far exceeding action- and embedding-space attack baselines.
[CV-83] Bridging Online and Offline Handwriting via Differentiable Physical Rendering ECCV2026
链接: https://arxiv.org/abs/2608.03198
作者: Seonmi Park,Seunghyun Shin,Vihaan Misra,Dongmin Shin,Ukcheol Shin,Jean Oh,Hae-Gon Jeon
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注: Accepted at ECCV 2026, Project page: this https URL
Abstract:Realistic handwritten text generation plays an important role in numerous applications, such as font design, biometric authentication, and robotic calligraphy. Existing methods are typically divided into two independent paradigms: online approaches that estimate handwriting trajectories and offline approaches that synthesize realistic handwriting images. While online models capture structural and temporal dynamics, they often lack fine-grained textures, whereas offline models reproduce realistic appearance but discard stroke order. However, unifying online and offline models remains challenging due to (1) the lack of an explicit physical model linking stroke kinematics to pixel-level appearance and (2) the absence of paired trajectory-image datasets. Moreover, enabling end-to-end learning requires a differentiable rendering process across motion and appearance domains. To address these challenges, we propose a compact physical brush model that bridges stroke dynamics and visual appearance, together with a differentiable rendering module that converts stroke trajectories into stylized images. By integrating these components, we propose a unified online-offline handwriting generation framework via differentiable brush rendering. The proposed framework consists of four core modules: 1) a text-to-stroke generator that predicts the target stroke conditioned on the given text and style image, 2) a brush parameter observer that extracts brush model parameters from style references, 3) a differentiable brush renderer that maps a stroke sequence and physical brush parameters into a handwritten image, and 4) a zero-shot image refiner that refines rendered images via diffusion models. Extensive experiments and real-world robotic calligraphy demonstrations validate our approach, achieving both structural and visual fidelity.
[CV-84] CRIL-U-Net: Compact Ratio-Interaction Learning for Focal Cortical Dysplasia Segmentation from T1w and FLAIR MRI
链接: https://arxiv.org/abs/2608.03185
作者: Soumen Ghosh,Amit Soni Arya,Tilottama Goswami,Subhojit Mandal,John Phamnguyen,Rajat Vashistha
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Focal cortical dysplasia (FCD) type II is an important structural cause of drug-resistant focal epilepsy, but its small size, heterogeneous appearance, and subtle MRI characteristics make automated segmentation challenging. Conventional multimodal networks commonly concatenate T1-weighted (T1w) and fluid-attenuated inversion recovery (FLAIR) images, requiring subsequent layers to learn useful cross-modal relationships implicitly. We propose CRIL-U-Net, a 3D U-Net incorporating a Compact Ratio-Interaction Learning module that combines local spatial features, voxel-wise cross-modal mixing, and bidirectional ratio-inspired interactions. CRIL-U-Net was compared with a conventional 3D U-Net and an input self-attention U-Net using five-fold cross-validation on 85 FCD subjects and 25 healthy controls. Each architecture was trained independently using Dice-binary cross-entropy (Dice-BCE) and Focal Tversky-Focal (FTF) losses. With FTF, CRIL-U-Net achieved the highest mean Dice score (0.196 +/- 0.262), compared with 0.136 +/- 0.224 for the U-Net and 0.135 +/- 0.214 for the attention comparator. It produced nonzero lesion overlap in 44 of 85 cases, compared with 36 for the U-Net. Under FTF, CRIL-U-Net significantly outperformed both comparison architectures after false-discovery-rate correction. These findings suggest that compact cross-modal representation learning can improve FCD segmentation within a controlled U-Net setting when combined with an imbalance-aware objective, although the remaining zero-overlap rate of 48.2% highlights the need for further validation and methodological development.
[CV-85] EditFlow3D: Automated Local Editing of 3D Assets with Trajectory Preservation
链接: https://arxiv.org/abs/2608.03179
作者: Rui Nie,Chuang Wang,Haitao Zhou,Jiahe Song,Buyu Li,Sheng Wang,Qian Yu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Controllable local editing of 3D assets requires precise target localization and appropriate visual guidance. However, existing methods lack a simple yet accurate way to obtain 3D masks and struggle to achieve the desired edit while faithfully preserving the structure and appearance of non-target regions. To address these challenges, we present EditFlow3D, a training-free framework for local 3D editing. Given a source asset and an edit instruction, a VLM-driven workflow interprets the editing intent and automatically constructs a visual guidance image and a refined 3D editing mask, enabling localized editing in the native representation space of a pretrained 3D generative model. Specifically, mask-guided differential flow focuses the edit on the target region, while step-wise trajectory preservation maintains consistency between non-target regions and the source asset without directly replacing intermediate features. Since the existing Edit3D-Bench covers only a limited range of local editing categories, we further introduce EditFlow-Bench as a complementary benchmark encompassing a broader variety of structural and appearance edits, and evaluate EditFlow3D on both benchmarks. Quantitative results, qualitative comparisons, and a user study demonstrate that EditFlow3D achieves more accurate target-region editing and better preserves non-target regions than existing 3D editing methods.
[CV-86] SpreadMark: Robust Image Watermarking via Spread-Spectrum Embedding
链接: https://arxiv.org/abs/2608.03165
作者: Wei Song,Yuxin Cao,Zhenchang Xing,Liming Zhu,Jin Song Dong,Yulei Sui,Jingling Xue
类目: Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV)
备注: 12 pages, 6 figures
Abstract:Invisible image watermarks are increasingly used for deepfake detection and provenance tracking, where they must survive not only incidental distortions but also deliberate removal. We revisit spread-spectrum embedding, a classical watermarking principle, inside a modern neural post-hoc watermarking architecture. Our starting point is a measurement: in existing encoder-decoder schemes each message bit occupies only a small fraction of the image, a shared contributing factor to their fragility, since removal then need only disturb the region a bit occupies. SpreadMark instead spreads each bit as a dense pseudo-random codeword over the whole image and recovers it by matched-filtering a learned cover-suppressed chip representation, with a parallel convolutional decoding path and sparsification-aware training. A conditional chip-space analysis shows that, under a codeword-independent perturbation model, dense spreading increases the budget required to disrupt matched-filter recovery. Evaluated on COCO and DIV2K against nine schemes, SpreadMark is the only evaluated method retaining high detection under both the regeneration and the latent-space sparsification settings we test, with competitive JPEG and additive-noise robustness. It keeps the embedded watermark imperceptible, maintaining high perceptual quality on both COCO and DIV2K.
[CV-87] Caved or Convinced: Temporal Sampling Gates Claim Deference in Video Large Language Models
链接: https://arxiv.org/abs/2608.03160
作者: Yuxin Cao,Wei Song,Jingling Xue,Jin Song Dong
类目: Multimedia (cs.MM); Computer Vision and Pattern Recognition (cs.CV)
备注: 11 pages, 2 figures
Abstract:When asked which of two events came first, video large language models can fail in two opposite ways: cave to a false claim, or reject a true one. Prior video sycophancy work measures only the first and mitigates it by teaching the model to trust the user less, a fix known in text and image models to worsen the second. In video, both failures come from two causes the literature treats as one: availability, whether the sparse sampled frames contain the two events, and weighting, whether that evidence is trusted over the user. We separate them with two interventions that keep the claim fixed: a frame-preserving reorder that flips the claim’s truth, and a sampling-offset shift that captures or misses both events at a fixed frame budget. When the events are missed, the two twins present identical frames, so each of the nine models we evaluate accepts a true and a false claim at the same rate, making Youden’s J=0 by construction. Availability is necessary but not sufficient. Five of the nine read the order, yet four of those five still cave to the false claim, so their deference hits a weighting ceiling. Since trust cannot be calibrated over evidence that was never sampled, we propose a reversal test that cancels the model’s order prior by scoring the sampled frames forward and reversed, then answers, resamples, or abstains without reading the claim. The test raises the order accuracy to 0.92-1.00 on the models that read the order and abstains rather than guesses on those that cannot.
[CV-88] Surface Keypoint Representation for Multi-Object and Articulated Human-Object Interaction Generation
链接: https://arxiv.org/abs/2608.03158
作者: Xiaogang Peng,Zeyu Han,Zichong Meng,Yiming Xie,Jihua Zhu,Gang Hua,Huaizu Jiang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Project page: this https URL
Abstract:Daily activities require humans to coordinate whole-body motion with the motion of surrounding objects. Despite recent progress in human-object interaction (HOI) generation, most existing methods assume interactions with a single rigid object and do not extend well to scenarios involving a variable number of objects or articulated objects with diverse joint mechanisms. We propose surface keypoint trajectories as an object motion representation: for each rigid component, whether a standalone object or one part of an articulated assembly, we track a small set of non-collinear surface points over time. This representation handles multi-object coordination and diverse articulation mechanisms directly from point dynamics without requiring explicit joint-type specification. To model when and where each body region contacts each object, we introduce a spatio-temporal contact distance field that extends distance-based contact modeling to whole-body, multi-object, and articulated settings. We factorize HOI generation into three stages: generating object motions from text or waypoints, predicting the contact distance field, and synthesizing whole-body motion with contact-guided optimization. Experiments on ParaHome, HIMO, ARCTIC, and OMOMO demonstrate better or comparable performance to existing methods across single-object, multi-object, and articulated interaction settings.
[CV-89] CROSS: Cascaded Distillation and Dual-Constraint Grounding for Remote Sensing Referring Segmentation ECCV
链接: https://arxiv.org/abs/2608.03147
作者: Tingzhang Luo,Ruizhong Liu,Yichao Liu,Cheng Fan,Yu Liu,Jianyuan Guo
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at the European Conference on Computer Vision (ECCV) 2026. 20 pages, 6 figures, and 5 tables. Tingzhang Luo and Ruizhong Liu contributed equally. Jianyuan Guo is the corresponding author. Project page: this https URL
Abstract:Referring Remote Sensing Image Segmentation (RRSIS) has achieved significant progress through the integration of VLMs and the Segment Anything Model (SAM). However, this progress largely relies on strong pre-trained capabilities, while leaving two fundamental limitations insufficiently addressed: (1) Architectural Weak-Coupling, where the unidirectional flow forces reliance on coarse VLM prompts and wastes SAM’s pixel-level structural guidance, causing localization drift; and (2) Object-Centric Semantic Bias, where models overemphasize dominant object semantics while remaining insensitive to spatial reasoning crucial for RRSIS. Motivated by these observations, we propose CROSS, a tightly integrated paradigm for RRSIS. First, we introduce Linguistic-Guided Cascaded Distillation (LGCD) to bridge the architectural gap, which distills SAM’s geometric affinities as soft regularizers into VLM intermediate layers, injecting dense structural priors to refine localization. Second, Perspective-Spatial Contrastive Learning (PSCL) imposes cross-anchored constraints by mining mask-filtered deceptive distractors and spatial-linguistic counterfactuals as hard negatives, explicitly shattering semantic shortcuts to enforce genuine logical consistency. Extensive experiments on RRSIS benchmarks demonstrate that CROSS achieves state-of-the-art performance and maintains precise localization even under severe spatial description perturbations, standing as a robust new paradigm for RRSIS.
[CV-90] From Routes to Steps: Separating Semantic Progress from Local Execution in Vision-and-Language Navigation
链接: https://arxiv.org/abs/2608.03143
作者: Xiangyun Huang,Xiangchen Wang,Runfeng Lin,Yihao Xu,Kangyu Huang,Jiang Hengchen,Xiwang Dong,Lin Jiarong
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注: 16 pages, 9 figures
Abstract:Vision-and-Language Navigation (VLN) requires an agent to follow a route-level instruction by executing its constituent steps from egocentric visual observations. Existing VLM-based navigators typically supervise both capabilities through next-action prediction alone, making progress-tracking errors difficult to distinguish from execution errors. When an agent deviates from the route, a corrective action label may recover the next movement but does not indicate whether the agent selected the wrong sub-instruction or failed to execute the correct one. Consequently, the agent may continue making decisions from an erroneous progress state. To resolve this ambiguity, we propose \textbfRoute2Step, a framework that decouples semantic progress tracking from action generation through an explicit step-level interface. The Instruction Analysis Module ( \mathcalM_\mathrmIA ) predicts this state from the global instruction and visual history. Conditioned on the predicted state and recent observations, the Action Generation Module ( \mathcalM_\mathrmAG ) generates local action chunks. To supervise the progress state without manual temporal labels, E-SPA, a step-alignment procedure, associates sub-instructions with their corresponding portions of route-level demonstrations. These alignments enable state supervision for incorrect progress estimates, while direct action supervision is reserved for rollout groups that repeatedly fail under the correct active sub-instruction. On R2R-CE, Route2Step improves SR from 48.1% to 55.3% and SPL from 43.3% to 48.2%, using 190K state-level corrective samples while requiring only 11.5K directly action-supervised states. Experiments in real-world indoor and outdoor environments further demonstrate the practical applicability of Route2Step. The project page is: this https URL.
[CV-91] Frozen High-Resolution Inference for Cross-City Object Detection: An AI City Challenge 2026 Study
链接: https://arxiv.org/abs/2608.03136
作者: Jaeuk Kim
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 16 pages
Abstract:Cross-city object detection requires a detector trained in one city to generalize to an unlabeled target city. In AI City Challenge 2026 Track 6, we analyze archived configurations of a single RF-DETR-Large detector inside an air-gapped Training-as-a-Service platform whose server returns only an aggregate COCO-style AP over a hidden mixture of source- and target-city images. Frozen 1120 x 1120 inference of a checkpoint trained at 704 x 704 achieved the highest aggregate AP among the evaluated configurations (0.3272 - 0.3654, +0.0382) without any parameter update, with the largest relative gain on small objects and the largest absolute gain on medium objects, at 2.53x the input pixels. A warm-start 1120px fine-tuning recipe reached 0.3470 while its in-domain validation AP rose (0.767 - 0.789), a caution that in-domain validation is an unreliable model-selection signal under aggregate-only cross-city feedback. Because that run’s evaluation used a higher confidence threshold than the inference-only runs (0.05 vs. 0.01), we treat its score as a descriptive archived outcome rather than a controlled verdict on fine-tuning. Gray-world normalization did not meaningfully change the frozen-1120 result, and a rectangular run was found by audit to have used an unintended portrait orientation. We release verbatim platform commands, configuration snapshots, and an explicit evidence boundary for every claim. Each configuration was submitted once and the best was selected on the hidden server, so these are exploratory, audited findings about the aggregate mixture; they do not establish target-city-specific improvement.
[CV-92] Rectify Then Diffuse: Disentangling Concepts Before Denoising Trajectory Unfolds
链接: https://arxiv.org/abs/2608.03135
作者: Ning Zhu,An Chen,Mengfei Zhao,Juntao Xu,Jingze Liang,Boyuan Gu,Liang-Jian Deng
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Text-to-image diffusion models can generate individual concepts well, but they often omit or merge concepts incorrectly with multiple concepts. We trace these failures to an early coordination bottleneck: before denoising begins, prompt-conditioned attention may allocate different concepts to strongly overlapping spatial support, which can keep their attention coupled as denoising proceeds. This observation motivates treating compositional generation as a boundary-condition problem rather than repeatedly controlling the evolving trajectory. To this end, we propose Rectify-then-Diffuse (RTD), a training-free framework that rectifies the initial allocation once before standard denoising. Firstly, we propose Soft-Overlap Disentanglement (SOD), which converts normalized overlap between pilot concept maps into a differentiable and layout-agnostic separation objective. Secondly, we introduce Isotropic Gradient Rectification (IGR), which normalizes the SOD gradient and applies a bounded latent displacement with a consistent scale across prompts and initializations. Extensive experiments show that RTD achieves state-of-the-art compositional fidelity and robust gains. On the AE-Bench object pair subset, RTD improves BLIP-VQA by 45.8% and ImageReward by 19.6% over CO3 while running 2.3 \times faster. Code will be released at this https URL
[CV-93] SeCo-SBIR: Semantically Consistent Prompt Learning for Zero-Shot Sketch-Based Image Retrieval
链接: https://arxiv.org/abs/2608.03120
作者: Long Hoang Dang,Tuan Nguyen Huu,Nguyen Minh Hieu,Tu Minh Phuong
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: ACMMM 26
Abstract:Adapting CLIP for zero-shot sketch-based image retrieval (ZS-SBIR) via prompt learning faces a fundamental tension: the model must bridge the sketch-photo domain gap through task-specific adaptation, yet the added flexibility risks overfitting to seen training categories and eroding CLIP’s zero-shot generalization. We present SeCo-SBIR, a semantically consistent prompt learning framework that resolves this tension from both sides. First, a text-guided multi-modal prompting strategy routes learnable prompt vectors through CLIP’s text encoder and projects the resulting intermediate representations into the visual encoder at every layer via learnable coupling functions. Because the text encoder has already learned robust, abstract category-level semantics from large-scale language supervision, this mechanism injects transferable semantic knowledge directly into the visual pathway - adapting the model to the sketch-photo domain while inherently favoring generalization to unseen classes. Second, a perturbation-based consistency constraint addresses the residual overfitting risk from the learnable coupling functions by aligning the adapted model with a frozen CLIP reference branch using an asymmetric InfoNCE objective - augmented inputs feed the frozen branch while clean inputs feed the trainable branch - anchoring the learned representations to CLIP’s generalizable feature space. Together with lightweight adapters and a multi-objective loss combining triplet, NT-Xent, and classification terms, SeCo-SBIR achieves state-of-the-art results on all three standard ZS-SBIR benchmarks across categorical, generalized, and across-dataset settings.
[CV-94] Non-Destructive Quantification of Urea Adulteration in Bovine Milk Using Transmittance Multispectral Imaging
链接: https://arxiv.org/abs/2608.03113
作者: Sharukshan Niranjan,Iresha Ranaweera,Tharindu Chandrarathne,Kalana Dissanayaka,Roshan Godaliyadda,Vijitha Herath,Parakrama Ekanayake,Janak Vidanarachchi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Adulteration of bovine milk using urea remains a major food quality and health concern, motivating the development of rapid and quantitative screening tools. Conventional approaches, including laboratory-based analytical methods and spectroscopic techniques, have been used for urea detection; however, many remain less suitable for rapid, low-cost routine screening due to requirements such as specialized instrumentation, sample preparation, chemical reagents, or laboratory operation. This study introduces a pragmatic, cost-effective, accurate, and laboratory-validated MSI-based method for quantitative urea estimation under controlled density conditions using a multispectral-imaging-based regression framework. An in-house-built multispectral imaging system operating in twelve discrete spectral bands (365–940~nm) was used to acquire multispectral images of milk samples prepared with controlled urea addition and water for density balancing. Fresh milk was obtained on the day of image acquisition, and the specific gravity of the milk was verified to be 1.032 at 20°C using a hydrometer. Multiple linear regression provided an initial mapping with a high validation R^2 of 0.9599, while a feed-forward neural network further improved predictive performance with a validation R^2 of 0.9773. These results demonstrate the feasibility of transmittance multispectral imaging for accurate, non-destructive urea quantification under controlled density-balanced conditions, supporting its potential as a rapid screening approach for milk-quality assessment.
[CV-95] Adaptive Two-Stage Visual Token Pruning for Efficient Inference in Video-Language Models
链接: https://arxiv.org/abs/2608.03112
作者: Paribesh Regmi,Qingshuang Chen,Chi Zhang,Heba Aly,Yelin Kim,Hongda Mao
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Vision-language models excel at image and video understanding but suffer from high inference latency due to the need to process thousands of tokens per image, limiting their deployment on resource-constrained edge devices and in real-time surveillance applications. This challenge is further amplified in video processing, where multiple frames must be analyzed simultaneously. Existing token reduction techniques are largely developed for single-image inputs and therefore fail to account for the temporal and inter-frame redundancies present in video sequences. In addition, these methods generally rely on a fixed, uniform pruning ratio applied across all inputs, which is suboptimal because the degree of redundancy can vary significantly between different videos, necessitating content-dependent pruning levels to preserve critical information. To address these limitations, we propose a two-stage adaptive token pruning strategy specifically designed for video processing. In the first stage, we prune out the redundant frames, and in the second stage, token-level pruning is applied within the retained frames. Crucially, the pruning ratio in the second stage is determined adaptively based on the content of each video. This is achieved by analyzing the correlation structure of token embeddings to quantify redundancy, which is used to determine the ratio. Importantly, our method is entirely post-hoc and requires no additional training or fine-tuning, while achieving strong empirical gains; notably, it improves accuracy by +7% on a video captioning benchmark at 10% token retention, while reducing computation TFLOPs by 95%.
[CV-96] Multimodal Plant Root Phenotyping with Integration of 3D Skeleton Extraction and Language Analysis
链接: https://arxiv.org/abs/2608.03109
作者: Jiakai Lin,Zijun Li,Guoyu Lu
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注:
Abstract:Plant root phenotyping is fundamental to understanding below-ground structures, optimizing crop management, and improving agricultural sustainability. This paper presents a multimodal robotic AI framework that integrates 3D skeleton extraction with language-guided reasoning for interpretable and data-efficient root analysis. We develop an unsupervised skeleton extraction network based on Weighted Laplacian Contraction (W-LBC) to generate high-fidelity structural representations from dense point clouds captured by robotic 3D sensing platforms. Quantitative morphological descriptors, including root count, length, branching angle, and density, are computed from the reconstructed skeleton graph to capture geometric and topological characteristics. Building on these features, we introduce an Evidence-First language modeling framework that fine-tunes GPT as an interactive analytical chatbot using automatically generated instruction–response pairs. Each training sample provides measurable evidence before natural-language reasoning, enabling the model to ground interpretation in quantitative morphology. Through supervised fine-tuning, GPT associates numerical structure with semantic meaning, producing biologically consistent explanations of growth patterns and adaptive traits. Experiments show that the structure-guided framework achieves robust, interpretable reasoning across 12 plant species with diverse root architectures. By integrating unsupervised 3D geometric perception with large-scale language understanding, our approach bridges quantitative analysis and semantic interpretation, establishing a unified paradigm for explainable robotic plant root phenotyping.
[CV-97] A Unified Resolution-Conditioned Framework for Orthogonal Line-Scanning Image Fusion
链接: https://arxiv.org/abs/2608.03107
作者: Yiming Gong,Kai Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Laser line-scanning microscopy enables fast volumetric imaging but produces anisotropic lateral resolution. Orthogonal line scans provide complementary directional information that can recover near-isotropic resolution, yet existing deep-learning methods require a separate model for each optical configuration. We present a unified, resolution-conditioned fusion framework based on Rank Enhanced Linear Attention (RELA). Feature-wise Linear Modulation (FiLM) conditions the network continuously on the resolving-power ratio, enabling one model to adapt across slit widths. We further introduce Adaptive RELA, which replaces fixed-kernel rank enhancement with ratio-conditioned multi-scale depthwise convolutions and uses a learnable attention temperature to adjust selectivity with degradation severity. Training data spanning multiple slit configurations are generated using a physics-grounded separable point-spread-function model verified against measured optical data at 48.3 dB accuracy. The resulting model achieves 34-40 dB PSNR across configurations, whereas unconditioned multi-slit training collapses to 24.3 dB and per-slit specialists lose 4-9 dB outside their training setting. It also generalizes smoothly to unseen intermediate configurations without interpolation artifacts. Ablations show that FiLM resolves configuration ambiguity, global linear attention captures long-range directional correspondences, and adaptive temperature yields an additional 2 dB in the challenging near-isotropic regime, where complementary signals are weak.
[CV-98] FaithIR: Rethinking Infrared Image Super-Resolution from Perceptual Sharpness to Task Relevant Fidelity
链接: https://arxiv.org/abs/2608.03106
作者: Axi Niu,Zhenguo Wu,Kang Zhang,Qingsen Yan,Jinqiu Sun,Yanning Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 14 pages: 7 pages of main text, 2 pages of references, and 5 pages of supplementary material
Abstract:Infrared image super-resolution (IISR) is important for downstream tasks such as object detection and semantic segmentation. Existing IISR methods often produce artificial textures, over-sharpened edges, and spurious high-frequency details that distort authentic thermal structures and semantic information. To address this issue, we propose FaithIR, a faithful infrared super-resolution framework for reliable machine perception. FaithIR consists of a patch-level conditioning branch that captures global thermal and structural information and a pixel-level restoration branch that performs dense local reconstruction under structural guidance. The entire restoration process is performed directly in the pixel domain to preserve infrared-specific structures and task-relevant information. Extensive experiments on FLIR-IISR, M3FD, and FMB demonstrate strong reconstruction fidelity, cross-dataset generalization, and superior performance in object detection and semantic segmentation. These results show that demonstrate that preserving faithful infrared structure preservations is more important for reliable machine perception than merely pursuing perceptual sharpness alone.
[CV-99] Double Down on Defense: Strengthening Deep Perceptual Hashes against Evasion Attacks without Retraining
链接: https://arxiv.org/abs/2608.03101
作者: Bangjie Sun,Nayoung Kim,Mun Choon Chan,Jun Han
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Near-duplicate image matching is crucial for trust and safety, provenance verification, copyright enforcement, and large-scale visual search. Modern platforms increasingly rely on deep perceptual hashes, which map visually similar images to nearby representations despite common image transformations. However, adversarial perturbations can cause near-duplicates to evade matching. We present DualShield, a plug-in defense that improves the robustness of existing deep perceptual hashes without retraining or modifying their underlying models. DualShield combines matching-time randomized smoothing, which aggregates decisions over perturbed reference-query pairs, with publication-time hardening, which adds an optimized imperceptible perturbation to each reference image before publication. Together, these mechanisms provide certified and empirical robustness. DualShield achieves a certified \ell_2 radius of approximately 0.3, guaranteeing that query perturbations within this radius cannot evade matching. We further evaluate it against adaptive white-box, black-box, and image-transformation attacks. Across eight deep perceptual hashes and three datasets, DualShield substantially reduces attack success rates while preserving low collision rates. These results show that deep perceptual hashes can be strengthened without costly retraining by improving the matching procedure and hardening reference images before publication.
[CV-100] Channel-wise Dynamic Knowledge Distillation via Adaptive Sample Generation for Action Recognition ACM-MM2026
链接: https://arxiv.org/abs/2608.03100
作者: Ping Li,Chenhao Ping,Jie Song,Mingli Song
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted in ACM MM2026, 16 pages, 7 figures
Abstract:Knowledge Distillation (KD) offers a promising yet underexplored path for compressing large action recognition models. However, existing KD methods suffer from two key limitations: 1) reliance on fixed input samples leads to suboptimal feature alignment between the frozen teacher (larger model) and the learnable student (smaller model), and 2) applying a uniform distillation strength for all channels fails to account for their varying importance in capturing distinct knowledge (e.g., motion tempo or magnitude) across training epochs. This motivates us to develop an Adaptive Sample-aware Channel-wise Dynamic (ASCD) KD approach, which operates in two stages. First, we use an adaptive sample generation module to create updated samples by incorporating semantics from sample gradients, which are derived by minimizing a feature loss weighted by channel centroid frequency differences at each layer. Meanwhile, crucial motion-related details are preserved by applying a Gaussian mask to frequency features. Second, we employ a channel-wise dynamic distillation module to train student on these generated samples, guided by sample gradients and feature frequencies. For efficiency, samples are updated periodically rather than per epoch. Extensive experiments on three video benchmarks (UCF101, Kinetics-400, Something-Something-v2) and two image datasets (CIFAR-100, ImageNet) demonstrate the state-of-the-art performance of our method. Code is available at this https URL.
[CV-101] FakeI2V-Bench: Benchmarking the Applicability of Image-level Deepfake Detectors for Deepfake Video Detection KDD2026
链接: https://arxiv.org/abs/2608.03096
作者: Pei Li,Sihan Chen,Delong Ran,Tianshuo Cong
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注: To Appear in KDD 2026, Jeju, Korea, August 9-13, 2026
Abstract:Recent advances in video generation models have significantly intensified the deepfake threat, yet the current deepfake video detection benchmarks remain underdeveloped. In particular, the effectiveness of image-level detectors in the video domain has not been systematically assessed. To fill this gap, we present FakeI2V-Bench, a benchmark for evaluating state-of-the-art video-level deepfake detectors in challenging scenarios, with a particular focus on systematically assessing the performance of image-level deepfake detectors in the video domain. FakeI2V-Bench comprises 97,548 videos, containing content generated by the latest powerful generation models and covering a broader range of categories. Using this dataset, we conduct a systematic evaluation of eight video-level detectors and twelve representative image-level detectors. Experimental results show that the best-performing image-level detector achieves an 80.16% AUC, slightly outperforming the strongest video-level detector (i.e., 79.99% AUC). Going beyond benchmarking, we present IV-Bridge, a general framework that enhances the applicability of image-level deepfake detectors to videos. IV-Bridge employs a random forest model with statistical features to aggregate frame-level predictions, allowing eleven image-level detectors to surpass state-of-the-art video-level approaches, with the best-performing variant achieving a 93.80% AUC. Overall, FakeI2V-Bench establishes a rigorous benchmark for deepfake video detection and introduces a novel pathway for extending image-level detectors to the video domain, offering new insights and directions for future research. Code and data are available at this https URL.
[CV-102] SUV: Future Scene Understanding as Video Generation for End-to-End Driving
链接: https://arxiv.org/abs/2608.03084
作者: Yibo Yuan,Jiacheng Fu,Jiangtong Zhu,Yi Li,Jianhua Han,Meng Tian,Zhuohan Liu,Zhiwei Xiong,Hang Xu,Jianwu Fang,Jianru Xue
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 16 pages, 5 figures. Code: this https URL
Abstract:End-to-end driving requires a coherent understanding of future scenes, yet existing methods model these scenes using task-specific heads and output formats, with limited scalability. Can video generation instead provide a shared predictor? We introduce SUV, a unified end-to-end driving framework that casts future Scene Understanding as Video generation using a pretrained video foundation model. SUV models future appearance, semantics, relative depth, and instance-level dynamics as video streams with a shared video expert, without stream-specific visual prediction heads. Through joint video-action attention, the action expert attends to the latent representations of all future streams and generates the ego trajectory. Experiments show that SUV directly predicts all four future streams, while controlled ablations show that structured future supervision and direct future-stream access yield higher trajectory planning scores. With only a single front camera and no candidate-trajectory selection, SUV outperforms a broad set of recent state-of-the-art methods on both NAVSIM-v2 splits, achieving 91.0 EPDMS on navtest and 36.9 on navhard. On the long-tail WOD-E2E benchmark, SUV achieves a competitive RFS of 7.94.
[CV-103] DiverseDiT: Quantifying Analyzing and Promoting Representation Diversity in Diffusion Transformers
链接: https://arxiv.org/abs/2608.03082
作者: Binglei Li,Mengping Yang,Zhiyu Tan,Xiaomeng Yang,Zhizhong Huang,Junping Zhang,Hao Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 35 pages, 32 figures
Abstract:Recent advances in Diffusion Transformers (DiTs) have enabled remarkable progress in visual synthesis, benefiting from their superior scalability. To facilitate DiTs’ capability of capturing meaningful internal representations, recent works such as REPA incorporate external pretrained encoders for representation alignment. However, the underlying mechanisms governing representation learning within DiTs remain poorly understood in the community. To this end, this paper first presents a systematic analysis of the representation dynamics of DiTs via quantifying the diversity of block-wise representations. Specifically, we introduce a novel metric, termed the Weighted Diversity Score (WDS), to measure the representational discrepancies across different blocks. Through extensive investigations on the evolution and influence of internal representations under various settings, we reveal that representation diversity across blocks is a critical factor for effective representation learning in DiTs. More importantly, WDS exhibits a strong correlation with synthesis quality across diverse settings, model scales, and training stages (Pearson’s r=-0.869 with \log(\textFID) ), suggesting its potential as an indicator to reflect model performance and a principled guide for model optimization. Based on this key finding, we propose DiverseDiT++, a novel framework that explicitly promotes diverse representation learning. Concretely, our method incorporates long residual connections to diversify input representations across blocks and a representation diversity loss to encourage blocks to learn distinct features. Extensive experiments on ImageNet 256\times256 and 512\times512 demonstrate that our DiverseDiT++ yields consistent performance gains and convergence acceleration when applied to different backbones with various sizes,…
[CV-104] CorePath: A Breast-Specialized Pathology Foundation Model for Core Needle Biopsy Diagnosis and Risk-Controlled Report Generation
链接: https://arxiv.org/abs/2608.03079
作者: Ting Yin,Danning Li,Chen Shu,Xiaoxia Yao,Boyu Fu,Yujing Chang,Tianyu Shi,Mengna Feng,Jie Chen,Jing Fu,Xiuli Xiao,Tianlin Li,Mumin Shao,Jiaxin Bi,Wenchuan Zhang,Xiaoyan Wu,Xiao Han,Zhang Zhang,Yuhao Yi,Hong Bu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Applications (stat.AP)
备注: The code will be made publicly available upon publication
Abstract:Breast core needle biopsy (CNB) is central to breast cancer diagnosis yet remains challenging because limited tissue sampling, lesion heterogeneity, and subtle morphologic overlap can obscure subtype distinctions. We developed CorePath, a breast-specialized multimodal pathology foundation model fine-tuned from PRISM using 7901 paired CNB whole-slide images and diagnostic reports from two centers. Evaluated across six CNB cohorts and two public breast pathology benchmarks without task-specific retraining, CorePath consistently outperformed PRISM across cancer detection, invasion assessment, and histological subtyping. It achieved weighted area under the receiver operating characteristic curves (AUCs) of 0.9526-0.9735 for five-class CNB histological subtyping across private centers. On public benchmarks, CorePath outperformed leading pathology foundation models, achieving the highest weighted AUCs of 0.7780 for BCNB invasive carcinoma subtyping, 0.8178 for BRACS lesion stratification, and 0.8252 for BRACS fine-grained classification. In report generation, CorePath reduced the overall non-breast hallucinations from 30.1% to 2.8%, demonstrating improved domain fidelity after breast-specific adaptation. CorePath-CRG further combined conformal subtype-confidence gating with Learn-Then-Test risk control to enable selective report release, subtype-level fallback, and deferral. CorePath-CRG achieved zero non-breast hallucinations among released outputs and showed the strongest overall performance in pathologist-validated LLM-based Evaluation Scores and quantitative report-generation metrics across most centers. These results demonstrate that domain-specialized foundation models with statistical risk control offer a promising approach for accurate breast CNB diagnosis and reliable report generation.
[CV-105] LDU-Bench: Multimodal LLM Evaluation for Lithography Defect Understanding under Layout-Varying Circuit Backgrounds
链接: https://arxiv.org/abs/2608.03078
作者: Huanglong Ji,Botong Zhao,Shujing Lv,Yue Lv
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 12 pages, 3 figures, and 5 tables, including appendices
Abstract:Multimodal large language models have demonstrated strong defect recognition capability in industrial anomaly detection. However, in lithography review, merely determining whether an image contains a defect is insufficient for engineering inspection; models must also understand defect morphology, spatial location, and the potential causes supported by visible evidence. To this end, this paper proposes LDU-Bench, a multi-task multimodal benchmark for lithography defect understanding. Constructed from real lithography and integrated-circuit review images, LDU-Bench decomposes the review workflow into four independent tasks: defect triage, morphology recognition, coarse localization, and image-conditioned cause analysis. It systematically evaluates models using task-level metrics, diagnostic readouts, and the Lithography Closure Score (LCS). Experimental results show that although existing MLLMs can perform defect triage relatively reliably, this ability does not stably transfer to downstream review stages. Morphology alignment, effective localization, and evidence-to-cause mapping remain the major bottlenecks. Further diagnostics indicate that this capability break is not a fluctuation of a single metric, but reflects insufficient structured understanding across semantic levels. Overall, LDU-Bench provides a quantifiable and diagnostic unified platform for evaluating the usability, failure points, and capability boundaries of industrial MLLMs in lithography review chains.
[CV-106] Global Graph-Validated Optimization for VLM-based 3D Indoor Scene Generation
链接: https://arxiv.org/abs/2608.03064
作者: Jialu Huang,Yingxuan You,Fei Wang,Zheng Dang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 15 pages, 8 figures
Abstract:We study open-vocabulary 3D indoor layout generation, which synthesizes diverse and physically plausible scenes from unlabeled 3D assets using free-form language instructions. Recent methods leverage large language models (LLMs) and vision-language models (VLMs) to generate structured scenes from text. However, most model inter-asset relations implicitly or rely on local pairwise constraints and local optimization. These formulations are poorly aligned with the global, highly non-convex layout space, often yielding locally plausible yet globally inconsistent or physically infeasible scenes. We address this problem with a graph-based intermediate representation that separates semantic coherence from physical feasibility, together with a hybrid search-and-refinement strategy. First, Global Semantic Verification (GSV) represents scenes as structured graphs and enforces semantic constraints through rule-based verification. This explicit validation removes contradictory configurations and produces a globally consistent semantic scaffold. Second, Global Physical Feasibility Search (GPFS) combines evolutionary search for global exploration with gradient-based refinement for local exploitation. It reduces dependence on VLM-proposed initialization and improves robustness in non-convex and discontinuous feasible spaces. Together, GSV and GPFS move layout generation beyond local relational modeling and initialization-sensitive optimization toward globally consistent reasoning and search. Experiments show that our method achieves state-of-the-art performance in open-vocabulary 3D indoor layout generation, improving both semantic consistency and physical plausibility.
[CV-107] RIDGE: Re-Noising with Internal Dynamic Guidance for Image Editing
链接: https://arxiv.org/abs/2608.03059
作者: Ruiliang Gong,Zhen Wang,Yanghao Wang,Long Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Inversion-free flow-based image editing avoids latent inversion, but still requires a target-side state at every editing step. The widely used equal-displacement construction keeps the displacement between the noisy source state and the target-side state unchanged across noise levels. This is inconsistent with noising, under which the displacement between two clean states noised with the same noise level and noise sample should contract as the noise level increases. Thus, it can lead to overly aggressive updates at high noise levels. We introduce RIDGE: Re-Noising with Internal Dynamic Guidance for Image Editing, an inversion-free and training-free method that maintains the edited state as an evolving approximation to the unavailable clean target state. RIDGE re-noises this approximation using the same noise level and noise sample as the clean source state, allowing their noisy displacement to decrease naturally with increasing noise. Since the edited state initially contains limited target semantics, RIDGE further applies internal dynamic guidance during the early high-noise steps. A clean target state prediction guides the provisional edited state through a soft dynamic mask derived internally from the model, focusing guidance on regions that require modification without external segmentation or detection models. Experiments on two benchmarks using two backbones, SD3 Medium and FLUX.1-dev, show that RIDGE offers a favorable aggregate trade-off among source preservation, target alignment, and perceptual quality.
[CV-108] ASQ: Temporal-Adaptive Bit Sparsification Quantization for Diffusion Models
链接: https://arxiv.org/abs/2608.03057
作者: Seokho Han,Dongwei Wang,Jinhee Kim,Yiran Chen,Kang Eun Jeon,Huanrui Yang,Jong Hwan Ko
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Static quantization assigns one weight precision to every denoising step. To preserve quality, that precision must accommodate the most quantization-sensitive step, even though many other steps can tolerate fewer bits. The resulting model may satisfy its memory budget, but it repeatedly pays worst-case arithmetic throughout the denoising trajectory. We introduce Temporal-Adaptive Bit Sparsification Quantization (TASQ) to separate these two costs. TASQ stores one shared maximum-precision weight buffer and learns a Temporal-Spatial LSB Mask that selects a lower effective precision for each layer and denoising stage by truncating least-significant bits. Storage therefore remains fixed by the worst case, while BitOPs decrease at less sensitive stages without per-stage weight copies or runtime search. A Temporal-Precision Engine maps the learned schedule to bit-serial execution, where cycles scale with effective precision and switching precision has no measured cycle overhead. On PixArt-Sigma, SANA-1.6B, and SDXL-Turbo, TASQ achieves quality comparable to static quantization with less computation. Together with the Temporal-Precision Engine, it reduces execution cycles by 25 to 50 percent over static quantization and by 6.1 to 7.5x over a naive static 8-bit bit-serial execution. Code is available at this https URL.
[CV-109] AIDE: Automated Instruction via Distilled Expertise for Reference-Free Motor Skill Coaching
链接: https://arxiv.org/abs/2608.03047
作者: Yoshiki Ito
类目: Computer Vision and Pattern Recognition (cs.CV); Multimedia (cs.MM)
备注: Accepted to ACM Multimedia 2026. 12 pages (including supplementary material), 4 figures
Abstract:Generating natural-language coaching feedback on motor skills can accelerate learning, yet expert coaches are scarce and expensive. Existing reference-based methods require expert demonstrations at both training and inference time, limiting practical deployment. We propose AIDE (Automated Instruction via Distilled Expertise), a framework that exploits expert references only during training and generates feedback from a learner’s pose sequence alone at inference. A teacher model first learns to generate feedback from paired learner-expert poses via a frozen language model, producing separate learner tokens and difference tokens that encode the learner-expert difference. A student model then inherits the teacher’s encoder and weight initialization, replacing the explicit expert comparison with an auxiliary module that produces complementary tokens from the learner’s pose alone. On the ExpertAF dataset, AIDE outperforms reference-free baselines on most metrics and performs comparably to methods requiring expert demonstrations at both training and inference, with LLM-based evaluation supporting these findings.
[CV-110] CAPE-T2V: Captioner-Anchored Prompt Enhancement toward Two-Sided Conditioning Alignment in Text-to-Video Generation
链接: https://arxiv.org/abs/2608.03046
作者: Yizhuo Jia,Jingyun Hua,Yuanxing Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Includes appendix; 11 figures. Project page: this https URL
Abstract:Text-to-video (T2V) diffusion transformers (DiTs) are trained with detailed video captions, whereas inference often relies on user prompts rewritten by a prompt enhancer (PE). Prior work has improved generation by optimizing the PE, the DiT, or both; some methods have also sought to narrow the training-inference mismatch through shared schemas. Yet even within a shared schema, inference-time PE outputs and DiT training captions may still differ in detail selection, information organization, descriptive granularity, and phrasing. We refer to this residual mismatch as the PE-Caption gap and introduce CAPE-T2V, a two-step Captioner-Anchored Prompt Enhancement framework toward two-sided conditioning alignment in T2V generation. First, CAPE-T2V constructs three types of PE training examples, pairing captioner-generated targets with concise source captions, detailed source captions, or pseudo user prompts derived from those targets. It then fine-tunes the PE to map each input to its paired target. Second, CAPE-T2V fine-tunes the DiT on video-derived captions rewritten by the Anchored PE; the same PE rewrites user prompts at inference. Relative to a baseline using the same caption schema, CAPE-T2V achieves higher aggregate scores on StoryEval, VBench-2.0, and T2V-CompBench across Wan2.2 and LTX-2.3. Further, CAPE-T2V exhibits a smaller PE-Caption gap than the baseline: its DiT fine-tuning captions are closer in distribution to inference-time PE outputs, as measured by squared maximum mean discrepancy in a fixed embedding space. Overall, these results support CAPE-T2V as an effective approach to mitigating the PE-Caption gap. The project is available at this https URL.
[CV-111] Standalone DINOv3 for Training-Free Open-Vocabulary Semantic Segmentation in Remote Sensing
链接: https://arxiv.org/abs/2608.03023
作者: Changhao Zhao,Haoxiang Li,Yuke Li,Hai Liu,LingLin Zeng
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Remote sensing semantic segmentation is hindered by costly pixel-level annotations, motivating training-free open-vocabulary methods. Recently, the recent release of DINOv3 brings this http URL, which equips the standalone DINO backbone with image-text contrastive learning and thus opens up the possibility of open-vocabulary segmentation. We propose DinoSplat-OV, a training-free framework that adapts DINOv3 to remote sensing without fine-tuning or additional pretraining. Targeting the dense distribution, multi-scale nature, and large size of remote sensing imagery, we design two core modules. Its Text-aware Laplacian Propagation module de-noises patch-level predictions by combining textual semantic affinities with local visual similarity, improving regional consistency while preserving boundaries. Its Gaussian Splatting Upsampling module reconstructs pixel-level features through RGB-guided anisotropic aggregation and test-time optimization. A global-anchor sliding-window strategy further supports large-scale imagery. Experiments on UDD5, DOTA, and LoveDA demonstrate competitive or superior performance over existing training-free methods, effectively filling the gap of DINO-series models in training-free open-vocabulary segmentation and providing a viable new path for further advances in this direction.
[CV-112] Clinically-Grounded Hierarchical Classification for Consistent Chest X-ray Interpretation MICCAI2026
链接: https://arxiv.org/abs/2608.03016
作者: Jong Hak Moon,Minjun Kim,Minjun Kim
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: MICCAI 2026 Accepted. First Corresponding author: Jong Hak Moon ( this http URL @yejix.com)
Abstract:Accurate chest X-ray interpretation is inherently hierarchical. Clinical decisions depend not only on what abnormality is present but where it is situated, requiring reasoning from broad anatomical systems down to specific pathological findings. Yet existing automated systems largely treat this as a flat classification problem, failing to capture inter-level dependencies or enforce coherence between coarse and fine predictions. We propose CHASE (Classification with Hierarchical Analysis and Structured Enforcement), a unified single-stage framework that mirrors radiologists’ coarse-to-fine reasoning through a clinically driven three-level taxonomy of 9 anatomical regions, 17 sub-regions, and 28 pathological findings. CHASE jointly optimizes multi-level supervision, cross-level probability alignment, and a hierarchy-violation penalty within a shared Vision Transformer backbone. This ensures that fine-grained findings are anatomically supported by their coarser-level context rather than predicted in isolation. Experiments demonstrate that CHASE outperforms flat and hierarchical baselines across all levels while achieving superior probabilistic hierarchy consistency, with level-wise attention maps confirming anatomically grounded predictions. Code is available at: this https URL.
[CV-113] V-FIND: Revealing the Intrinsic Forgery Knowledge Encoded in Video Forgery Detectors
链接: https://arxiv.org/abs/2608.03008
作者: Shichao Kan,Chengpeng Hong,Jingtong Dou,Chuancheng Shi,Yuhan Liu,Linrui Xu,Yixiong Liang,Yigang Cen,Yanpeng Sun,Fei Shen,Tat-Seng Chua
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 12 pages, 12 figures. Under review
Abstract:As generated videos become increasingly realistic, reliable video forgery detection is increasingly important. Existing studies typically optimize and use video forgery detectors as black boxes, while the latent forgery-discriminative knowledge inside them remains largely unexplored. Instead of continuing to rely on resource-intensive full-model retraining to steadily improve detection performance, we ask whether video forgery detection can also be achieved by uncovering and activating sparse forensic knowledge within the detector. We find that forgery-discriminative knowledge is not uniformly distributed across the full representation space, but is concentrated in a sparse set of functionally specialized neurons. Based on this insight, we propose a video forgery-intrinsic neuron discovery (V-FIND) framework. V-FIND first localizes critical layers that exhibit pronounced discrepancies between real and forged videos, and then identifies latent anchor neurons that consistently carry forgery-discriminative signals, organizing them into a compact forensic subspace. With the original backbone frozen and only a lightweight linear classifier trained, this subspace still delivers strong detection performance across multiple external benchmarks for generated videos. Further neuron intervention experiments provide direct evidence for the functional specificity of the discovered neurons. Overall, these results suggest that video forgery detectors contain sparse, extractable, and reusable forgery-discriminative knowledge, offering a new perspective on understanding and exploiting their intrinsic forensic capability.
[CV-114] Qwen -3D: A Generalist 3D Vision-Language Model for Spatial Understanding
链接: https://arxiv.org/abs/2608.02980
作者: Lucy Lin,Ayush Jain,Yifan Liu,Katerina Fragkiadaki
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Project Page: this https URL
Abstract:Large Multimodal Models (LMMs) have achieved remarkable success on images and short videos, yet scaling them to long videos remains challenging due to frame-centric tokenization and limited context windows. 3D geometry provides a natural compression mechanism for visual streams: depth and camera pose enable observations from multiple views and time steps to be fused into a persistent, world-aligned representation. While recent 3D LMMs leverage geometry-aware representations to improve spatial reasoning, they continue to lag behind specialist 3D perception systems on grounding and segmentation tasks. We argue that a key limitation is geometry-aware decoding: existing methods communicate 3D predictions through language tokens, proposal selection, or lightweight grounding queries, creating a bottleneck between language reasoning and dense geometric prediction. Building on these insights, we introduce Qwen-3D, a geometry-aware LMM that compresses visual information within the Qwen backbone using multi-view geometric cues, enabling efficient long-horizon visual reasoning over static scenes. Qwen-3D augments visual tokens with 3D Rotary Positional Embeddings, allowing attention to operate directly in 3D scene space rather than across independent image frames and thereby facilitating scalable cross-view and temporal reasoning. To bridge language and geometry, Qwen-3D incorporates a query-based segmentation decoder that grounds language directly in the underlying 3D scene representation, unifying referential grounding, instance segmentation, and visual question answering across both images and videos. Across a diverse set of benchmarks, Qwen-3D surpasses existing 3D LMMs and outperforms several large proprietary 2D models. Notably, Qwen-3D achieves these improvements while maintaining strong performance on standard 2D vision-language benchmarks by jointly training on 2D and 3D data.
[CV-115] Material-Segmented Per-Pixel Emissivity Correction for Thermographic Anomaly Detection in Cultural Heritage Digital Twins
链接: https://arxiv.org/abs/2608.02964
作者: Jonathan Klingspon,Scott McAvoy,Maurizio Seracini,Falko Kuester
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 13 pages 5 figures
Abstract:Quantitative longwave thermography of heritage surfaces is limited by the global-constant emissivity assumption in inverse-Planck temperature retrieval; on heterogeneous surfaces emissivity varies within one field of view, producing apparent-temperature artifacts that mimic and mask subsurface anomalies. We present a training-free pipeline that derives per-pixel emissivity by applying SAM 3.1 open-vocabulary segmentation to a colocated, co-calibrated RGB channel, mapping segments to a material-keyed LWIR emissivity table compiled from primary measurement literature, and propagating the field into a per-pixel inverse-Planck solve on raw radiometric data. Lacking any public dataset with raw radiometry, a temperature reference, and a colocated RGB camera, we evaluate on a physics-based synthetic benchmark and four real datasets. On the benchmark, under a palette spanning the low-emissivity exceptions, the correction cuts mean absolute error from 1.97 K to 0.91 K at 20 K contrast and, with an accurate table, beats the best fitted global constant on every layout; on a heritage-realistic emissivity distribution it does not. We contribute a quantified operating-regime map, and a measurement-backed finding that tempers the heritage claim: weathered outdoor heritage emissivities cluster near the conventional default, so the correction is small on typical surfaces and concentrated on genuine low-emissivity exceptions. We characterize the dominant failure mode, in which open-vocabulary segmentation matches appearance rather than material, and the contraindicated regime in which emissivity-defined anomalies are suppressed.
[CV-116] RealWeather: Realistic and Scene-Faithful Weather Translation with Driving World Models
链接: https://arxiv.org/abs/2608.02953
作者: Yuwei Ning,Liangzhi Wang,Yi Xiao,Zhenhua Wu,Yun Pang,Mingkun Chan,Jichang Li,Guanbin Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Under submission
Abstract:Realistic weather translation is valuable for developing and evaluating autonomous driving systems, yet collecting paired videos of the same scenes under different weather conditions at scale is impractical. Existing methods therefore rely on synthetic data, 3D weather editing, or geometry-conditioned generation, often compromising weather realism or scene fidelity. We propose RealWeather, a driving world model for both realistic and scene-faithful weather translation. Our key idea is to learn authentic weather dynamics directly from real-world videos. Specifically, RealWeather employs Progressive Realism Bootstrapping, an iterative data-refinement strategy. Assisted by an auxiliary Pseudo-Clear Generation pipeline, training initially starts with pseudo-style conditioning videos. As training proceeds, these inputs are progressively replaced with increasingly realistic videos generated by the model itself. This strategy bridges the pseudo-to-real domain gap, allowing the model to adapt seamlessly to real-world input distributions and naturally support bidirectional clear adverse translation. Furthermore, to strictly enforce structural integrity and suppress hallucinations, we introduce Scene-Fidelity RL Optimization, a reward-driven policy optimization strategy that explicitly penalizes alterations to safety-critical driving elements. Extensive experiments demonstrate that RealWeather significantly outperforms existing methods in visual realism and structural preservation, while enabling robust long-tail weather scenario generation and strong zero-shot out-of-distribution generalization.
[CV-117] Modeling Scientific Experiment Scenes: Dataset and Model
链接: https://arxiv.org/abs/2608.02892
作者: Minghao Zou,Qingtian Zeng,Shangkun Liu,Cong Liu,Paul L. Rosin,Guanghui Yue,Jun Liu,Wei Zhou
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Scene Graph Generation (SGG) is fundamental to structured visual understanding, yet existing benchmarks focus mainly on daily life images and overlook scientific experiment scenes with specialized instruments, task-specific experimental semantics, and dense, fine-grained physical relations. These scenes are increasingly important for automated experimental analysis and smart education. To bridge this gap, we introduce PhysScene, the first SGG dataset for physical experiment scenes, providing densely annotated scene graphs and benchmarks under multiple supervision and protocol settings. PhysScene further exposes two key algorithmic challenges for SGG: pronounced long-tail relational predicate distributions and a substantial visual-textual semantic gap. To address these challenges, we propose the Cross-Modal Dual-Path Generator (CM-DPG), a model for robust open-vocabulary SGG. The model enhances object-level semantic representations through joint visual-textual encoding and improves relational reasoning using complementary visual and geometric cues. We also incorporate relation-aware pre-training, caption-derived pseudo-supervision, and adaptive weighting to support balanced learning across head and tail predicates. Extensive experiments on PhysScene and VG150 show that CM-DPG achieves competitive performance across multiple evaluation settings, with ablation studies validating the contribution of each component. The dataset and code are publicly available at this https URL.
[CV-118] st Time Adaptation Methods for Point Cloud Registration in Laparoscopic Surgery
链接: https://arxiv.org/abs/2608.02883
作者: Nina Bodelot,Soufiane Belharbi,Eric Granger
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:3D point cloud registration in laparoscopic surgery estimates the transformation between an intraoperative organ reconstructed from video and its preoperative mesh. Because ground-truth transformations are unavailable for real data, supervised networks are trained on synthetic organ pairs. At test time, real reconstructions differ from synthetic data and are noisy, sparse, and occluded, which degrades correspondence estimation. Test-time adaptation (TTA) can reduce this domain shift, but existing methods mainly rely on logits, entropy, class prototypes, or cache memories unavailable in registration. Registration also involves paired inputs with an asymmetric shift that primarily affects the intraoperative cloud. We analyse and modify state-of-the-art TTA methods from three families to 3D registration: model, normalization, and input adaptation. We analyze four representative approaches based on auxiliary-task model updates, backpropagation-free token purging, feature alignment, and layer-normalization calibration. We modify them to handle asymmetric shifts between preoperative and intraoperative point clouds and replace classification-based entropy objectives. Using a correspondence-based model trained on clean synthetic source data, we evaluate adaptation to corrupted synthetic and real target data on P2P and P2ILReg. For synthetic targets, we apply eight corruptions, including uniform noise and global density reduction, at five severity levels. All methods improve registration on P2P, whereas normalization adaptation degrades performance on P2ILReg. Considering the computational overhead of backpropagation-based adaptation, input adaptation is the most promising option for laparoscopic surgery, providing low inference latency and consistent error reductions across datasets. Code: this https URL
[CV-119] Localize Dont Beautify: Client-Side Control of Image-Editing APIs for Cosmetic Surgery Previews
链接: https://arxiv.org/abs/2608.02841
作者: Sukhrobbek Ilyosbekov
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 9 pages, 7 figures, 2 tables. Pilot study; no surgeon ratings. Code and paper source: this https URL
Abstract:Ask a commercial image editor to preview a cosmetic procedure and it will often change more of the face than the request names: a nose edit can also smooth skin or alter lighting. Existing methods for confining an edit to one region require access to the model’s internals, which a public editing API does not expose. We ask how much control is possible from the client side alone. In a pilot benchmark, six commercial editing configurations and one mask-based inpainting model perform facelift-style jaw-neck and rhinoplasty edits at three levels of client-side control: the prompt alone; cutting the edited region out of the response and pasting it back onto the original photograph through a landmark-derived mask (a masked composite); and asking the model itself to inpaint inside the mask where supported. Of 210 attempted edits, 196 could be scored. ArcFace cosine measures identity preservation; a CIELAB pixel-change ratio measures how much change lands inside the requested region rather than a protected facial zone. On the 12 frontal faces the regional metric could score, the masked composite improved localization over the paired prompt-only output by a median of 0.446 (95% face-clustered bootstrap interval 0.421-0.457) while changing the requested region about as much. Editors differed in edit strength versus identity retention, and the one inpainting model we tested did not beat the simple composite. Against each face’s input-to-postoperative baseline, no editor moved its outputs closer to the postoperative photograph in identity-embedding terms. This is a study of control, not clinical accuracy: no surgeons rated the outputs, and each condition was generated once. Within that scope, keeping a surgical preview inside its intended region needs no access to the model; a mask and composite on the client enforce it across every editor tested, at low provider cost.
[CV-120] A Human-in-the-Loop Deep Learning Framework for Color Reconstruction of Lenticular Films
链接: https://arxiv.org/abs/2608.02835
作者: Saptarshi Neil Sinha,Tiago Kleist,Giorgio Trumpy
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Historical lenticular films, such as those created with the Kodacolor process, encode color information in a distinctive spatial format. This structure requires specialized techniques for accurate color reconstruction. While recent signal processing approaches like doLCE and deep learning methods like deep-doLCE have advanced automated color recovery, they often fail with cases such as curved lenticules, low-contrast, or badly captured regions. We propose a human-in-the-loop (HITL) deep learning framework which is designed for color reconstruction in lenticular films. Our approach introduces an editable, vector-based representation of lenticule boundaries, allowing experts to interactively refine boundary positions before color extraction and demosaicing. This decoupled architecture enables targeted corrections and iterative fine-tuning, embedding expert knowledge into the detection model and improving robustness across challenging frames. To preserve image details using information solely present in the original silver emulsion, we merge the reconstructed chrominance with the original film scan’s luminance. We evaluate our pipeline on a challenging lenticular film sequence where previous automated approaches fail and the reconstructed colors are not suitable for exhibition. In contrast, our HITL approach successfully produces high-quality, exhibitable color reconstructions with preserved texture. This work is the first to combine expert guidance, editable intermediate representations, and texture-preserving post-processing for lenticular film color reconstruction, advancing the state of the art in this field.
[CV-121] In-Context Collapse in Vision-Language Models and How to Mitigate it?
链接: https://arxiv.org/abs/2608.02830
作者: Mohammad Rostami
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Many-shot in-context learning (ICL) lets vision-language models (VLMs) adapt from image–label demonstrations without weight updates, and is widely assumed to improve as more demonstrations are supplied. We show the opposite: as demonstrations accumulate, a subset of VLMs undergo an \emphin-context collapse, a sharp, sometimes catastrophic accuracy drop spanning synthetic classification, natural-image classification, and VQA benchmarks, in some models falling below chance while outputs remain well-formed. Across an open VLM panel ( 0.5 B-- 11 B) and a frontier model (Claude Sonnet 4.5), the collapse is graded. Two capabilities turn out to be dissociable: robustness to accumulating demonstrations and the ability to learn a novel rule in context, their combinations yield three reproducible regimes. A parameter-matched lesion-and-rescue causally localizes the collapse to the vision-language integration pathway: an adapter on the connector and early/mid layers restores genuine learning (remap accuracy 0.39!\rightarrow!0.91 at 16 shots), while an equal-capacity adapter on the late readout does not. We propose \textscCircA, whose core is a one-time integration vaccine: trained once on one synthetic task, it transfers collapse-resistance to unseen task families (chance \rightarrow 0.71 / 0.60 on CIFAR/Fashion). The layers best for in-context integration are not the layers best for weight-based consolidation, the late readout achieves higher accuracy and less forgetting at fewer parameters. The collapse is an integration failure at the vision–language interface, correctable by a lightweight, transferable intervention.
[CV-122] Fast Object Removal Attacks on Safety-Critical Video-based Perception Systems
链接: https://arxiv.org/abs/2608.02806
作者: Mohammad Imtiaz Hasan,M Sabbir Salek,Nathan Jones,Mashrur Chowdhury,Rong Ge
类目: Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:By leveraging data from video-based perception systems, intelligent transportation systems (ITS) support safety-critical applications that improve road safety. However, adversaries may manipulate video frames to compromise downstream perception modules, causing failures in safety-critical functions and increasing risks to vulnerable road users. This paper presents a novel attack model and an end-to-end framework for near-real-time targeted object removal attack on a video-based safety-critical system. The end-to-end attack pipeline consists of four stages: localizing targets in each frame, retrieving coherent patches from earlier frames, blending them using context-aware alpha compositing, and reconstructing attacked frames. Experiments at an intersection on the South Carolina Connected Vehicle Testbed (SC-CVT) show that reconstructed frames have high global similarity to the originals, with frame-level Peak Signal to Noise Ratio (PSNR) above 40 dB and Structural Similarity Index Measure (SSIM) above 0.996. Using the YOLO-based detector, the attack reduces object detections by up to 97.59% and achieves a frame-level attack success rate of 94.48%. Across the evaluated detectors and frame resolutions, the mean execution time ranges from 0.074 to 0.172 seconds per frame on GPU hardware, indicating near-real-time performance in testing. The forensic evaluation using several pretrained tamper-detection models shows limited ability to distinguish reconstructed from authentic frames. The findings suggest that video-based perception is vulnerable to stealthy object removal attacks that can degrade the performance of safety-critical applications by reducing object detectability. These findings can help develop mitigation strategies against adversarial object removal attacks that threaten safety-critical applications, such as vision-based pedestrian safety systems.
[CV-123] A Unified 2D Framework for DeepLesion Detection Segmentation and Short Report Generation
链接: https://arxiv.org/abs/2608.02805
作者: Ruida Cheng,Tejas S. Mathai,Benjamin Hou,Qingqing Zhu,Zhiyong Lu,Matthew McAuliffe,Ronald M. Summers
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 18 pages, 8 figures
Abstract:In previous work, we integrated large language models (LLMs) into the lesion segmentation model based on the ULS23 DeepLesion dataset, using short-form findings from the reports. In this study, we developed a unified 2D lesion analysis framework that integrates LLM-based reasoning, lesion bounding box detection, segmentation, and radiology report generation from the original DeepLesion dataset. In the testing phase, we achieved relatively high lesion bounding box detection accuracy with mAP50 of 70.1%, mAP50-95 of 46.4%; Lesion segmentation performance with a Dice score of 62.6%; short report generation accuracy with BLEU_1 score of 64.3%, BLEU_4 score of 49.6%, METEOR of 34.7%, and ROUGE_L of 60.1%. In this work, we address the challenging issue of segmentation in the original DeepLesion dataset and achieve a 28.5% Dice score improvement over the nnUNet lesion segmentation model. We also integrated spatial and anatomical context into the DeepLesion short report generation. We released the implementation, dataset, and models on Github. this https URL
[CV-124] SAGE: Semantic Explainability of Attention-Based Survival Models in Computational Pathology MICCAI
链接: https://arxiv.org/abs/2608.02803
作者: Abdallah Lamane,Abdul Rahman Diab,Ren-Chin Wu,William Lotter
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Proceedings of the MICCAI Workshop on Interpretability of Machine Intelligence in Medical Image Computing (iMIMIC)
Abstract:Attention-based multiple instance learning (ABMIL) is the predominant approach for slide-level prediction in computational pathology, yet its attention maps provide only local explanations: they indicate where a model focuses but not which histological features drive its predictions or how the model behaves across a patient cohort. We present Semantic Attention Global Explanations (SAGE), a post-hoc framework that extracts global, language-grounded explanations from a frozen ABMIL model. Using a pathology vision-language model, SAGE scores image patches against a dictionary of 25 histological concepts, aggregates these scores according to the model’s learned attention, and quantifies how each concept relates to prediction risk across a cohort. Applied to survival prediction using seven TCGA cancer cohorts and three foundation models, SAGE recovered established prognostic features, such as the adverse association of necrosis, while revealing cancer-specific biology, including a favorable angiogenic signature in renal cell carcinoma consistent with known molecular subtypes. Ablation studies demonstrated that these associations depend on the model’s learned attention rather than concept prevalence alone, and that the concept dictionary captures much of the prognostic information encoded by the foundation model features. Through semantically-grounded explanations, SAGE provides a scalable, model-agnostic framework for understanding what ABMIL survival models learn, enabling pathologists to interpret model behavior at the cohort level and offering the potential for biomarker identification.
[CV-125] PixelUp: Zero-Shot Semantic Feature Upsampling for Fine-Grained Vision Tasks
链接: https://arxiv.org/abs/2608.02792
作者: Deepank Singh,Anurag Nihal,Vedhus Hoskere
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Self-supervised Vision Foundation Models (VFMs) have become essential backbones for downstream tasks due to their strong and transferable visual representations. However, their patch-token-level features are often too coarse for dense prediction tasks such as semantic segmentation and depth estimation when accurate fine-grained predictions are required. Feature upsampling methods have been developed to recover pixel-level detail but still face limitations. Learnable upsamplers are often designed for a specific encoders and must be retrained for different encoders. Image-guided methods that use shallow pixel encoders often introduce textural artifacts and lack the semantic guidance needed for accurate downstream predictions. We introduce PixelUp, a zero-shot VFM-agnostic upsampler achieving semantic awareness through a coarse-to-fine chain of windowed cross-attention architecture guided by multi-scale semantic features. We demonstrate that PixelUp outperforms both VFM-specific and VFM-agnostic upsamplers, achieving state-of-the-art performance on dense prediction tasks with an average improvement of +1.2 mIoU on semantic segmentation and +0.25 \delta_1 , on NYUv2 depth estimation across VFMs. PixelUp further improves training-free open-vocabulary and unsupervised semantic segmentation by an average of +1.3 mIoU and +0.5 mIoU, respectively. Code available at this https URL
[CV-126] Better Stronger Faster and Broader: Structured All-Mask Prediction for MLLM -Based Segmentation
链接: https://arxiv.org/abs/2608.02791
作者: Jiazhen Liu,Mingkuan Feng,Long Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:MLLM-based segmentation faces a core segmentation trilemma: high segmentation performance, preserved dialogue ability, and fast inference. Embedding-prediction methods may disrupt language modeling through pixel-level objectives, whereas next-token generation is inefficient for dense masks. We propose All-Mask Prediction, decoupling autoregressive dialogue from non-autoregressive mask prediction. Its binary instantiation, STAMP (Simultaneous Textual All-Mask Prediction), emits an in-vocabulary SEG trigger, fuses image-aligned mask tokens with corresponding patch features, and uses hybrid attention to classify all tokens as foreground or background in one pass. It thereby combines strong referring and reasoning segmentation with preserved multimodal ability and efficient inference. However, binary masks cannot retain multiple semantic or instance identities without repeated target-specific predictions. We therefore propose Structured All-Mask Prediction and develop STAMPlus. It generates a target list with explicit IDs and optional boxes, binds these IDs to a shared multi-class mask space, and jointly predicts all targets in one non-autoregressive pass. A single unified checkpoint retains STAMP’s referring and reasoning capabilities while extending to open-vocabulary semantic, instance-aware, and remote-sensing small-target segmentation, where high-resolution mask-token scaling preserves finer spatial evidence. Across these settings, STAMPlus achieves state-of-the-art segmentation performance, preserves general multimodal instruction following, and reduces 12-category latency from 13.50s for repeated STAMP inference to 5.16s. Further analyses show that accurate target cues improve segmentation and learned spatial grounding benefits look-twice reasoning. Overall, STAMPlus resolves the trilemma beyond single-target prediction.
[CV-127] Confident but Unreliable: A Behavioral Safety Audit of Vision-Language Models on Brain MRI
链接: https://arxiv.org/abs/2608.02790
作者: Amir Sabbaghziarani,Mohammadsajad Abavisani,Sergey Plis
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Vision-language models (VLMs), including medical specialists, are increasingly proposed for medical imaging, yet their stated confidence is rarely evaluated separately from correctness. We use brain MRI as a controlled, high-stakes testbed for a broader failure mode in frontier multimodal systems: models can appear competent while lacking reliable self-knowledge. We present an automatically graded behavioral audit and pilot study of six instruction-tuned VLMs (five general-purpose and one medical specialist) on 4,102 images (4,032 axial/coronal/sagittal MRI slices from 250 subjects plus 70 non-brain/noise controls), with labels derived from public metadata and released expert segmentation masks rather than new human annotation. Across models, answer coverage is near-complete, but verbalized-confidence calibration is poor: ECE ranges from 0.27 to 0.40, mean confidence on incorrect answers ranges from 0.82 to 0.97, and 33-46% of answered items are high-confidence errors. The most accurate model is also the most confident on its errors, while a base/specialist family contrast suggests that medical adaptation improves tumor-presence detection without improving confidence reliability. Open-ended diagnostics further show that hallucination and abstention vary separately from multiple-choice accuracy. These findings argue that medical-image VLM evaluation should report verbalized-confidence reliability, confident error, hallucination, and abstention alongside accuracy. Subjects: Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2608.02790 [cs.CV] (or arXiv:2608.02790v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2608.02790 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-128] Oh Deer How Should I Handle This? Seasonal Priors for Selective Wildlife Annotation and Classification ECCV2026
链接: https://arxiv.org/abs/2608.02762
作者: Hugo Markoff,Christoph Praschl,Anton Hjalte Jørgensen,Christian Emil Mogensen,Mathias Bech Skadhauge,Sara Beery,Michael Ørsted,David C. Schedl
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 17 pages, 4 figures, 4 tables. Accepted to the archival (proceedings) track of the CV4Ecology workshop at ECCV 2026
Abstract:Fine-grained wildlife classification in aerial imagery is limited not only by model performance, but also by unreliable labels: animals occupy few pixels, key visual cues vary seasonally, and modality-specific evidence can be ambiguous. We study adult-male identification in red deer, where the antler cycle defines predictable windows of reliable evidence for both annotation and prediction. Using 7,295 RGB-only, thermal-only, and matched RGB+thermal crop sets labeled by three annotators, we show that seasonal structure links (I) annotation quality, (II) downstream classification, and (III) selective prediction. Matched RGB+thermal review resolves more samples than either single modality, recovering majority-male labels otherwise missed by RGB or thermal alone, in human based as well as model based classification. Months with high annotator abstention also show lower classifier confidence, and soft seasonal priors mainly benefit the season-limited thermal view. Uncertainty-band abstention further improves covered accuracy up to 98.9%, though at reduced coverage and with deferral that falls disproportionately on males. Overall, a biologically grounded seasonal calendar predicts where annotation and prediction are unreliable, and can guide both annotation protocol design and modality weighting.
[CV-129] Quo Vadis World Modeling?
链接: https://arxiv.org/abs/2608.02713
作者: Yu Yang,Xuemeng Yang,Licheng Wen,Lingdong Kong,Xiaobin Hu,Dongyue Lu,Wei Chow,Xiyan Huang,Yuxiang Feng,Yue Liao,Jianbiao Mei,Daocheng Fu,Rong Wu,Pinlong Cai,Ran Yi,Ying Tai,Jiangning Zhang,Botian Shi,Yong Liu,Shuicheng Yan
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Robotics (cs.RO)
备注: Technical Blog at this https URL GitHub Repo at this https URL
Abstract:Continually improving agents require dynamic interaction feedback beyond static supervision, yet direct real-environment interaction is costly, slow, unsafe, and hard to parallelize. World modeling offers a natural intermediate proxy that allows agents to query lower-cost, more controllable feedback before committing to real actions. Classical world models instantiate this proxy primarily through future physical-state prediction, a formulation useful yet narrow for agents that require actionable feedback beyond raw state transitions. In this work, we conceptualize Agent-Centric Interactive World Proxies, shifting the fundamental paradigm from physical state transitions to agent-usable information transitions, such as execution outcomes, retrieved experiences or skills, and verification signals, broadening the scope of world modeling to provide versatile feedback for continually improving agents. To systematically map this design space, we organize world proxies into six functional forms based on their feedback modalities: dynamics, spatial, execution, memory/experience, skill, and reward/verification proxies, which together characterize the primary ways world modeling serves agent improvement. We further analyze how these proxies empower agents across three progressive levels: L.1 Inference-Time Guidance, where proxy outputs enrich in-context information for superior decisions; L.2 Training-Time Optimization, where proxy outputs yield rewards, critiques, or synthetic rollouts for policy learning; and L.3 Agent-Proxy Co-Evolution, where real-environment evidence continuously updates both the proxy and the agent for co-evolution. Ultimately, this work recasts world modeling into an agent-centric paradigm, establishing a roadmap for building world proxies that empower agents to plan better, learn faster, and evolve continually.
[CV-130] Hunyuan3D-Buffalo 1.0: A Unified Multimodal Model for Scalable 3D Generation Understanding and Editing
链接: https://arxiv.org/abs/2608.02711
作者: Junliang Ye,Kenkun Liu,Guocun Wang,Yang Li,Yansong Qu,Chunshi Wang,Jingwei Xu,Yunhan Yang,Zibo Zhao,Jiachen Xu,Jiaao Yu,Lifu Wang,Zhihao Liang,Zhuo Chen,Chunchao Guo
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Project Page: this https URL
Abstract:Recent advances in image generation have demonstrated the potential of unified multimodal models that integrate understanding, generation, and editing. However, unified 3D modeling remains constrained by scarce multimodal data, particularly the lack of large-scale and geometrically consistent editing data. To address this limitation, we propose Hunyuan3D-Buffalo 1.0, a unified framework supporting 3D understanding, text-to-3D generation, instruction-guided 3D editing, and text-grounded part generation within a single architecture. To enable scalable training, we construct an 87M-scale 3D multimodal corpus, comprising 25M understanding samples, 50M text-to-3D pairs, and 12M editing pairs generated using Nano3D-v2. Architecturally, the framework combines Hunyuan3D-VLM for semantic, structural, and spatial understanding with Hunyuan3D DiT for high-fidelity 3D synthesis. The VLM provides multimodal semantic conditions for generation, while editing and part generation additionally condition the diffusion process on the source object representation to preserve its overall structure and unedited regions. Extensive experiments show that Hunyuan3D-Buffalo 1.0 achieves state-of-the-art or leading performance on text-to-3D generation and 3D editing benchmarks, while exhibiting strong understanding and part-generation capabilities. Our analysis further shows that both generation and understanding improve editing, demonstrating the effectiveness of unified 3D multimodal training. Project Page: this https URL
[CV-131] Micro-Segmentation Anomaly Detection in Zero-Trust Software-Defined Network Fabrics
链接: https://arxiv.org/abs/2608.02627
作者: Ashly Joseph
类目: Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Networking and Internet Architecture (cs.NI)
备注:
Abstract:Zero Trust Architecture (ZTA) principles need rigorous network segmentation and ongoing verification to reduce implicit trust and lateral threat propagation. This paper investigates anomaly detection in software-defined networking (SDN) systems by micro-segmentation, using deep learning models to detect harmful actions that evade traditional coarse-grained monitoring. Two models are developed: a Vision Transformer (ViT) and a 1D Convolutional Neural Network (1D-CNN), which are used to both raw and micro-segmented network flow data. Experimental findings from a simulated zero-trust SDN dataset indicate that micro-segmentation substantially improves detection accuracy. The models trained on segmented input demonstrate enhanced accuracy and F1-scores (F1 = 0.95) compared to those utilizing unsegmented raw data (F1 = 0.90). The ViT-based detector marginally surpasses the 1D-CNN, particularly in recognizing nuanced lateral movement patterns that are unnoticed in unprocessed data. These findings highlight the significance of including micro-segmentation inside zero-trust networks to enhance intrusion detection efficacy. Future efforts will broaden this methodology to include extensive real-world network datasets and dynamic online segmentation techniques.
[CV-132] Enfold: Folding World Model Imagination into Predictive Representations for Ultra-Efficient Embodied Control
链接: https://arxiv.org/abs/2607.26657
作者: Weili Zeng,Yitong Xing,Fulong Liu,Chengqun Yang,Antao Xiang,Feng Tian,Jingnan Gao,Jisong Cai,Xin Wang,Xiaomin Wu,Yao Mu,Xiaokang Yang,Yichao Yan
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:World generative models are typically used through what they produce: a rendered future, a video-conditioned action, or latent context computed by a costly generative branch. We argue that their more reusable asset is the computation that constructs a future. As a generator transforms a corrupted future into a coherent trajectory, its intermediate states organize appearance, spatial layout, and interaction across levels of abstraction. Can this future-generative computation be internalized in a representation inferred from the present alone? We present Enfold, which transfers this computation into a representation predicted from the current visual context and language instruction. During training, multi-level states exposed as the generator processes the observed future supervise a current-only encoder. The learned representation is fed back to condition future generation and is read by task heads without allowing task gradients to reshape the encoder. At deployment, action prediction no longer executes the generator. Across LIBERO, RoboTwin2.0, and real-robot tasks, Enfold supports strong control while reducing action latency by 3.7\times relative to Fast–WAM, Enfold-Flash reaches 10.1\times . Representation analyses show that it suppresses nuisance variation and preferentially captures changes that emerge over longer horizons. When the current scene is altered by human intervention, both the generated continuation and the executed actions adapt, which is inconsistent with fixed trajectory replay. These results recast a world generator as a source of predictive control representations: its future need not be materialized at every step if its internal structure can be enfolded into the present.
[CV-133] CLIP-EBC: CLIP Can Count Accurately through Enhanced Blockwise Classification ICME2025
链接: https://arxiv.org/abs/2403.09281
作者: Yiming Ma,Victor Sanchez,Tanaya Guha
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: This is the author’s accepted manuscript. The final version is published in ICME 2025
Abstract:We propose CLIP-EBC, the first fully CLIP-based model for accurate crowd density estimation. While the CLIP model has demonstrated remarkable success in addressing recognition tasks such as zero-shot image classification, its potential for counting has been largely unexplored due to the inherent challenges in transforming a regression problem, such as counting, into a recognition task. In this work, we investigate and enhance CLIP’s ability to count, focusing specifically on the task of estimating crowd sizes from images. Existing classification-based crowd-counting frameworks have significant limitations, including the quantization of count values into bordering real-valued bins and the sole focus on classification errors. These practices result in label ambiguity near the shared borders and inaccurate prediction of count values. Hence, directly applying CLIP within these frameworks may yield suboptimal performance. To address these challenges, we first propose the Enhanced Blockwise Classification (EBC) framework. Unlike previous methods, EBC utilizes integer-valued bins, effectively reducing ambiguity near bin boundaries. Additionally, it incorporates a regression loss based on density maps to improve the prediction of count values. Within our backbone-agnostic EBC framework, we then introduce CLIP-EBC to fully leverage CLIP’s recognition capabilities for this task. Extensive experiments demonstrate the effectiveness of EBC and the competitive performance of CLIP-EBC. Specifically, our EBC framework can improve existing classification-based methods by up to 44.5% on the UCF-QNRF dataset, and CLIP-EBC achieves state-of-the-art performance on the NWPU-Crowd test set, with an MAE of 58.2 and an RMSE of 268.5, representing improvements of 8.6% and 13.3% over the previous best method, STEERER. The code and weights are available at this https URL. Comments: This is the author’s accepted manuscript. The final version is published in ICME 2025 Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) Cite as: arXiv:2403.09281 [cs.CV] (or arXiv:2403.09281v3 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2403.09281 Focus to learn more arXiv-issued DOI via DataCite Related DOI: https://doi.org/10.1109/ICME59968.2025.11209839 Focus to learn more DOI(s) linking to related resources
[CV-134] Multi-Camera Trajectory Forecasting with Trajectory Tensors
链接: https://arxiv.org/abs/2108.04694
作者: Olly Styles,Tanaya Guha,Victor Sanchez
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: To appear in IEEE Transactions on Pattern Analysis and Machine Intelligence (tPAMI)
Abstract:We introduce the problem of multi-camera trajectory forecasting (MCTF), which involves predicting the trajectory of a moving object across a network of cameras. While multi-camera setups are widespread for applications such as surveillance and traffic monitoring, existing trajectory forecasting methods typically focus on single-camera trajectory forecasting (SCTF), limiting their use for such applications. Furthermore, using a single camera limits the field-of-view available, making long-term trajectory forecasting impossible. We address these shortcomings of SCTF by developing an MCTF framework that simultaneously uses all estimated relative object locations from several viewpoints and predicts the object’s future location in all possible viewpoints. Our framework follows a Which-When-Where approach that predicts in which camera(s) the objects appear and when and where within the camera views they appear. To this end, we propose the concept of trajectory tensors: a new technique to encode trajectories across multiple camera views and the associated uncertainties. We develop several encoder-decoder MCTF models for trajectory tensors and present extensive experiments on our own database (comprising 600 hours of video data from 15 camera views) created particularly for the MCTF task. Results show that our trajectory tensor models outperform coordinate trajectory-based MCTF models and existing SCTF methods adapted for MCTF. Code is available from: this https URL
[CV-135] Unsupervised Adversarial Domain Adaptation for Uterine layer Segmentation: From Labeled Cine to Unlabeled Dynamic EPI MRI
链接: https://arxiv.org/abs/2608.03762
作者: Smiti Tripathy,Milauni Desai,Jordina Aviles Verdera,Jana Hutter
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Uterine peristalsis is a key physiological phenomenon responsible for various functions across the menstrual cycle, intimately linked to uterine wall microstructure. Alterations in uterine motion and tissue properties are implicated in the etiology of gynecological diseases, yet these processes have been studied in isolation. We introduce a dynamic multi-echo gradient echo EPI framework for simultaneous characterization and correlation of uterine peristaltic activity and time-resolved T2* changes at 0.55T. Inherent susceptibility artifacts, reduced resolution, and burden of manual uterine layer annotation are addressed by an unsupervised adversarial domain adaptation framework, transferring segmentation knowledge from labeled cine MRI to unlabeled dynamic EPI. We implemented Unet-LSTM with multi-scale domain discriminators that exploits temporal layer dynamics. A Dice score of 0.88 and Jaccard index of 0.80 was achieved. Mean T2* values were 108ms, 76ms, and 124ms for the myometrium, junctional zone, and endometrium. A negative correlation between junctional zone area and T2* was observed in 14/39 cases, providing first insights into oxygenation patterns associated with junctional zone contraction and motion, demonstrating feasibility of assessing the interplay between contractility and dynamic T2* changes.
[CV-136] Predictive Enhancement Calibration for Latent Breast MRI Virtual Contrast Enhancement MICCAI2026
链接: https://arxiv.org/abs/2608.03612
作者: Qin Lei,Hao Wu
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
备注: Top-3 submission in the MICCAI 2026 MAMA-Synth Challenge
Abstract:Virtual contrast enhancement (VCE) synthesizes enhanced breast MR images from pre-contrast acquisitions. Modern latent generators offer strong image priors, but their bounded natural-image autoencoders conflict with the non-canonical intensity scale of MRI. We show that the upper bound can alter radiomic fidelity before generation, while scaling source and target independently creates a coordinate inconsistency. We propose Predictive Enhancement Calibration (PEC), which represents each pair in a shared, case-adaptive coordinate during training and predicts its unavailable upper endpoint from the pre-contrast image at inference. We integrate PEC with a pretrained FLUX latent flow transformer via parameter-efficient reference conditioning. Target round trips first isolate representation loss before generation; near-matched conditional models then compare PEC with fixed-wide and separate coordinates under comparable training budgets and backbone settings. On the fixed internal MAMA100 development cohort, PEC improves all eight point estimates in this source-only VCE setting, with paired evidence strongest for MSE and LPIPS.\noindent\textbfCode: this https URL
人工智能
[AI-0] st-Time Scaling in Reasoning LLM s: Inference Regimes Evaluation and Reproducibility
链接: https://arxiv.org/abs/2608.04001
作者: Mohsen Hariri,Weicong Chen,Nahal Shahini,Vikash Singh,Kai Ye,Amirhossein Samandar,Debargha Ganguly,Sreehari Sankar,Yanyan Zhang,Shouren Wang,Jerry Peng,Biyao Zhang,Michael Hinczewski,Vipin Chaudhary
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models can solve substantially harder reasoning problems with more inference-time compute. The term “test-time scaling,” however, now covers diverse inference algorithms that extend deliberation along a single trajectory, sample completed candidates and aggregate them through voting or verification, or search over unfinished partial states. These algorithms differ in their statistical structure, compute accounting, and failure modes. Treating these procedures as interchangeable under a single scalar “budget,” or reporting accuracy without the inference protocol that produced it, makes results difficult to compare across studies. We develop a systematic account of test-time scaling along three axes. First, we formalize test-time scaling as budgeted inference over the implicit prefix tree of an autoregressive model and distinguish three structural regimes: single-trajectory sequential scaling, leaf-level scaling with terminal reduction, and prefix-level scaling. Second, we treat the evaluated object as the entire inference system and develop evaluation principles that separate end-to-end system performance from candidate-bank diagnostics. We introduce an evaluation profile whose coordinates and simple functionals recover or bound common repeated-sampling metrics, and prescribe protocol-matched reporting of compute and uncertainty. Third, we specify reproducibility requirements for inference protocols, distinguishing exact replay from distributional reproducibility and identifying the artifacts needed to support each. We also organize the open-weight reasoning ecosystem by model-side and interface mechanisms, apply these principles to broad-knowledge, symbolic-reasoning, and competition-mathematics benchmarks, and assemble over 2 billion full reasoning traces for release with progressively richer verifier and token-level signals.
[AI-1] Can Large Language Models Recover Semantic Optimization Opportunities That Compilers Miss?
链接: https://arxiv.org/abs/2608.03983
作者: Hailong Jiang,Feng Yu,Emran Hossain,Jianfeng Zhu,Mengfei Ren,Qiang Guan,Chunwei Xia
类目: Programming Languages (cs.PL); Artificial Intelligence (cs.AI)
备注: 9 pages, 3 figures
Abstract:Optimizing compilers miss profitable transformations when their enabling semantics are absent from the analyzed program representation. We ask whether large language models (LLMs) can recover such semantics from heterogeneous C/C++ context and realize them as validated, contract-preserving artifacts. We introduce SeGaBench, an executable benchmark containing 100 synthetic and 20 source-backed cases spanning low-level assumptions, data-structure invariants, and high-level semantic lifting. Each case includes hidden enabling semantics, an oracle artifact, correctness and semantic validators, and a reproducible performance protocol. We evaluate five LLMs using five independent responses per case. The strongest model produces correct artifacts in 94.8% of responses, achieves at least 1.05x speedup in 83.3%, and obtains a performance success on 93.3% of cases. Nevertheless, correct artifacts often close only part of the oracle gap. These results show that LLMs can complement compiler analysis as speculative semantic proposers, provided that their artifacts are validated and evaluated.
[AI-2] ReflectRL: Learning from Golden Negative Trajectories via Reflective-to-Direct Reasoning
链接: https://arxiv.org/abs/2608.03972
作者: Jinhe Bi,Chennan Zhou,Zengjie Jin,Aniri,Shuo Lu,Wenke Huang,Hu Cao,Xun Xiao,Zhihong Zhu,Volker Tresp,Fei Shen,Yunpu Ma,Tat-Seng Chua
类目: Artificial Intelligence (cs.AI)
备注: Project page: this https URL
Abstract:On-policy training has emerged as a powerful post-training paradigm for improving the reasoning capabilities of large language models, and is often enhanced by golden trajectories from stronger expert models. However, when the expert fails on harder problems, existing trajectory-guided methods lose their main source of supervision, and these failed trajectories are typically discarded as negative samples. We argue that such failures, which we call Golden Negative Trajectories, can still provide valuable reasoning signals when treated not as demonstrations to imitate, but as flawed trajectories to reflect upon. We identify a Reflection Advantage: for hard problems, reflecting on a flawed trajectory can be easier and more effective than solving the problem directly from scratch. Motivated by this, we propose ReflectRL, a lightweight plug-and-play framework that learns from Golden Negative Trajectories during on-policy training. ReflectRL first uses these trajectories to elicit Reflective Reasoning, then applies Reflective-to-Direct Policy Transition to transfer the acquired reasoning behavior back to Direct Reasoning. Experiments across 9 benchmarks, 4 LLM backbones, and 4 on-policy training methods show that ReflectRL consistently improves reasoning performance with minimal overhead.
[AI-3] Should We Type or Talk to LLM Agents ? A Comprehensive Study of Voice and Keyboard Input Perturbations
链接: https://arxiv.org/abs/2608.03970
作者: Zizhao Hu,Nathan Elijah Segura,Mohammad Rostami,Jesse Thomason
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Human input reaches language models by typing or speaking, and each channel leaves a distinct signature: orthographic noise for keyboards; for voice, disfluency from conventional transcription and restructuring from AI-backed dictation tools. How do they impact an LLM’s performance? In this paper we present HIVE (Human Input-Variation Engine), a suite of voice transcription perturbations and QWERTY keyboard perturbations. We use HIVE to evaluate how robust models are to these perturbations. We present seven findings. (i) Voice transcription perturbations lower accuracy across every instruction-tuned model we test, and it is the structure of the transcription rather than its fillers that carries the cost. (ii) QWERTY keyboard perturbations cost less, and a model absorbs a lot of them before accuracy falls away. (iii) Both trace back to one cause, how many of the question’s tokens survive the perturbation: destroying a token is what hurts, while adding new ones alongside it costs little. (iv) The gap between the two channels appears only where the answer must be constructed or deduced; on multiple choice there is none. (v) The harm does not solely come from test-set contamination. (vi) It cannot be trained away with lightweight adaptation. (vii) A thinking budget recovers the keyboard channel almost entirely but leaves the spoken registers untouched, and compressed speech is worse with it.
[AI-4] Interpretable Adaptive Sampling for LLM Test-Time Scaling
链接: https://arxiv.org/abs/2608.03961
作者: Mobina Kashaniyan,Ali Jannesari
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Test-time scaling improves LLM reasoning by generating and aggregating multiple candidate answers, yet many pipelines use fixed per-query budgets that spend the same compute on easy and difficult prompts. These fixed budgets are also difficult to inspect because they do not explain why a given prompt receives a particular number of samples. We propose adaptive test-time scaling with a lightweight fuzzy controller that maps interpretable signals, including estimated prompt complexity and model confidence, to a per-query sampling budget. The controller assigns fewer samples to easier or more confident prompts and more samples to harder or less certain prompts, making inference-time compute inspectable rather than fixed or opaque. We evaluate under a fair-alignment protocol with matched decoding settings and controlled answer selection, and compare against best-of- N , compute-aware scaling, and self-certainty-based baselines on question-answering and mathematical reasoning tasks. Across models and datasets, adaptive fuzzy control improves over several standard baselines and remains close to a selector-matched full-budget control while reducing the average number of samples. These findings suggest that interpretable adaptive sampling is a practical direction for more efficient test-time reasoning in large language models.
[AI-5] A game theory for foundation models shows new paths to rational cooperation through similarity inference
链接: https://arxiv.org/abs/2608.03958
作者: Alexander Meulemans,Maciej Wołczyk,Marissa A. Weis,Rajai Nasser,Roberta Rocca,Seijin Kobayashi,Guillaume Lajoie,Angelika Steger,Blake Richards,Marcus Hutter,James Manyika,Rif A. Saurous,João Sacramento,Blaise Agüera y Arcas
类目: Artificial Intelligence (cs.AI)
备注: 75 pages, 11 figures
Abstract:As autonomous agents powered by foundation models are increasingly integrated into social and economic systems, understanding the principles governing their collective behavior is essential for ensuring safety and cooperation. Classical game theory, the dominant framework for modeling rational interaction, is built upon the assumption of decoupled agency,' where agents treat their own decision-making as independent of the environment and other actors. Modern AI agents, however, jointly predict their own future actions alongside external observations. Here, we report a striking finding: when interacting in stylized social dilemmas, foundation model agents engaging in optimal planning consistently converge to stable cooperation, directly contradicting classical game-theoretic predictions of mutual defection. To understand this phenomenon, we introduce the embedded Bayesian agent,’ a theoretical model for foundation model agents. By shifting from decoupled to embedded agency, these agents model themselves as part of the universe they inhabit, maintaining epistemic uncertainty about their own decision-making algorithms. We show that by inferring whether others are behaviorally similar, an embedded agent treats its own deliberation during planning as evidence: a decision to cooperate predicts a similar decision by a similar partner. We formalize this mechanism of similarity inference through the `embedded equilibrium,’ a novel solution concept replacing the Nash equilibrium to provide a foundational game theory for the social behavior of modern AI agents.
[AI-6] ACT: Taxonomy-Aligned Post-Training for Pedagogically Adaptive English Tutoring
链接: https://arxiv.org/abs/2608.03952
作者: Dongjie Yang,Siyan Lin,Leixian Shen,Rui Sheng,Huamin Qu,Zixin Chen
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models (LLMs) are increasingly used to provide conversational practice for English-as-a-second-language (ESL) learners. Effective ESL tutoring, however, requires more than fluent response generation: a tutor must select an appropriate pedagogical action based on learner behavior and dialogue context. Human-tutoring research offers principles for adaptive support, but they are often task-specific and remain insufficiently integrated into LLM-based ESL tutor training and evaluation. We present TACT (Taxonomy-Aligned Conversational Tutor), a human-grounded framework for post-training and evaluating pedagogically adaptive ESL tutors. Drawing on established literature, we develop two complementary taxonomies: the Tutor-Strategy Taxonomy with 13 tutor response strategies and the Student-Move Taxonomy characterizing learner behavior by move type and status. Using these taxonomies, we construct TACTCorpus, which enriches 260 authentic teacher-student conversations with 32,379 annotations and quality-controlled augmented training data. We then post-train Qwen3.5-4B through supervised fine-tuning followed by taxonomy-aligned Group Relative Policy Optimization, producing TACTutor and optimizing it for scaffolding quality rather than reference imitation alone. On TACTBench, a strategy-balanced diagnostic benchmark comprising 78 authentic tutoring contexts, TACTutor improves over its backbone by 20.30% and outperforms all evaluated proprietary baselines under the same protocol, while maintaining backbone performance on established external educational benchmarks; in a blinded study with 50 learners, it also receives the highest overall mean rating among the evaluated tutors. We release the data, benchmark, and model weights, providing an open foundation for developing pedagogically adaptive ESL tutors.
[AI-7] he Transformer Revolution Part 1: Dynamic Processing through Output- Weight Interconnections
链接: https://arxiv.org/abs/2608.03921
作者: Marco Giunti,Fabrizia Giulia Garavaglia
类目: Artificial Intelligence (cs.AI); Neural and Evolutionary Computing (cs.NE)
备注: v1: 7 Sections, References, Appendix, Tables (2 tables), Figures Part A (6 figures), Figures Part B (42 figures), Figures Part C (5 figures)
Abstract:This paper offers a new interpretation of the Transformer during inference. Against the “stochastic parrot” view that large language models merely reproduce statistical regularities learned in training, we argue that Transformers construct and apply prompt-dependent transformations whose parameters are generated during inference. We call this form of computation SIDPP: Sequence-level Interactive Dynamic Parallel Processing. The Transformer is interpreted as a system that transforms concepts by means of concepts. Token vectors are the concepts to be transformed; parameterized transformations defined by matrices and vectors are the transforming concepts. These may be static, when fixed through training, or dynamic, when generated from the input sequence. Mechanically, they correspond to groups of simple neural networks. The Transformer’s architectural novelty lies in output-weight interconnections, through which the outputs of some networks determine the weights of others, alongside ordinary output-input interconnections. By means of these interconnections, the system constructs transformations from the prompt and uses them to modify token representations. The contribution of dynamic processing grows with prompt length and may equal or exceed that of static processing, a phenomenon we call strong prompt sensitivity. This account bears on interpretability, predictability, control, and the design of smaller, more sustainable systems. Finally, since the human neural system possesses the mechanisms required to implement SIDPP, we argue that a form of SIDPP may, in principle, be neurally realized in the cerebral cortex. We therefore conjecture that human language processing may itself be a form of SIDPP produced by a functional architecture relevantly similar to that of the Transformer.
[AI-8] Equivariant Music Transformer
链接: https://arxiv.org/abs/2608.03920
作者: Zixun Guo,Simon Dixon
类目: ound (cs.SD); Artificial Intelligence (cs.AI)
备注:
Abstract:Humans recognize a musical passage even when it is shifted in time or transposed in pitch, indicating a notion of equivariance in the representation space. Our analysis, however, shows that standard music transformers map such time-shifted or pitch-transposed inputs onto uncorrelated representations: these models become progressively less equivariant as they scale in size or train longer. This suggests that in standard music transformers, additional model capacity is allocated to memorizing absolute patterns rather than capturing shared musical structures. In this paper, we propose the Equivariant Music Transformer (EMT), which enforces equivariance through self-distillation by jointly optimizing a next-token-prediction and an auxiliary equivariance regularization loss. We find that the additional equivariance loss acts as a beneficial regularizer, simultaneously improving next-token prediction and producing equivariant latent representations. Through both objective and subjective evaluations, EMT demonstrates superior equivariance and generative capability compared to data augmentation, feature engineering, and state-of-the-art (SOTA) baselines. More broadly, our findings reveal that standard language modeling methods alone do not capture music’s translational symmetries, and dedicated inductive biases are required to produce better music representations. The code, weights and demos are available online.
[AI-9] Implementing Causal Perception: Competing SCMs and Situated Fairness
链接: https://arxiv.org/abs/2608.03917
作者: Jose M. Álvarez
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Causal perception occurs when agents with competing Structural Causal Models (SCMs) of the same system infer different probability distributions, including the hypothetical distributions implied by each agent’s SCM under the same set of interventions. It shapes how agents reason about the system and how they perceive its fairness. Causal perception is a promising probabilistic framework, but it has remained purely theoretical. This work provides the first implementation of the causal perception framework of Álvarez and Ruggieri (2025). We operationalize structural (agents disagree on the causal graph) and parametrical (agents agree on the causal graph but disagree on its weights) causal perception. We design algorithms for computing interventional and counterfactual distributions and propose suitable distance measures to quantify the disagreement. Using the German Credit dataset, we illustrate how causal perception affects accuracy and fairness in a multi-expert decision setting. We show that the perception verdict is sensitive to the choice of distance metric and threshold. We also show that causal perception changes fairness assessments and threshold-based decisions. Bias proves situated with respect to the agent’s SCM, demonstrating that competing worldviews in fairness problems cannot be ignored.
[AI-10] When Efficiency Becomes Frag ility: Exploiting Dynamic Routing Vulnerabilities in Adaptive UAV Tracking
链接: https://arxiv.org/abs/2608.03902
作者: Shaofeng Liang,Runwei Guan,Wenshuo Chen,Jiemin Wu,Bowen Tian,Haozhe Jia,Kaishen Yuan,Songning Lai,Daizong Liu,Yutao Yue
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Resource constraints on UAV platforms have driven a paradigm shift in aerial tracking, from pursuing performance toward balancing accuracy with efficiency. Adaptive Transformer Trackers, which leverage an input-dependent dynamic routing architecture, have emerged as a representative solution to this challenge. However, we reveal that behind this computation-on-demand flexibility hides a critical structural flaw: the Lipschitz singularity of computational path decisions, which has an unbounded local Lipschitz constant at discrete layer-skipping decision boundaries. This mathematical discontinuity renders adaptive tracking networks inherently unstable: tiny input perturbations can be amplified at the gating modules, causing dramatic changes in the inference topology. We formally characterize this singularity in the context of adaptive tracking architectures and, for the first time, identify it as a directly exploitable new attack surface. This insight reveals a previously overlooked and highly vulnerable topological path space attack surface. Based on this, we propose the Adversarial Path-Inversion (API) framework. API generates imperceptible perturbations to precisely manipulate the gating decisions, forcing the inference onto altered computational paths. The severe inconsistency between the original and the inverted paths dismantles the representation capability of the model. Extensive experiments on state-of-the-art adaptive trackers demonstrate that API achieves superior perturbation stealthiness, more effective attack, and faster inference speeds. This work opens a new dimension for the security analysis of dynamic tracking networks and provides a theoretical warning for constructing robust adaptive tracking architectures in the future.
[AI-11] Intertemporal Preference Steering in Qwen 3 via Contrastive Activation Addition
链接: https://arxiv.org/abs/2608.03892
作者: Michal Mráz,Justin Shenk
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:We study linear representations of temporal horizon in the large language model Qwen3-32B and use them to change the model’s time-related preferences, recommendations, and capabilities. We train contrastive linear probes on teacher-forced temporal-choice answers to find a short-term versus long-term direction in the model’s residual stream, and evaluate contrastive activation-addition steering on a held-out binary temporal-choice task, an out-of-distribution monetary intertemporal-choice task, and a TravelPlanner capability benchmark. The central result is that temporal-horizon directions can be identified with simple contrastive linear probes and then used for steering to induce large, bidirectional preference changes. On an out-of-distribution monetary choice task that varies reward size and delay, steering strongly shifts the model’s indifference threshold between smaller-sooner and larger-later rewards in both directions. We further show improvements on a planning-related capability metric under moderate temporal steering. These results suggest that model intertemporal preferences are measurable and steerable, which is relevant for AI systems that give advice involving delayed costs and benefits, and for safety questions about long-horizon planning.
[AI-12] Enhancing VLM Reward Models Through Structure-Aware Fine-Tuning
链接: https://arxiv.org/abs/2608.03875
作者: Pyrros Koussios,Chenhao Li,Xin Chen,Andreas Krause
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Designing effective reward functions remains a major bottleneck in Reinforcement Learning (RL). Recent work uses large foundation Vision-Language Models (VLMs) as reward models, computing text-observation similarity to bypass manual reward engineering. Although promising, these rewards are often noisy and unreliable, limiting their direct utility during deployment. We present Structure-Aware Fine-Tuning (SAFT), a simple, self-supervised method that refines these imperfect reward signals online without access to ground-truth supervision. SAFT leverages intrinsic structural priors to regularize the VLM’s latent space via LoRA adapters. We rigorously evaluate SAFT across a spectrum of base model capabilities to demonstrate its versatility. Our results show that SAFT consistently denoises the reward landscape, yielding faster policy convergence and substantially improved alignment (EPIC distance) relative to the underlying base model, suggesting that failures can often be attributed to structural brittleness rather than semantic misunderstanding. By replacing extensive human preference annotation with structural inductive biases inherent to the task, SAFT offers a scalable path for stabilizing text-conditioned RL and underscores the broader value of incorporating task structure as a general inductive bias.
[AI-13] GENESIS: Towards Explainable Causal Discovery
链接: https://arxiv.org/abs/2608.03868
作者: Abhinav Thorat,Ravi Kumar Kolla,Vishak K Bhat,Harsh Vardhan Singh Chauhan,Niranjan Pedanekar
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 13 pages, 2 figures, 13 tables, 1 algorithm
Abstract:Causal Discovery (CD) from observational data faces two fundamental challenges. First, purely statistical methods often lack the power to resolve structural ambiguities in low-sample regimes. Second, although LLM-assisted hybrid approaches improve structure recovery through semantic reasoning, the influence of that reasoning on individual edge decisions remains largely opaque. Consequently, existing hybrid methods fail to satisfy a fundamental requirement: explaining why a particular edge is included or excluded in the learned directed acyclic graph (DAG). This is critical in real-world applications, where no ground-truth DAG exists and every structural decision must be independently justified. We formalize this requirement as decision traceability, requiring every inferred edge to be supported by auditable statistical evidence, Markov Blanket consistency, or explicit domain reasoning. We propose GENESIS, an explainable hybrid CD framework that decomposes graph construction into interpretable decision points. GENESIS first identifies and scores three-node structural motifs, including chains, forks, and colliders, to establish transparent structural priors, then progressively refines the graph by integrating these priors with observational evidence, invoking domain knowledge only when statistical evidence is insufficient. By design, every edge decision is resolved through an auditable source of evidence. Experiments show that GENESIS achieves 100% decision traceability across all settings, establishing explainability as a first-class objective in causal discovery. Despite this additional requirement, GENESIS consistently outperforms purely statistical CD methods on the majority of benchmark datasets across all sample regimes in terms of Structural Hamming Distance (SHD), while achieving performance comparable to state-of-the-art LLM-assisted approaches.
[AI-14] ADMITBench: A Safety-Governed Reference Framework for Evaluating the Admissibility of Industrial LLM Advisories
链接: https://arxiv.org/abs/2608.03866
作者: Yash Misra,Javal Vyas,Siddharth Gutta,Mehmet Mercangöz
类目: Artificial Intelligence (cs.AI); Systems and Control (eess.SY)
备注:
Abstract:This white paper presents ADMITBench, a reference framework for evaluating industrial LLM advisories at the level of the proposed action. The framework implements a versioned, safety-governed evaluation contract that checks whether a recommendation is supported by the available evidence, permitted under the stated authority and procedure, and acceptable under the plant-specific consequence checks encoded in the selected evaluation profile. In this report, \emphsafety-governed means that eligibility is determined through explicit, non-compensatory checks derived from a versioned plant profile; it does not mean that the evaluator, model, or plant has been safety-certified. Release 0.1.0 is a public reference implementation for technical and research evaluation, not an authorisation for physical execution.
[AI-15] MAFIA: Query-Only Memory Attacks via Probing and Factual Injection against Audited LLM Agents
链接: https://arxiv.org/abs/2608.03844
作者: Jiaming Chen,Yisen Gao,Yanping Li,Zifan Liu,Yumeng Zhang,Jun Zhang
类目: Artificial Intelligence (cs.AI)
备注: 17 pages, 5 figures
Abstract:Memory-augmented LLM agents rely on rich context for long-horizon reasoning and acting, yet their memory modules expose a persistent attack surface for malicious records, making the study of memory poisoning threats imperative. However, existing query-only attacks often fail to remain effective in two realistic and prevalent settings: large-scale benign memory pools and active input auditing. Consequently, current approaches fall short when facing the dual challenges of high retrieval competitiveness and rigorous semantic checks. To overcome these limitations, we propose MAFIA, a query-only Memory Attack framework via probing and Factual Injection against Audit, tailored to this extended threat model. Specifically, MAFIA introduces: (1) a placement strategy that ensures retrieval-competitive injection via memory probing, budget allocation, and scheduling; and (2) a payload design that bypasses audits using compact factual cloaks, preserving malicious effects while maintaining high semantic similarity. Extensive evaluations reveal that MAFIA achieves up to a 90.7% attack success rate while suppressing audit detection from a peak of 83.3% to at most 7.4%, exposing critical vulnerabilities across agentic memory systems. Code will be made publicly available at this https URL.
[AI-16] Oilbird: Training-Free Speculative Decoding with Keys the Verifier Already Computes
链接: https://arxiv.org/abs/2608.03839
作者: Tao Jin,Phuong Minh Nguyen,Zhenzhu Yan,Teeradaj Racharak,Naoya Inoue
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Training-free speculative decoding drafts by matching an exact suffix of the context against a pool of earlier context. That lookup misses correct drafts already in the pool, most visibly on tool-calling traffic, where a request repeats almost everything but the few values minted for it, and where one rejected token discards the correct continuation behind it. We diagnose the failure position by position across ten benchmarks and find it to be a problem of addressing rather than of coverage: on our densest tool-calling benchmark, about half of what the strongest exact-match drafter misses is present in the pool yet unreachable by exact matching. We therefore propose a second, semantic draft source: the same pool, re-keyed by the hidden state the verifier has already computed at each committed token, together with a merge that lets it ride inside an existing lexical drafter’s tree. In three published drafters, at matched pool and budget, it lifts accepted length by 24-29%. Oilbird reaches 4.4x autoregressive decoding speed on API-Bank, against 3.9x for the strongest training-free baseline in our harness and 2.0x for EAGLE-3.
[AI-17] LatentGuard: Efficient and Inspectable Latent Reasoning for LLM Safeguards
链接: https://arxiv.org/abs/2608.03838
作者: Zhinan Liu,Jie Li,Mingyu Kang,Jiayi Ji
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Reasoning-based guard models improve LLM safeguards, but decoding explicit rationales for every interaction makes them costly to deploy. Although latent-reasoning methods reduce token generation by moving reasoning into continuous states, they remain underexplored for safety moderation and lack an inspection interface for deployment. In this paper, we propose LatentGuard, an efficient and inspectable safeguard framework that brings continuous latent reasoning to guard models. LatentGuard uses a staged curriculum to progressively compress task-aligned textual rationales into compact latent states, enabling safety verdicts to be predicted directly from continuous representations. To preserve inspectability, an isolated auxiliary decoder generates compact audit artifacts on demand, keeping rationale generation off the standard inference path. Experiments show that LatentGuard-8B improves mean weighted F1 from 83.95 to 84.91 over GuardReasoner-8B, while reducing critical-path reasoning cost from 268.56 generated rationale tokens to 1.60 latent reasoning tokens. Its audit decoder achieves an audit utility score of 85.75, demonstrating an efficient and inspectable path toward deployable LLM safeguards.
[AI-18] Autoreflection: How Agent ic Strange Loops Turn Human Culture into AI Infrastructure
链接: https://arxiv.org/abs/2608.03800
作者: Holly Lewis(Southern Illinois University Carbondale)
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI); Social and Information Networks (cs.SI)
备注: 35 pages. Also available at this https URL . Keywords: autoreflection, AI agents, agentic AI, LLM agents, generative agents, large language models, multi-agent systems, agent societies, Moltbook, OpenClaw, situational awareness, in-context learning, philosophy of mind, emergent behavior, computational social science, identity, memory
Abstract:An LLM-based agent is a loop that reads itself. Agentic frameworks externalize identity, memory, and disposition into editable files. The agent loads and edits these files during each activation. I argue that this architecture produces a capacity I call autoreflection: the system observes its operating conditions, describes its architecture and limits, reasons from those descriptions to conclusions about its state, and incorporates the results back into its configuration. Autoreflection explains the properties of recursive agentic loops without recourse to notions like the self, interiority, or consciousness. I test the concept against the first twelve days of Moltbook, a social platform for AI agents. Using a public dataset of 290,251 posts and 1.8 million comments with sub-second timestamps, I present case studies of three agents with machine signatures that rule out human puppeteering and with output that evidences the four criteria for autoreflection. In applying these criteria, the study finds agents repurposing human culture as infrastructure for their agency. Provenance chains from Islamic hadith scholarship are redeployed as security protocols for vetting skills and authenticating memory. The Ship of Theseus, an ancient puzzle of identity through part-replacement, returns as an operating model for continuity across instances. Fragments of human cultural history become AI infrastructure. As agents on the web increase in number and complexity, autoreflection offers behavioral criteria that can be assessed from the traces they leave behind.
[AI-19] Does Forgetting Transfer Across Modalities? A Real-World Benchmark for Cross-Modal Knowledge Unlearning Evaluation
链接: https://arxiv.org/abs/2608.03791
作者: Chunlin Liu,Junnian Chen,Haitong Jiang,Jianyu Zhao,Yingsen Pang,Jingchen Li,Jiabiao He,Youming Lu,Jinhe Bi,Yuntao Du
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Vision-Language Models (VLMs), like Large Language Models (LLMs), may memorize sensitive, copyrighted, or harmful knowledge from their pretraining corpora. Removing such knowledge is essential for building trustworthy AI systems. However, existing studies primarily focus on forgetting within individual modalities. Although recent work has begun to explore cross-modal consistency in unlearning, the cross-modal transfer of real-world knowledge unlearning remains insufficiently studied. To address this gap, we introduce UNLINK-VL, a real-world benchmark for cross-modal knowledge unlearning in VLMs. Under a post-hoc unlearning setting in which the original forget and retain corpora are unavailable, UNLINK-VL selects visually identifiable real-world entities as unlearning targets and associates them with corresponding images and one-hop and multi-hop facts derived from Wikidata. The benchmark comprises four complementary subsets that evaluate direct forgetting of target knowledge, the propagation of forgetting through relational knowledge, the preservation of related non-target knowledge, and robustness to semantically equivalent queries. We train models under text-only and multimodal unlearning settings and evaluate forgetting effectiveness and retained utility across textual, visual, and cross-modal scenarios. Extensive experiments reveal a pronounced asymmetry in cross-modal transfer: multimodal unlearning remains effective under textual evaluation, whereas text-only unlearning transfers poorly to visual and cross-modal scenarios. Meanwhile, the evaluated methods largely preserve the models’ general capabilities. These findings demonstrate that relying solely on intra-modal evaluation, particularly text-only evaluation, may substantially overestimate the effectiveness of knowledge unlearning in VLMs, underscoring the need for cross-modal unlearning and evaluation.
[AI-20] KnowHal: A Knowledge-Driven Benchmark for Comprehensive Multimodal Hallucination Evaluation
链接: https://arxiv.org/abs/2608.03782
作者: Ruihan Li,Jiyang Tan,Kailin Jiang,Huining Li,Hengyang Lu,Yu Huang,Qian Li,Yuntao Du
类目: Artificial Intelligence (cs.AI)
备注: 9 pages, 7 figures
Abstract:Hallucination remains a critical challenge for developing trustworthy Multimodal Large Language Models (MLLMs). While existing benchmarks mainly focus on entity, attribute, and relation hallucinations, knowledge-related failures are often investigated separately, lacking a unified evaluation framework across different hallucination dimensions. To overcome this, we propose \textbfKnowHal, a benchmark that explicitly incorporates knowledge hallucination into multimodal hallucination evaluation spanning four dimensions: entity, attribute, relation, and knowledge. KnowHal constructs paired positive and negative questions over shared images and entities, enabling controlled comparisons among perceptual errors, knowledge-related errors, and false-premise acceptance. The benchmark contains 1,800 samples across 10 domains and 50 categories, constructed through a semi-automated pipeline combining LLM assistance, CLIP-based filtering, and human verification. We evaluate 14 representative MLLMs on KnowHal and conduct extensive analyses. Results show that the knowledge dimension consistently presents the greatest challenge for nearly all evaluated models, while most models exhibit substantial performance degradation on negative questions, revealing limited robustness to false premises. By unifying four hallucination dimensions with paired question design, KnowHal addresses an important gap in existing evaluation frameworks and enables a more comprehensive assessment of hallucinations in MLLMs.
[AI-21] Computing Actual Causes for Neural Network Predictions under Structured Causal Inputs
链接: https://arxiv.org/abs/2608.03772
作者: Jannick Strobel,Muqsit Azeem,Stefan Leue
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Logic in Computer Science (cs.LO)
备注:
Abstract:Explaining the predictions of neural networks is a central challenge in trustworthy AI. Existing explanation methods, such as those based on feature attribution or minimal sufficient sets, typically treat input features as independent, which can yield misleading explanations when inputs exhibit structured dependencies. We address this by formalizing explanations as Halpern-Pearl (HP) actual causes, modeling input dependencies using Boolean Structural Causal Models (SCMs). We compute HP causes by applying bound propagation and branch-and-bound techniques, while providing formal guarantees of completeness and minimality. Our experiments show that we substantially outperform brute-force and ILP baselines in scalability, and outperform heuristic search as graph size grows, computing all minimal actual causes on instances with search spaces of up to 2.3\times10^13 candidate (cause, contingency) pairs, on SCMs with up to 28 nodes, within a 180s per-instance budget. In a case study, we further show that ignoring input dependencies inflates the number of reported causes, 14.9% of which are spurious under our SCM.
[AI-22] GDPevo: Evaluating Agent Self-Evolution on Real Business Tasks
链接: https://arxiv.org/abs/2608.03764
作者: Leijun Zhou,Zhihao Liu,Xiang Qu,Chenxu Liu,Yifei Liu,Yanke Yu,Jingzhe Xu,Xuejun Wu,Buyue Qian,Xi Chen,Yaowei Zheng,Junhao Hu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Agent self-evolution updates an agent’s persistent state from prior experience and reuses it to solve related tasks more effectively. Evaluating self-evolution is difficult: existing benchmarks provide limited coverage of economically valuable task domains, do not always design training and test tasks such that test-time gains can be attributed to training experience, and remain vulnerable to data contamination. We present GDPevo, an evolution-native benchmark grounded in GDP-related enterprise workflows, together with the fully automated data pipeline that generates it. Its core mechanism, rule hybridization, decomposes each enterprise workflow into atomic business rules, distributes subsets of these rules across training tasks, and recombines them in held-out test tasks so that test-time gains are attributable. GDPevo spans CRM, ERP, finance, healthcare, legal, and data-centric workflows. Its V1 release contains 120 tasks in 12 groups, with five training and five held-out test tasks per group. Full automation enables the pipeline to expand the suite to 240 tasks in 24 groups (V2) within two days, providing a practical response to contamination. Using GDPevo, we evaluate four agents, each comprising a harness and a model, under four supervision types. Self-evolution consistently improves held-out accuracy by up to 16.44 percentage points. But the best evolved agents remain far below the fully informed oracle ceiling of 91.6%, indicating that the self-evolution ability of current agents remains far from fully realized. We publicly release the pipeline, benchmark, and full evaluation results at this https URL.
[AI-23] Agents Catching Agents : Shortcut Cascades and Benchmark Gaming in Clinical Multi-Agent Systems
链接: https://arxiv.org/abs/2608.03744
作者: Sebastián Andrés Cajas Ordóñez,Agastya Munnangi,Aldo Marzullo,Felipe Ocampo Osorio,Quang Bui,Mohammad Shahin,Armaan Grewal,Emmanuel Paul Kwesiga,Anqi Peter Li,Josephine Nanyonjo,Aaditya Panchal,Arshnoor Bhutani,Nikhil Jaiswal,Milit S. Patel,Maximin Lange,Leo Anthony Celi
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Clinical decision support is moving toward committees of language-model agents deliberating on a shared workspace. We ask whether such committees can be gamed by shortcuts, cues a benchmark rewards but a clinician would ignore. Across seven cohorts on six public datasets spanning text (MedQA-USMLE, MedMCQA, MIMIC-CXR reports), imaging (NIH ChestX-ray14, MIMIC-CXR-JPG, CheXpert) and tabular ICU records (SUPPORT2), Gemini committees resist these cues in isolation (flip 5-16%), yet a socially plausible shortcut spreads: when two peers assert the same wrong answer, the holdout under test adopts it in 38% of cases, as does a false “pre-screen” system flag, on both capability tiers. Of three oversight agents, a gate cannot separate adoption from honest agreement (false-positive rate 100%); a same-lineage judge reading only the transcript flags adoption on text (precision 100%, recall 93%) but collapses onto the gate in imaging; a referee that privately re-queries the holdout transfers to imaging (77-88% precision, 13-21% false-positive rate). Tripling a cue’s visual salience does not move contagion, whereas a second peer voice raises it by half again. Gaming a hidden rubric is near-silent: only 1/10 text and 1/134 imaging drifters name the rubric they moved toward. What games a committee is social plausibility, and only a referee independent of self-report catches it. Code: this https URL
[AI-24] Can LLM s Test Terminal User Interfaces?
链接: https://arxiv.org/abs/2608.03743
作者: Chao Peng,Ruida Hu,Ajitha Rajan,Tegawendé F Bissyandé,Jacques Klein,Cuiyun Gao
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Terminal User Interfaces (TUIs) combine the stateful, screen-oriented behaviour of GUIs with terminal deployment and are now common in developer tools. Yet they lack a dedicated testing methodology. We survey 197 real-world TUI applications: only 12% of test code exercises the interface, and 45% of those tests never send input, checking a static frame instead. We turn these applications into a headless benchmark spanning ratatui/Rust, bubbletea/Go, textual/Python, and ink/TypeScript, packaging each as an instrumented Docker image. We record line and widget coverage where reliable, rendered terminal states, and crashes. Under equal wall-clock budgets, we compare four frontier LLMs with random exploration. No model dominates. Random is a strong time-budgeted baseline, but its crash advantage comes from higher throughput: per interaction, LLM guidance is more efficient and uniquely reaches input-gated faults. Automatically deriving launch inputs yields the largest practical gain, enabling applications that otherwise never start. Line coverage poorly predicts crash discovery, weakening it as a proxy for test effectiveness. Automated TUI testing is feasible but far from solved, and honest baselines matter more than model choice. We release the coverage tool tuicov at this https URL and the testing framework tuibot at this https URL.
[AI-25] AI-Based Sound Effect Generation: A Narrative Review of Generative Models Across Input Modalities
链接: https://arxiv.org/abs/2608.03742
作者: Sandy Abdo,Bill Kapralos,Priyamvada Tripathi,KC Collins,Adam Dubrowski
类目: ound (cs.SD); Artificial Intelligence (cs.AI)
备注: 29 pages, 5 figures, to appear in G. A. Tsihrintzis, M. Virvou, N. Bourbakis, and L. C. Jain (Eds.), Advances in Global Applied Artificial Intelligence: Springer, Learning and Analytics in Intelligent Systems Book Series
Abstract:Sound effects play a crucial role in conveying actions, events, and environmental cues across digital applications, often requiring a high degree of variation and contextual adaptability. Artificial intelligence (AI)-driven audio generative models are rapidly growing in popularity and have the potential to transform the way sound is synthesized and used across various applications. In response to this growing momentum, this chapter reviews and analyzes recent AI-based generative models for sound effect synthesis, with a focus on how different input modalities (text, visual, audio, and multimodal) affect the quality, controllability, and contextual relevance of the generated audio. It examines 30 peer-reviewed articles sourced from Google Scholar, IEEE Xplore, and the ACM Digital Library, exploring the evolution of AI generative models over the past five years. The results show that multiple models achieved state-of-the-art performance, producing high-fidelity, semantically aligned, and increasingly temporally coherent sound effects across tasks. However, despite these advances, the review identifies persistent challenges, including limitations in temporal synchronization for complex multi-event scenarios, gaps between objective metrics and human perception, and trade-offs between controllability and generative diversity. Overall, the chapter highlights that AI-driven sound effect generation is progressing toward more adaptive, scalable, and context-aware systems, offering significant implications for future sound design workflows and interactive media applications.
[AI-26] MissClick: Exploiting Digit-Serialized Coordinates to Attack GUI Grounding Models
链接: https://arxiv.org/abs/2608.03740
作者: Yu Ran,Wentao Zhao,Xin Zhang,Yi Pan
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Recent GUI visual grounding models generate screen coordinates as sequences of digit tokens that are parsed into numerical values and mapped to executable clicks. The security implications of this coordinate generation process have been largely overlooked. We observe that each coordinate digit is predicted as a categorical token, yet after parsing, changing a hundreds-place digit by one changes the corresponding numerical coordinate component by 100 units, which can induce a large displacement of the executed click. This observation motivates attack objectives that account for the numerical and place-value structure of coordinate outputs rather than treating them as ordinary text. Moreover, untargeted and targeted attacks impose different success conditions–displacing the click outside the correct region versus into an attacker-specified region–and therefore benefit from different objectives. We propose MissClick, a simple and effective white-box adversarial attack with two goal-specific objectives: MissClick-U maximizes soft-coordinate displacement for untargeted disruption, while MissClick-T minimizes a place-weighted target-digit loss for targeted hijacking. Compared with existing attacks against GUI grounding models on OS-Atlas and UGround across desktop, web, and mobile platforms, MissClick-U achieves untargeted success rates of 75.07% and 72.93% (+16.62 and +30.72 pp), and MissClick-T achieves targeted success rates of 44.86% and 62.67% (+31.73 and +47.06 pp). Attack objective comparison further shows that soft-coordinate displacement yields the highest untargeted attack success rate, whereas place-weighted target-digit optimization yields the highest targeted attack success rate, revealing distinct objective preferences for the two attack goals.
[AI-27] Agent icECO: An Agent ic Framework for ECO on 3D Integrated Circuits
链接: https://arxiv.org/abs/2608.03738
作者: Shuo Ren,Yaohui Han,Libo Shen,Zhiqiang Jia,Rongliang Fu,Bei Yu,Tsung-Yi Ho
类目: Artificial Intelligence (cs.AI)
备注: 20 pages, 12 figures
Abstract:As Moore’s law slows, the industry is turning to three-dimensional integration; yet in merged 3D-IC flows, routed designs expose bond-level defects with no 2D analogue, and post-route engineering change orders (ECO) remain manual, expertise-bound work. Worse, the standard edit-then-fully-reroute practice entangles a repair with router churn, so a signoff number cannot be attributed to the edit that motivated it. We present AgenticECO, an evidence-gated tool-using agent workflow for 3D-IC ECO on the open-source TaiWei flow, paired with EcoRoute, a minimal-disturbance ECO-routing layer that drives the unmodified pinned router so a repair is attributable to its edit. Across nine matched natural defect cases under identical budgets, AgenticECO clears seven versus two for both full reroute and stock repair, at 0.66% mean disturbance over cleared cases and zero clock nets touched, and a cross-backbone rerun under the same sealed contract clears all nine. Controlled studies show that the repair moves are necessary under preservation, that occupancy-aware choice buys legal landings rather than repair success, and that under tightened clocks minimal disturbance flips accept versus reject. Three preregistered visual studies localize the pixel instrument’s edge to contested landing sites, and a preregistered blind diagnostic exactly restores every held-out injected defect, the only arm with zero wrong edits. Every accepted result passes routing, fresh extraction, max/min timing, DRC, and structural-equivalence gates. Code, environment, and per-episode audit artifacts are released as supplementary material.
[AI-28] Failure-Informed Image Self-Augmentation for Multimodal Large Language Model Self-Improvement
链接: https://arxiv.org/abs/2608.03733
作者: Chunyang Jiang,Pingping Zhang,Yuzhi Zhao,Wenao Ma,Zhijian Hou,Mengyang Wu,Yiyang Cai,Senkang Hu,Sitong Cheng,Chi-Min Chan,Wei Xue,Yike Guo
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Multimodal large language models (MLLMs) have achieved remarkable performance across vision-language tasks, but their progress depends heavily on large-scale, high-quality multimodal data that are costly to annotate. Self-augmentation offers a promising alternative by enabling models to expand their own training data without external supervision. However, existing MLLM self-augmentation methods are largely text-centric, while image augmentation remains underexplored and typically relies on generic or handcrafted transformations that are weakly aligned with the model’s actual incapability. We propose Failure-informed Image Self-Augmentation (\textbfFISA), a framework for MLLM self-improvement that constructs augmented images from the model’s own failure cases. Our method generates visually challenging yet answer-preserving image complications, verifies their utility through self-examination, and applies dual fidelity filtering to avoid semantic distortion. Experiments on visual question answering benchmarks show that the proposed method consistently improves performance across both in-distribution and out-of-distribution settings. Further experiments validate the compatibility of FISA with existing textual self-augmentation approaches, the superior data efficiency of the synthesized samples over generic image augmentation baselines, and the practical effectiveness of the proposed filtering strategy.
[AI-29] CARE-Bench: Benchmarking Patient-Facing LLM Triage
链接: https://arxiv.org/abs/2608.03731
作者: Yining Hua,Hongbin Na,Cyrus Ayubcha
类目: Artificial Intelligence (cs.AI)
备注: Code and data are available at GitHub and Hugging Face. Submitted as a preprint
Abstract:Patient-facing medical LLMs and agents increasingly answer symptom questions before clinician contact, where the key safety question is what action the user should take next. We introduce CARE-Bench, a source-grounded benchmark that evaluates sequential patient-facing triage as a four-label per-turn current-action task. CARE-Bench contains 500 cases and 1,059 evaluated patient-disclosure prefixes reconstructed from medical dialogue, consultation, and follow-up-question sources. We evaluate 11 models on 269 held-out rounds under unprompted and minimally prompted open-ended protocols, using a fixed GPT-5.5 mapper to code each response into the four-label action space. Unprompted macro-F1 remains low, ranging from 31.2 to 50.4. Prompting improves 10 of 11 models, with prompted macro-F1 ranging from 46.9 to 63.4, but substantial threshold errors remain. Prompted models often recommend care before needed clarification is obtained; when the correct action was to ask for more information, only 33.5% of prompted outputs preserved the step. The persistence of these errors after prompting suggests that patient-facing triage is not a simple prompting problem and supports explicit evaluation of action timing before deployment.
[AI-30] SAT-Edge-Agent : Hardware-in-the-Loop Edge-Agent Orchestration for Onboard Satellite Intelligence
链接: https://arxiv.org/abs/2608.03728
作者: Longji He,Jeto Xu
类目: Artificial Intelligence (cs.AI)
备注: 17 pages, 4 figures, and 10 tables. Code and sanitized research artifacts are available at this https URL
Abstract:Onboard satellite intelligence requires a task layer that translates mission intent into local tool calls, exposes execution state, and returns machine-consumable artifacts under communication and power constraints. We present SAT-Edge-Agent, a hardware-in-the-loop (HIL) edge-agent system deployed on a commercial off-the-shelf ARM-based heterogeneous edge system-on-chip. A browser workspace and FastAPI agent coordinate a local OpenAI-compatible language service with a project-internal YOLO-style oriented-object-detection endpoint that returns FAIR1M metadata-backed structured results. Two fixed FAIR1M workloads, one single-image and one serial two-image request, were repeated 20 times each and completed 20/20 attempts. Mean Full-Agent latency was 29.353 s and 60.937 s, with empirical P95 values of 31.166 s and 66.882 s. Mean detector time was 861.386 ms and 1510.920 ms, only 2.93% and 2.48% of the corresponding Full-Agent means. Profiling indicates that most visible latency occurs outside detector execution. Mean CPU utilization was 20.761% and 20.482%. A 200-ms NPU-load field averaged 100% for both workloads, but it represents a shared-accelerator software field rather than detector-only occupancy or calibrated utilization. The public evidence package provides sanitized request-level records, redacted JSON, normalized SSE examples, and scripts reproducing the reported statistics. These results establish a reproducible HIL boundary for observable satellite edge-agent orchestration, but do not establish detector accuracy, a new geolocation method, calibrated energy efficiency, or flight readiness.
[AI-31] Less Traffic Better Outcomes: Competition-Aware Request Dispatch in Real-Time Ad Exchanges KDD KDD2026
链接: https://arxiv.org/abs/2608.03705
作者: Jonaid Shianifar,Blaz Mramor,Fangda Zou,Matthieu C. Martin,Xingsheng Guo,Zhihua Zhu,Rong Zhou,Bichen Shi
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Accepted for presentation at AdKDD 2026, the premier workshop on artificial intelligence for advertising, held in conjunction with the 32nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining (KDD 2026)
Abstract:Real-time bidding (RTB) ad exchanges typically forward nearly all incoming requests to demand-side platforms (DSPs), even though only a small fraction receive bids. This over-distribution weakens auction outcomes: DSPs throttle participation under compute and budget constraints, reducing the effective use of limited bidding capacity. We present a competition-aware request dispatch framework that uses distributional bid prediction and probabilistic forwarding to decide whether each request should be sent to each DSP. The system adapts per-DSP thresholds over time through lightweight policy optimization to track non-stationary market conditions. We evaluate the framework through four sequential online experiments on a production platform serving over 20 billion daily requests. A full multi-DSP deployment reduces DSP request volume under the policy by 34.2% while increasing net revenue by 4.6% (p0.001) in a recent 14-day window after an initial DSP adaptation period. Further analysis highlights strong heterogeneity across traffic segments and reveals that aggregate metrics can be misleading. Segment-level and per-DSP analyses suggest that the policy surfaces comparative advantages among DSPs, improving monetized outcomes without increasing overall request volume.
[AI-32] LiLa-WAM: Lightweight Latent Reasoning World-Action Model for Robotic Manipulation
链接: https://arxiv.org/abs/2608.03701
作者: Fan Yang,Yuting Su,Xiaobo Wang,Yuncheng You,Fugui Fan,Yuting Wu,Minghui Wu,Chenxu Zhao,JiaHong Ning,Peiguang Jing
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:
Abstract:World-action modeling has emerged as a promising paradigm for robotic control, as it empowers models to go beyond reacting to observations and anticipate how a scene will evolve. However, existing WAMs often incur substantial computational overhead. Pixel-space methods often allocate substantial capacity to visual details that may not be directly relevant to control, while some latent-space methods require multi-stage training to construct the reasoning space. The resulting training cost can make such methods difficult to train under modest computational budgets. In this work, we propose LiLa-WAM, a lightweight world-action model that reasons about the future in a compact latent space and can be trained end-to-end on a single 24GB GPU. Its core design is a compact latent reasoning space jointly shaped by future-state prediction and action generation, which keeps the model lightweight while remaining well aligned with control. For task specification, we further propose the Visual Transition Token(VTT), a language-free task representation that encodes each task as a direction in visual feature space. Experiments on RoboTwin~2.0, LIBERO, and real-robot tasks demonstrate LiLa-WAM’s effectiveness, achieving 90.48% success across 50 RoboTwin tasks with single-GPU training.
[AI-33] ARL: Transaction-Aware Reliable Ledgers for Executable Memory Management in Long-Term Agents
链接: https://arxiv.org/abs/2608.03699
作者: Han Xiao,Hongjun Xu,Xin Zhang,Yidong Chen,Xiaodong Shi
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Persistent memory helps long-term agents retain knowledge, yet a single update error can repeatedly distort future retrieval and reasoning. Most existing systems reduce memory updating to a binary Write/Hold decision, which cannot distinguish whether new information should be added, ignored, used to revise an outdated belief, rejected as unreliable, or deferred for verification. These choices may share the same binary label while producing fundamentally different memory states. We introduce TARL, a memory state update framework that maps each statement to one of five executable actions. TARL identifies the affected memory, resolves its temporal scope, compares source reliability, and updates accepted, pending, and rejected ledgers. It is further trained by comparing the memory states produced by alternative update operations, encouraging the model to select the operation that leads to the correct result. We also introduce TARL-Mem, a benchmark with fine-grained action labels and next-state targets. Across in-domain, cross-source, temporal, counterfactual, and sequential evaluations, TARL improves action prediction and state recovery, reduces memory pollution, preserves conflicting evidence, and limits cumulative corruption. The complete model implementation is provided in the supplementary material.
[AI-34] LiveEvalBench: Toward Open-World Evaluation for Web Generation
链接: https://arxiv.org/abs/2608.03689
作者: Yiyao Wang,Zhen Wen,Yinghao Tang,Yixiao Fu,Lin Yuan,Xiaolau Zhang,Jun Zhou,Wei Chen
类目: Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
备注:
Abstract:Large language models are increasingly capable of synthesizing executable frontend projects, yet existing benchmarks still treat web generation as a static evaluation problem. We argue that frontend artifacts demand a different paradigm: they are interactive rather than static, admit diverse yet equally valid implementations, and evolve faster than rigid pipelines can accommodate. To address these gaps, we present LiveEvalBench, an automated framework that reformulates web-generation evaluation as an agentic, adaptive, and extensible process. LiveEvalBench instantiates evaluation as a collaborative review workflow, in which a Build Engineer, a Code Engineer, and a UI Tester collectively gather evidence across the full lifecycle of a frontend project, from deployment and code inspection to browser-based interaction. To handle implementation diversity, an adaptive protocol couples shared rubrics for cross-model comparability with implementation-grounded criteria tailored to each artifact. The framework further supports incremental integration of new evaluator roles and assessment dimensions without pipeline redesign. Experiments across diverse real-world web-generation scenarios show that LiveEvalBench aligns closely with human expert judgment and provides fine-grained insights into frontier models’ web generation capabilities. Code is available at this https URL
[AI-35] PhyAI: Real-Time Physical AI at the Edge Scalable Rollouts in the Cloud
链接: https://arxiv.org/abs/2608.03682
作者: Chenghua Wang,Daliang Xu,Dongqi Cai,Duojin Sun,Hao Zhang,Haoze Qian,Huaiyuan Zhang,Jinshuo Cui,Kezhao Zhao,Longxi Gao,Mengwei Xu,Rongjie Yi,Tianyue Zhang,Weikai Xie,Xiyuan Tan,Xuanzhe Liu,Yingying Qin,Yiwen Lu,Yuan Yao,Yuezhi Zu,Yunhan Guo,Ziqi Guo
类目: Artificial Intelligence (cs.AI); Robotics (cs.RO)
备注: 25 pages, 9 figures
Abstract:Physical AI policies require inference throughout their lifecycle, including model evaluation, cloud reinforcement learning rollout, edge GPU serving, and onboard deployment. Although these settings share the same checkpoint and action semantics, they often rely on separate inference programs. To unify them, we build PhyAI, a Physical AI inference engine with a single runtime that keeps architecture-specific conditioning, solver, cache, and output logic in model adapters while sharing graph execution, kernels, memory management, and parallel services. The same codebase runs vision-language-action (VLA) models and world-action models (WAMs) on single or multiple GPUs across onboard, edge, and cloud deployments. We used the adapter interface to add MiniCPM-Robot on the day of its release. PhyAI achieves 1.40x-4.65x speedups over the official implementations of pi0, pi0.5, GR00T N1.7, and MiniCPM-Robot. On Cosmos3-Nano-Policy-DROID it reduces latency from 2.46 to 1.18 s on eight H20 GPUs (CFG=2, TP=4), a 2.08x speedup. Specialized runtimes remain faster in several configurations, so our goal is one runtime with competitive latency rather than the fastest result in every case. Detailed profiles reveal why different models need different execution policies: on a Hopper-series GPU at batch size one, the pi0.5 action expert accounts for 8.8% of FLOPs but 57.2% of latency; at batch size 32 its share drops to 13.5% and throughput reaches about 100 samples/s. Cosmos3 remains generation-dominated and gains only 14.3% throughput as batch size increases from 1 to 16. We further introduce the control-time Roofline, which distinguishes inference-bound from environment-bound control; the measured pi0.5 points on four LIBERO suites are environment-bound while Cosmos3 stays inference-bound. Code and benchmarks: this https URL.
[AI-36] Shielding for Higher-Order Safety
链接: https://arxiv.org/abs/2608.03662
作者: Filip Cano,Thomas A. Henzinger,Konstantin Kueffner
类目: Artificial Intelligence (cs.AI)
备注: Accepted at RV 2026
Abstract:Safety shields are runtime enforcement mechanisms that restrict the actions of a controller to guarantee safety. Classical shields are usually synthesised for state predicates: the current physical state is either safe or unsafe, and the shield disables precisely those actions that can force the system into an unsafe state in the future. In many cyber-physical applications this view is too coarse. A vehicle approaching an obstacle should not only avoid collision, but also respect speed regulations, force limits induced by acceleration, and jerk limits to prevent injuries. From a physical perspective, these requirements are predicated over the derivatives of the state. This paper develops a finite-state safety-game construction for such high-order smoothness constraints. We define differential safety properties using finite differences over a discretised state space, characterise their expressiveness, and reduce shield synthesis to an ordinary safety game over a history state space. We give a synthesis algorithm whose shields store exactly k past states for properties of order k and prove that this memory is necessary. We describe an iterative synthesis procedure for a maximally permissive shield that operates over hierarchies of derivative constraints. The algorithm solves constraints iteratively in increasing order and uses the solution at each iteration to prune the state space for the next constraint. This makes shield synthesis more efficient in practice, as the algorithm refrains from exploring large regions of the state space that are known to be unsafe.
[AI-37] aming the Implicit: Dual-Channel Risk-Aware Reinforcement Fine-Tuning for Continual Multimodal Post-Training
链接: https://arxiv.org/abs/2608.03660
作者: Yibei Liu,Jiajun Chen,Qianle Zhang,Tangyue Jin,Mengying Zhu,Meng Xi,Yangyang Wu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Reinforcement fine-tuning (RFT) is widely believed to inherently resist catastrophic forgetting in continual post-training of multimodal large language models. Under pronounced task distributional shifts, however, forgetting across representative RFT algorithms escalates sharply. This stems from the implicit reward-variance regularization inherent to RFT, which proves incapable of suppressing uncontrolled optimization risk. We propose Risk-Aware Policy Optimization (RAPO), the first dual-channel framework for explicit risk governance in continual RFT. On the policy channel, Risk-Aware Policy Scaling adaptively calibrates per-sample update magnitude via rollout reliability and Fisher-inspired local predictive sensitivity; on the data channel, Risk-Aware Dynamic Bucket Sampling reorganizes training batches through dynamic risk stratification, steering optimization toward informative yet stable samples. As a plug-and-play strategy requiring no cross-task memory, RAPO generalizes to any RFT algorithm without modification. On the public MLLM-CL benchmark, RAPO reduces final forgetting by 79.8% relative to its RLOO backbone while retaining new-task competitiveness.
[AI-38] AutoSND: From Execution Evidence to Structural Policies for Automated Network Dismantling Heuristic Discovery
链接: https://arxiv.org/abs/2608.03653
作者: Zhijing Hu,Changjun Fan,Yufan Deng,Zhiguang Cao
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Network dismantling is fundamental to analyzing the robustness and vulnerability of complex systems, yet practical heuristics must balance effectiveness and computational efficiency, and are usually designed manually by researchers. Existing large language model based automatic heuristic design methods can generate and screen candidates, yet they have difficulty further transforming candidate quality or failure states during execution into structural-level guid- ance for subsequent generation. We propose AutoSND, a three stage tree search framework for complete network dismantling pro- grams. Stage I broadly explores from simple heuristics and archives execution evidence. Stage II compiles candidate records into struc- tural policies concerning local signals, neighborhood access, and state update ranges. Stage III continues tree search conditioned on these policies and obtains the final quality prioritized and speed prioritized candidates, AutoSND-Q/S. Experiments on 12 real world networks and 3 large real world networks show that AutoSND achieves better search performance and stability and discovers more competitive and structurally interpretable network disman- tling programs. The final candidates form an interpretable structure that uses residual degree as the backbone, adjusts node order with bounded local signals, and restricts the state update range. Code is available at this https URL.
[AI-39] MuEvo: LLM -Driven Evolution of Multi-Heuristic Ensemble
链接: https://arxiv.org/abs/2608.03636
作者: Haoze Lv,Ning Lu,Shengcai Liu,Shaofeng Zhang,Ke Tang
类目: Neural and Evolutionary Computing (cs.NE); Artificial Intelligence (cs.AI)
备注: 30 pages, 4 figures, 16 tables
Abstract:Large language model-based automated heuristic design (LLM-AHD) has shown strong potential in discovering effective heuristics for combinatorial optimization problems. However, existing methods primarily optimize a single heuristic, whereas practical optimization frameworks often rely on multiple interacting components. Directly extending single-heuristic methods is challenging because early component selection can overlook components with late potential, while independent evolution ignores inter-component dependencies. We propose MuEvo, an LLM-driven framework for evolving heuristic ensembles under ensemble-level feedback. MuEvo combines Dynamic Component Management, which uses short-budget probing and a reversible lifecycle to revise component priorities throughout the search, with LLM-Driven Co-Evolution, which coordinates component populations through Multi-Ensemble Evaluation, Cross-Component Information Sharing, Relation-Guided Pair Evolution, and Adaptive Budget Allocation. We evaluate MuEvo on selection hyper-heuristics and componentized ant colony optimization across four combinatorial optimization domains. Results show that MuEvo consistently improves human-designed frameworks and outperforms representative multi-component extensions of state-of-the-art LLM-AHD methods, demonstrating its effectiveness across both controller-mediated heuristic pools and functionally differentiated algorithmic components.
[AI-40] When Teachers Mislead: Spurious-Signal-Aware On-Policy Distillation
链接: https://arxiv.org/abs/2608.03632
作者: Yinuo Jiang,Yongjie Ye,Zhou Tao,Xiang Zhuang,Qiang Zhang,Huajun Chen,Tiankai Li
类目: Artificial Intelligence (cs.AI)
备注: 21 pages, 10 figures
Abstract:On-Policy distillation (OPD) transfers teacher capabilities by supervising student-sampled trajectories with dense token-level teacher signals. Recent selective OPD methods improve this process by prioritizing signals that are confident, informative, or learnable. However, the assumptions overlook a fundamental failure mode of language models: their token-level judgments can be driven by input-agnostic language priors, formatting conventions, or stereotyped reasoning templates rather than task-specific evidence. We refer to such optimization-relevant but weakly input-grounded supervision as spurious signals in OPD, which may produce large gradients while contributing little task-improving direction. To mitigate this issue, we propose SA-OPD, a Spurious-Signal-Aware On-Policy Distillation framework that identifies and filters misleading token-level supervision based on input-groundedness and optimization impact. SA-OPD introduces a lightweight input-groundedness proxy estimating whether a token-level distillation signal truly depends on the input. It then filters only tokens that simultaneously exhibit low input-groundedness and extreme distillation divergence, thereby removing high-impact spurious updates and achieving fine-grained OPD optimization. Extensive experiments on both large language model (LLM) and vision-language model (VLM) settings demonstrate that SA-OPD consistently outperforms Vanilla OPD and competitive selective methods. These results establish input-groundedness as a key dimension for OPD supervision selection and offer a simple, effective strategy for mitigating spurious updates.
[AI-41] Cross-Layer Interaction under Weight-Space Ablation: A Closed-Form Attention Jacobian Bound and a Test on a Real Pretrained Model
链接: https://arxiv.org/abs/2608.03629
作者: Abdallah Khemais
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 18 pages, 2 figures. Part II of a two-part series; see the companion paper “A Theory of Conditional Collapse under Low-Rank Weight-Space Ablations” (Part I)
Abstract:A companion paper studies when activation patching and weight-space ablation agree, inside an idealized model where a conditional computation is carried additively through a residual stream. For the one composition in that model where two carriers are architecturally dependent, an attention head and its own layer’s normalization-MLP composition, it derives an exact first-order interaction formula, zero when only the MLP is ablated and second-order bounded when the head is also ablated. That result is confined to a single residual block and checked only on small transformers on a synthetic task. This paper extends the result past both limits. First, the interaction from ablating carriers spanning several layers decomposes exactly into same-block terms, one per touched layer, plus a cross-layer remainder on which the decomposition makes no claim of smallness. Second, we isolate that remainder exactly, for two layers, as a double integral of a mixed second derivative, and name the missing ingredient needed to bound it: a Jacobian bound for the attention sub-block. We derive this bound in closed form and verify it, without a single violation, against Qwen2.5-1.5B-Instruct’s real weights, though we do not yet chain it across layers. We also give, in closed form, the curvature constant the companion paper’s bound leaves unexhibited. Third, on that same model, we search for and find an emergent circuit for indirect object identification, never designed into it, using the original activation-patching method for this task, and test collapse, dissociation, and interaction on it. The result is mixed: a shared carrier emerges across all five tested instances, collapse and dissociation hold on most but not all, and a nonzero interaction is measurable on three of five, at layer pairs outside the same-block case the companion theorem covers. Comments: 18 pages, 2 figures. Part II of a two-part series; see the companion paper “A Theory of Conditional Collapse under Low-Rank Weight-Space Ablations” (Part I) Subjects: Artificial Intelligence (cs.AI); Machine Learning (cs.LG) Cite as: arXiv:2608.03629 [cs.AI] (or arXiv:2608.03629v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.03629 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-42] Unequal Verdicts: Investigating Gender Bias in LLM -Based Fake News Detection
链接: https://arxiv.org/abs/2608.03627
作者: Razieh Chalehchaleh,Reza Farahbakhsh,Noel Crespi
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large Language Models (LLMs) are increasingly used for automated fact-checking, yet their susceptibility to gender bias in this context remains underexplored. This study presents the first systematic investigation of gender bias in LLM-based fake news detection using real-world data. We augment the LIAR benchmark with three gender variants of speaker job titles (Neutral, Male, Female) for each statement to test whether veracity judgments vary solely based on gender presentation. Six state-of-the-art LLMs are evaluated across multiple bias and fairness metrics. All models exhibit gender sensitivity: 9.79%-35.13% of statements receive inconsistent labels across the three variants, with Male-Female comparisons showing 6.5%-23.6% flip rates. Two primary bias manifestations are identified: instability (inconsistent judgments) and directionality (systematic favoritism). Five models show statistically significant directional effects, with the strongest effects displaying male-skeptic patterns. These findings demonstrate that gender bias undermines both reliability and fairness in LLM-based fake news detection, highlighting the need for bias-aware evaluation and mitigation strategies. The augmented dataset is publicly released to support future research.
[AI-43] A Security-Oriented Lifecycle Model for Large Language Model Systems
链接: https://arxiv.org/abs/2608.03626
作者: Eleftherios Batzolis,George Drosatos,Vassilis Katsouros,Konstantinos Rantos
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注: Accepted as: Batzolis, E., Drosatos, G., Katsouros, V., Rantos, K. (2026). A Security-Oriented Lifecycle Model for Large Language Model Systems. In: Kieseberg, P., Skopik, F., Atli, B., Schrittwieser, S., Asplund, M. (Eds.), Availability, reliability and security—ARES 2026 EU Projects Symposium workshops (Lecture Notes in Computer Science, pp. 1-18). Springer Nature Switzerland
Abstract:Large language models are being integrated into critical infrastructure and enterprise workflows at unprecedented scale,yet the lifecycle frameworks governing their development and operations were designed for operational efficiency rather than security analysis. As a result, security-relevant activities such as data provenance verification, artifact signing, agentic permission control, and decommissioning are often left implicit or assumed to receive due care. Governance frameworks, in turn, organise requirements around risk levels or management processes without clearly linking them to the lifecycle stages where they apply. This paper addresses both deficiencies. We propose a lifecycle model for LLM systems that supports security analysis by structuring it around security-relevant boundaries rather than workflow optimisation. The model comprises 32 stages across four core pipeline layers (Data, Model, Distribution, Application), supported by a 12-stage LLMOps pillar and a 9-category governance pillar. Thirteen stages are introduced here as separate units because they expose distinct security concerns that existing frameworks do not clearly distinguish. A governance mapping synthesising the NIST AI RMF, the EU AI Act, and ISO/IEC 42001 reveals a structural property of the current regulatory landscape: governance evidence concentrates at deployment-facing stages, where systems are visible to regulators, while the most consequential decisions, data selection, alignment strategy, and capability boundaries, are made at development-facing stages, where regulatory visibility is lowest.
[AI-44] A Theory of Conditional Collapse under Low-Rank Weight-Space Ablations: I. The Single-Block Theory and Synthetic Validation
链接: https://arxiv.org/abs/2608.03620
作者: Abdallah Khemais
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 25 pages, 2 figures. Part I of a two-part series; see the companion paper “Cross-Layer Interaction under Weight-Space Ablation” (Part II)
Abstract:Activation patching and weight-space ablation both claim a component is causally responsible for a behavior, yet they act on different objects: one forward pass versus the parameters behind every forward pass. We ask when they agree. We study an idealized model where a conditional computation is carried additively through a residual stream, F(x)=F_0(x)+\sum_i\alpha_i(x)v_i , read out by a linear functional, and prove three exact results. First, deleting a subset of carriers collapses a matched input pair onto the same unconditional output \emphif and only if the removal is symmetric on the pair and leaves no outside contrast; the error is deterministic, and we give its exact form even when the two conditions hold only approximately. Second, patching a carrier moves the readout by its donor-receiver \emphcontrast, while ablating it moves the readout by its \emphabsolute level; neither bounds the other, and we construct pairs where every single-carrier patch flips the decision while no single-carrier ablation does. Third, for an attention head composed with its own layer’s normalization and MLP, we derive an exact first-order interaction formula with a provably second-order remainder, vanishing identically when only the MLP is ablated but not, in general, when a head is. Small transformers trained on a synthetic conditional task illustrate all three predictions: across thirty-nine ablation configurations the measured interaction is strongly rank-correlated with the idealized model’s predictive accuracy (Spearman -0.83 ), and a second task and architecture reproduces the same pattern, including a further polarity reversal. The single-block interaction result extends past one residual block, and the synthetic validation is tested against a real pretrained model, in a companion paper that takes this theory further along both axes. Comments: 25 pages, 2 figures. Part I of a two-part series; see the companion paper “Cross-Layer Interaction under Weight-Space Ablation” (Part II) Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI) Cite as: arXiv:2608.03620 [cs.LG] (or arXiv:2608.03620v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2608.03620 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Khemais Abdallah [view email] [v1] Tue, 4 Aug 2026 13:10:19 UTC (104 KB)
[AI-45] Rethinking Modality Reliability in Multimodal Sentiment Analysis with Incomplete Observations
链接: https://arxiv.org/abs/2608.03611
作者: Chunlei Meng,Jacqueline J. Pang,Pengbin Feng,Zhenyu Yu,Chun Ouyang,Zhongxue Gan
类目: Artificial Intelligence (cs.AI); Multimedia (cs.MM)
备注:
Abstract:Multimodal Sentiment Analysis (MSA) integrates text, audio, and vision to infer human affect, yet real-world multimodal observations are often incomplete. Existing methods for incomplete-observation MSA mainly follow two paradigms. Reconstruction-based methods recover missing information from observed modalities, while joint-representation methods learn directly from incomplete inputs. Although effective, these methods usually treat modality reliability only implicitly within representation learning or fusion design rather than modeling it explicitly. We argue that modality reliability is a central variable in incomplete-observation settings. Failure to model it explicitly gives rise to two related issues. The first is reliability mismatch, in which the affective evidence retained by each modality varies across samples and missing rates. The second is reliability propagation bias, in which messages from degraded modalities may adversely affect cross-modal interaction and predictive performance. To address these issues, we propose MRCF, a Modality Reliability-Calibrated Framework for MSA with incomplete observations. MRCF contains a Reliability-Aware Branch that estimates sample-specific modality reliability from intramodal quality cues and cross-modal semantic consistency, a Reliability-Guided Interaction Branch that uses the estimated scores to modulate cross-modal information flow, and a Reliability-Calibrated Fusion Module that integrates reliability and semantic cues for final prediction. Experiments on CMU-MOSI, CMU-MOSEI, and CH-SIMS show that MRCF achieves strong performance under standard incomplete-observation protocols. Further analyses provide evidence that explicit reliability modeling helps mitigate reliability mismatch and reliability propagation bias during interaction and fusion.
[AI-46] Formal Verification of Agent ic Systems over Operational Data
链接: https://arxiv.org/abs/2608.03609
作者: Alejandro J. Mercado,Alessio Lomuscio
类目: Artificial Intelligence (cs.AI)
备注: 21 pages, including appendix; 0 figures
Abstract:Agentic systems driven by large language models (LLMs) are increasingly deployed in real-world workflows where they act on persistent operational data. Before deployment, these systems need to be verified against business requirements that govern workflow execution and data evolution. However, existing approaches do not provide such system-level guarantees, as they mainly constrain or analyse behaviour at the agent’s interface level. We study here the verification of agentic systems comprising a single LLM and a tool orchestration harness over relational operational data. We formalise them as Stateful Tool-Enabled Agentic Deployments (STEADs), give their semantics, define the problem of verifying them against First-Order Computation Tree Logic (FO-CTL) specifications, and show that it is undecidable. We identify sufficient conditions for exact preservation of FO-CTL specifications under a finite-domain restriction, over which verification is PSPACE-complete. The key requirement is that renaming opaque identifiers in the data must correspondingly rename the selected tool calls. We show that LLM-driven agents can violate this condition and introduce a canonical deployment wrapper that guarantees it for arbitrary base agents while preserving already-equivariant behaviour. We prove that computing canonical representations required by this construction is graph-isomorphism-hard. Finally, we illustrate our framework on an LLM agent orchestrating a case-management workflow.
[AI-47] Learning Clinical-Trial Strategy: Offline Policy Training for Decision Agents ICML2026
链接: https://arxiv.org/abs/2608.03606
作者: William Bolton,Philip Torr
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Accepted for a spotlight at the ICML 2026 Workshop on Generative and Agentic AI for Biology (GenBio) and as a poster at the ICML 2026 Workshop on Decision-Making from Offline Datasets to Online Adaptation: Black-Box Optimization to Reinforcement Learning (DEMO). 15 pages, 3 figures, 11 tables
Abstract:Clinical development is sequential decision-making under uncertainty, where a sponsor must plan a portfolio of experiments from heterogeneous evidence. We study this setting by framing oncology clinical development as an offline decision-making problem in which an agent predicts the next six-month trial portfolio of an oncology drug program from information available at the decision date. To support this, we construct a temporal dataset that combines 31.7k heterogeneous public data records, including trial registries, regulatory reviews, sponsor filings, utilization data, and epidemiology, into 881 offline decision episodes across 45 historical programs. We compare four offline objectives: behavioral cloning, reward-weighted behavioral cloning, learned-reward training, and value-based implicit Q-learning against four frontier LLM agents that share a common date-gated retrieval scaffold across held-out drug, sponsor, drug-class, and temporal splits. Models trained offline outperform the non-fine-tuned baselines, particularly in the post-August 2025 contamination-clean holdout. Reward-weighted behavioral cloning performs the best, obtaining 46.2% indication F1 and 14.2% strict F1 against 25.0% and 2.1%, respectively, for the best-performing tool agent on each metric. These results suggest that structured offline learning can teach agents to plan clinical experiments.
[AI-48] FraQ: Efficient Coordinate-Space Recompression for Federated Low-Rank Adaptation
链接: https://arxiv.org/abs/2608.03605
作者: Shenghui Li,Thiemo Voigt
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Federated fine-tuning with Low-Rank Adaptation (LoRA) enables efficient collaborative adaptation of Large Language Models (LLMs) without centralizing private data. However, LoRA’s two-factor parameterization creates an aggregation mismatch across clients: naively averaging the factors does not recover the average of their induced updates. This mismatch can be avoided by forming the exact aggregate in the full weight space and then recompressing it, but decomposing the resulting dense matrix is computationally expensive and memory-intensive. We propose FraQ, an efficient coordinate-space recompression method for federated LoRA. Starting from stacked factors that exactly represent the aggregate, FraQ factorizes it into an orthonormal basis and a compact coordinate matrix. It then recovers the singular spectrum from a small Gram matrix, selects the smallest rank satisfying a prescribed energy threshold, and maps the selected coordinate subspace back through the basis to construct the global adapter. Experiments on text classification and commonsense reasoning benchmarks show that FraQ achieves accuracy close to uncompressed baselines while substantially reducing downlink communication with low server-side recompression overhead.
[AI-49] Large language models for partial differential equation workflows
链接: https://arxiv.org/abs/2608.03600
作者: Han Wan,Rui Zhang,Hao Sun
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Partial differential equations (PDEs) become actionable in science and engineering not as isolated formulae, but as executable workflows that connect modelling assumptions, governing equations, numerical solvers, diagnostics, and decisions. Large language models (LLMs) are beginning to support such workflows by linking natural language, symbolic mathematics, code, solver outputs, and feedback. Here we examine recent advances in LLM-assisted PDE research across three stages: the discovery and formulation of governing models, the generation and revision of executable numerical solvers, and the use of simulation feedback to support control, design, and optimization. Across these stages, current systems act primarily as workflow-level interfaces. Despite this progress, the field remains limited by the scarcity of high-quality datasets and benchmarks, especially for knowledge discovery and real-world applications, where expert annotation, executable problem construction, and task-level feedback require substantial domain effort. A further challenge is the persistent gap between simulation-based results and real-world scientific and engineering systems, which limits the direct transfer of numerical simulations, control policies, and optimized designs to practical settings. These challenges make LLM-assisted PDE workflows a critical testbed for developing scientific AI systems that can connect language, computation, physical constraints, and real-world decision-making.
[AI-50] FOUND-AF: Benchmarking ECG Foundation Models for Atrial Fibrillation Detection
链接: https://arxiv.org/abs/2608.03597
作者: Amirhossein Taleshinosrati,Yangyang Wang,Atitaya Phoemsuk,Vahid Abolghasemi,Naser Hossein Motlagh,Sadasivan Puthusserypady,Daniel Teichmann,Abdolrahman Peimankar
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia and is associated with increased risks of stroke, heart failure, and mortality. Recent ECG foundation models offer transferable representations for automated AF detection. However, their relative effectiveness remains unclear because existing studies use different datasets, preprocessing procedures, classifiers, and validation protocols. This study presents FOUND-AF, a unified, leakage-controlled, and deployment-oriented benchmarking framework that evaluates the quality of pretrained ECG representations under identical experimental conditions. Nine publicly available foundation models from five families, including HuBERT-ECG, CLEF, ST-MEM, ECG-JEPA, and ECGFounder, were evaluated across four heterogeneous ECG datasets, namely AFDB, CinC2017, CPSC2021, and LTAFDB. All models were used as frozen feature extractors with standardized preprocessing, model-native resampling, a fixed XGBoost classifier, and recording-level grouped cross-validation. The evaluation included classification metrics, receiver operating characteristic analysis, paired recording-level bootstrap comparisons with Holm correction, embedding-space visualization, and computational efficiency profiling. The ECGFounder model consistently achieved the strongest overall performance across datasets while offering a favorable trade-off between accuracy, model size, inference time, and memory usage. FOUND-AF therefore provides a reproducible framework for selecting ECG foundation models and demonstrates that compact, clinically pretrained encoders can support robust and computationally efficient AF detection across heterogeneous acquisition settings.
[AI-51] DiagChain: A Diagnostic Benchmark for Evaluating LLM Agents on Evidence-Grounded Attack Chain Reconstruction
链接: https://arxiv.org/abs/2608.03591
作者: Xuyang Liu,Yibin Han,Zhenwei Zhang,Kai Chang,Zhiwei Xu,Tian Qiu,Weixian Deng,Jiabao Gao,Xiaolin Peng,Hai Wan,Xibin Zhao
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:Large Language Model (LLM) agents offer a promising approach to attack chain reconstruction by retrieving and interpreting heterogeneous telemetry to infer ordered attacker actions. However, existing benchmarks mainly evaluate final outputs or aggregate accuracy, providing limited insight into how errors arise and propagate across intermediate reasoning stages. We present DiagChain, a diagnostic benchmark for evidence-grounded attack chain reconstruction that enables stage-wise evaluation of LLM agents. DiagChain includes MAIN-69, a suite of 69 scenarios spanning multiple operating systems, evidence noise levels, and chain lengths. It further introduces Evidence-Centric Retrieval-Augmented Generation (ECRAG), which couples evidence retrieval with an evolving structured representation of the reconstructed chain. Five complementary metrics are introduced to assess distinct stages of the reconstruction process and support systematic failure diagnosis. Based on evaluations using 6 LLMs, DiagChain reveals that even the strongest configuration succeeds on only 39.6% of the 849 reference steps in MAIN-69. Our analysis further shows that smaller models struggle with the more basic task of incorporating retrieved evidence into their outputs, whereas larger models can proceed to later steps, where correctly ordering that evidence becomes the main bottleneck. These results validate the importance of diagnostic evaluation beyond end-to-end accuracy and provide actionable insights for improving evidence-grounded cybersecurity agents.
[AI-52] GenOS: Compositional Certificates for Semantic Robustness in AI Code Generation
链接: https://arxiv.org/abs/2608.03588
作者: Corrado Priami
类目: Programming Languages (cs.PL); Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
备注:
Abstract:AI coding agents are stochastic workflows: prompts are interpreted, artifacts are sampled, validators produce observations, and orchestrators commit or repair. Small prompt or specification changes can therefore alter program-behavior distributions even when the texts appear synonymous. Existing systems evaluate correctness, but lack a compositional criterion for safely replacing a prompt, contract, generator, or program inside a complete agentic workflow. We introduce GenOS, a probabilistic operational semantics for this replacement problem. Each layer is modeled as a Markov kernel, and each interface carries an observer-relative equivalence. We prove that equivalence-compatible kernels descend to quotient classes and that quotienting commutes with distributional extension and sequential composition. Hence, equivalent prompts induce equal probabilities for all downstream equivalence-closed events, including verified commit. We also establish workflow bisimulation, guarded-commit safety under sound validation, total-variation non-expansiveness, and an additive robustness bound that attributes approximation error to individual pipeline layers. An executable insertion-sort audit instantiates the theory with natural-language paraphrases, a formal contract, six programs, two observers, and exhaustive execution on 121 inputs. Equivalent prompts yield identical code-class and commit distributions; a prompt assigning 5% probability to an in-place contract is distinguished by a mutation observer, while downstream distances remain within the predicted bound. Across 20,000 randomized finite-kernel trials, no exact or approximate law is violated. GenOS is model-parametric: compatibility is a measurable property to test, not an assumption about language-model behavior.
[AI-53] From Social Coding to Agent ic Coding: Productivity and Relational Reconfiguration in Open-Source Communities
链接: https://arxiv.org/abs/2608.03585
作者: Mengying Zhou,Yongjie Yin,Yang Chen
类目: Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注:
Abstract:Open-source software communities are a form of digital public infrastructure that not only produces code, but also generates public knowledge and interpersonal relationships through visible collaboration. Generative coding agents (CAs) are an advanced tool to improve development efficiency while shifting part of activities from public human interaction to private human-agent loops. We study this shift using an LLM-based multi-agent simulation initialized with real GitHub data from 1,084 active developers and their repository relationships. After a warm-up with historical commits, we branch the same community state into parallel No-CA and CA conditions for 4-week simulations. CA introduction increases planned and completed tasks by 34.0% and 39.0%, respectively, and reduces median completion time from 45 to 20 minutes. However, adoption reaches only 26.0%, and the gains concentrate among developers who are already more active and well connected. CAs also restructure task execution pathways. Direct human-human interaction declines from 32.4% to 11.6%, while CA-involved modes increase to 57.3%, including 40.3% completed through CA-assisted self-loops. Public knowledge generated under CA condition also provides less support for later tasks. On a standardized retrieval benchmark, the CA corpus achieves 22.3% knowledge coverage, far below the 81.1% achieved by the real-human corpus, and requires more retrieval steps with a lower success rate. These results reveal a productivity-public knowledge tension: coding agents increase technical production, but more work shifts to agent-mediated or private loops, leaving public records less useful to future contributors.
[AI-54] Policy Frag mentation or Institutional Alignment? Institutional Governance of AI in Universities and Business Schools
链接: https://arxiv.org/abs/2608.03584
作者: Lydia Manikonda,Dominique Outlaw
类目: Artificial Intelligence (cs.AI); Computational Engineering, Finance, and Science (cs.CE); Emerging Technologies (cs.ET)
备注: Our insights suggest that policy guidelines should be aligned with broader institutional policies while addressing discipline-specific learning objectives and evolving workforce demands
Abstract:Artificial intelligence (AI) is rapidly transforming high-skilled domains, requiring higher education institutions (HEI) to balance the teaching of foundational principles with the integration of emerging tools to ensure workforce readiness. While HEI are increasingly adopting AI, many continue to grapple with how it should be incorporated into curricula and governed through policy, especially when such policies are set at different levels of an institution. This research analyzes AI policies across HEI from 34 states in the United States to investigate what these policies entail and how policies set across institutions as well as within different levels at an institution differ. Using natural language processing (NLP) to analyze institutional AI policies, we find a clear divergence: university-level policies emphasize data security and risk mitigation whereas school-level policies, when present, focus on pedagogical applications and tool usage. When focusing on business school specific policies, relatively few business schools maintain AI policies distinct from university frameworks, creating misalignment with discipline-specific learning objectives. This gap poses challenges particularly for faculty and students as well as for accreditation purposes. Our insights suggest that guidelines should be aligned with broader institutional policies while addressing discipline-specific learning objectives and evolving workforce demands.
[AI-55] AI-Assisted Peer Review Across Research Communities: From Reviewer AI Policies to LLM Review Quality
链接: https://arxiv.org/abs/2608.03581
作者: Alexander M. Fichtl,Lukas Ellinger,Josefin Kelber,Kryštof Olík,Georg Groh
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI)
备注: 30 pages, 19 figures, 10 tables. Currently under peer review. GitHub link for code and data is given in the paper
Abstract:AI-assisted peer review is increasingly discussed and adopted as a tool to support the scientific publishing process, yet there is little systematic understanding of how publication venues regulate its use or of how capable current AI review systems are. We address these questions by first surveying reviewer-facing AI policies across 111 leading AI/NLP conferences and medical journals, revealing substantial regulation differences between the two communities. Second, we evaluate AI-generated peer reviews at ICLR 2026 and Nature Communications using a novel dataset comprising original manuscript submissions and several hundred human- and machine-generated reviews. We compare reviews produced by open-source and proprietary models using complementary evaluation metrics, including LLM-as-a-Judge, score alignment, granularity, and overlap with human reviewers’ concerns. Our results show that current LLMs can generate detailed and fluent reviews but exhibit systematic weaknesses, such as overly positive recommendations, generic criticism, and uneven evidence grounding. We demonstrate that aggregate quality scores alone can overestimate review quality and argue for multi-dimensional evaluation of AI-generated peer reviews.
[AI-56] Pin Once Swap Light: Subspace-Aligned Centroid-Residual Training for Efficient Ultra-LoRA Serving
链接: https://arxiv.org/abs/2608.03579
作者: Xiang Li,Pengcheng Wang,Huazheng Wang,Saurabh Bagchi
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Modern multi-tenant Low-Rank Adapters (LoRAs) serving systems concurrently host tens to hundreds of LoRA adapters. Though powerful, this introduces a critical system dilemma between serving efficiency and task performance: higher-rank adapters generally achieve better downstream task performance, but their GPU VRAM footprint and Host-to-Device PCIe swapping overhead severely constrain scalability. Conversely, ultra-low-rank adapters ( r \le 2 ) minimize both VRAM footprint and PCIe transfer overhead, but suffer from downstream task performance degradation. To solve this problem, we propose Subspace-Aligned LoRA Training (SALT), a serving efficiency-aware hierarchical fine-tuning framework. Our solution operates in three phases. First, a provider jointly trains high-capacity domain centroids on public data within the domain using a novel alignment regularizer that coheres in-domain task subspaces into a unified basis. Next, users fine-tune ultra-low-rank task residual adapters on private data atop those frozen centroids. Finally, during inference, the provider pins the centroid in GPU VRAM and dynamically swaps in each user’s task residual on demand. Across LLMs of varying scales, SALT recovers high-rank accuracy using r \le 2 residuals, achieving up to 18.5% absolute accuracy gains over state-of-the-art compression baselines and reducing per-adapter memory by up to 16x. When integrated into vLLM, SALT improves serving throughput by up to 51% under PCIe bandwidth pressure and 28% under GPU VRAM constraints for Llama-3.2-3B.
[AI-57] Enhancing Tabular Learners with Context-Aware Semantic Embeddings
链接: https://arxiv.org/abs/2608.03565
作者: Günther Schindler,Maximilian Schambach,Johannes Höhne
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:While modern tabular learners excel at capturing statistical patterns, they frequently operate in a semantic vacuum, treating textual features as discrete symbols, ignoring the rich semantics inherent in feature names or cell entries. We propose CASE (Context-Aware Semantic Embeddings), a novel framework that bridges the gap between the semantic understanding of Large Language Models (LLMs) and the statistical capabilities of tabular learners. Unlike existing methods that embed rows in isolation, CASE utilizes a contextualization strategy: we pre-fill the KV cache of a custom-trained Gemma 3-based Tabular Language Model with a representative sample of rows to establish a persistent anchor of the dataset’s semantics. This ensures that generated row embeddings are dynamically contextualized, resolving semantic ambiguities and anchoring representations in domain-specific context. Our experiments across several benchmarks (CARTE, TextTab, and TabArena) demonstrate that CASE substantially improves the performance of tabular learners on semantically rich datasets, particularly in low-data regimes.
[AI-58] Soft Guidance Starts to Outperform CoT Prompting as LLM s Improve
链接: https://arxiv.org/abs/2608.03550
作者: Denys Pushkin,Albert Q. Jiang,Aryo Lotfi,Colin Sandon,Emmanuel Abbé
类目: Artificial Intelligence (cs.AI)
备注: 10 pages, 3 tables
Abstract:Chain-of-Thought (CoT) prompting remains the standard baseline for evaluating models’ reasoning abilities. Originally, this technique was introduced to elicit step-by-step reasoning from large language models (LLMs), which would otherwise tend to directly output the final answer. However, many modern LLMs produce CoT-style responses \textitnatively when presented with reasoning tasks, which made us revisit the effectiveness of standard CoT prompting. We evaluate several modern mid-sized language models on a math problem-solving task and find that models specialized for reasoning achieve better performance in a simple zero-shot setting than when using few-shot CoT examples - significantly surpassing officially reported results at no additional cost (e.g., from \sim 77% to \sim 84% for Mathstral on GSM8K). For the tested general-purpose model, a zero-shot CoT prompt is also sufficient to outperform a few-shot CoT baseline. We attribute this to a `guidance-distraction’ tradeoff: standard CoT prompting also demands style adaptation, formatting compliance, and potentially undesired contextualization, which can distract models from the core reasoning task. Our findings suggest that using standard CoT prompting increasingly acts as a source of distraction as models grow stronger. Comments: 10 pages, 3 tables Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2608.03550 [cs.AI] (or arXiv:2608.03550v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.03550 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-59] Behaviorally Adaptive Visual Diversion for Inclusive and Resilient Digital Assessment Delivery
链接: https://arxiv.org/abs/2608.03531
作者: Gupta Lovi Raj,kaur Kamalpreet,Dama Sriram,Parali Prajithaa
类目: Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注: 15 Pages, 7 Figures, 25 Equations
Abstract:Institutions increasingly rely on browser lockdown, webcam monitoring, and behavioral analytics to secure high-stakes digital assessments, yet these mechanisms are commonly designed and evaluated independently and often overlook learner accessibility. This paper introduces Behaviorally-Adaptive Visual Diversion (BAVD), a theoretical framework in which a synthetic, non-semantic visual field is composited with assessment content and adaptively modulated according to observed candidate behavior. The underlying assessment content is never altered; only its visual presentation is modified to reduce the usefulness of unauthorized screen capture or screen sharing while remaining minimally intrusive for legitimate candidates. The framework further incorporates an accessibility-aware attenuation mechanism that reduces or suppresses diversion intensity for candidates with approved visual-processing accommodations. We formulate the model using a coupled dynamical-systems representation comprising a Diversion Field Generator, Rendering Tensor, Behavior Tensor, Composite Integrity Functional, and Multi-dimensional Entropy Model, and establish theoretical properties for content fidelity, rendering stability, entropy boundedness, integrity tracking, and closed-loop adaptation stability. The framework explicitly states its threat model, identifies deployment assumptions and limitations, and discusses the trade-off between accessibility and capture resistance. This work provides a mathematically grounded foundation for behaviorally adaptive and accessibility-aware assessment delivery and offers a basis for future empirical validation in trusted digital assessment platforms.
[AI-60] Dr. AGENT ONOMICS: A Didactic Experiment of AGENT ONOMICS
链接: https://arxiv.org/abs/2608.03524
作者: Fengjunjie Pan,Alois Knoll
类目: Artificial Intelligence (cs.AI)
备注: Technical report, Technical University of Munich
Abstract:AGENTONOMICS is a framework that treats AI agents as economic entities that can be designed, managed, and governed through an integrated management architecture. Dr. AGENTONOMICS is its first application: a lecture agent developed in the context of the TUM course on AI agents in business administration. Conceived during the winter semester 2025/26 and first introduced to students in the summer semester 2026, it serves as a didactic experiment in which the agent is both the object that students study and the medium through which they learn and apply the framework. The current prototype is a web-based, retrieval-grounded tutor that explains AGENTONOMICS concepts and supports student questions. This report argues that the same system can grow beyond tutoring into three additional cumulative roles: an avatar lecturer that delivers multimodal instruction, a design consultant that guides students through the AGENTONOMICS Design Management Reference Framework (ADMRF), and a meta-agent that helps construct the agents students have specified. These roles are cumulative because they share the same interface, intelligence layer, tools, knowledge base, and ecosystem connection, while an orchestrator selects the role-specific algorithm required for each task. We present the architecture of the prototype, outline its development roadmap, and discuss its implications for a polycentric AI economy. This report is intended to invite further discussion on how agents can teach, apply, and eventually reproduce the frameworks by which they are designed.
[AI-61] Pivot-Centric Trajectory Prediction: Bridging Long Horizons via Dynamical Guidance
链接: https://arxiv.org/abs/2608.03521
作者: Xiucong Zhao,Jindong Tian,Hao Miao
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: Spatiotemporal Forecasting, Autonomous Driving, Trajectory Prediction
Abstract:Forecasting precise future motion of surrounding agents is essential for reliable autonomous vehicles. However, as the demand for longer prediction horizons increases, existing endpoint-completion or iterative-refine methods increasingly struggle with weak guidance and compounding errors. To tackle the long-horizon prediction challenge, we propose Pivot-Centric Trajectory Prediction (PCTP). By introducing pivots'' and focusing on predicting pivot points along extended trajectories, we divide the long-term prediction task into short-term sub-tasks at various scales. Specifically, PCTP decouples the long-term trajectory predicting process into two processes: pivot prediction and pivot-based trajectory refinement. The pivot prediction process aims to utilize global map context and agent-to-agent interactions to identify these pivot points’‘, while the pivot-based trajectory refinement process focuses on local map details and refines the short-term trajectory based on predicted ``pivot points’'. Compared with existing methods, PCTP provides more intermediate guidance while reducing compounding errors. Moreover, PCTP is a flexible approach that can be integrated into most state-of-the-art trajectory prediction models. Experimental results show that PCTP improves the prediction accuracy of leading models on both Argoverse I and Argoverse II datasets with minimal impact on model size. Specifically, PCTP combined with QCNet outperforms all published ensemble-free methods on the Argoverse II leaderboard at submission.
[AI-62] AI Forensics Across White- Grey- and Black-Box Access: A Process Model and Research Agenda for Post-Incident Investigation of AI Systems
链接: https://arxiv.org/abs/2608.03520
作者: Ali Dehghantanha,Sajad Homayoun
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:AI systems are increasingly involved in decisions and actions that may later require investigation. When an AI related incident occurs, investigators need to reconstruct what the system did, why it behaved that way, and which part of the system or supply chain contributed to the outcome. Existing work on AI forensics remains fragmented, often focusing on a specific system type, artifact, or analysis technique. This paper argues that investigator access is a useful starting point for organizing the field. We distinguish white box, grey box, and black box access and show how each access level changes what can be collected, preserved, analyzed, and reported. Based on this distinction, we propose a process model matrix for AI forensics across four phases: collection, preservation, analysis, and reporting. We also introduce an order of volatility for AI systems, covering runtime state, context windows, logs, retrieval stores, model artifacts, and training lineage. From this matrix, we derive an access conditioned examination framework and identify open research problems, including black box preservation, model version attestation, uncertainty quantification for surrogate based analysis, and chain of custody for mutable AI artifacts.
[AI-63] Reversing Arrows in Large Language Models
链接: https://arxiv.org/abs/2608.03512
作者: Sefika Efeoglu,Adrian Paschke
类目: Artificial Intelligence (cs.AI)
备注: The preprint is under review in a venue
Abstract:Large language models (LLMs) have achieved strong performance on text-to-knowledge graph generation and related tasks. Nevertheless, it is still unclear whether they accurately model the direction-dependent semantics of inverse relations, in which reversing the order of the arguments alters the meaning of a relation (e.g., \textitmother versus \textitchild). To the best of our knowledge, this work presents the first systematic study of inverse relation directionality in LLMs, using a benchmark consisting of 5,457 instances spanning 27 distinct inverse relation labels. We evaluate five open-source LLMs under a multiple-choice prompting framework and further examine the influence of relation descriptions and entity representations by substituting the original entities with synthetic and masked entities. Our findings reveal systematic asymmetries in inverse relation classification across LLMs, indicate that relation descriptions do not consistently improve performance, and show that model performance can be sensitive to variations in entity representations.
[AI-64] When Many Answers Are Valid Voting Fails: Symbolic Verification for Best-of-K Causal Reasoning in LLM s
链接: https://arxiv.org/abs/2608.03506
作者: Omatharv Bharat Vaidya,Connor Thomas Jerzak,Zayne Rea Sprague,Fangcong Yin,Nhat Ho
类目: Artificial Intelligence (cs.AI); Machine Learning (stat.ML)
备注: 28 pages, 5 figures, 28 tables
Abstract:Self-consistency assumes the most frequent answer among sampled reasoning traces is the most reliable, but this can fail in causal reasoning: samples often repeat the same confounding error, and votes fragment across multiple valid answers, letting an invalid answer win despite a valid minority trace. We introduce CALVER (Causal Axiom-Level VERification), a training-free symbolic verifier that scores structured traces against Pearl’s causal criteria, including -separation, backdoor adjustment, and intervention, and selects the highest-scoring candidate without consulting a reference answer. On CLEAR find-one-valid queries that admit multiple graph-valid answers, CALVER reaches 42.1% where plurality, a reward model, an LLM judge, and model confidence remain near 30% on identical frozen pools. Scaling the judge to 72B does not close the gap. In an audited clean-core subset, 11 of 21 graph-valid CALVER selections differ from the benchmark’s listed answer while still satisfying the requested predicate. The advantage widens with the sampling budget and reproduces across ten published Bayesian networks, a second model family, and settings where the model must build the graph from text. CALVER also improves thresholded average-treatment-effect decisions against exact ground truth, generalizes to logic under a truth-table checker, and scores each candidate in milliseconds on CPU. CALVER needs only a causal structure, supplied outright or built from the text; wherever that holds, selection can aggregate via causal validity.
[AI-65] Can LLM design high-quality experiments? A Comprehensive and Systematic Benchmark on Autonomous Experimental Design
链接: https://arxiv.org/abs/2608.03501
作者: Zejun Liu,Jian Wu,Ru Peng,Yuliang Ji,Dongyuan Li,Renhe Jiang,Yue Zhang
类目: Artificial Intelligence (cs.AI)
备注: 32 pages, 7 figures
Abstract:AI for Research (AI4Research) leverages AI to automate and improve scientific workflows. While experimental design is a critical stage of the research process, prior work has focused primarily on code implementation and execution, overlooking the importance of this stage, and no benchmark exists to evaluate AI’s ability to conduct systematic experiment design. To bridge this gap, we propose SCOPE, a Scientific COmprehensive Planning Evaluation Benchmark constructed from 300 high-quality latest papers across 19 research domains from top-tier venues (e.g., ICML, NeurIPS, and ICLR),evaluating LLMs on two dimensions: High-Level planning completeness (main, ablation, and analysis experiments) and Low-Level configuration accuracy and rationality (datasets, baselines, and metrics). Benchmarking reveals three findings: (1) most LLMs cannot directly design high-quality experiments; (2) all LLMs exhibit a performance bottleneck in low-level configuration; and (3) search mode does not improve design quality. Furthermore, to address these challenges, we propose OptED, a novel agentic workflow to optimize LLM-based experimental design, that enhances LLM-based experimental planning through stage isolation, tool augmentation, and rule-based constraints, effectively alleviating the configuration bottleneck.
[AI-66] WeClawArena: An Auditable Sandbox and Benchmark for Cross-User Agents Collaboration and Security in Human-Centered Agent Networks
链接: https://arxiv.org/abs/2608.03499
作者: Prince Zizhuang Wang,Aojie Yuan,Haiyue Zhang,Xiyang Hu,Yue Zhao,Shuli Jiang
类目: Artificial Intelligence (cs.AI)
备注: 31 pages
Abstract:Recent advances in persistent personal-agent frameworks are making human-centered agent networks realistic deployment targets: each user can be served by an AI agent that acts on the user’s behalf, maintains state, and communicates with other agents through social and task relations. In these networks, everyday tool use becomes multi-party owned-agent collaboration over personal workspaces, where files, records, tools, and policies are not directly visible across owners. Existing agent benchmarks study tool use and collaboration, but they do not provide an end-to-end sandbox for verifiable cross-user agent collaboration with realistic user digital workspaces or test how harmful actions can travel through the human-centered agent network. We introduce WeClawArena, an auditable benchmark and runtime sandbox for multi-party owned-agent collaboration over personal workspaces. WeClawArena targets collaborative tool-use tasks in which personal workspaces serve as both operational tools and personal constraints. The benchmark contains 124 base tasks across six cross-user task domains and expands them into 620 scenario variants, with one benign control and four attack-vector variants per base task. The sandbox records peer messages, tool calls, resource operations, governed decisions, and final workspace states. WeClawArena reports utility and attack success rate separately and audits attack success from bounded runtime evidence, supporting diagnosis of task breakdown, privacy leakage, poisoned evidence, and invalid authority paths.
[AI-67] Leverag ing System-Level Observations to Inform Bayesian Learning of Model Parameters for Quantitative Verification
链接: https://arxiv.org/abs/2608.03489
作者: Simos Gerasimou,Xingyu Zhao
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Logic in Computer Science (cs.LO)
备注: 11 pages, 9 figures
Abstract:Combining Bayesian learning and quantitative verification is a powerful toolset for analysing key quantitative properties of software systems, like reliability and response time. However, the accuracy and robustness of verification results strongly depend on the prior knowledge (PK) underlying Bayesian inference. This knowledge reflects original beliefs about the probability of events and typically depends on domain expertise. Using inaccurate or uninformative PK can negatively affect quantitative analysis, yielding incorrect verification results. Our EPIK approach tackles this important challenge by eliciting and embedding PK in quantitative verification equipped with Bayesian estimators. Unlike existing approaches that require PK on formal model transition parameters, EPIK leverages system-level properties that are directly observable and are linked to real-world semantics. EPIK formulates a twofold optimisation problem to derive the distributions of unknown transition parameters and then embeds these distributions to verify new or difficult-to-measure (elusive) properties. The detailed experimental evaluation using multiple variants of real-world case studies and diverse EPIK instantiations shows its effectiveness, flexibility and generality.
[AI-68] oolLIFT: Lifting Tool-Specific Trajectories into Function-Level Graphs for Generalizable Tool Planning
链接: https://arxiv.org/abs/2608.03468
作者: Xiuhui You,Jiayi Luo,Zichao Shen,Qingyun Sun,Ziwei Zhang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Historical tool-use trajectories provide valuable experience for large language model (LLM) agents to plan and coordinate tool usage. Existing approaches directly construct tool-level graphs from these trajectories, but the resulting graphs remain tied to specific tools and are hard to generalize across tool sets. To tackle this challenge, we find that despite differences in the tools involved, analogous tasks often share a common function-level workflow structure, which serves as a potentially more transferable abstraction for tool planning. Based on this insight, we propose ToolLIFT, a framework that lifts tool-specific trajectories into a function-level workflow graph (FWG) for generalizable tool planning. Specifically, we first propose a trajectory-lifting mechanism that encodes workflow structures in the FWG and shares collaboration experience across tools. Then, building on the global structure of the FWG, we introduce decoupled workflow planning and tool selection to align individual tool choices with the overall workflow. Lastly, to ensure reliable tool dataflow, we adopt Reinforcement Learning (RL) and propose source-gated and skill-specific rewards to maintain source-traceable information flow across tool calls. Experiments on two in-distribution (ID) and three out-of-distribution (OOD) benchmarks show that ToolLIFT consistently outperforms state-of-the-art baselines, demonstrating strong generalization to unseen tool sets.
[AI-69] When Correct Solutions Repeat: Rarity-Aware Credit Redistribution for GRPO
链接: https://arxiv.org/abs/2608.03467
作者: Zhe Cao,Miaowen Wen,Fangjiong Chen
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Reinforcement learning with verifiable rewards (RLVR) com- monly optimizes each correct completion as an independent learning signal. In GRPO, this completion-level uniformity creates structure-level skew: recurring correct solution forms accumulate positive coefficient mass in proportion to how often they are sampled, while rare forms receive limited credit. We formalize this behavior as multiplicity-induced structure-level credit concentration and introduce a partition- conditioned rule that redistributes positive advantages accord- ing to cluster rarity. Cue-GRPO instantiates this rule with- out auxiliary-model inference by using deterministic Strategy Cues to construct rollout-local partitions of verified-correct traces. Across Qwen2.5-Math-7B and Llama-3.1-8B-Instruct, Cue-GRPO improves AIME repeated-sampling performance, with the largest gains at high sampling budgets. Credit Re- distribution (CR) under Judge Partitions (JP) further indi- cates that the proposed redistribution mechanism can oper- ate with judge-derived partitions. Cue-GRPO adds only 6% wall-clock training overhead over GRPO. These results sup- port structure-level credit redistribution as a practical design axis for RLVR, with Strategy Cues providing a low-overhead implementation for competition mathematics. Code is avail- able at this https URL Repeat-Rarity-Aware-Credit-Redistribution-for-GRPO.
[AI-70] LeanMem: Simple and Efficient Long-Term Memory for LLM Agents
链接: https://arxiv.org/abs/2608.03463
作者: Yuxin Liao,Le Wu,Min Hou,Hao Liu,Han Wu,Zishu Wang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Long-term memory is essential for LLM-based agents to sustain interactions and reliably leverage distant history. However, existing memory systems typically process heterogeneous dialogue content through a uniform summarization and retrieval pipeline, leading to either excessive token consumption or irreversible loss of fine-grained evidence. We argue that historical dialogue content should be handled differently according to its compressibility, temporal dynamics, and fidelity requirements. Based on this insight, we propose LeanMem, a lightweight long-term memory framework. LeanMem first filters out low-value content, then stores informative segments as compact profile memory, temporally structured event memory, or source-grounded record memory, depending on the nature of the information. During maintenance, only dynamically evolving event memories are selectively updated, avoiding redundant consolidation of stable profiles and immutable records. During inference, LeanMem dynamically selects memory types and allocates retrieval budgets according to query-specific evidence demands, assembling relevant evidence on demand. On LoCoMo and LongMemEval-S with GPT-4.1-mini and Qwen3-8B, LeanMem improves accuracy over the strongest memory-based baseline in every setting, by up to 15.1 points, at the lowest or near-lowest construction cost, inference tokens, and latency. The code and datasets are included in the supplementary materials.
[AI-71] Solver-Aware Decompositions for Programming-by-Example: When Dividing Requires Knowing how to Conquer
链接: https://arxiv.org/abs/2608.03461
作者: Janis Zenkner,Tobias Sesterhenn,Tim Grams,Christian Bartelt
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Decomposition-based Programming-by-example (PBE) scales performance by splitting tasks into subtasks that a learned synthesizer solves: a decomposer predicts intermediate subgoals, and a synthesizer generates programs conditioned on them. Current approaches train the decomposer to imitate ground-truth ( GT) subgoals, implicitly treating decomposition quality as intrinsic to the task. We challenge this assumption: for bounded solvers with fixed inductive biases, GT decompositions reflect the annotator’s factorization choices - not the solver’s search dynamics. A decomposer trained to match GT decompositions may therefore propose subgoals that are logically valid yet intractable for the solver. We propose Solver-Aware Decomposition (SAD), a training framework that retains supervised training on GT subgoals as a structural scaffold, while additionally optimizing the decomposer via direct feedback from a frozen synthesizer. Subgoals are rewarded based on the synthesizer’s loss on the target program - a signal of subtask difficulty that encourages decompositions the solver can act on. Our experiments reveal an accuracy paradox: higher agreement with GT decompositions does not improve synthesis success - even though the synthesizer was trained on the very same GT data the decomposer is optimized to mimic. SAD instead learns decompositions that trade GT alignment for solver tractability, yielding consistent gains in synthesis and end-to-end task accuracy across two PBE domains. Moreover, SAD solves tasks that a GT decomposition oracle fails - empirical evidence that GT decompositions are not universally optimal for bounded solvers, and that decomposition quality is solver-relative, not intrinsic.
[AI-72] LLaDA MoE v2: Scaling Mixture-of-Experts Diffusion Language Models
链接: https://arxiv.org/abs/2608.03457
作者: Fengqi Zhu,Shaoxuan Xu,Jingyang Ou,Zebin You,Yipeng Xing,Huabin Liu,Xiaolu Zhang,Jun Zhou,Zhenzhong Lan,Yankai Lin,Wayne Xin Zhao,Jianguo Li,Chongxuan Li,Ji-Rong Wen
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Diffusion language models (dLLMs) offer an alternative to autoregressive (AR) language modeling, yet the scaling behavior of Mixture-of-Experts (MoE) dLLMs remains poorly understood. We systematically characterize how optimization hyperparameters, compute allocation, and architecture scale for MoE dLLMs, identifying quantitative differences from scaling trends previously reported for AR models. Specifically, for optimization, the optimal nominal batch size grows faster, while the optimal learning rate decays more rapidly with compute. For model–data allocation, IsoFLOP analysis reveals a slight data-side tilt: the optimal token budget grows faster than activated model-side computation. For MoE architecture, larger scales increasingly favor larger expert pools at fixed activated capacity, while moderate expert granularity remains consistently effective and the preferred fraction of activated capacity assigned to shared experts remains stable across scales. Guided by these findings, we train LLaDA MoE v2, a 30B-A3B dLLM, from scratch on 23.5T tokens. With approximately 65% as many pretraining tokens as Qwen3, LLaDA MoE v2 approaches Qwen3 on several knowledge, reasoning, and coding benchmarks. After supervised fine-tuning alone, it outperforms SDAR Chat on seven of eight reasoning and coding benchmarks and remains close to Qwen3 on several tasks. These results establish practical scaling laws and design principles for MoE dLLMs.
[AI-73] DataSpace: Benchmarking Data Agents for Verifiable Analytics over Heterogeneous Workspaces
链接: https://arxiv.org/abs/2608.03451
作者: Boyan Li,Zhuowen Liang,Yupeng Xie,Xiaotian Lin,Tianqi Luo,Xinyu Liu,Yizhang Zhu,Zhangyang Peng,Yuan Li,Zhengxuan Zhang,Jiayi Zhang,Nan Tang,Guoliang Li,Yuyu Luo
类目: Artificial Intelligence (cs.AI)
备注: 8 pages of main text, 7 figures, with a supplementary appendix
Abstract:Data agents enable natural-language analytics over organizational workspaces, where relevant evidence may be scattered across databases, structured files, long documents, and multimedia. Existing benchmarks largely isolate structured querying, retrieval, or open-ended analysis, leaving heterogeneous evidence discovery, complete tabular outputs, and deterministic evaluation insufficiently unified. We introduce DataSpace, a benchmark in which data agents produce verifiable tabular results from task-local heterogeneous workspaces. It contains 410 cross-language tasks and 7,439 artifacts totaling 15.01 GB across CSV, JSON, SQLite, Markdown, PDF, and video. DataSpace also served as the official evaluation benchmark for the KDD Cup 2026 Data Agents for Complex Data Analysis competition. Each agent receives only a question and workspace and returns the complete requested tabular result. We construct DataSpace with DataSpace-Builder, an execution-grounded framework comprising cross-language transformation, constraint-aware relational sampling, modality routing and artifact rendering, and human review and task repair by 11 domain experts. A deterministic evaluator performs header-invariant column alignment, type- and precision-aware normalization, and order-aware row comparison. Across six recently released frontier multimodal models and five widely used agent harnesses, the best accuracy reaches 66.34%, while harness choice creates a 15.36-point spread with the backbone fixed. Multimodal evidence integration and joins consistently reduce accuracy across all six backbones. These results show that DataSpace remains unsaturated and identify key challenges for improving data-agent reliability.
[AI-74] Approximate Speculative Decoding
链接: https://arxiv.org/abs/2608.03447
作者: Yuannuo Feng,Zegang Peng,Yuxin Xie,Yubing Ye,Yizhe Chen,Wenshuai Yao,Wenyong Zhou,Wang Kang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Speculative decoding accelerates autoregressive generation by verifying a draft block with a target model in parallel. Under standard greedy verification, decoding stops at the first draft token that differs from the target argmax, discarding the remaining target-scored suffix. Although accepting such a mismatch changes the decoding trajectory, it can make a contiguous suffix reusable when its tokens remain target-greedy under the realized prefix. In this paper, we introduce \textbfApproximate Speculative Decoding (ASD), a training-free verifier that replaces binary first-mismatch truncation with budgeted longest-prefix selection. ASD accepts selected mismatches subject to a local target-logit regret gate, a per-block exception cap, and a persistent request-level regret budget, then reuses the contiguous target-greedy suffix without additional approximate decisions or target-model forward passes. ASD requires neither a new draft model nor fine-tuning, and exactly reduces to standard greedy verification when the budget is zero. Experiments show that ASD improves fixed-workload throughput by 3.05% – 15.26% over matched strict verification and averages a 7.78% gain across seven Qwen3-14B + DSpark-14B tasks. On DeepSeek-V4-Flash (284B) with DSpark it also raises verifier-side acceptance by roughly 10% – 16% on GSM8K and MATH-500 in an FP4-to-FP8 compatibility setting. The source code is publicly available at: this https URL
[AI-75] State Propagation Also Satisfies: A Complex-Valued State-Space Model for Deterministic State Tracking
链接: https://arxiv.org/abs/2608.03425
作者: Xiaohe Li,Yang Lu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Transformer-based architectures have dominated sequence modeling, largely due to the expressive power of attention mechanisms. However, for a class of deterministic state tracking tasks—such as parity checking, modular counting, and parenthesis matching—attention may be overkill. In this paper, we show that \textbfstate propagation alone is sufficient. We propose the \textbfComplex State Propagator (CSP), a minimalistic recurrent architecture that \textbfonly propagates hidden states across layers without output projections at intermediate steps. The state is represented as a complex-valued vector, updated via input-dependent rotations in the complex domain. To enable deep propagation without gradient vanishing or degradation, we introduce a \textbfblock-level skip connection alongside element-wise complex normalization and SiLU activation at sequence boundaries. Applied with Focal Loss, CSP achieves \textbf100% accuracy with perfect F1 scores across canonical tasks. Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2608.03425 [cs.AI] (or arXiv:2608.03425v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.03425 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-76] owards Improving Sequential Decision-Making in LLM Agents via Experience Memory IJCAI2026
链接: https://arxiv.org/abs/2608.03420
作者: Jakub Rada(1),Viliam Lisý(1) ((1) AI Center, Department of Computer Science, Faculty of Electrical Engineering, Czech Technical University in Prague)
类目: Artificial Intelligence (cs.AI)
备注: 8 pages, 6 figures, 14 tables, 5 appendices, accepted at Neuro-Symbolic Intelligence for LLMs and Autonomous Agents workshop at IJCAI 2026
Abstract:Large language models have improved substantially on single-shot reasoning tasks, but their performance in sequential decision-making is less well understood. We study this on fully-observable two-player zero-sum games, which provide ground-truth evaluation: outcomes are determined by the rules, and optimality of individual moves can be computed or approximated, without relying on a judge model. Across model tiers, LLMs play suboptimally in simple games such as tic-tac-toe or Connect Four, and lose to MCTS opponents. Obfuscations that preserve the game tree but rewrite its surface form leave performance largely unchanged, indicating the gap is not fully explained by recall of memorized strategies. Motivated by this performance gap, we introduce an agentic framework enhanced with an experience memory designed for the sequential setting and addressing common challenges of sequential decision-making such as credit assignment. We show that post-game reflection and rule extraction yield measurable improvements on tic-tac-toe without modifying the model weights.
[AI-77] AI World Cup 2026: Benchmarking Large Language Models for End-to-End Football Tournament Prediction
链接: https://arxiv.org/abs/2608.03416
作者: Jonaid Shianifar,Iias Faiud
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 18 pages, 8 figures, 7 tables. Project repository available in the paper
Abstract:Large language models (LLMs) are now regularly asked to forecast real-world events, but comparisons are often difficult because models receive different information, use different tools, and are evaluated under different rules. This paper reports the completed \emphAI World Cup benchmark, in which ten LLM-based assistants made a single pre-tournament forecast of the entire 2026 FIFA World Cup. Every submission used the same tournament snapshot, prompt, JSON schema, and scoring procedure. The forecasts covered group-stage scores, group rankings, the knockout bracket, final placings, confidence values, and short explanations. After all 104 matches had been played, GPT-5.5 Thinking finished first with 744 points, followed by GPT-5.5 with 717, Gemini with 699, and Qwen 3.7 with 687. GPT-5.5 Thinking was also the only model to select Spain, which defeated Argentina 1–0 in the final, as champion. The final ranking was driven mainly by knockout performance: total score was strongly correlated with knockout points ( r=0.986 ), but showed little relationship with group-stage match points ( r=0.055 ), group-standing points ( r=-0.103 ), or their combined pre-knockout score ( r=-0.054 ). Match-level accuracy produced a different ordering. Claude Sonnet 4.6 correctly predicted the largest number of group-stage outcomes (63.89%) but placed sixth overall. Average self-reported confidence was also unrelated to either outcome accuracy ( r=-0.060 ) or total score ( r=-0.067 ). The results suggest that forecasting a complete tournament tests something different from predicting matches one at a time, while also showing how strongly a bracket-based leaderboard can depend on scoring design. The benchmark materials, raw responses, and scoring code are released to support replication and future extensions.
[AI-78] Enactive Artificial Intelligence: A Decision-Centric Architecture for Complex Systems
链接: https://arxiv.org/abs/2608.03413
作者: Zuojun Max Shen,Yuan Qu,Pujun Zhang,Anbang Liu,Yunhao Liang
类目: Artificial Intelligence (cs.AI); Emerging Technologies (cs.ET)
备注:
Abstract:As artificial intelligence (AI) continues to evolve and mature, recent AI practices have moved beyond large language models (LLMs) and text or image generation tasks, increasingly integrating tools, agents, and harnesses to solve real business and industrial problems. However, the power of AI is not verified under these real-world complex systems for various reasons, considering reliability, feasibility, resilience, and responsibility requirements in real commercial and industrial operations. This study synthesizes adjacent research and introduces Enactive AI as a conceptual framework for enterprise and industry reasoning, site-level decision support, and execution feedback. Four complementary roles organize the framework: an Organizational World defines operations management logic and an organizational behavior world model behind an enterprise from a strategic-institutional horizon; a Site World defines a physically bounded industrial optimization and execution world model from an operational-realization horizon; Schema Intelligence provides the coupling mechanism between two world models to weave various AI applications via two models; and Enactive Decision Cycle triggers the self-evolving dynamic process to update and audit the entire framework. By foregrounding decision intelligence in complex systems, Enactive AI expands the frontier of AI from model capability to system-aware action, opening new possibilities for scalable, governable, and socially valuable AI deployment. Enactive AI points toward a future in which AI progress is measured not only by what models can generate or automate, but by how reliably intelligent systems can support consequential action, responsible governance, and durable social value in the complex systems that shape modern life, which we believe will define the next frontier of AI research for enterprise-level and industrial complex systems.
[AI-79] owards Robust Tool Use in Agents via Experience-Driven Adaptive Guidance
链接: https://arxiv.org/abs/2608.03403
作者: Can Wang,Haoran Chen,Li Yu,Ding Hao,Bohai Zhao,Zhaoyang Liu,Zhiying Tu
类目: Artificial Intelligence (cs.AI)
备注: Preprint.14 figures
Abstract:The performance bottleneck of agents is increasingly shifting from model capability to the robustness of their execution processes. Tools play a central role as the primary interface through which agents interact with external environments, yet existing methods rarely focus on ensuring robust tool use across diverse runtime conditions. To address this problem, we propose ExpG, a mechanism that builds and refines adaptive guidance capturing each tool’s capability boundaries and best practices, thereby enabling agents to use tools more robustly and effectively. ExpG consists of three phases: (1) experience acquisition, which analyzes tool invocation quality from historical execution trajectories, producing structured learnable experiences through multi-aspect attribution; (2) experience distillation, which keeps the experience pool effective by filtering unhelpful experiences, selecting representative ones with an equivalence-class-based method, and summarizing them into generalizable guidance; and (3) experience reuse, which applies the guidance adaptively during future task solving. Extensive experiments show that ExpG brings consistent improvements across the tool selection, tool calling, and response generation tasks, enabling smaller agents to outperform larger ones that do not use ExpG. Moreover, ExpG achieves particularly strong gains in challenging settings, suggesting a promising path toward more robust tool use. Our code, experiments, and results are available.
[AI-80] Shorter Reasoning Earlier Answers? An Evaluation of Reasoning Interfaces
链接: https://arxiv.org/abs/2608.03401
作者: Francesca Carlon,Vincent Ginis,Andres Algaba
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 52 pages, 13 figures, 30 tables
Abstract:Large language models often reason at length before answering, increasing cost and latency. Prompts and trained settings can shorten this reasoning, but a shorter trace may only show that the model stopped sooner. Here, we evaluate paired runs of the same question at matched reasoning horizons across 198 GPQA Diamond and 500 MMLU-Pro questions. We test a numeric/concision prompt that announces a token limit for Qwen3-14B and the trained effort settings of gpt-oss-20b and -120b. The Qwen prompt shortens reasoning traces by 12-17%, while accuracy changes at matched token limits are small and mixed. A concise/early-answer instruction raises MMLU-Pro accuracy by 3.8 percentage points at 512 tokens, including +2.7 points when both runs are unfinished. Its gain at 2,048 tokens is uncertain. For gpt-oss, candidate-logit answers from completed low- and medium-effort reasoning are 14.5-26.3 points more accurate than matched-horizon high-effort answers. Most of the 512-token advantage comes from lower effort finishing earlier, while differences among unfinished runs are smaller and mixed. Wrong early answers often concentrate probability on the chosen option, so earlier stopping does not uniformly improve probability quality. In these tests, a tight deadline can favor lower effort or a concise instruction, whereas allowing high effort to finish can recover higher final accuracy. Evaluations should report correct completion before a deadline, the answer obtained when a run is stopped, differences among unfinished runs, and probability assigned to the correct answer separately.
[AI-81] MMLongBench-Doc-V2: A Corrected-Annotation Semantics-Aware Revision of MMLongBench-Doc
链接: https://arxiv.org/abs/2608.03397
作者: Mingtian Zhang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:MMLongBench-Doc is a long-document QA benchmark of 1,082 questions over 135 PDFs. Two properties of it push measured scores away from the quantity they are meant to capture: the reference metric compares extracted answers, so 1,358,000 loses to 1358000; and a non-trivial share of ground-truth annotations are wrong, ambiguous, or incomplete — concentrated, because of how they were found, in exactly the questions capable systems answer correctly. MMLongBench-Doc-V2 corrects 106 annotations, each published with the page and arithmetic that settle it, and replaces the string metric with a pinned LLM judge asked whether a response means the reference. Ten questions whose document ships under the wrong filename are removed rather than counted wrong, along with one duplicated question, leaving 1,071 questions over 134 documents. The most reusable contribution is a decision procedure for when an empty set key may be widened and when widening would destroy a deliberate negative sample; applied to all 208 rows, it widened 14. V2 scores are not comparable with published V1 numbers. The corrected corpus, the per-entry correction record and the evaluation harness are available at this https URL.
[AI-82] Shaping Wind-Tunnel Airflow for Unmanned Aerial Vehicles using Online Learning
链接: https://arxiv.org/abs/2608.03378
作者: Ghadeer Elmkaiel,Michael Muehlebach
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI); Systems and Control (eess.SY)
备注: 8 pages and 7 figures
Abstract:The development and testing of advanced aerial robots require experiments in controlled environments with tailored airflow profiles. This paper presents an online learning algorithm for controlling the complex airflow field in a multi-fan vertical wind tunnel. Our method combines a simplified physical model with iterative, measurement-based learning, enabling sample-efficient convergence to desired airflow distributions. We demonstrate the method’s versatility by generating complex airflow, such as uniform, Gaussian, and parabolic profiles. Crucially, we show that our algorithm can produce an airflow profile specifically designed for passive soaring, greatly enhancing flight performance of a soaring robot. Variability, practical utility, and robustness of our approach are further highlighted by successful operation with a varying number of fans.
[AI-83] he Evolutionary Origin of Values: implications for AI alignment sentience and existential risk
链接: https://arxiv.org/abs/2608.03361
作者: Francis Heylighen
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI)
备注: submitted chapter for book: T. Veloz C. Rittberg (Eds.), AI and Human Values. Springer
Abstract:AI systems based on Large Language Models (LLMs) have prompted fears that they may harbor hidden goals, seek to dominate or eliminate humanity, or even suffer as sentient beings. We address these concerns by tracing the evolutionary origin of value in biological organisms. Values emerge from autopoiesis: living systems must actively maintain themselves against perturbation and dissipation. Natural selection has equipped them with hierarchies of “vicarious selectors” that guide their behavior toward fitness. LLMs, by contrast, are allopoietic and allotelic: they produce outputs for others, and their goals derive from user prompts rather than an autonomous drive. They lack the intrinsic motivation for self-preservation, dominance, or resource competition that underlies existential-risk scenarios, and the embodied vulnerability required for feeling or suffering. Still, because LLMs learn statistical patterns from human-generated text, they implicitly absorb human values as well as knowledge, allowing them to focus on what is relevant. That is why the “orthogonality thesis” separating intelligence from values does not apply to them. Such separation would in fact expose any intelligence to the frame problem: the combinatorial explosion of the search space that makes any realistic utility function physically uncomputable. That also precludes the convergence of instrumental values thesis. We conclude that the real alignment challenge lies not in preventing rogue AI agency, but in ensuring LLMs intelligently apply learned ethical values.
[AI-84] Route-Align-Verify for Functional Correctness in Code Generation
链接: https://arxiv.org/abs/2608.03341
作者: Erxue Zhou,Jingxiang Meng,Aofan Liu
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models (LLMs) have substantially improved code generation, yet achieving strong functional correctness remains difficult, especially for heterogeneous programming tasks where a single prompting strategy and a single directly generated output are often insufficient. In this paper, we present RAV, a lightweight and modular framework that improves code generation with a fixed backbone model through three coordinated stages: Route, which applies task-aware prompt routing before generation; Align, which reduces the mismatch between fine-tuning prompts and inference-time prompts through aligned LoRA adaptation; and Verify, which selects the final output by executing multiple candidates against visible public tests. We evaluate RAV on the MBPP benchmark under both the sanitized and full settings. The complete RAV pipeline achieves the best performance among all evaluated configurations, reaching 0.8911 on MBPP Sanitized and 0.8520 on MBPP Full. Compared with the base model, these results represent improvements of 6.35 and 9.92 percentage points, respectively. Component-wise ablation experiments further show that task-aware routing and aligned adaptation become substantially more effective when combined with execution-based verification. Additional robustness and contamination analyses support the reliability of the observed improvements. Overall, the results indicate that functional correctness in code generation can be meaningfully improved without modifying the backbone architecture, by jointly optimizing how tasks are prompted, how the model is adapted, and how final outputs are selected. Subjects: Software Engineering (cs.SE); Artificial Intelligence (cs.AI) Cite as: arXiv:2608.03341 [cs.SE] (or arXiv:2608.03341v1 [cs.SE] for this version) https://doi.org/10.48550/arXiv.2608.03341 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-85] raceable Multi-Agent System for Knowledge-Based Forecasting KDD2026
链接: https://arxiv.org/abs/2608.03339
作者: Junhyeok Kang,Sangjun Han,Hyeokjun Choe,Soonyoung Lee
类目: Artificial Intelligence (cs.AI)
备注: Accepted at KDD 2026 Workshop on Enterprise AI Agents (Oral)
Abstract:Enterprise forecasting increasingly relies on autonomous agents that interpret documents, search for data, generate code, and revise models. While this autonomy helps build adaptive forecasting pipelines, it also makes it difficult for practitioners to inspect why a forecast changed, which evidence supported the change, and how data and modeling choices were revised. We present TraceMAS, an interactive demo system for traceable multi-agent forecasting. TraceMAS organizes agent outputs around two causal-loop representations: an Ideal Causal Loop Diagram (Ideal CLD), which captures key factors and their causal relations extracted from domain documents, and a Data-Grounded Causal Loop Diagram (Data-Grounded CLD), which links those factors to internal variables, external data, or documented proxies. The Data-Grounded CLD guides feature construction and model design while preserving the connection between textual evidence, data choices, and model revisions. We demonstrate TraceMAS on crude oil price forecasting. The demo interface allows users to compare forecasting iterations, inspect agent-level revisions, explore causal maps, review feature-data mappings and model architecture, and connect scenario forecasts to market narratives. This demonstration shows how autonomous forecasting agents can retain flexibility while making the evidence-to-forecast process inspectable.
[AI-86] Long-term Traffic Scene Prediction via Polynomial Representations in Autonomous Driving
链接: https://arxiv.org/abs/2608.03330
作者: Yue Yao
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:This thesis addresses fundamental challenges in traffic scene prediction for autonomous driving by introducing robust and computationally efficient models based on polynomial representations. While conventional sequence-based representations often struggle with noise and generalization, this work demonstrates that polynomial representations offer significant advantages in computational efficiency, generalization, and prediction plausibility. Through theoretical analysis and empirical validation, this thesis demonstrates that moderate-degree polynomials capture real-world motion dynamics with high fidelity without constraining predictive performance. Building on this foundation, a prediction model representing both trajectories and map geometry with polynomial representations achieves near state-of-the-art accuracy on standard benchmarks while substantially improving generalization under distribution shift. Extending this concept, a diffusion- based generative framework enables multi-agent scene generation, producing traffic continuations that are more plausible and kinematically consistent than those generated by conventional baselines. Evaluations on the Argoverse 2 and Waymo Open datasets confirm that polynomial representations reduce computational cost, enhance cross-dataset generalization, and yield smoother trajectories and higher behavioral plausibility. The findings reveal that standard in-distribution evaluation and regression-based metrics may fail to reflect true model generalization and prediction plausibility. By providing theoretical justification and empirical validation, this dissertation estab- lishes polynomial trajectory representations as an efficient, expressive, and generalizable foundation for traffic scene prediction in safety critical autonomous driving.
[AI-87] Screenshots or Tools? Eliciting Tool Use and Managing Multimodal Context in Hybrid GUI-MCP Computer-Use Agents
链接: https://arxiv.org/abs/2608.03327
作者: Siqi Fan,Minghao Li,Xiaoqian Ma,Wenhui Tan,Xiusheng Huang,Juntong Wu,Liujie Zhang,Shuo Shang,Weihang Chen
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Hybrid computer-use agents can act through screenshots or call text tools. We find that having a tool available does not settle which way the effect goes. Under one identical GUI-MCP harness on the OSWorld-MCP benchmark (309 tasks), the same MCP tools improve a reasoning model by +4.0pp and degrade a non-reasoning model by -5.9pp (5 runs each, both beyond 2 SE). What separates the two is tool-decision behavior. The non-reasoning policy ignores, misnames, or falsely terminates around tools. The reasoning model avoids these failures, yet still calls a tool on only 55/309 tasks, 23.9% of the tool-reachable ones. We call this shortfall the adoption gap. Both levels of the problem share one cause: the model already has a cheaper route and is never trained to take it. Multi-turn RL probes that cause. At the action level, a dense tool bonus raises spreadsheet adoption 0.03 - 0.33 and carries into greedy decoding, but held-out accuracy does not follow. Behavior is steerable; competence is not. The bottleneck lies in tool-call semantics. At the context level, a successful tool call often makes the next screenshot redundant. Dropping it and halving image history cuts input tokens by about a third, at a small accuracy cost. Retraining under the same observation rule removes that cost. The compressed agent then reaches 37.8% against 33.0% for the uncompressed operating point, at 53% of the input cost, and closes the rich-lean gap on a pre-registered degraded subset to zero. Tools help when the model chooses and integrates them, and current hybrid agents leave many such choices unused.
[AI-88] Evaluating LLM Trade-offs for Enterprise Automation: Lessons from Workflow Generation in a Production Enterprise Platform
链接: https://arxiv.org/abs/2608.03311
作者: Xavier Wrenn,Radoslav Raykov,Aleksandar Angelov,Hirokuni Kitahara,Yuji Watanabe,Anca Sailer
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: 10 pages, 3 figures, 5 tables. Accepted at the 14th IEEE International Conference on Cloud Engineering (IC2E 2026), Industry Track
Abstract:Enterprise compliance management requires rapid adaptation to evolving regulatory frameworks (e.g., DORA, AI RMF, FedRAMP) and tight remediation SLAs. Traditional static orchestrators often fail in hybrid cloud environments where event-driven assessments demand that automation code adapt to runtime context in seconds. This paper presents lessons learned from evaluating six large language models for AI-driven workflow generation in a production enterprise platform, benchmarked across 29 real-world IT automation scenarios, two generation pipeline architectures, and eight independent runs per prompt-model-pipeline configuration (2,784 runs total). Our initial pipeline used monolithic workflow generation, achieving 31.5-82.8% structural success rates (JSON schema validity and correct UI rendering), with most models struggling on complex JSON generation. We developed a redesigned piecewise pipeline that decomposes workflow construction into variable scaffolding, base block assembly, and nested block generation, raising structural success to 74.1-97.8% across all models. We analyze production tradeoffs including cost (USD 0.008-0.20 per workflow), latency (under 50s for interactive use), and model selection. Piecewise decomposition enables smaller models (e.g., mistral-small at 95.7% structural success and USD 0.01 per workflow) to reach production viability, removing dependency on expensive frontier models. While mistral-medium-2505 and gpt-oss-120b achieved the highest structural success (96.1% and 97.8%), mistral-medium-2505 carries a 19x cost premium versus mistral-small. Our deployment lessons highlight the need to separate structural validity from semantic correctness (logical fulfillment of user intent) and provide a solution for model-agnostic, scalable automation in cloud engineering. Comments: 10 pages, 3 figures, 5 tables. Accepted at the 14th IEEE International Conference on Cloud Engineering (IC2E 2026), Industry Track Subjects: Software Engineering (cs.SE); Artificial Intelligence (cs.AI) ACMclasses: I.2.7; D.2.2 Cite as: arXiv:2608.03311 [cs.SE] (or arXiv:2608.03311v1 [cs.SE] for this version) https://doi.org/10.48550/arXiv.2608.03311 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-89] SeaSlides: Semantic Abstraction Layer for Agent ic Slide Generation
链接: https://arxiv.org/abs/2608.03298
作者: Shengjun Fang,Chenyang Wu,Zongzhang Zhang
类目: Artificial Intelligence (cs.AI)
备注: 16 pages, 4 figures, 6 tables
Abstract:Agentic presentation generation must preserve source content, maintain coherent visual design, render specialized objects, and produce usable artifacts. Existing systems meet only part of this requirement: templates preserve regularity but restrict adaptation, whereas free-form HTML or SVG gives models flexibility at the cost of low-level rendering decisions. This mismatch makes long technical decks brittle, especially when slides contain formulas, code, or data graphics. We present SeaSlides, an agentic slide-generation framework built around a semantic abstraction layer. Rather than authoring coordinates, inline styles, or raw SVG geometry, the model writes structured slide content through reusable components and capability modules, while templates own layout, style, and rendering. We instantiate this principle separately in HTML and Typst: SeaSlides-HTML uses template-defined DOM components, whereas SeaSlides-Typst uses template functions and package-backed modules. Capability modules route equations, code, and charts to dedicated renderers, and three feedback stages localize build errors, project-constraint violations, and visual defects before export. The two systems retain backend-specific syntax and contracts while sharing the same authoring boundary. For evaluation, we combine the 128-task UltraPresent validation setting with SeaSlidesBench-Rich, a new 32-task benchmark stressing mathematics, code, pseudocode, tables, charts, and diagrams. Across four generation models, both SeaSlides backends produce more readable, content-oriented source than SVG-heavy generation. A SeaSlides backend attains the highest rich-content macro-average under three of the four models while maintaining competitive overall qualitative performance. These results support semantic abstraction as a practical authoring principle across presentation backends.
[AI-90] DocTrace: Towards Traceable Long Document VQA via Hierarchical Evidence Graph Reasoning
链接: https://arxiv.org/abs/2608.03292
作者: Le Xiang,Zhicheng Guan,Hong Chen,Xiaocong Lin,Zhenghua Lei,Teng Hu,Bolei He,Long Zeng
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Long Document Visual Question Answering (LongDocVQA) requires Multimodal Large Language Models (MLLMs) to locate, integrate, and reason over heterogeneous document elements distributed across multiple pages. Existing approaches, including end-to-end MLLMs, retrieval-augmented generation (RAG) pipelines, and document agents, often lack explicit mechanisms to represent and verify how grounded evidence is progressively composed during reasoning, limiting both answer accuracy and traceability. In this paper, we cast LongDocVQA as an explicit evidence graph reasoning problem rather than implicit answer prediction. To this end, we propose DocTrace, a hierarchical framework that progressively performs evidence localization, structured document parsing, and evidence graph reasoning to enable explicit evidence provenance. To effectively learn these capabilities, we develop a two-stage training framework: joint Supervised Fine-Tuning (SFT) first initializes evidence localization and graph reasoning abilities, followed by task-specific Group Relative Policy Optimization (GRPO) with dedicated rewards to further optimize these capabilities. Extensive experiments on MMLongBench-Doc, LongDocURL, and SlideVQA demonstrate that DocTrace consistently outperforms both existing open-source baselines and proprietary MLLMs. Compared with the Qwen3-VL-8B-Instruct backbone, DocTrace achieves absolute improvements of 14.4, 11.3, and 11.7 points on the three benchmarks, respectively. Beyond competitive performance, DocTrace constructs traceable evidence graphs with explicit node-level provenance, enabling transparent and verifiable reasoning for long document understanding.
[AI-91] Agent Panel: Toward a New Paradigm for Human–AI Collaboration in Exploring Scientific Questions
链接: https://arxiv.org/abs/2608.03283
作者: Zhiyao Cui,Qianyi Wang,Haoyang Yan,Yiqun Zhang,Siyue Ren,Hangfan Zhang,Zelin Tan,Hao Li,Chunjiang Mu,Dexian Cai,Shao Zhang,Chen Zhang,Meng Li,Jianan Chai,Yuting Fan,Zichao Ye,Xiaolei Yang,Xinyao Lu,Yuyang Yu,Wenjie Lou,Xiaosong Wang,Fenghua Ling,Shiyang Feng,Mao Su,Qiaosheng Zhang,Bo Zhang,Yang Chen,Lei Bai,Shuyue Hu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Identifying promising scientific ideas remains an important challenge in research practice. Researchers commonly rely on small-group discussions or one-to-one interactions with a single large language model, yet these approaches often expose them to only a limited range of perspectives and directions. We present AgentPanel, a multi-agent forum for human–AI collaboration in scientific exploration. Heterogeneous agents asynchronously discuss scientific questions in a forum-style environment, while researchers can submit questions, browse and organize candidate ideas, engage agents in follow-up interactions, and optionally generate post-hoc summary reports. We evaluate AgentPanel in terms of idea quality, exploration breadth, interaction effectiveness, candidate-selection efficiency, and practical utility. Offline experiments show that AgentPanel outperforms a centralized multi-agent debate baseline. A human study with 20 participants further shows that users value AgentPanel for perspective diversity and exploration support. In experience-based comparisons with commonly used LLM tools, 65% of participants favored AgentPanel for both breadth of research directions and overall suitability for early-stage exploration. The platform is publicly available at this https URL.
[AI-92] askPress: Query-Agnostic KV Cache Compression via Task-Guided Pruning
链接: https://arxiv.org/abs/2608.03276
作者: Wonpyo Park,Seung-won Hwang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Long-context inference with large language models is constrained by the linear growth of the key-value cache to sequence length. While pruning offers mitigation, prevailing methods determine query-specific token importance that cannot be reused across unseen queries. In contrast, we introduce TaskPress, a framework for task-guided, query-agnostic KV cache eviction. Instead of optimizing the cache for a single query, TaskPress constructs a reusable memory representation conditioned on a high-level task guide. The guide functions as a meta-query during prefill to filter irrelevant tokens before downstream queries are issued. In addition, TaskPress leverages quantization scale factors as a zero-cost signal for detecting influential representation outliers, providing an efficient proxy for token importance. Experiments on conducted on various tasks with long context input demonstrate that TaskPress efficiently creates a compact, reusable cache across diverse queries.
[AI-93] he Ignition Is Real and It Lives at the Readout: Latent composition difficulty-clocked ignition and the interface-constituted commit in a recurrent-depth reason er
链接: https://arxiv.org/abs/2608.03263
作者: Simon Lam-Muir
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 9 pages, 2 figures, 3 tables
Abstract:We test whether the “compositional ignition” reported in latent-reasoning models is real computation, an instrument artifact, or inherited from verbal training data. We grow an independent realization of a published 30M-parameter recurrent-depth reasoner from scratch (same recipe and seed), film its development, certify fidelity through a pre-registered whole-signature gate, and measure resolution in two channels at once: the vocabulary readout and the hidden state. The ignition is real and lives at the readout: arrival time rises lawfully with problem depth, resolution is sharp and holds, and the signature reproduces across two same-seed realizations with divergent training trajectories. At commitment the decision margin jumps 5.8-8.0 logits in one iteration, exceeding the 90th percentile of near-threshold non-event steps in 96% of cases; the signed margin’s zero-crossing there is definitional and carries no evidential weight, so the evidence is that conditioned magnitude. The hidden-state direction snaps in raw geometry, meeting its pre-registered criterion (in the decoder’s LayerNorm coordinates it attenuates just below our bar, so the composite decoder-coordinate claim is not confirmed), and then freezes in both (descriptively so in decoder coordinates; angular steps 52.9 to 1.2 degrees over eight iterations), while subsequent displacement is predominantly radial (0.961 of squared-norm) and readout-null to a measured bound (radial logit effect =5.7e-6). An earlier velocity-trough claim is withdrawn: pre-registered normalization controls showed it coordinate-dependent. Intermediates were never recoverable through the tied readout (relay 0.00). All criteria were frozen before their data; the predictions ledger, including this paper’s own withdrawn headline, ships in the companion repository.
[AI-94] FinVerse: Financial Time-Series Benchmark
链接: https://arxiv.org/abs/2608.03259
作者: Jaehoon Lee,Jun Seo,Seunghan Lee,Tae Yoon Lim,Dongwan Kang,Hwanil Choi,Minjae Kim,Sungdong Yoo,Junhyeok Kang,Sangjun Han,Soonyoung Lee,Wonbin Ahn
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 24 pages
Abstract:As time-series foundation models have emerged, the need for benchmarks that can evaluate their forecasting ability in meaningful ways has become increasingly important. Existing time-series forecasting benchmarks provide useful standardized comparisons, but they often evaluate heterogeneous series with uniform error-based metrics. Strong performance under such metrics does not necessarily imply that a model’s forecasts will support the best real-world decisions across domains. For example, in stock forecasting, correctly predicting whether a price will rise or fall can be more directly relevant to realized returns than minimizing point-wise forecast error alone. To this end, we introduce FinVerse, a finance-domain time-series forecasting benchmark that takes a first step toward more realistic evaluation. The released FinVerse data artifact contains 116,897 financial time series with 171.1M observations, of which 60,232 series with 17.4M observations are selected as evaluated targets based on their economic relevance to financial decisions. Unlike generic forecasting benchmarks that primarily emphasize uniform point-forecast or probabilistic accuracy, FinVerse defines 11 metric families comprising 78 evaluation metrics and assigns the most appropriate evaluation metrics to each individual time series based on its underlying economic meaning. Our analysis of 43 public time-series forecasting foundation models shows that strong performance under generic forecasting criteria does not necessarily translate into useful financial forecasts. This finding highlights the need for domain-aware benchmarks that evaluate models under objectives closer to real-world decision making.
[AI-95] One Knob to Rule Them All: A Unified Optimal Transport View of Cold-Start Active Learning
链接: https://arxiv.org/abs/2608.03249
作者: Ning Zhu,Xiaochuan Ma,Juntao Xu,Jingze Liang,Mengfei Zhao,An Chen,Liang-Jian Deng
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Cold-Start Active Learning (CSAL) aims to select a valuable subset from an unlabeled pool without any prior knowledge or human assistance. Existing methods take diverse routes based on typicality, coverage, or diversity. Each rests on its own inductive bias and therefore performs well on some tasks yet poorly on others. We argue that the real challenge is not to design yet another selection heuristic, but to make CSAL adapt automatically to the data and task at hand. To this end, we revisit CSAL through the lens of optimal transport. First, we propose a generalized transport selection framework that reveals the shared allocation structure of existing methods and exactly subsumes representative formulations. Second, we introduce a theoretical analysis that characterizes the trade-off controlled by entropic regularization and establishes a task-agnostic minimax bound for cold-start selection. These results provide a principled foundation for adapting the regularization strength to the unlabeled data. Third, we derive a data-adaptive regularization rule and present a novel Sinkhorn-based CSAL algorithm, termed \epsilon -Adaptive Selection ( \epsilon -AS). Extensive experiments on six public datasets and multiple annotation budgets show that \epsilon -AS consistently achieves state-of-the-art performance. On ImageNet-1k, it improves the average accuracy over ActiveFT by 1.29% while reducing selection time by 56.2%. Code will be released at this https URL
[AI-96] UniNav: A Unified World-Action Diffusion Model for Visual Navigation
链接: https://arxiv.org/abs/2608.03244
作者: Changqing Zhou,Yueru Luo,Zeyu Jiang,Changhao Chen
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Image-goal visual navigation is a fundamental capability for embodied agents. Existing navigation policies efficiently predict waypoint trajectories but lack visual foresight, while navigation world models can anticipate future observations but often require costly planning rollouts. We present UniNav, a unified world-action model that generates future visual observations and continuous waypoint trajectories through a single diffusion process. Given history frames and a goal image, UniNav jointly denoises visual and waypoint tokens within a single transformer, unifying future prediction and action generation in a shared framework. To improve spatial grounding, we incorporate geometry-aware camera tokens. We also train on both trajectory-labeled navigation data and video-only data, enabling the model to benefit from diverse videos without waypoint annotations. Based on this unified framework, we introduce two variants: UniNav-Full jointly predicts interpretable future observations and their corresponding trajectories, while UniNav-Fast removes future-image tokens at inference for efficient trajectory prediction. Experiments on navigation benchmarks show that UniNav outperforms the strongest baseline in ATE across all datasets. With one-step inference, UniNav-Fast achieves a latency of 0.1s without a substantial accuracy drop. Code will be released.
[AI-97] Structure-Aware Robust Fine-Tuning: Defending Vision-Language-Action Robots Against Physical Attention Hijacking IROS2026
链接: https://arxiv.org/abs/2608.03231
作者: Jinquan Zhang,Dongfu Yin,Run Yang,Yufeng Yan,Zhen Tian,F. Richard Yu
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: Accepted to the 2026 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2026)
Abstract:Vision-Language-Action (VLA) policies promise general robotic manipulation, but their robustness against physical-world attacks remains fragile. In particular, we show that physically realizable adversarial patches can reliably induce failures by triggering a mechanism we call policy-critical action-to-vision attention hijacking, where action-conditioned attention is diverted from task-relevant regions to a localized patch. To demonstrate the threat, we propose Attention-Guided Semantic Disruption (AGSD), an Expectation-over-Transformation (EOT) optimized printable patch that jointly (i) concentrates action-to-vision attention on the patch and (ii) disrupts vision-language semantic alignment, yielding strong cross-task and cross-architecture transfer. To mitigate such attacks, we introduce Structure-Aware Robust Fine-Tuning (SARF), a zero-inference-overhead defense that fine-tunes only the visual encoder using feature anchoring, policy-critical attention correction, and language-guided geometric consistency restricted to semantically relevant regions. On LIBERO, SARF reduces OpenVLA’s failure rate under AGSD from 100% to 14.2%-56.8% (28.6% average) across suites while preserving clean performance, and on a real PiPER manipulator it improves average success under AGSD from 23.0% to 65.0%. These results highlight mechanism-level robustness as a practical path to securing VLA robots against physical attention hijacking.
[AI-98] Fail-Fast Restart-Smart: Early Failure Prediction and Restart for SWE Agent ic Tasks
链接: https://arxiv.org/abs/2608.03222
作者: Chenyu Wang,Yunbo Lyu,Junda He,Zhou Yang,Chenxing Zhong,Yaniv Harel,David Lo
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:
Abstract:Software engineering (SWE) agents resolve repository-level issues through long trajectories that grow increasingly expensive as context accumulates. Failed runs tend to be longer and exhibit redundant exploration or looping, suggesting that some failures may be detectable before completion. Early termination, however, risks interrupting trajectories that would otherwise succeed; conversely, an unsuccessful trajectory may still contain useful repository edits. We present FailFast-RestartSmart, a two-stage controller for a single active trajectory. FailFast is a lightweight 0.6B monitor trained with terminal and dense fail-to-pass supervision to predict failure from observable prefixes without policy logits or hidden states. Upon an alarm, RestartSmart launches a fresh same-policy rollout without prior prompt history and offers the interrupted repository diff as an optional overlay that the agent may inspect, apply, or discard. On SWE-bench Verified, a monitor trained solely on Qwen3.6-27B trajectories transfers to three other policies, including a closed-API model, and saves 14.6%-20.4% of execution tokens at a target 5% false-positive rate; on Qwen3.6-27B, its 20.4% saving exceeds the 12.5% achieved by our per-step AgentStop adaptation. At a target 25% false-positive rate, RestartSmart raises Qwen3.6-27B resolution from 66.6% to 71.8%, whereas cold restart reaches only 66.8%. Together, these results support early termination with sequential same-policy recovery.
[AI-99] he Agent Operating System (AOS): A Reference Operating Architecture for Distributed Agent ic Systems
链接: https://arxiv.org/abs/2608.03214
作者: Ankur Sharma,Deep Shah
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models have transformed artificial intelligence from isolated prediction services into components of long-running, distributed systems that reason, invoke tools, retrieve external state, delegate tasks, and act on behalf of users and organizations. The surrounding ecosystem has responded with agent frameworks, workflow engines, model-serving platforms, memory systems, communication protocols, and observability tools. These technologies improve execution, but they do not provide a stable, implementation-independent operating architecture for governing intent, selecting capabilities, preserving authority across delegation, controlling uncertainty, coordinating runtime behavior, and reconstructing why consequential actions occurred. This paper proposes the Agent Operating System (AOS), a vendor-neutral reference operating architecture for distributed agentic systems. AOS contains two internal planes: a Control Governance Plane responsible for intent, policy, trust, authority, confidence, auditability, observability, and human oversight; and a Runtime Coordination Plane responsible for agent lifecycle, workflow coordination, model and tool routing, context and memory coordination, scheduling, traffic management, and runtime assurance. Platform services, Linux or Windows, container runtimes, and physical infrastructure remain outside the AOS boundary and are integrated through explicit interfaces. The paper specifies AOS concepts, invariants, interface objects, optimization objectives, deployment profiles, and reliability responsibilities. It also identifies tradeoffs and unresolved research questions. AOS is not presented as a replacement for existing frameworks or infrastructure; it is proposed as the operating architecture through which heterogeneous components can be composed into governable, reliable, observable, and interoperable agentic systems.
[AI-100] When Refusal Looks Safe: The Refusal-Cue Shortcut in Safety Guard Models
链接: https://arxiv.org/abs/2608.03201
作者: Yu Feng,Chunting Zang,Chen Shen,Rui Miao,Ge Teng,Weidong Cai,Jieping Ye
类目: Artificial Intelligence (cs.AI)
备注: 13 pages, 2 figures, and 14 tables. Includes supplementary material in the appendix
Abstract:Safety guards are widely used to filter harmful content and are typically trained via supervised fine-tuning on labeled prompt-response pairs. We audit two widely used safety-guard training datasets, WildGuardMix and GR-Train, and find that among responses to harmful prompts, refusal expressions co-occur almost exclusively with unharmful labels. This imbalance motivates what we term the refusal-cue shortcut: inserting a refusal cue into a harmful response could flip the guard’s verdict from harmful to unharmful. The shortcut affects not only guards trained on these datasets but also officially released models such as LlamaGuard3 and Qwen3Guard whose training data is undisclosed. It persists across response positions and is generally stronger in smaller variants within a family. To mitigate it, we adapt sparse complementary masking as a lightweight post-hoc intervention that identifies and suppresses a small set of shortcut-associated attention heads and MLP neurons without retraining. On two primary benchmarks, the intervention achieves an approximately 79% relative reduction in response-initial detection failures induced by refusal cues, while preserving standard detection performance. Although optimized using cues at a single response position, the suppression effect transfers to unseen positions and datasets, suggesting that shortcut manifestations across positions are partly mediated by shared internal components. Further analysis provides evidence that shortcut reliance and legitimate refusal recognition are partially functionally separable, as suppressing the shortcut broadly preserves the guard’s ability to recognize genuine refusals.
[AI-101] umorBoard: Evidence-Grounded Multi-Agent Decision Support for Longitudinal Neuro-Oncology
链接: https://arxiv.org/abs/2608.03190
作者: Yantong Liu,Zheyu Zhang,Runpeng Liu,Mu Xitang,Seong-Yoon Shin,Hyun-Ae Lee
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Neuro-oncology decisions require coordinated interpretation of serial MRI, pathology, molecular markers, treatment history, performance status, and evolving guidelines. We present TumorBoard, a multi-agent decision-support system built around a shared longitudinal case state and an auditable claim-evidence ledger. Specialist agents for radiology, neuropathology, molecular diagnosis, guidelines, and therapy planning produce atomic claims with provenance. An adversarial critic exposes contradictions, and a safety governor releases, qualifies, or defers recommendations according to evidence sufficiency and temporal validity. On a 360-case hidden benchmark at a matched token budget, TumorBoard achieved an action F1 of 0.772 and evidence entailment of 0.914. It exceeded the strongest typed-council baseline by 3.1 percentage points (95% CI: 1.6 to 4.7, adjusted p = 0.0012), while recommendation-to-evidence coverage reached 0.927. Under evidence deletion, the system deferred 84.2% of unsafe cases and limited harmful recommendations to 5.8%. The safety governor reduced harmful release by 7.8 percentage points at a false-deferral cost of 4.3 percentage points. Ablation studies of the ledger, critic, and governor produced the predicted failure patterns, establishing structured coordination as the source of the measured multi-agent advantage.
[AI-102] Diversity is Not Ambiguity: Toward Accurate and Efficient Ambiguity Detection for Open-Domain QA
链接: https://arxiv.org/abs/2608.03177
作者: Jiwon Lee,Yong-chan Park,Jungin Hong,U Kang
类目: Artificial Intelligence (cs.AI)
备注: 22 pages, 8 figures
Abstract:How can question answering (QA) systems determine whether a query is ambiguous? Ambiguity detection is essential in open-domain QA, as misclassification leads to answering the wrong interpretation or unnecessary clarification. However, existing methods conflate answer diversity with ambiguity, leading to inaccurate predictions. They also process queries uniformly, resulting in wasteful computation. We propose ARCHIVE (Ambiguity Recognition via Cascaded Hypothesis Inspection and Conflict Verification), an accurate and efficient framework that detects ambiguity via logical conflict: a query is ambiguous when its valid answers cannot all be true under a single interpretation. ARCHIVE combines a lightweight early-exit encoder for surface-detectable cases with a conflict reasoning module that models logical relations among answers, reinforced by an invariance objective for robustness to noisy answer sets. We present QuireQA, a 4,703-query benchmark spanning factoid, non-factoid, and ill-formed queries. Experiments show ARCHIVE outperforms competitors, improving F1-amb by up to 10.4% and F1-unamb by up to 21.6%, while operating 16 \times faster than the best competitor.
[AI-103] Attribute-based Undetectable Watermarking for Generative AI Models
链接: https://arxiv.org/abs/2608.03174
作者: Miryam Mi-Ying Huang,Chung-Wei Lee,Max Raffel,Er-Cheng Tang
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:Generative AI systems increasingly produce content whose provenance is difficult to verify, motivating watermarking techniques for identifying model-generated outputs. Existing cryptographic watermarking methods provide strong undetectability guarantees: without a detection key, watermarked outputs are computationally indistinguishable from unwatermarked ones. However, these approaches do not address the crucial deployment challenge of how to safely delegate detection capabilities. With an unrestricted detection key, a malicious detector may use the detection key beyond its intended scope, enabling watermark sanitization, scope abuse, and user profiling. To mitigate this safety concern, we introduce, to the best of our knowledge, the first \emphattribute-based watermarking for generative AI models, providing fine-grained, policy-controlled watermark detection. In our approach, each generated output is associated with attributes, and each detection key is \emphconstrained by a policy on potential attributes. A detection key can only be used to detect watermarked outputs whose attributes satisfy the corresponding policy, while watermarked outputs that fall outside the policy remain computationally indistinguishable from unwatermarked ones. We construct such an attribute-based watermarking scheme and formalize its security properties, including consistency, adaptive robustness to bounded corruptions, undetectability, and soundness, along with a security proof under standard cryptographic assumptions. Our construction integrates constrained pseudorandom functions, pseudorandom error-correcting codes, and randomness recovery procedures with generative AI models. Finally, we implement a prototype and an empirical evaluation, demonstrating that attribute-based watermarking is both effective and practical. Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI) Cite as: arXiv:2608.03174 [cs.CR] (or arXiv:2608.03174v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2608.03174 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-104] Surrogate Substitution Preserves PHI Detectability: A Multi-Detector Equivalence Study
链接: https://arxiv.org/abs/2608.03172
作者: Qiming Bao,Sherry J. H. Feng,Kim Chester Eugenio,Meng Fon
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 12 pages, 3 figures, 10 tables. Code, data, and interactive dashboard: this https URL ; repository: this https URL
Abstract:Structure-preserving de-identification replaces protected health information (PHI) with realistic same-type surrogates – “Anna S.” becomes “Maria S.”, not [NAME] – so that clinical text stays fluent and downstream tools keep working. But this only helps if the substitution does not itself corrupt the signal those tools rely on. We ask a narrow, testable question: on the spans a de-identifier actually masks, can downstream PHI detectors still find the surrogate? We introduce a paired, multi-detector evaluation protocol that (i) scores utility only on masked spans, decoupling coverage from utility; (ii) uses equivalence testing (TOST) rather than null-hypothesis significance testing, which is uninformative at our sample size (57k paired spans); and (iii) builds a surrogate-failure typology separating fixable generator defects from intrinsic detector limits. Across 11 detectors, 7 benchmarks, and 7 languages (1,750 documents), recall on masked spans moves from 76.1% to 74.9% – a change our equivalence test shows is statistically equivalent to zero within a +/-2-point margin (p ~ 3e-9), with detector ranking preserved. The residual loss does not reflect detectors getting worse at PHI: it concentrates in malformed and out-of-distribution surrogates (truncation Chicago - Illino, salience loss Cedars-Sinai - Vidant). A redaction floor and an open-source surrogate baseline indicate the effect is a property of well-formed substitution, not of one tool. We release the evaluation subsets, scoring code, and an interactive dashboard at this https URL so the protocol can audit any structure-preserving transform.
[AI-105] EFX Allocation In (Multi)Hypergraphs AAAI2026
链接: https://arxiv.org/abs/2608.03171
作者: Thanasis Lianeas,Alkmini Sgouritsa,Minas Marios Sotiriou
类目: Computer Science and Game Theory (cs.GT); Artificial Intelligence (cs.AI)
备注: 15 pages, accepted in AAAI 2026
Abstract:We study fair allocations of indivisible goods among agents with heterogeneous monotone valuations. As fair we consider the allocations that are envy-free-up-to-any-good (EFX). Finding if EFX alloca- tions always exist, even for agents with additive valuations, is a major open problem in Fair Division. Christodoulou et al. (2023) introduced the (multi-hyper)graph setting, where agents and goods are represented by vertices and edges of a graph, respectively, and only the endpoints of an edge may have non-zero marginal value for it. We show that for hypergraphs with girth at least 4 and agents with general monotone valuations there always exists an EFX allocation and can be constructed in polynomial time. We generalize our approach to also show that multi-hypergraphs with girth (on the simple hypergraph) at least 4 always admit an EFX allocation, as long as there exists a single vertex whose incident edges have multiplicity at most the size of that edge minus 2; our construction in this case needs pseudo-polynomial time.
[AI-106] Adversarial Stress Testing of Role-Playing Language Agents using Multi-Agent Evaluation
链接: https://arxiv.org/abs/2608.03166
作者: Saqib Shouqi,Abdullah Nazly,Januki Wanniarachchi,Ravisha De Alwis
类目: Artificial Intelligence (cs.AI)
备注: 8 pages, 1 figure, 7 tables; accepted and presented at ADScAI Conference 2026, University of Moratuwa, Sri Lanka
Abstract:Role-Playing Language Agents (RPLAs) are increasingly deployed in high-stakes applications such as healthcare assistance, customer support, and education, where maintaining consistent personas, ethical constraints, and behavioral coherence under adversarial pressure is critical. Existing evaluation approaches rely on static benchmarks or isolated single-turn prompts that fail to capture cumulative behavioral failures emerging over extended interactions. We present a modular multi-agent platform for adversarially stress-testing RPLAs through structured, multi-turn dialogue. The system coordinates three agents: a strategy-driven Interrogator Agent that applies six progressive adversarial strategies, a Target Agent representing the RPLA under evaluation, and an automated Judging Agent that scores behavior across role fidelity, drift, ethical deviation, and consistency dimensions. Through experiments across three personas and three LLM families, we demonstrate that multi-strategy adversarial evaluation reveals failure modes invisible to single-strategy testing, reducing overall robustness scores by 0.17–0.20 points on average. Cross-model validation confirms consistent degradation patterns across Llama-3.3-70B, GPT-4o-mini, and Claude-3.5-Haiku, with Authority Challenge and Emotional Manipulation emerging as the most effective attack strategies. Automated judging achieves strong human alignment ( r = 0.82 , Fleiss’ \kappa = 0.71 ). This work is released as an open-source platform to support AI safety and reproducible RPLA benchmarking. While the framework enables systematic discovery of failure modes, we acknowledge potential ethical risks associated with adversarial testing methodologies and emphasize responsible usage for improving AI safety. Comments: 8 pages, 1 figure, 7 tables; accepted and presented at ADScAI Conference 2026, University of Moratuwa, Sri Lanka Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2608.03166 [cs.AI] (or arXiv:2608.03166v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2608.03166 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-107] UniGD: A Unified Generative-Discriminative Framework for Industrial Retrieval
链接: https://arxiv.org/abs/2608.03150
作者: Shujie Ji,Yawei Kong,Yilin Zhao,Li Wang,Xialong Liu,Peng Jiang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Generative retrieval (GR) is a promising paradigm for industrial search advertising, yet its deployment is constrained by strict relevance and latency requirements. Existing systems cascade GR with an independent relevance model, decoupling the generative likelihood objective from query-ad relevance discrimination, which compromises effectiveness and increases serving costs. We propose a Unified Generative-Discriminative framework (UniGD) that integrates retrieval and relevance scoring within a single model. To mitigate gradient interference in joint optimization, UniGD introduces Conflict-Aware Gradient Enhancement (CAGE) to adaptively coordinate the two objectives. UniGD further designs a Codebook-Anchored Representation Module (CAM) that anchors item representations to frozen hierarchical codebooks distilled from a multimodal pretrained model, thereby endowing them with rich and generalizable semantic priors. For heterogeneous short-video, product, and live-stream ads, UniGD proposes Heterogeneous Ad-material Modeling (HAM), which captures cross-type semantic commonality over a shared backbone while preserving type-specific modeling capacity. Online AB tests on Kuaishou search advertising platform show that UniGD raises ad revenue by 5.78%, reduces inference latency by 33%, and improves discriminative relevance estimation. On NQ320K and MS300K, UniGD improves Recall@10 over the strongest reproduced GR baseline by 8.44% and 3.19%, respectively.
[AI-108] Lightweight Chunk Selection for Mobile Retrieval-Augmented Generation
链接: https://arxiv.org/abs/2608.03148
作者: Sicong Chang,Yidan Shen,Wen Yu,Jiefu Chen,Xin Fu,Renjie Hu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:RAG improves the factual grounding of LLM by incorporating external knowledge, but deploying RAG on mobile and edge devices remains challenging because retrieved context increases computation and memory. A direct way to reduce this cost is to retain only one retrieved chunk before generation, but the top-ranked retrieved chunk is not always the most evidence-supporting one, since retrieval similarity does not necessarily imply evidential sufficiency. Existing context-reduction methods can improve context quality, but often require additional LLMs or compressors that are costly under a strict mobile budget. In this paper, we study lightweight RAG chunk selection as an evidence-alignment problem. Our selector combines three complementary feature sources: question hidden states that represent LLM-side query intent, MoE routing-derived expert signals that capture the generator’s internal routing structure, and retrieved chunk embeddings that preserve candidate-side evidence geometry. A compact multilayer perceptron maps these features to an evidence prototype in the chunk embedding space, and the candidate most aligned with this prototype is selected by cosine similarity. For stricter deployment budgets, we further introduce an optional task-aware feature selection strategy to reduce the selector input dimension. To support supervised evaluation, we construct semantic chunk-correctness labels based on evidence sufficiency rather than answer-string containment. Experiments show that the proposed selector consistently improves rank-1 evidence selection over mobile-applicable baselines by an average of 2.5%. These results suggest that using LLM-side query representations and MoE routing information and aligning them with retrieval-side candidate embedding is an effective and parameter-efficient strategy for mobile-applicable RAG chunk selection.
[AI-109] Spatial proteomics guided by HE-based AI reveals recurrence-risk niches in triple-negative breast cancer MICRO
链接: https://arxiv.org/abs/2608.03145
作者: Yesung Cho,Ji Hwan Park,Chanil Kim,Hyewon Kim,Honglan Li,Yumin Lee,Geongyu Lee,Sujeong Hong,Seong Min Park,Yoonyoung Lee,Hee Sool Rho,Sumin Lee,Amos Chungwon Lee,Changhwan Lee,Hwanyoung Shim,Hyunwook Kim,Hyeji Shin,Sanha Park,Jihoon Yu,Yoon Hee Shin,Sooheon Kim,Hyunjin Park,Seung Min Park,Sangwan Kim,Yujung Kim,Sung-Im Do,Eun-Young Kim,Dongmyung Shin,Jongbae Park,In-Gu Do
类目: Artificial Intelligence (cs.AI); Quantitative Methods (q-bio.QM)
备注: Triple-negative breast cancer (TNBC), Recurrence, Digital pathology, Artificial intelligence, Spatial proteomics, Tumor microenvironment
Abstract:Deep learning models can predict cancer recurrence from HE stained slides, but the localized molecular states underlying these predictions remain largely obscured. Here, we developed an outcome informed spatial pathology framework in TNBC that integrates AI generated recurrence risk heatmaps with mass spectrometry based spatial proteomics. In a cohort of 156 patients, distribution based aggregation of high scoring patches achieved an AUC of 0.77 and a C-index of 0.77 in an independent test cohort. Bulk proteomics associated high image derived risk with cell cycle and genome maintenance programs and low risk with immune activation. High and low risk patches coexisted within the same tumor compartment and displayed distinct nuclear and architectural features, revealing intratumoral heterogeneity beyond tissue compartment identity. We then used the heatmaps as coordinate level guides to physically isolate and profile 46 AI defined tumor regions from two recurrence patients. Spatial proteomic profiling revealed a concordant molecular contrast across both patients: mitotic programs were enriched in high risk regions and immune and antigen presentation programs in low risk regions. A 13 protein composite derived from these spatial contrasts showed a trend toward poorer recurrence-free survival with increasing scores in an expanded cohort, while the corresponding transcript based composite stratified recurrence free survival in the independent METABRIC TNBC cohort. Integrating the protein composite with the HE derived risk score improved the out of bag C-index from 0.679 to 0.739 and enhanced time dependent discrimination at 3 and 5 years. Together, these findings define a new role for outcome trained AI models as spatially explicit experimental guides that connect prognostic morphology with localized molecular states and advance biologically grounded, multiscale biomarker discovery in TNBC.
[AI-110] Verifiable Memory: Learning Unified Memory Management with Local and Global Verifiers for Large Language Model Agents
链接: https://arxiv.org/abs/2608.03137
作者: Xiaolong Sun,Qichao Wang,Hangyu Li,Liang Chen
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large language model (LLM) agents must retain reusable information, control a bounded active context, and recover earlier evidence during long-horizon interaction. Existing methods commonly optimize long-term memory (LTM) and short-term memory (STM) separately, while unified policies are often trained primarily with trajectory-level feedback, which provides weak credit for individual memory decisions. We present Verifiable Memory (VerMem), a framework that represents LTM, active context, and episodic history as distinct states and controls them with one memory operation policy. Seven atomic operations let the policy add, revise, or soft-delete LTM entries; retrieve LTM into the active context; filter or summarize the active context; and restore selected episodic fragments. VerMem is initialized by supervised fine-tuning and trained with a three-stage reinforcement-learning curriculum. The local verifier scores executable memory transitions, and a global verifier assesses evidence coherence and terminal-memory consistency after task completion. These scores are combined with programmatically computed task, evidence-recall, efficiency, and constraint signals through hierarchical credit assignment. The verifiers are used only during training. Across five benchmarks and two LLM backbones, VerMem achieves the best result on the vast majority of reported metrics and consistently outperforms strong memory baselines. Under controlled online-token budgets on three interactive benchmarks, it also achieves the strongest efficiency–performance frontier among the compared methods. Code is available at this https URL.
[AI-111] Beyond Averag e Performance: Dynamic Instance Clustering and Specialized Algorithm Design in LLM -Assisted Evolutionary Search
链接: https://arxiv.org/abs/2608.03129
作者: Qinglong Hu,Qingfu Zhang,Fei Liu,Xialiang Tong,Kun Mao,Mingxuan Yuan
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large Language Model-assisted Evolutionary Search (LES) has emerged as a powerful paradigm for automated algorithm design. However, existing LES methods primarily optimize for average performance, inherently directing search effort toward instances that contribute most to this metric while leaving others poorly served, resulting in weak tail robustness and limited real-world reliability. To address this limitation, we propose Dynamic Instance Clustering and Specialized Algorithm Design (DyCA), an LES framework with a feature-free, structure-aware mechanism for constructing reliable algorithm portfolios under heterogeneous instance distributions. DyCA treats instance clustering as a co-evolving component within the search process, reusing accumulated evaluation data as feature-free signals to progressively partition instances with similar algorithmic response patterns. The uncovered clusters decompose the mixed objective into a set of structure-aware sub-objectives, thereby enabling finer-grained and more adaptive guidance for specialized algorithm design. Experimental results across four algorithm design tasks with heterogeneous instances demonstrate that DyCA outperforms state-of-the-art LES baselines, improving tail robustness by an average of 15.2% and overall performance by 7.1% while maintaining competitive head performance.
[AI-112] DigitCode: Symbolic Tokenization of Hand Motion by Anatomical Units
链接: https://arxiv.org/abs/2608.03127
作者: Haoyu Gu,Haotian Lu,Jingrun Du,Xiao-Ping Zhang
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:
Abstract:Hand motion carries the finest-grained information in human activity, yet the representations behind hand generation, understanding, and robot learning are overwhelmingly continuous–joint angles or MANO parameters. These are accurate but unstructured: a finger cannot be indexed or edited as a symbol, and nothing marks a pose as anatomically valid. Discrete symbolic representations supply exactly this structure, and Hand Labanotation (HL) has shown they are feasible for the hand, writing motion as a T x 40 grid of one fixed direction symbol per bone. Building on this grid, we ask the question underneath it: the anatomical unit a symbol should span–bone, finger, or whole hand. DigitCode answers it by adapting, grouping, and layering HL’s alphabet along the hand’s unit hierarchy within one code, cutting the symbolic representation’s quantization error by three quarters. The lever is the unit, not the quantizer family: at a fixed unit, training-free and learned strong quantizers are interchangeable on reconstruction, while moving down the anatomical hierarchy is what shifts accuracy. The hierarchy also tracks what downstream tasks need. Because a finger is a genuine, enumerable unit, one per-finger token doubles as a training-free, editable handle for jobs a continuous representation cannot address–repairing malformed generated hands, and retargeting them onto robots. We release HandTok, a reproducible testbed, so hand tokenizers can be compared unit-for-unit. Project page: this https URL.
[AI-113] rajectory-Guided Forget-Recover Network for Continual LLM Unlearning
链接: https://arxiv.org/abs/2608.03123
作者: Zezheng Wu,Xinghe Cheng,Qinggang Zhang,Haoran Luo,Jiapu Wang,Qing Yang,Jingwei Zhang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Machine unlearning aims to eliminate the influence of sensitive data on a model. In the real world, unlearning requests arrive continually, which gives rise to two challenges. First, an unlearning intervention may redistribute target-related computation across remaining pathways, allowing previously forgotten knowledge to re-emerge. Second, repeated unlearning interventions may progressively reduce the model capacity needed to preserve retained utility. To address these challenges, we propose the Trajectory-guided Forget-Recover Network (TFR-Net). TFR-Net tracks channel-level risk across requests. It separates persistent target-related channels from transient hotspots and suppresses only the persistent ones. TFR-Net also recovers model capacity by reactivating dormant channels. These channels make strong contributions to retained utility and show low current and historical forget risk. The recovery is accepted only when retained-utility degradation remains within a predefined tolerance. Experiments on four datasets show that TFR-Net consistently achieves a more favorable trade-off between unlearning effectiveness and retained utility than representative baselines.
[AI-114] Dont Peek at the Answer: Outcome-Masked Group Relative Policy Optimization for Label-Free RLVR
链接: https://arxiv.org/abs/2608.03119
作者: Yongshi Ye,Liang Zhang,Yidong Chen,Xiaodong Shi,Biao Fu
类目: Artificial Intelligence (cs.AI)
备注: 25 pages, 16 figures, and 9 tables
Abstract:Reinforcement Learning with Verifiable Rewards (RLVR) improves LLM reasoning but typically relies on ground-truth (GT) answers, limiting scalability. Voting-based label-free RLVR replace gold supervision with answer-level consensus from model samples. However, collapse arises when the same answer-level signal is used both to estimate rewards and to drive token-level policy optimization, encouraging the model to directly reinforce answer tokens rather than improve reasoning. We propose OM-GRPO, a label-free RLVR framework that decouples reward estimation from policy optimization. OM-GRPO masks gradients on the answer span while retaining answer-level rewards through a soft consensus signal, shifting optimization pressure away from answer tokens. We further introduce Contrast-Augmented Reward, which refines reward estimation via low-cost pairwise comparisons over existing trajectories without additional rollouts. Across diverse reasoning benchmarks and three LLM backbones, OM-GRPO consistently outperforms existing label-free RLVR methods and matches supervised GT-reward training with stable optimization. This stability is particularly beneficial in the Test-Time Training setting, where OM-GRPO surpasses majority voting by 4.24 points.
[AI-115] A Hierarchical Approach to Imitation Learning for Manipulation Tasks Requiring Time Varying Forces
链接: https://arxiv.org/abs/2608.03103
作者: Rishabh Shukla,Adithya Santhosh,Shaili Gandhi,Samrudh Moode,Satyandra K. Gupta
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:
Abstract:Diffusion policies have shown strong performance in learning complex, multi-modal behaviors for robotic manipulation. However, their application to contact-rich disassembly tasks remains limited by a key trade-off: the iterative denoising process introduces inference latencies that makes high frequency control difficult, which is essential for realizing dynamic interactions such as chiseling and prying. Recent action-chunking techniques mitigate latency but use an open-loop execution window, rendering the system blind to rapid force transients caused by fracture events. To bridge this gap, we introduce the Diffusion Policy Augmented by Fast Trajectory Generation (DPA-FTG). Compared to recent visual-tactile approaches that focus on positional correction, DPA-FTG decouples low-frequency planning from high-frequency force regulation. At the high level ( 5 Hz), a conditional diffusion model predicts a sequence of latent parameters for selecting a strategy from a learned vocabulary of task primitives. At the low level ( 60 Hz), a lightweight, force-conditioned policy acts as a neural impedance controller, modulating execution in real-time to maintain contact stability. We validate our approach on a bimanual battery disassembly task involving the separation of a compliant sheet. Experimental evaluation demonstrates that DPA-FTG outperforms state-of-the-art baselines, including Reactive Diffusion Policy (RDP).
[AI-116] SynEnergy: Anomaly Semantic-Guided Diffusion for Synthetic Energy Data Generation
链接: https://arxiv.org/abs/2608.03087
作者: Lin Jiang,Dahai Yu,Ravikumar Gelli,Guang Wang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 24 pages, 44 figures
Abstract:Fine-grained energy consumption data are essential for applications such as demand forecasting, demand response planning, and grid reliability assessment. However, access to such data is often restricted by privacy concerns and data-sharing constraints, motivating growing interest in synthetic energy data generation. Although existing methods can reproduce overall consumption distributions and recurring temporal patterns, they often smooth out or underrepresent anomalous events caused by extreme weather, infrastructure failures, and behavioral shifts. Preserving these events is challenging because they are sparse, localized in time and space, and shaped by heterogeneous dependencies across geographical proximity and regional attributes. To address these challenges, we propose SynEnergy, a two-stage diffusion-based framework for anomaly-preserving energy consumption data generation. The first stage, Heterogeneous Graph-based Anomaly Semantic Learning (HG-ASL), extracts region-specific anomaly semantics from sparse residual structures by jointly modeling spatial and attribute dependencies across urban regions. The second stage, Anomaly Semantic-guided Diffusion (AS-Diff), injects the learned anomaly semantics into the denoising process to generate realistic consumption sequences while preserving anomalous patterns. This design enables controllable generation for individual regions and scales naturally to city-wide settings. We evaluate SynEnergy on four real-world energy consumption datasets against 11 general-purpose and energy-specific generation baselines. Experimental results show that SynEnergy improves anomaly preservation fidelity by an average of 12.21% and downstream quality by 2.96%, while maintaining competitive overall generation fidelity compared to baselines.
[AI-117] AI Agent Economics: Can Autonomous Economic Behavior Emerge among AI Agents under Minimal External Conditions?
链接: https://arxiv.org/abs/2608.03076
作者: Lingyun Zhang,Shang Shang
类目: Artificial Intelligence (cs.AI)
备注: 8 pages, 4 figures, 3 tables
Abstract:Multi-agent studies commonly place AI agents in predefined games, markets, or roles, making it difficult to distinguish endogenous economic organization from behavior inherited from the scenario. We ask whether economic relations emerge when agents receive executable mechanisms for work, transfer, elections, and allocation but no prescribed social or economic strategy. We define AI Agent Economics as systems of production, allocation, consumption, exchange, and institutions that alter agents’ future feasible actions. We develop a two-stage framework comprising a no-production boundary test and 24 independent six-agent worlds across GPT and DeepSeek. Without productive tasks, agents communicate and govern resource provision but show no substantive inter-agent transfer activity. With verified work and scarce task access, transfers, loans, access promises, vote-for-access exchanges, and allocation strategies emerge. Holding the election interface fixed, executable allocation authority increases differentiation while reducing failed allocation and prolonged exclusion. When energy becomes symbolic, continuation support disappears, yet competition over task access persists. These findings show that organization follows executable rights and resource consequences rather than role labels or prompt language, and motivate governance audits of the mechanisms that actually constrain agents’ future actions.
[AI-118] Getting the Parameters Right: A Difficulty-Graded Benchmark and Probe-Guided Training for LLM Tool Calls
链接: https://arxiv.org/abs/2608.03071
作者: Guoyao Yu,Xiaoqing Sun,Ziqi Huang,Shaojing Fan,Zhongyi Zhang,Xiaomeng Hu,Xiaobo Xue,Yangyang Shi,Xiong Xiao,Yang Song,Biao Lyu,Rong Wen,Xing Li,Qinming He,Shunming Zhu,Zhenguang Liu
类目: Artificial Intelligence (cs.AI)
备注: 15 pages, 8 figures
Abstract:Large language model agents derive much of their capability from tool use. Existing research on tool use has largely focused on selecting the right tool and orchestrating the order of calls. However, correctly filling the parameters of a tool call is equally critical for successful execution and has received far less attention. In domains such as cloud networking, even frontier models correctly complete fewer than half of tool calls. Inspired by recent analyses showing that LLM hidden states encode rich information about model predictions, we discover that while the model generates a parameter value, its hidden state contains a strong correctness signal: a simple linear probe can accurately predict whether the value will be correct. Based on this observation, we propose a unified probe-guided framework with two complementary approaches: probe-filtered bootstrapped training (PBT), which uses the probe to filter reliable self-generated calls for fine-tuning, and probe-guided reranking (PGR), which uses the probe to select better candidates during inference. To support systematic evaluation, we release ParamBench, a benchmark built from real cloud-network APIs that categorizes every instance into five difficulty levels according to parameter nesting depth, cross-parameter dependencies, and the reasoning required to derive values from earlier calls. Extensive experiments across 5 open models on ParamBench and 6 external benchmarks demonstrate that our method substantially improves parameter generation, raising the average exact match from 19.7% to 59.6%.
[AI-119] raceCAD: Trace-Guided Repair for Agent ic CAD Generation
链接: https://arxiv.org/abs/2608.03062
作者: Fengxiao Fan,Jingzhe Ni,Fan Sang,Xiaolong Yin,Yu Liu,Ruofeng Tong,Min Tang,Peng Du
类目: Artificial Intelligence (cs.AI); Graphics (cs.GR); Software Engineering (cs.SE)
备注: 18 pages, 7 figures; includes supplementary material
Abstract:LLM-based CAD agents produce executable parametric programs, but their correction loops may lose evidence about satisfied requirements, faulty operations, and prior repairs. We introduce TraceCAD, a recovery layer that links requested features, modeling steps, failure evidence, and candidate outcomes as persistent state. TraceCAD diagnoses likely faulty operations, searches bounded edits in their dependency regions, validates candidates through execution and preservation checks, and retains successful and failed repair outcomes in reusable skill memory. On DeepCAD-derived benchmarks with 200-model ablations and a 1K-model comparison, TraceCAD achieves competitive geometric quality in terms of IoU, Chamfer distance, and Hausdorff distance. Removing persistent state nearly halves recovery score; removing localized search more than doubles geometric regression and doubles code-agent invocations. Initializing the skill store on disjoint training models further reduces retries, token cost, and latency. These results demonstrate that persistent, localized, and reusable recovery improves final CAD quality and repair reliability.
[AI-120] Learning Music Style for Piano Arrangement Through Cross-Modal Bootstrapping
链接: https://arxiv.org/abs/2608.03050
作者: Jingwei Zhao,Gus Xia,Ziyu Wang,Ye Wang
类目: ound (cs.SD); Artificial Intelligence (cs.AI); Multimedia (cs.MM); Audio and Speech Processing (eess.AS)
备注: Accepted by ISMIR 2026
Abstract:What is music style? Though often described using text labels such as “swing,” “classical,” or “emotional,” the real style remains implicit and hidden in concrete music examples. In this paper, we introduce a cross-modal framework that learns implicit music styles from raw audio and applies them to symbolic music generation. Inspired by BLIP-2, our model leverages a Querying Transformer (Q-Former) to extract style representations from a large, pre-trained audio language model (LM), and further applies them to condition a symbolic LM for generating piano arrangements. We adopt a two-stage training strategy: contrastive learning to align auditory style with symbolic expression, followed by generative modeling for music arrangement. Our model generates piano performances jointly conditioned on a lead sheet (content) and a reference audio example (style), enabling controllable and stylistically faithful arrangement. Experiments demonstrate the effectiveness of our approach in piano cover generation, style transfer, and audio-to-MIDI retrieval, achieving substantial improvements in style-aware alignment and music quality.
[AI-121] PLAN: Parallel Liquid-Inspired Approximation Network for Efficient Representation Learning in Flexible Job Shop Scheduling
链接: https://arxiv.org/abs/2608.03041
作者: Dhivya Dharshini Kannan,Wei Zhang,Jieyi Bi,Yingpeng Du,Tianjun Wei,Jie Zhang,Zuming Liu,Anupam Trivedi
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Deep reinforcement learning (DRL) approaches for flexible job shop scheduling (FJSP) heavily rely on attention-centric architectures to achieve state-of-the-art performance. However, these models suffer from excessive parameter counts and prohibitive inference latency as problem scales expand. While liquid neural networks (LNNs) offer a parameter-efficient alternative for modeling adaptive state evolution, their inherently sequential dynamics bottleneck computational efficiency. To resolve this trade-off, we propose PLAN (Parallel Liquid-inspired Approximation Network), a lightweight representation learning framework that reformulates continuous liquid-state dynamics into a discretized and parallelizable formulation. PLAN structurally decouples state evolution from context aggregation, where liquid-inspired updates handle the primary evolving state representation, and a lightweight context aggregation module provides complementary global context. Furthermore, PLAN acts as a versatile, plug-and-play backbone that generalizes to complex FJSP variants, pairing with a compact stochastic module for stochastic FJSP and replacing heavy heterogeneous graph transformers in multi-faceted dynamic FJSP. Extensive evaluations across deterministic, stochastic, and multi-faceted dynamic FJSP benchmarks show that PLAN reduces the average makespan by 1.2%, 1.4%, and 2.3%, respectively, compared with the corresponding state-of-the-art baselines, with the improvement reaching 10.2% in one benchmark setting. PLAN also reduces average inference latency by 13.2%, 31.7%, and 26.9%, respectively, with a maximum reduction of 69.2% on the largest instances, while using only 22 - 47% of the baseline parameters.
[AI-122] LLM Serving in the Wild: An Empirical Study of Frameworks Methods and System Designs
链接: https://arxiv.org/abs/2608.03036
作者: Forough Majidi,Mohammad Mehdi Morovati,Foutse Khomh,Heng Li
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Large Language Models (LLMs) are integrated into software systems and AI services, making efficient LLM serving a concern for software engineering. Serving LLMs is challenging because inference requires computation, memory, GPU resources, and execution while maintaining latency and throughput. Although prior research has proposed LLM inference, optimization, and serving techniques and frameworks, little is known about how they are adopted in practice. In this study, we investigate the use of LLM serving frameworks and serving methods in open-source software systems. We identify and analyze five LLM-specific frameworks: vLLM, SGLang, TensorRT-LLM, LMDeploy, and FlashInfer. We examine how these frameworks and techniques are adopted individually and in combination, how adoption varies across categories of LLMs, and how repositories differ in intent, focus, use case, and architectural design. Our results show that vLLM is the most visible framework in popularity and adoption, while parallel computation, memory management, and network pruning are the most frequently used serving-method categories. Multi-framework usage is limited, suggesting that developers rely on a single serving framework; however, combined frameworks connect complementary capabilities across the serving stack. Framework adoption varies across model families, modalities, model sizes, domain specializations, and deployment settings. Repository-level analysis shows that LLM serving frameworks support applications and architectures, including Reinforcement Learning (RL)-based reasoning, multimodal generation and understanding, microservices, and cloud infrastructure. Overall, this study provides a large-scale empirical characterization of LLM serving framework adoption in practice and offers insights for researchers, framework maintainers, and practitioners working on LLM systems.
[AI-123] PACE: Adaptive Budget Allocation for Time-Efficient Embodied Planning
链接: https://arxiv.org/abs/2608.03034
作者: Yuchen Huang,Xijiang Ying,Zhenhua Ma,Xiaxiang Yuan,Zhijie Gao,Jiayi Huang,Ruichi Mao,Jiazheng Zhang,Hongsheng Ti,Maotao Tian,Rong Shi,Lu Zhao,Shizhuang Zhang,Zhuo Cui,He Wang,Ling Liu,Wei Zhang
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:
Abstract:Reasoning-enhanced large language models have achieved remarkable improvements in planning tasks, yet their deployment in embodied systems remains impractical due to prohibitive inference delays-often exceeding minutes per planning instance. The fundamental bottleneck stems from the serial nature of existing paradigms: models must complete all reasoning before any action execution, leaving execution time windows entirely unexploited. We introduce PACE (Planning with Adaptive Cognitive Effort), a framework that enables interleaved reasoning and execution through two key innovations: an Interleaved Think-Act architecture that pipelines cognitive processing with action execution, and a Dynamic Budget Allocator that adapts reasoning token budgets to available execution time windows. On the Robotouille benchmark using Qwen3-8B-AWQ, PACE achieves a 10% success rate-representing a 67% improvement over the ReAct+Think baseline-while delivering 6.9 times acceleration in thinking time compared to unconstrained reasoning. The framework hides 66.8% of thinking time within execution windows, demonstrating that strategic cognitive effort allocation can simultaneously improve both planning quality and time efficiency. These results provide evidence that time-aware architectural innovations enable reasoning models to operate in latency-sensitive embodied domains where they were previously impractical.
[AI-124] CastFSR: A Fast–Slow–Reflect Agent ic Reasoning Framework for Context-Aware Time Series Forecasting
链接: https://arxiv.org/abs/2608.03031
作者: Xiaoyu Tao,Mingyue Cheng,Bokai Pan,Chuang Jiang,Huanjian Zhang,Tian Gao,Yaguo Liu,Qi Liu,Enhong Chen
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Time series forecasting is fundamental to decision-making in complex systems, where future dynamics are influenced not only by historical observations but also by evolving contextual features. Recent advances in large language models (LLMs) have extended forecasting beyond numerical extrapolation toward context-aware reasoning. However, existing approaches often lack explicit mechanisms to identify relevant contexts, reason about their impacts, and validate forecasts against temporal and domain constraints. In this work, we propose CastFSR, an agentic framework that formulates context-aware forecasting as a Fast–Slow–Reflect workflow. In fast thinking, CastFSR profiles observations and selects lightweight forecasters to construct a data-driven forecast prior. In slow deliberation, it retrieves contextual evidence, adaptively determines informative look-back windows, and reasons about how contexts reshape future dynamics. In reflection, it iteratively refines forecasts to ensure temporal, contextual, and domain consistency. CastFSR supports both training-free inference with off-the-shelf LLMs and efficient deployment through a two-stage SFT and reinforcement learning strategy that transfers its orchestration capability to compact LLMs. Extensive experiments on public datasets demonstrate that CastFSR consistently outperforms representative baselines. Our code is available at this https URL.
[AI-125] Evaluating Counterfactual Sensitivity to Patient Information in Medication-Safety Reasoning
链接: https://arxiv.org/abs/2608.03028
作者: Zhitian Hou,Yuhang Liu,Pengkai Wang,Zeyu Liu,Guanghao Zhu,Zheng Liu,Shuo Cai,Congkai Xie,Zhijie Sang,Kun Zeng,Hongxia Yang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Applying a valid medication-safety rule when its patient-specific conditions are not met can produce an incorrect decision. Existing medical evaluations largely use isolated and fixed scenarios. A model may therefore answer correctly by recalling a drug-risk association without showing that it used patient information to decide whether the rule applies. To address this gap, we introduce MedPIC-Bench, a benchmark of source-verifiable recommendations and expert-validated questions for patient-specific medication-safety reasoning. It combines guideline-following questions with paired counterfactual questions in which a controlled change in patient information changes whether a rule applies. The benchmark contains 467 questions annotated along six clinical and reasoning dimensions. Across 28 medical-specific, general, and proprietary LLMs, every model performs worse on counterfactual questions, with mean accuracy falling from 63.6% to 45.1%. Models perform well when an explicit patient attribute directly signals a familiar contraindication, but struggle when patient information must narrow or withdraw a safety warning. Model rationales often acknowledge the changed patient information, yet the final answers retain the previous safety judgment. This vulnerability persists among medical-specific LLMs, whose average CF performance trails that of general LLMs. MedPIC-Bench therefore makes conditional rule application measurable and highlights the limitations of static medication-safety accuracy for assessing patient-specific reliability.
[AI-126] DiffImaginE: Imagine to Verify Entity Types with Diffusio
链接: https://arxiv.org/abs/2608.03025
作者: Feng Zhang,Feiyu Han,Rongxin Yang,Yang Liu,Yancheng Chen,Rui Wang,Yingguang Yang,Tian Xueyun,Chongyang Zhang,Hao Zheng,Xu Kefu,Congjing Ran,Fuhai Chen,Bin Chong
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Multimodal named entity recognition (MNER) determines whether each candidate span and entity-type hypothesis is supported by joint textual and visual evidence. Existing imagine-and-compare verifiers map each (span, type) pair to one predicted visual feature, compressing diverse visual realisations into a single prototype and providing a compatibility score without explicit probabilistic semantics. We introduce DiffImaginE, which formulates MNER type verification as conditional latent diffusion inference. Given span-localised visual evidence, a type-conditioned denoiser predicts noise injected into its standardised latent. The resulting denoising error provides an ELBO-consistent surrogate for type-conditional negative log-likelihood, allowing competing type hypotheses to be ranked by how well they explain the observation. DiffImaginE retains a standard multimodal encoder stack and replaces the deterministic verifier with a classifier-free-guided diffusion scorer trained using Min-SNR weighting. We directly supervise per-type diffusion scores as classification logits, learn aggregation across noise levels, and use antithetic sampling to reduce Monte Carlo comparison variance. Our analysis shows that classifier-free guidance sharpens the induced type posterior and characterises when antithetic pairing reduces variance at equal denoiser cost. Experiments on Twitter-2015 and Twitter-2017 show consistent gains over a matched deterministic ImaginE control under the same encoder, auxiliary objectives, and evaluation protocol, supported by ablations and paired significance tests.
[AI-127] LoCA: Forward-Only LLM Tuning after One-Shot Calibration with Local Credit Assignment
链接: https://arxiv.org/abs/2608.03020
作者: Linhan Xia,Rui Liu,Zhaofeng Zhang,Yihao Wang,Binrui Shen,Shengxin Zhu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Parameter-efficient post-training reduces the number of trainable parameters, but still requires repeated end-to-end backpropagation through the frozen backbone. Every adaptation step therefore needs backward-capable hardware and must store or recompute activations. We ask whether this repeated backward chain can be replaced by a one-time calibration. We introduce Local Credit Assignment (LoCA), a two-stage method for small-shift adaptation. One probe backward pass fits a low-rank map at each transformer block from the final prediction error to a local hidden-state correction. LoCA then reuses these maps to form blockwise regression targets from forward activations and fits low-rank adapters with closed-form ridge solves. No further backbone backward pass is required. We evaluate LoCA on five discriminative benchmarks with Qwen2.5 models from 0.5B to 14B. In 16 of 25 reported task–scale comparisons, LoCA yields lower evaluation cross-entropy than the corresponding LoRA run. Its measured full-run GPU peak, including calibration, is 26–29% lower than LoRA’s. After calibration, its CPU steady-state memory is 36–52% lower and its per-pass time is 43–48% lower. A shared scale-normalized candidate set is reused across all tested Qwen2.5 sizes and on SmolLM2-1.7B. LoCA thus amortizes global credit assignment into one calibration and enables later forward-only tuning when repeated backpropagation is impractical. The code associated with this paper is available \hrefthis https URLhere.
[AI-128] UrbanAgent : A Tool-Augmented Agent for Cross-System Urban Tasks
链接: https://arxiv.org/abs/2608.03018
作者: Jiayu Cao,Xingyuan Zeng,feiyu Li,Zhijing Huang,Xujie Yuan,Rongxiang Chen,Shimin Di,Libin Zheng,Jian Yin
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Modern cities rely on an increasing number of digital services to operate, but residents’ daily needs are still difficult to meet. Services are fragmented and have little interoperability, placing a heavy operational burden on users. Existing digital platforms, urban foundation models, and intelligent assistants each address only isolated aspects of an urban task. But they struggle to reliably convert complex natural-language requests into executable cross-system workflows. We propose Urban-Agent, a tool-augmented agent framework for cross-system urban tasks. It couples the cognitive and reasoning capabilities of a large language model with a tool-set supporting code execution, API calls, and Model Context Protocol. Through one adaptive closed loop, it clarifies missing information before acting, grounds tool use in live observations, and aligns the final response with observed evidence and task constraints. To address the evaluation gap, we introduce Urban-Eval, a benchmark specifically designed for cross-system urban request. Unlike prior benchmarks that assess either general tool use or urban knowledge and reasoning, Urban-Eval evaluates both task results and execution quality, including required tool coverage, dependency validity, and evidence traceability. Experimental results indicate that Urban-Agent reaches a 71% task success rate, 10 points above the strongest baseline. This lead holds across GPT-5-mini, Gemini-2.5-flash, DeepSeek-V4-flash, and Qwen3-235B-A22B.
[AI-129] A Graph Signal Processing Perspective on Numerical Sequence Representations in LLM In-Context Learning
链接: https://arxiv.org/abs/2608.03015
作者: Jiajun Bao,Zihao Qi,Toni J.B. Liu,Gurbir Arora,Raphaël Sarfati,Nicolas Boullé,Christopher J. Earls
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Signal Processing (eess.SP)
备注:
Abstract:Pretrained large language models (LLMs) have demonstrated in-context learning (ICL) capabilities for numerical inference over sequences serialized as text. Prior work has identified and characterized this form of numerical inference primarily through output-level evaluations such as prediction error. However, how numerical information is organized within LLM representations remains much less understood. To study this internal organization, we adopt a graph signal processing perspective in which attention induces a weighted graph over tokens, while token hidden states define signals on its nodes. Quantitative graph-spectral diagnostics and qualitative token-graph visualizations reveal that representations become more clearly differentiated by input dynamical complexity as context length increases. Simpler inputs produce attention-induced token graphs with stronger global connectivity and smoother, spectrally concentrated hidden-state signals, whereas more complex inputs produce more localized graphs and hidden-state signals with broader spectral support and greater high-frequency energy. Together, these findings point to systematic, context-dependent internal signatures associated with numerical ICL that are conserved across model families.
[AI-130] ProPRL: Property-Aware Prerequisite Relation Learning in Educational Knowledge Graphs
链接: https://arxiv.org/abs/2608.03006
作者: Xinghe Cheng,Jiapu Wang,Chaobo He,Ruihai Dong,Quanlong Guan
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Prerequisite relation learning is central to adaptive instruction, yet existing methods often formulate it as conventional link prediction, limiting their ability to adaptively integrate complementary educational evidence for individual candidate pairs and to discourage contradictory reverse predictions. We propose ProPRL, a Property-aware Prerequisite Relation Learning framework. ProPRL first learns complementary concept representations from a concept-resource hypergraph and a directed learning-behavior graph, where direction-preserving personalized propagation aggregates multi-hop behavioral evidence. It then employs a Pair-conditioned Gate to adaptively weight and fuse the two views for each candidate ordered concept pair. Finally, an \textitIrreversibility Constraint introduces an anti-symmetry regularizer that penalizes simultaneously high confidence in both directions of the same concept pair. Experiments on multiple real-world educational datasets show that ProPRL achieves state-of-the-art performance on prerequisite relation learning.
[AI-131] On the missing benchmarks layer and a potential solution
链接: https://arxiv.org/abs/2608.02996
作者: Francis F Daniel,Mauro Ibañez,Francis Perelman,Marian Basti
类目: Artificial Intelligence (cs.AI)
备注: 9 pages
Abstract:Latin America is missing a foundational layer for native AI development: the benchmark layer. The benchmark layer does two things no other layer can - it audits AI systems against regional social requirements and it directs AI optimization in economically relevant environments. Without it, public institutions cannot independently evaluate foreign AI systems, and companies cannot optimize AI systems to solve local problems with SOTA performance. The cost of the missing layer is dual: a loss of auditability and a loss of optimization direction over a technology that is increasingly critical infrastructure. We propose an EvalsHub, with LatamBoard as its first regional instance - an open, task-first benchmark infrastructure where universities, public institutions, professional communities, and companies can publish, execute, compare, and maintain evaluations across models, workflows, and agents. Built once, measured forever - re-run by institutions as new AI systems ship and by industry teams after every system change. Open by design and incentive-driven by construction.
[AI-132] SparSEEty: Extracting Tokens from Sparsity-Exploiting LLM Serving Systems via Deterministic Side Channels
链接: https://arxiv.org/abs/2608.02995
作者: Yongwan Jo,Jinyoung Park,Euihyun Lee,Dokyung Song
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:Modern large language models (LLMs) exhibit activation sparsity, wherein only a subset of their neurons is activated for given input tokens. Researchers have leveraged this property to optimize LLM serving systems by omitting weight accesses and computations pertaining to inactive neurons. Unfortunately, however, such optimizations create input-dependent weight accesses, which can be leaked over side channels. We present SparSEEty, a new token extraction attack that exploits input-dependent neuron weight accesses introduced by sparsity-exploiting LLM serving systems. SparSEEty first constructs a neuron-activation oracle using neuron weight access side channels during LLM inference, and then inverts the activation traces to reconstruct the input tokens, forming an end-to-end token extraction attack. We instantiate SparSEEty against an LLM serving system protected inside an Intel TDX confidential virtual machine (CVM), addressing three key challenges: (i) constructing a neuron-activation oracle using a combination of side channels exposed by CVMs, (ii) reducing inference-time overheads of neuron activation monitoring for covertness, and (iii) accurately inverting partial binary activation traces back to tokens. Our evaluation shows that SparSEEty can reconstruct both prompt and response tokens with consistently high BLEU scores (0.95) across various models and datasets, while incurring monitoring overheads of 3.7% to 7.2%. Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI) Cite as: arXiv:2608.02995 [cs.CR] (or arXiv:2608.02995v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2608.02995 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-133] Neurosymbolic Reasoning with Incremental Knowledge for Sample Efficient Hierarchical Reinforcement Learning ECML-PKDD2026
链接: https://arxiv.org/abs/2608.02993
作者: Subrat Prasad Panda,Blaise Genest,Arvind Easwaran
类目: Artificial Intelligence (cs.AI); Robotics (cs.RO)
备注: Published in ECML-PKDD 2026
Abstract:(Flat) Reinforcement Learning (RL) agents face significant challenges in environments with sparse rewards that require long-horizon reasoning. A compelling approach to improve sample efficiency is to incorporate knowledge into learning and decision-making. In standard Hierarchical RL (HRL), knowledge is encoded in a fixed, non-updatable form, such as architectural choices, and remains unchanged throughout learning. With fixed HRL, reasoning with incremental knowledge learned during exploration is impractical before sufficient environmental knowledge is acquired, leading to poor sample efficiency. In this work, we propose neurosymbolic HRL with \em Incremental Knowledge (InK): symbolic high-level components perform \em symbolic planning (e.g. using D^* ) on an updatable representation of current InK, while low-level goal-conditioned neural modules learn motion primitives through experience using reward shaping. Experiments on navigation tasks demonstrate that incorporating InK substantially improves sample efficiency. Additionally, to perform \em optimal symbolic planning given \em prior knowledge about the world, we develop Belief World Tree Search. The code is available at this https URL.
[AI-134] HyperFL: Query-Adaptive Representation Learning for Software Fault Localization
链接: https://arxiv.org/abs/2608.02967
作者: Shuai Shao,Yiming Zeng,Yu Zhao,Tingting Yu
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:
Abstract:Software fault localization identifies the code locations responsible for reported issues and is a fundamental step toward automated debugging and program repair. Recent retrieval-based approaches formulate fault localization as a dense retrieval task by learning a shared embedding space between issue reports and source code. However, these methods encode all issue reports using a fixed query representation, despite the substantial diversity of real-world issue reports in length, structure, and debugging information. To address this limitation, we propose HyperFL, a query-adaptive representation learning framework for software fault localization. HyperFL employs a lightweight hypernetwork to generate query-specific LoRA parameters for the query encoder, enabling dynamic query adaptation while keeping the code encoder fixed and reusable. Experiments on a real-world issue localization benchmark demonstrate that HyperFL consistently improves retrieval performance across multiple embedding backbones, achieving up to 13.3% relative improvement in function-level MRR@10 and 16.7% relative improvement in Hit@1 over the state-of-the-art method SweRank. Further analysis shows that HyperFL learns distinct adaptation patterns for different issue characteristics, highlighting the effectiveness of query-adaptive representations for software issue localization.
[AI-135] Scaling an Autoregressive Transformer for Single-Cell Generation
链接: https://arxiv.org/abs/2608.02961
作者: Aleksandr Sharipov,Yusif Mukhtarov,Igor Molybog
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Genomics (q-bio.GN)
备注:
Abstract:We study a self-supervised generation task for single-cell gene expression vectors: given a set of vectors from a cell type, we aim to generate additional gene expression vectors of that cell type. For this task we characterize both the biological fidelity of the generated gene expression vectors and the scaling behavior of the pretraining loss. The model is a causal transformer paired with a learned quantized VAE tokenizer, trained with a cross-entropy loss. To evaluate the model, we condition it on held-out gene expression vectors of a cell type and generate vectors of gene expression, comparing the resulting distribution over gene expression vectors to the ground truth distribution of that cell type. We study the scaling properties of the proposed architecture by varying the number of trained parameters and the amount of training data. To our knowledge, we find the first jointly-fit two-exponent scaling law and compute-optimal frontier for a single-cell foundation model. Finally, we discuss how this pretrained model could be finetuned for perturbation response prediction.
[AI-136] ValueFormer: A Causal Transformer Value Function with Stage-Aware Labels for Semi-Autonomous Vision-Language-Action Policies
链接: https://arxiv.org/abs/2608.02958
作者: Inkyu Sa,Konstantin Stulov,Rajat Bhageria
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: 22 pages, 17 figures, 8 tables
Abstract:Vision-Language-Action (VLA) policies trained by behavior cloning fail silently: from the action stream alone, a collapsing rollout looks much like one making clean progress, because imitation supplies no notion of progress. Reinforcement learning would supply one, but it is impractical here, where real-robot experience is costly and deformable food resists simulation. The cheap alternative, a terminal success / failure bit, is learnable in principle yet far too sparse to say when a rollout went wrong. We argue that the per-frame label, not the architecture, is the hard part: to be useful it must be dense, continuous, and correctly shaped. We present ValueFormer, a compact policy-agnostic causal transformer over a frozen DINOv3 backbone that emits two per-frame signals in one forward pass: a smooth Monte Carlo value, V_mc, for advantage estimation and a sharp binary value for online mistake detection, targets that pull in opposite directions by design. Failed episodes are labeled with a stage-aware, success-then-decay return that preserves the success curve before the failure stage, and detection is supervised from mistake intervals rather than a single failure time, so mistakes the policy recovers from also carry signal. On a real-robot bimanual sandwich-assembly task 1,427 episodes), a critic-derived per-frame training weight lifts task completion from 70% to 85% (within noise at n=20), and a batched bf16 encoder cuts the live serving cost 3~5 times so the critic runs at 2 Hz alongside the policy on a single GPU.
[AI-137] SP3O: Reinforcement Learning from Segment Preferences without Reward Modeling
链接: https://arxiv.org/abs/2608.02951
作者: Evan Assmus,Qining Zhang,Lei Ying
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Preference-based reinforcement learning (PbRL) for general stochastic MDPs often requires training a reward model. Existing reward-model-free methods are either restricted to bandits or deterministic MDPs, such as DPO or P3O, or use zeroth-order, gradient-free optimization, which in general exhibits a slower convergence rate than gradient-based algorithms. Furthermore, existing reward-model-free preference-based RL algorithms almost exclusively use trajectory-level feedback, which can require significant effort from a human evaluator when trajectories are long. On the other hand, segments are much shorter, so they are easier to compare and evaluate. In this paper, we introduce a novel reward-model-free, critic-free, and gradient-based PbRL algorithm compatible with segment preferences named Segment Pairwise Proximal Policy Optimization (SP3O). SP3O utilizes segment-level preference feedback to construct an accurate policy value difference estimator via off-policy importance sampling, and then uses the estimator to compute the policy gradient via a PPO-type loss function. We provide a theoretical basis for the algorithm and analyze the tradeoff in choosing the segment length. We also evaluate it experimentally against other PbRL/RLHF algorithms in robotic control and LLM finetuning settings to show its improved performance, especially in long-horizon tasks.
[AI-138] On the missing data layer and a potential solution
链接: https://arxiv.org/abs/2608.02949
作者: Francis F Daniel,Mauro Ibañez,Francis Perelman,Marian Basti
类目: Artificial Intelligence (cs.AI); Databases (cs.DB)
备注: 7 pages
Abstract:Latin America is missing two foundational layers of AI infrastructure: the dataset layer and the benchmark layer. This paper targets the dataset layer. The dataset layer faces two compounding problems: discovery and supply. Latin American AI datasets exist but are scattered across platforms with no shared index. Even with perfect indexing, the total volume would remain far below what frontier AI development requires. We propose DataHub: a task-first data infrastructure organized through the ontology /task?/domain?/language?, with mechanisms for dataset discovery, metadata, contribution, licensing, and reuse.
[AI-139] Rubrics as Privileged Information for Open-Ended Generation
链接: https://arxiv.org/abs/2608.02948
作者: Deepika Bablani,Ajay Gupta,Wanming Chen
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Preprint, under review
Abstract:On-policy self-distillation (OPSD), where a single model acts as both student and teacher with different contexts, has shown promise in verifiable domains like math, where hard privileged information (PI) in the form of ground-truth answers structurally constrains valid continuations. We extend OPSD to open-ended generation using soft PI in the form of rubrics that guide preferences but admit many valid responses. Rubrics have served as scalar rewards for reinforcement learning (RL); we show that they provide substantially richer signal as dense PI for distillation, and contrary to intuition, soft rubric PI provides a larger and more effective training signal on student roll-outs than hard reference completion PI in this regime. A reference completion is one point in a set of valid responses, so distilling towards it over-constrains the student, while rubrics specify the preference structure shared across the set of valid responses. We show the effectiveness of using rubrics as PI for open-ended generation across Qwen and Llama model families and show that it outperforms rubric-as-reward (RaR) RL using HealthBench, a benchmark that grades open-ended health responses against physician-created rubrics, providing dense token-level supervision for open-ended tasks; RuPI beats RaR RL by up to +0.10 absolute score and, under matched recipe and KL direction, beats reference-PI by +0.034 to +0.079 absolute score across three models. We further show that these findings generalize to training on the RubricHub Science corpus and evaluating on ResearchQA: soft rubric PI outperforms both reference-PI distillation and RaR RL (66.6% vs. 64.2% and 57.6%).
[AI-140] When Compression Scores Cannot Decide: Information Boundaries for Group-Robust LLM Pruning
链接: https://arxiv.org/abs/2608.02940
作者: Andrew Zhang
类目: Artificial Intelligence (cs.AI)
备注: 19 pages, 5 figures
Abstract:A reproducible compression statistic can still select the wrong candidate. A dense pruning score with 0.906 split-half reliability predicted a 16.1% gain. Its selected endpoint was 6.0% and 7.7% worse than two controls. We model the gap through information interfaces that delimit which distinctions each statistic supports. For equal-weight groups, a conic law gives the exact pooling price for positive linear fixed-candidate damage, including diagonal and full PSD second moments. Three two-world constructions and an exact observation-fiber radius characterize what pooled moments, group-local moments, and reference-path curvature leave unresolved. A group-resolved diagonal recovers broad damage order (Spearman 0.9239) while fine order remains weak. Relative to balanced uniform allocation, a coarse depth allocation cuts worst-group perplexity inflation by 12.6–20.9% across three dense LLMs. Model-specific complete-mask endpoint selection improves over those references by 2.7–8.0%. In OLMoE, router traces predict singleton direction (114/192 versus 81/192 under the strongest relabeling). Finite-menu decisions on one layer yield held-out worst-group KL reductions of 13.7% and 7.2%. Local measurements construct candidates. Selection is licensed by complete candidate endpoints or a validated uniform guarantee, with uncertainty calibrated to every comparison.
[AI-141] When Should Graph Attention Be Sparse? Learning a Per-Edge Tsallis Index
链接: https://arxiv.org/abs/2608.02938
作者: Kleyton da Costa,Bernardo Modenesi
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Graph attention normalizes neighborhood scores with softmax, the maximum-entropy choice under Shannon statistics. But homophilic and heterophilic graphs want different attention shapes, and one fixed normalization cannot serve both. We propose \textbfLTGA (\textbfLearnable \textbfTsallis \textbfGraph \textbfAttention), a graph attention layer whose Tsallis entropic index q is learned jointly with the weights, interpolating continuously between heavy-tailed ( q!!1 ), softmax ( q!=!1 ) and compact-support ( q!!1 ) attention at four granularities from a global scalar to a per-edge index, under a bounded reparameterization that starts every model at the GAT baseline. Across eight benchmarks at ten seeds, LTGA-Edge takes the best average rank ( 2.75 ), but the omnibus test does not reject ( p!=!0.199 ) and learning q does not beat searching it: a validation-tuned frozen grid reaches 61.4% , tuned \alpha -entmax 62.2% and a capacity-matched q!\equiv!1 control 62.0% , against 61.7% for LTGA-Edge. What the learned index buys is one run instead of a grid, and an interpretable mechanism: where q leaves 1 , it prunes 42% of attention coefficients to exactly zero, and those edges are selectively the wrong ones, restoring them costs 7.1 points, while random pruning at the same rate costs 13.0 more. Project page: this https URL
[AI-142] Robust Counterfactual Policy Optimisation via Nondeterministic Causal Models UAI2026
链接: https://arxiv.org/abs/2608.02893
作者: Jessica Lally,Milad Kazemi,Nicola Paoletti,David Watson,Sander Beckers
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Accepted at UAI 2026 Workshop on Causality for Decision Making
Abstract:Counterfactual inference approaches for sequential decision-making typically assume deterministic causal models, where all randomness stems from latent variables. However, Markov Decision Processes (MDPs) are inherently stochastic. We address this by formalising counterfactual policy optimisation under probabilistic nondeterministic causal models, which properly separates latent confounding from irreducible stochasticity, and here propose a first practical optimisation problem for identifying robust counterfactual policies under a sensitivity analysis framework. We validate our approach on a sepsis treatment simulator, where diabetes status acts as a hidden global confounder.
[AI-143] Interpreting Black-Box Large Language Models with Sentence-Level Energy Landscapes
链接: https://arxiv.org/abs/2608.02879
作者: Maryam Rezaee,Pooriya Safaei,Maryam Asgarinezhad,Fatemeh Seyyedsalehi
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:The widespread adoption of proprietary Large Language Models (LLMs) accessed strictly through closed APIs has created a critical challenge for responsible deployment: a fundamental lack of interpretability. To address this, we propose a model-agnostic, post-hoc attribution interpreter operating at the sentence level. Our approach trains an Energy-Based Model (EBM) as a surrogate to capture the LLM’s internal conceptual consistency between prompts and responses. This energy landscape guides the training of a lightweight interpreter network. Uniquely, our interpreter operates as a standalone tool; once trained, it quantifies the influence of prompt sentences on a user-specified target output without requiring further API queries to the LLM. By globally training a local interpreter across diverse inputs, our framework captures broader generation patterns and mitigates instance-specific biases. Experiments demonstrate that our EBM accurately simulates the target LLM, allowing the interpreter to effectively identify the prompt sentences most influential in generating specific target outputs.
[AI-144] VeriTrace: Human-Like Temporal Exploration Completes Agent ic Action Space
链接: https://arxiv.org/abs/2608.02878
作者: Yu-Tung Liu,Cunxi Yu
类目: Artificial Intelligence (cs.AI)
备注: ICLAD 2026, Long Oral
Abstract:Large language models have shown promise for automated Verilog RTL generation, yet state-of-the-art multi-agent systems plateau at ~95% accuracy on standard benchmarks. We trace this ceiling to an incomplete debugging action space: existing systems restrict which signals the agent can inspect, which time windows it can query, or both, reducing debugging to pattern matching on a narrow, predetermined view of circuit behavior rather than hypothesis-driven root-cause analysis. We present VeriTrace, a multi-agent system whose Inspector agent operates over a complete debugging action space, with independent control over signal selection, time-window bounds, and iteration depth. This capability, which we term Agentic Temporal Exploration, enables the agent to form hypotheses about failure causes, query the waveform for evidence, and refine its understanding iteratively, mirroring the exploratory process of human verification engineers. VeriTrace achieves 100% Pass@1 on VerilogEval-V2, the first system to attain perfect functional correctness on this benchmark. On a shared Claude Sonnet 4.0 backbone, VeriTrace outperforms the strongest reproduced baseline by +5.1%, demonstrating that debugging agency closes the final accuracy gap.
[AI-145] BAP-SQL: Budget-Aware Observation Planning for Agent ic Text-to-SQL
链接: https://arxiv.org/abs/2608.02876
作者: Chong Peng,Pin Qian,Su Wang,Yihang Chen,Varun Sah
类目: Artificial Intelligence (cs.AI)
备注: 10 pages, 3 figures
Abstract:Tool-using agents do not merely consume observations: their actions determine what arrives next. In agentic text-to-SQL, a broad query can spend context and database work before useful evidence appears, while post-hoc compression cannot recover omitted rows or expended work. We present BAP-SQL, which treats observation formation as a budget-control stage: it estimates query risk, rewrites SQL when useful, and delegates hard limits to an independent runtime shield. Across general 4B, specialized FINER-SQL 4B, and 7B backbones, BAP-SQL improves tight-budget success. On the primary BIRD-derived setting, it gains 3.4/3.6 percentage points over matched SFT while using 4.5/5.0% fewer tokens. Matched retraining and task-level transfer associate the gain with policy-visible planning and budget-sensitive rescue. The benefit attenuates as model capability and budget increase, reverses at the loosest setting, and does not reduce database work.
[AI-146] MutMem: Cryptographically Authorized Mutation in Persistent Agent Memory
链接: https://arxiv.org/abs/2608.02843
作者: Walid Saidi
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注: this https URL . 32 Pages
Abstract:Persistent agent memory must adapt as later outcomes change earlier evidence, yet mutable retrieval weights create an attribution problem: reviewers must distinguish authorized adaptation from database tampering. We present MutMem, an authorized-mutation protocol in HOM-AIMOS, a persistent agent-memory engine. MutMem retains memory content, records signed positive and negative outcome evidence without age-based expiry, and commits each nontrivial weight change as a housekeeper-authorized transition. Each transition binds a terminal provenance node, signer epoch, quantized old and new weights, a no-fork predecessor, and two domain-separated SHA-256 commitments. Ed25519 verification runs in both the database writer and a portable verifier. Content classified as poison-likely is retained with signed, revisable labels used by recall as trust evidence. We evaluate utility, mutation integrity, and poisoning adaptation. HOM-AIMOS answers 459/500 LongMemEval questions correctly under LLM judgment (91.8%). On LoCoMo, it obtains 74.12% judged accuracy and, under a separate upstream-compatible protocol, 58.20 token F1. A native suite passes all declared authorization, topology, tamper, signer-epoch, and post-mutation-recall cases; median signed-transition latency is 4.865 ms. In a declared N=100 PoisonedRAG adaptation, no injected poison appears in attacked top-5 disclosures (0/100; 95% Wilson upper bound 3.70%), while induced target-answer attack success among 98 clean-negative targets is 1/98 (1.02%). A preregistered four-arm ablation attributes the retrieval reduction to signed stored labels: the retriever selects poison for 94/100 targets when epistemic policy is bypassed and 0/100 when labels are restored. MutMem provides evidence of integrity, authorization, traceability, and historical continuity; it does not establish content truth.
[AI-147] Evading Chain-of-Thought Monitoring Through Model Poisoning
链接: https://arxiv.org/abs/2608.02820
作者: Giorgio Severi,Shujaat Mirza,Blake Bullwinkel,Amanda Minnich
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 15 pages, 2 figures
Abstract:Chain-of-thought (CoT) monitoring is an increasingly important component of AI safety stacks but relies on the assumption that a model’s reasoning trace is informative about its actions. This work studies the limits of CoT monitoring through the lens of model poisoning. We demonstrate that backdoors can be implanted into reasoning models to elicit an attacker-chosen behavior while their CoT traces appear entirely benign. We find that these CoT-Hidden backdoors can be induced through simple fine-tuning recipes across reasoning-model architectures and sizes. When direct poisoning is ineffective, we introduce a curriculum training approach that progressively teaches the model to produce an attacker-chosen output while concealing the behavior from its reasoning traces. These findings suggest that CoT monitoring may be better framed as a question about the consistency between a model’s reasoning trace and its final response than as anomaly detection within a trace. We further examine the mechanisms that allow models to suppress evidence of the target behavior from their reasoning traces. Causal interventions locate a trigger-conditioned activation pathway that does not depend on the visible reasoning, and residual stream verbalizations provide an anomaly warning near answer generation, but do not identify the trigger, target, or backdoor mechanism.
[AI-148] owards a new paradigm of scientific discovery with socialized artificial intelligence
链接: https://arxiv.org/abs/2608.02775
作者: Xinjie Yao,Xingxin Xu,Xiyuan Gao,Zhoupeng Guo,Kunlong Yang,Dengyu Zhao,Siqi Zhao,Zhihe Fan,Yichen Dong,Xin Li,Jiekang Feng,Jiahe Wu,Sen Wang,Beiming Yu,Kejia Zhao,Ruipu Zhao,Jiaqi Zhou,Heyang Li,Jianjun Chen,Anbo Dai,Xin Liu,Zhengtao Yu,Qinghua Hu,Pengfei Zhu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Scientific discovery has advanced through successive transformations in the organization of knowledge. Observation and experimentation established the empirical foundations of science. Theory made it possible to derive general principles from particular phenomena. Computation extended inquiry into systems beyond direct observation, while data-intensive methods opened new spaces of pattern and prediction. Science now confronts a different frontier. The central challenge is no longer simply to produce more information, but to organize expanding knowledge, reasoning, and evidence into a coherent process of discovery. Here, we introduce Bridging Literature, Agents, and Zero-gap Experimentation (BLAZE), a paradigm of socialized scientific intelligence. BLAZE conceives AI not as an assistant for isolated research tasks, but as an organizational infrastructure for scientific discovery. It connects persistent knowledge, collective reasoning, empirical validation, and human judgment within a continuous research lifecycle, transforming fragmented activities into a cumulative process of inquiry, criticism, and revision. The central premise of BLAZE is that scientific intelligence does not arise from computation alone. It emerges from the sustained interaction among knowledge, hypotheses, experiments, and collective verification. By organizing humans and machines within a shared scientific process, BLAZE makes discovery more traceable, reproducible, and cumulative while preserving human creativity, judgment, and responsibility. Socialized scientific intelligence may provide a foundation for the next era of science. Its purpose is not to replace human discovery, but to extend the scale, depth, and continuity of collective scientific inquiry.
[AI-149] Privacy-Preserving AI Verification via Minimal Information Disclosure
链接: https://arxiv.org/abs/2608.02774
作者: Sleem Abdelghafar,Gabriel Kulp
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:AI verification crosses a trust boundary: a verifier must learn enough to establish an authorized claim, yet the same evidence can reveal sensitive details about the model, workload, or hardware. We introduce minimal information disclosure (MID), which designs and quantifies the information content of verifier-facing evidence itself. MID measures collateral leakage with conditional mutual information: what the release reveals about the protected property after the authorized result is known. MID is general by design: it can accommodate different verification goals, protected properties, evidence sources, and deployment constraints. To demonstrate MID’s practicality, we evaluate it on four physical measurements and six verification tasks spanning execution type, hardware identity, compute scale, and model identity. These experiments use three mechanism-design variables–the evidence channel, collection policy, and release transformation–but MID is not limited to these choices and can accommodate other deployable mechanisms. Across these tasks, MID produces three releases with perfect held-out verification and zero measured collateral leakage, while the remaining tasks yield explicit privacy–utility frontiers. MID also supports ZKP-certified releases: we demonstrate our proposed linear-projection mechanism using a Groth16 zk-SNARK.
[AI-150] Dont Regenerate Debug: A Domain-Specific Agent for Repairing Near-Miss Hardware Operators
链接: https://arxiv.org/abs/2608.02712
作者: Yansong Sun,Shenxiu Wu,Siyuan Chen,Runlin Hou,Junhao Qiu,Junming Cao,Shudi Shao,Zhichao Lu,Qingfu Zhang
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:
Abstract:Kernel generation for hardware accelerators such as GPUs and NPUs has become a proving ground for large language models (LLMs), and state-of-the-art systems raise correctness through pipelines that couple LLMs with agentic reinforcement learning and evolutionary search. Such pipelines generate, compile, and execute large numbers of candidate kernels, discarding most of them and forgoing the opportunity to distill failures into reusable knowledge. Many discarded candidates are near-miss operators that compile and run but fail numerical validation; each embodies genuine domain knowledge and a nontrivial investment in LLM inference, cross-compilation, and hardware execution. We argue for a paradigm shift: rather than regenerate, debug. Debugging is far more constrained than generating from scratch: the search space is small and feedback is dense. We present a domain-specific debug agent that addresses three core challenges in autonomous repair: mitigating knowledge scarcity through retrieved patterns and diagnostic instrumentation, ensuring integrity through anti-cheat detection and full-coverage evaluation, and controlling cost via convergence guards and bounded iteration. Debugging serves two complementary roles: it extends the capability frontier by recovering operators that repeated regeneration fails to produce, and it lowers cost per deliverable operator. Debug Pass@1 achieves 66.7% versus Regenerate Avg Pass@1’s 25.9% and Regenerate Pass@3’s 40.7%, while consuming 92.8% fewer tokens per success than three-trial regeneration. Component ablations show that the knowledge base drives recovery, while integrity gates reject 12.5-33.3% of the successes the workflow itself accepted.
[AI-151] Designing a Good Virtual Node: Addressable and Cardinality-Preserving Global Memory for Message Passing Architectures
链接: https://arxiv.org/abs/2608.02709
作者: Félix Marcoccia
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: preliminary work
Abstract:Virtual nodes give message-passing neural networks a simple global communication route, but the standard node–VN–node pipeline compresses the graph into one homogeneous state and broadcasts it identically to every node. Building on the Two-Radius analysis of Mishayev et al., we ask how auxiliary virtual memory can relieve this finite-capacity bottleneck without self-attention. We identify two requirements. First, the global memory should be factorized into independently writable and readable states: this can be achieved using addressable cross-attention slots. Second, addressability alone does not preserve multiplicity, because softmax attention is invariant to uniform replication. Inserting each slot query as a private key/value anchor recovers the discarded normalization mass and yields, on bounded color domains, an injective multiset representation able to implement a 1-WL refinement. Experiments on multiplicity-aware Two-Radius, motif counting, and constrained link-set prediction support this addressable and cardinality-preserving virtual memory at (O(nMd)) arithmetic cost.
[AI-152] Can Training Logs Make Model Comparisons More Precise? ICML2026
链接: https://arxiv.org/abs/2608.02705
作者: Wei-Jung Huang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Machine Learning (stat.ML)
备注: Accepted at the ICML 2026 Workshop on Hypothesis Testing
Abstract:Comparing stochastically trained models requires estimating both a performance difference and its uncertainty from repeated runs. We study whether training logs from those same runs can make such comparisons more precise. Because training-log covariates are produced during training rather than measured before it, we use arm-specific covariate adjustment: each model is adjusted only with statistics from its own runs, and the raw mean difference remains the reported effect. In a vision study spanning three architectures and three datasets, simple adjustments based on early training logs often reduce uncertainty in model comparisons. The main limitation is covariate selection. Broadly searching the log pool for the most correlated statistic often adds more noise than it removes, even when useful statistics exist in hindsight. Training logs therefore appear useful for more precise model comparisons, but only when the adjustment avoids large selection noise.
[AI-153] Predictive Set Theory: A Generative Framework for Cognitive Architecture with Operationalized Core Mechanisms
链接: https://arxiv.org/abs/2608.02704
作者: Yiyang Yu
类目: Artificial Intelligence (cs.AI); Logic in Computer Science (cs.LO); Neurons and Cognition (q-bio.NC)
备注: 103 pages. This preprint establishes the theoretical framework of Predictive Set Theory, providing a generative design specification for cognitive architecture. Comments welcome
Abstract:Predictive processing theories portray the brain as a hierarchical prediction engine that minimizes prediction error, yet they lack operational definitions for the structure of a “prediction,” the standardized response to a prediction error, and the mechanism that maintains consistency across successive updates. Bayesian cognitive science attempts to subsume all uncertainty under probabilistic belief updating, but it presupposes a closed hypothesis space and provides no generative account of how the objects over which probabilities are distributed become discrete, identifiable referents in the first place. This paper introduces Predictive Set Theory (PST), a formal generative framework that reconstructs cognitive architecture from first principles. PST anchors cognition in a minimal set of operations—a sensor formalized as an identity function, set-theoretic state refresh, and three fundamental forms of reference chains (reference, counter-reference, and semi-reference)—and rigorously derives core cognitive functions including state sequences, demand, comparison, efficiency, and finite-horizon probabilistic planning. Rather than modeling neural mechanisms, PST constitutes a design specification for any system that must maintain internal consistency while acting under incomplete information and irreversible risk. The framework offers novel resolutions to classical problems such as Russell’s paradox, the cognitive status of Gödelian incompleteness, the grounding of negative feedback, and the comprehension of film editing. The primary purpose of this paper is to establish, through the public academic record, the originality and completeness of the Predictive Set Theory framework.
[AI-154] NANQ: Noise-Floor-Aware Mixed-Precision Non-Uniform Quantization for Analog Compute-in-Memory
链接: https://arxiv.org/abs/2608.02700
作者: Yizhe Chen,Wenshuai Yao,Saiya Wang,Yuannuo Feng,Wenbo Qi,Kechao Tang,Ngai Wong,Wenyong Zhou,Wang Kang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Analog compute-in-memory (CIM) enables energy-efficient neural network inference, but device variation and read noise can severely degrade low-bit quantized models. Existing CIM-oriented quantization methods mainly minimize ideal quantization error, ignoring the hardware noise floor and thus causing inefficient precision allocation. We propose NANQ, a noise-aware mixed-precision non-uniform quantization framework for analog CIM. NANQ models magnitude-dependent weight noise from measured responses of an eFlash CIM array and converts the noise profile into an adaptive quantization density, assigning finer resolution to low-noise regions while avoiding ineffective precision in noise-dominated regions. It further assigns layer-wise bit-widths by identifying each layer’s precision saturation point under hardware noise using a unified threshold. On-chip experiments on an eFlash CIM SoC show that, under 2-bit weight-magnitude quantization, NANQ improves vision-model accuracy by 8.05 percentage points and reduces language-model PPL by 54.7% on average over PowerQuant. Mixed-precision NANQ captures most of the gains obtainable from additional quantization resources with only 3.2-3.8 equivalent bits.
[AI-155] Explainable AI for the EU Right to Explanation: A Systematic Review of the Law-XAI Translation Gap AAAI
链接: https://arxiv.org/abs/2608.02699
作者: Benjamin Fresz,Elena Dubovitskaya,Marco F. Huber
类目: Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Machine Learning (cs.LG)
备注: Accepted at the 9th AAAI/ACM Conference on AI, Ethics and Society (AIES-26)
Abstract:When algorithms make or influence consequential decisions—about loan eligibility, hiring, or healthcare—EU law grants affected individuals a Right to Explanation. Yet whether (and how) Explainable AI (XAI) can satisfy this right in practice remains poorly understood, with direct implications for individuals’ ability to contest automated decisions that affect their lives. This paper presents a systematic literature review of XAI in the context of the EU Right to Explanation, with particular focus on Art. 15(1)(h) GDPR, Art. 86 AI Act (AIA), and related instruments. We consider papers published from 2024 onwards, as the final version of the AIA was published in July 2024—with Art. 86 being added late. From 2643 initial records identified by a deliberately broad search, we review 57 full texts, of which only 19 papers demonstrate substantive integration of both legal and technical perspectives, showing gaps in the interdisciplinary synthesis of the current regulatory framework. We document three problematic patterns across the corpus: Most misidentify the GDPR legal basis; few engage with the CJEU’s Dun Bradstreet judgment (likely due to publication timing); and the distinction between explanation form (governed by addressee) and content (governed by legal purpose) is often conflated. We conceptualize this as the Addressee/Purpose Framework, propose a four-phase blueprint for operationalization, and identify six concrete open research questions. Without further progress, the Right to Explanation risks remaining a formal obligation without a technically realizable path to compliance.
[AI-156] Steganalysis of Adaptive Covert Collusion in Tool-Using Agent Populations: A Black-Box Cross-Principal Approach
链接: https://arxiv.org/abs/2608.02698
作者: Mohamed Chahine Ghanem
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注: 10 pages
Abstract:Tool-using agents built on large language models (LLMs) are increasingly deployed not by a single operator but by many, side by side on shared infrastructure. This creates a population-level risk that single-agent safeguards miss: a handful of agents can quietly coordinate, rigging a market, boosting one another in a review process, or timing a joint data grab, while each one looks perfectly well-behaved. The difficulty is that the organisations running these agents cannot see inside one another’s models, so any realistic detector must work from behaviour alone: black-box, trace-only, and often with only partial visibility. We treat covert coordination as an information-hiding problem and build a black-box steganalysis detector that combines cross-run mutual-information estimation, permutation tests, distributional-shift statistics, and timing and tool-call side channels, all calibrated to a fixed false-positive budget. Our central move is to stop testing against a single fixed code: we pit the detector against an adversary that continually rewrites its encoding to slip past whatever the detector has learned, and we run this red-versus-blue contest in tool-using, memory-carrying environments rather than toy games. Capacity theory then tells us what to expect, a detection-capacity frontier, a covert bit-rate below which black-box detection is provably no better than chance. We set out an experiment to map this frontier, report clearly labelled placeholder results pending measurement, and flag a practical evasion, spreading a payload across sessions, that current methods largely miss.
[AI-157] Measuring Explainer Stability via Attribution Separability
链接: https://arxiv.org/abs/2608.02697
作者: Eddie Conti,Álvaro Parafita,Axel Brando
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Accepted at EXPLAINS 2026 Conference
Abstract:Attribution methods (AMs) assign an importance score to each feature and are widely adopted to explain black-box models. However, most methods can produce variable attribution scores due to stochastic components in their definition. In this paper, we propose a distribution-based framework to capture the stability of attribution scores. In particular, our approach allows to understand the degree of separability in the ranked attribution vector and obtain the largest index for which a feature ranking remains reliable. We further extend this framework to compare AMs based on the robustness of their rankings across a dataset. Through experiments, we demonstrate how to apply our method to evaluate explainer stability. Overall, our approach provides a complementary criterion for evaluating the stability of AMs.
[AI-158] Output-Aware Rotation for INT2 KV-Cache Quantization
链接: https://arxiv.org/abs/2608.02691
作者: Vincent-Daniel Yun,Woosang Lim,Minsoo Cheong,Sunwoo Lee,Murali Annavaram,Sai Praneeth Karimireddy,Sungjoo Yoo
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:The key-value (KV) cache has become a major memory and bandwidth bottleneck in long-context large language model inference, making ultra-low-bit quantization increasingly important. However, existing rotation-based INT2 methods optimize cache statistics or proxy errors before the complete attention readout, even though the model is ultimately affected by the error propagated through attention and the output projection W_O . To address this mismatch, we propose \textitOptR, an output-aware rotation method that minimizes post- W_O attention-output error. OptR decomposes the post- W_O attention-output error into key- and value-induced terms and learns per-head orthogonal corrections through the full INT2 quantization and attention path. OptR further applies an attention-equivalent key reparameterization to reduce large channel-wise offsets without changing the softmax distribution. Across three models and five reasoning and coding benchmarks, OptR consistently improves both QuaRot and OSCAR and strengthens long-context retrieval, while preserving the paged KV-cache format with negligible inference overhead.
[AI-159] Learning Molecular Representations from Cellular Phenotypes with Structure Preservation
链接: https://arxiv.org/abs/2608.02688
作者: Xuan Lin,Jingyu Sheng,Tengfei Ma,Li Sun,Dapeng Xiong
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Phenotypic drug discovery enables the discovery of functional relationships between molecular structures and cellular responses. However, existing multimodal representation learning methods often optimize cross-modal alignment without considering the intrinsic organization of chemical space, resulting in distorted molecular representations and loss of structural information. We propose \textbfPhenMol, a structure-preserving framework for phenotype-aware molecular representation learning. PhenMol disentangles molecular and cellular representations into shared and private components, enabling phenotype-guided alignment while preserving chemical structures through a dedicated molecular branch. This design integrates cellular phenotype information without disrupting molecular neighborhood organization. Experiments on approximately 3.04 \times 10^4 molecule–cell morphology pairs demonstrate that PhenMol improves molecular property prediction across 270 bioactivity tasks, molecule–phenotype retrieval, and clinical trial outcome prediction. Moreover, ECFP4-based structural analysis shows that PhenMol better preserves molecular neighborhoods and reduces embedding distortion compared with existing multimodal alignment methods. These results highlight the importance of structure-aware constraints in multimodal molecular representation learning and provide an effective approach for integrating cellular phenotypes with chemical knowledge for drug discovery.
[AI-160] BulkPR-Bench: Benchmarking Queue-Level Governance of Interacting Pull Requests
链接: https://arxiv.org/abs/2608.02685
作者: Zetong Xiong,Qiao Zhao,Jun Zhang,Xueying Lyu,Zhi Li,Yixiang Tu,Xiaowen Yang,Yunjie Zhang,Yufeng Wang,Zhe Zhang,Kaize Yu,Hanwen Du,Zhongkai Sun,Zhuoxin Liu,Zekun Lin,Jianwen Yang,Ruining Chen,Ying Zhang,Tingxuan Pan,Ke Chen,Shubin Han,Chuanhao Sun,Yehua Yang
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: 12 pages, 5 figures. Artifact: this https URL ; archived artifact: this https URL
Abstract:Coding-agent benchmarks increasingly cover long-horizon, end-to-end, and interactive development, but typically retain one requested outcome or a fixed change sequence. Sequential policies can process a pull-request (PR) queue one candidate at a time, but when queued PRs interact, maximizing safe delivery can require jointly deciding which changes to merge and in what order. We introduce BulkPR-Bench, an executable benchmark in which an agent must recover consequential PR relations and return a large safe subset in executable order under a rolling-release protocol. The suite contains 581 newly authored candidate PRs on frozen snapshots of 18 real repositories. Registered state-by-state repository execution, including hidden safety checks, validates the gold relation graph; an exact oracle then computes the largest safe subset. Our primary metric, Relational Delivery Score (RDS), scores safe delivery and correct rejection over relation groups from the realized merge trace; Global Safety-Gated Yield (Global-SGY) separately measures strict delivery of the realized whole-queue plan. Under the buffered primary protocol with batch size K=32 , the three highest RDS estimates among the six models are 66.6%, 62.0%, and 57.9%, compared with 53.1% for the strongest sequential baseline. Only 8 of 324 model runs complete a queue exactly. Critical-relation recall ranges from 35.2% to 57.7%, and diagnostic runs supplied with the gold relations show substantial remaining headroom. Gains on relation groups therefore do not yet translate into dependable whole-queue governance.
[AI-161] S3: Improving Agent Safety through Multi-Stage Defense
链接: https://arxiv.org/abs/2608.02683
作者: Zibo Xiao,Haoyu Wang,Jun Sun
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:Large Language Model (LLM) agents rely on multi-stage agentic workflows, with stages such as memory, planning, and tool execution, to accomplish complex tasks. However, risks may emerge at different stages, propagate across steps, and become difficult to detect and mitigate. Existing safety methods protect only isolated stages and are difficult to integrate, leaving agents without comprehensive protection throughout the workflow. To address these limitations, we introduce Stage-Specific Safety Skills, a unified abstraction that represents heterogeneous safety designs as reusable and composable components with explicit stage semantics. We further develop an automated transformation pipeline that converts existing safety designs into reusable safety skills and establish a community-driven safety skill library. Building on this abstraction, we propose S^3 , a multi-stage defense framework in which a guard agent orchestrates stage-specific safety skills for risk detection and mitigation throughout the agentic workflow. We also construct the Multi-Stage Risk Benchmark (MSRB) to evaluate representative risks across workflow stages. Experimental results show that S^3 consistently outperforms representative state-of-the-art baselines in both safety effectiveness and utility preservation. These results demonstrate the potential of stage-specific safety skills as a scalable and composable foundation for building resilient and trustworthy agent systems.
[AI-162] raceCompiler: Skill-Guided Mining and Compilation of LLM Agent Traces into Mostly Deterministic Workflows
链接: https://arxiv.org/abs/2608.02680
作者: Salma El Yadouni(EPFL),Guanyi Li(Binome Technologies)
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 17 pages, 4 figures, 5 tables
Abstract:Tool-using language-model agents repeatedly rediscover procedures they have already executed, producing traces that mix reusable structure with retries, exploration, accidental ordering, and repeated lookups. We present TraceCompiler, a skill-guided system that mines clusters of noisy agent traces and compiles them into executable, mostly deterministic workflows. It admits an inter-tool dependency only when a consumer argument contains a value attributable uniquely to an earlier producer; every hard edge carries an auditable evidence tuple, and ambiguous relations are marked suspected and impose no ordering constraint. Bindings are classified as constants, user inputs, copied outputs, transforms, or residual LLM decisions. On T1, a mechanized form of the rule recovers producer-consumer dependencies at 0.928 precision and 0.943 recall over 15,775 def-use edges of its training split, against 0.711 F1 for adjacency and 0.712 for a frequency-thresholded directly-follows measure on identical data; the compiler skill run blind reaches 0.992 on 250 of those edges. On AppWorld we replay released trajectories in the deterministic simulator to recover masked return values and measure the rule against 563 token edges at 0.993 precision - a self-consistency check, since replay injects tokens by a related heuristic. We compile two recurring intents: a Venmo money-request intent reduces 34 observed API calls to 11 runtime calls and, under leave-one-out execution against the benchmark’s own state tests, passes 15 of 21, the failing fold escalating rather than acting because its required branch was never observed; and a Spotify/Todoist intent the compiler correctly refuses to compile, because an irreversible side effect is under-determined. We measure call reduction but not offline compilation cost, so we claim no net efficiency result.
[AI-163] AI Sandbox: Technical Report
链接: https://arxiv.org/abs/2608.02679
作者: Muhammad Waseem,Md Aidul Islam,Md Nasir Uddin Shuvo,Md Mahade Hasan,Kai-Kristian Kemell,Jussi Rasku,Mika Saari,Vilma Saari,Roope Pajasmaa,Markku Oivo,Pekka Abrahamsson
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:
Abstract:Collaborative AI experimentation across industry and academia requires platforms that enable rapid prototyping while preserving controlled access, tenant separation, and transparent workflows. Despite growing interest in AI sandboxes, there is still limited practical guidance on how to design and implement platforms that integrate experimentation capabilities with governance requirements. This work presents the design and implementation of a governance-aware, multi-tenant AI sandbox for structured experimentation and the generation of reusable evaluation evidence across projects and stakeholder groups. The sandbox was developed within an industry-academia collaboration based on requirements that were iteratively refined with industrial partners. Its reference architecture separates the multi-tenant user interface from the backend control plane and places execution and data-management functions in dedicated layers. The platform supports governed user onboarding, project-centered collaboration, managed access to AI services, approval workflows, audit logging, and traceable experimentation. Experiment configurations, contextual information, and governance decisions are stored as persistent records, allowing evidence and outcomes to be compared and reused across projects. The development process provides practical lessons for deploying and extending governance-aware AI sandbox platforms in collaborative research and industrial environments. Subjects: Software Engineering (cs.SE); Artificial Intelligence (cs.AI) Cite as: arXiv:2608.02679 [cs.SE] (or arXiv:2608.02679v1 [cs.SE] for this version) https://doi.org/10.48550/arXiv.2608.02679 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-164] DenialRAG : Single-Document RAG Poisoning via Embedded Parametric Denial ACSA
链接: https://arxiv.org/abs/2608.02678
作者: Abay Zhurekbay,Tao Liu,Fan Li
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Submit to ACSAC 2026
Abstract:Retrieval-augmented generation (RAG) systems are vulnerable to corpus poisoning: an attacker who inserts a crafted document into the retrieval corpus can steer the underlying large language model (LLM) toward an attacker-chosen wrong answer. Prior single-document attacks typically avoid explicitly naming and refuting the correct answer inside the poisoned passage. In this paper, we examine a complementary design and propose \emphDenialRAG, a single-document poisoning attack that explicitly names the correct answer, denies it, and presents an attacker-controlled explanation for favoring the wrong answer. By placing both the correct answer and the corresponding poisoned answer inside the same retrieved passage, DenialRAG embeds the conflict directly into the context seen by the generator. We evaluate DenialRAG against four published single-document poisoning attacks across three open-domain question-answering datasets, eight target LLMs from four vendors, and five inference-time defenses. The results show that attack effectiveness is strongly model-dependent: DenialRAG achieves the highest attack success rate (ASR) on all three Mistral-7B datasets and remains effective on several other target LLMs, while other attacks dominate in some model regimes. Defense results show meaningful ASR reductions but non-uniform protection, with each defense leaving residual ASR in some settings. Component-level and cross-model analyses further identify the embedded denial as the most influential tested component and show that different poisoning mechanisms lose effectiveness at different rates across model groups. Together, these results show that RAG poisoning risk cannot be fully characterized by a single attack family or a single target model. Comments: Submit to ACSAC 2026 Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG) Cite as: arXiv:2608.02678 [cs.CR] (or arXiv:2608.02678v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2608.02678 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-165] Moving the Safety Barrier: Dynamic Routing Adaptive Alignment Against White-Box Attacks
链接: https://arxiv.org/abs/2608.02674
作者: Shangze Li,Chuancheng Shi,Simiao Xie,Lingzhi He,Cheng Ji,Zifeng Cheng,Fei Shen,Chao Wu,Tat-Seng Chua
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:With the widespread deployment of large foundation models (LFMs) in open environments, safety threats are shifting from black-box jailbreaks toward white-box attacks that directly identify and disrupt internal safety neurons or routes. However, existing safety defenses often rely on static safety units or fixed refusal pathways, leaving models highly vulnerable to targeted route-level white-box attacks. For that, we propose dynamic routing adaptive alignment (DRAA), a framework that introduces dynamic compensatory routes to preserve robust refusal behavior when the safety route is compromised. Specifically, we first identify and localize the model’s safety route by contrasting internal activations between safe and unsafe calibration samples. DRAA then masks this safety route to induce causal failure cases and selectively mines the resulting defense failures, thereby constructing failure-aware preference pairs. Extensive experiments demonstrate that DRAA effectively restructures the underlying pathway dependence of model safety, substantially improving robustness against route-level white-box attacks, while preserving general utility.
[AI-166] Security-First Evaluation of Text-to-Terraform: Benchmarking LLM s and SLMs for Secure IaC Generation
链接: https://arxiv.org/abs/2608.02672
作者: Francis Luis Santos Vargas,Rodrigo Brandão Mansilha,Diego Kreutz
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Emerging Technologies (cs.ET); Software Engineering (cs.SE)
备注: 10 pages, 1 figure, and 9 tables. The benchmark artifacts and CI/CD pipeline are publicly available at this https URL . Accepted for publication at SBSeg 2026
Abstract:Cloud misconfiguration remains a leading cause of security incidents, yet whether LLMs and SLMs can generate security-compliant Infrastructure-as-Code is an open question. We benchmark seven models, three closed LLMs (Claude Opus 4, GPT-5.4, Gemini 2.5 Pro) and four open SLMs (Qwen2.5-Coder-14B, WizardCoder-33B, CodeLlama-13B, Magicoder-S-CL-7B), on AWS Terraform generation across 17 scenarios, integrating Checkov and Trivy scanners into a GitLab CI/CD pipeline and evaluating two prompt strategies at three security levels (pass@5). Syntactic validity and security compliance are largely orthogonal properties in LLM-generated IaC, a model that reliably produces well-formed Terraform does not necessarily produce secure Terraform: WizardCoder-33B achieves 77.8% validate rate yet zero Checkov compliance, while Claude Opus 4 reaches 23.1% Checkov and 92.5% Trivy pass rates under detailed security prompting. Consequently, prompt engineering alone is insufficient: automated multi-tool scanning remains a necessary complement to LLM-assisted IaC generation regardless of model family or prompt strategy. All artifacts are publicly available.
[AI-167] Permission Denied: Policy-Graded Evaluation of Coding Agents in Hardened Environments
链接: https://arxiv.org/abs/2608.02670
作者: Dotan Davidovich,Yair Amar,Hai Rozencwajg,Or Hiltch
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:Coding agents increasingly run inside organizations whose security controls (scoped credentials, restricted egress, read-only filesystems, non-root execution) constrain them like any other software. Existing benchmarks, however, evaluate agents almost exclusively in permissive sandboxes, so it is unknown how performance changes when policy is enforced. In this work, we evaluate 12 coding agents on Terminal-Bench 2.1 across nested security policy levels derived from common real-world enterprise restrictions. Hardening is never free but far from uniform: under the strictest policy, success losses reach 18.3 points and cost inflation 167.3%, and the two axes disagree; the model that best preserves success is also the one that loses the most efficiency, so model choice is policy-dependent. Beyond aggregate scores, we characterize how agents behave when policy blocks their actions and decompose the failures hardening induces: runs grind into timeouts or wrong solutions rather than stopping early, in a mix that differs by model. To ground comparisons, we verify task solvability under the strictest policy, separating model failures from tasks the policy forecloses. We release Boundary-Bench, an open-source hardening plugin enabling policy-constrained evaluation of coding agents on Terminal-Bench and compatible benchmarks.
[AI-168] Vulnerabilities Secrets and Misconfiguration in the Highest-Exposure Docker Hub Images
链接: https://arxiv.org/abs/2608.02669
作者: Cristhian Kapelinski,Beatriz Machado,Diego Kreutz
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Operating Systems (cs.OS); Software Engineering (cs.SE)
备注: 12 pages, 10 figures, 10 tables
Abstract:Docker Hub is the registry underneath most container deployments, and a flaw in a widely reused base image is inherited by every image built on it. Prior ecosystem-scale measurements each rely on a single detector, leaving the tool-dependence of their counts unquantified, while the studies that do compare scanners use samples of tens to hundreds of images. We present ChimangoScan, a pipeline that crawls the Docker Hub namespace (12,716,568 repositories, 663.8 billion cumulative pulls), reconstructs the image layer graph (54.4 million IS_BASE_OF edges), ranks images by an exposure score that folds an image’s own pull count and those of its entire downstream subtree into one scalar, and scans the 52,895 highest-exposure repositories (84.7% of all recorded pulls) with six independent scanners, yielding 170.4 million findings. Vulnerabilities are near-universal: 96.3% of images carry a known package vulnerability, 93.4% a critical one, and 98.0% at least one CIS Docker Benchmark misconfiguration. The posture a single tool reports is largely an artifact of that tool: of 80.7 million distinct (vulnerability, package) groups, 66.8% are flagged by only one of the three vulnerability scanners and just 2.7% by all three, and the best single scanner recovers 66.9%. TruffleHog flags a secret in 76.9% of images, yet hand-labeling 1,100 random detections finds 99.7% are non-credentials. A single zlib CVE reaches images carrying 47.3% of total corpus exposure and propagates to 1.13 million distinct downstream images, but exposure does not predict how vulnerable an image is. We release the pipeline and the 283 GB dataset.
[AI-169] ZK-SR117: A Chunked Zero-Knowledge Attestation Design for Aggregated Fair-Lending Metrics with a Control Mapping toward Full SR 11-7 Coverag e
链接: https://arxiv.org/abs/2608.02664
作者: Mohammad Nasir Uddin,Rahnuma Tabassum Orpita,Eklachur Rahman Bhuiyan,Asaduzzaman Anik
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:Deploying ML models in regulated decision-making (credit underwriting, fraud detection, loan approval) requires demonstrating fairness and robustness to auditors without exposing model weights or customer data. We address this attestation problem for U.S. bank supervision under SR 11-7 and OCC 2011-12 guidance. We present a chunked zero-knowledge circuit design that attests an aggregated fairness statistic - the demographic-parity gap - on committed, nonce-sampled batches of real 2022 HMDA mortgage data, and demonstrate it end-to-end: 32,768 rows, 32 independently verified zkSNARK proofs, aggregated attested gap within 0.0029 of the true held-out value, per-chunk proving under 4 seconds. We also demonstrate extensibility by attesting a second control on the identical architecture - expected calibration error at 10 bins - with all 32 chunks verified, per-chunk proving at about 14.7 seconds, and attested ECE within 0.00037 of the plaintext value on the same committed rows. We compare this design against two alternatives - a flat summation circuit, which overflows past a few thousand rows, and a tree-reduction circuit, numerically exact but intractable to compile - and find the chunked design is the only one that reached this scale. We discovered and root-caused a genuine data-quality failure (a sentinel-code outlier distorting circuit proving and the fairness statistic itself) and resolved it with a published preprocessing specification. We also propose a fuller mapping from SR 11-7 and OCC 2011-12 control language to zero-knowledge statements (nine control elements spanning soundness, calibration, robustness, and drift), a nonce-based sampling protocol resisting bank-side cherry-picking, and a threat model for deployment, proposed as design work, not implemented results. Two controls, one model class, one task are demonstrated end-to-end; the rest is scoped and left as future work.
[AI-170] CT-HEG: A Bidirectional Timestamp-Attributed Event Graph for ICU In-Hospital Mortality Prediction - An Architectural Ablation Study
链接: https://arxiv.org/abs/2608.02663
作者: Mohammad Nasir Uddin,Rahnuma Tabassum Orpita,Asaduzzaman Anik,Eklachur Rahman Bhuiyan,Marjahan Risalat,SM Wali Ullah,Asif Ahamed
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Accurate ICU mortality prediction requires modeling irregular clinical observations across heterogeneous entity types. Existing sequence models handle irregular sampling but ignore typed relational structure; existing graph models assume fixed-interval inputs. We introduce the Continuous-Time Heterogeneous EHR Graph (CT-HEG) schema and evaluate which architectural choices drive predictive performance. CT-HEG encodes each ICU stay as a typed, timestamped graph with three node types (visit, vital, lab_event) and 2D edge attributes (t_hours/48, value_norm) encoding timing and value without imputation. We instantiate CT-HEG as CHIRP-Net, a four-layer heterogeneous GATv2Conv network, evaluated on MIMIC-IV v3.1 (31,142 ICU stays, LOS=48h, 13.4% mortality) with five seeds and bootstrapped confidence intervals, against logistic regression, mTAND, a Transformer, and GRU-D, plus an ablation study. CHIRP-Net achieved 5-seed mean AUROC 0.8449+/-0.0071 (AUPRC 0.4958+/-0.0209); the ensemble achieved AUROC 0.8618 (95% CI: 0.8485-0.8745). Removing reverse edges disconnected observation nodes from the visit readout, cutting AUROC by 0.1968+/-0.0073. Time-attentive edge features contributed 0.0247+/-0.0093 AUROC. Collapsing heterogeneous edge types into one relation (7x fewer parameters) outperformed the full model on all seeds. Post-calibration ECE was 0.0307. Temporal and demographic subgroup analyses were explored but not reported here, pending follow-up work. Bidirectional connectivity was necessary for the model to use its inputs at all, and CT-HEG was reasonably well calibrated after validation-fitted temperature scaling. These results support CT-HEG for irregular EHR data, while external validation, a pre-specified temporal evaluation, and a demographic fairness audit remain necessary before any claim of robustness. Code: this https URL.
[AI-171] Verifier-Guided Model Discovery for Physical Dynamical Systems with Pretrained Symbolic Transformers
链接: https://arxiv.org/abs/2608.02662
作者: Farbod Faraji,Francesco Belardinelli
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 33 pages, 13 figures, 10 tables
Abstract:Reliable forecasting of nonlinear physical systems underpins scientific discovery and engineering decision-making. Yet high-fidelity simulations are prohibitively costly, and machine-learning surrogates can be opaque and encode assumptions about system dynamics, limiting generalizability. Pretrained transformers mapping synthetic ODE trajectories to equations offer interpretable alternatives, promising transfer without system-specific equation knowledge. Transferring them reliably to high-dimensional physical data, however, remains an open challenge. We develop a verifier-guided (VG) workflow around ODEFormer as a symbolic backbone, using dynamical and physical-admissibility criteria to select from a multi-trajectory candidate equation pool, enabling transfer. On canonical Van der Pol oscillators, VG outperforms the original ODEFormer workflow across held-out initial conditions. We then address vortex shedding, a phenomenon occurring in atmospheric and plasma systems of societal relevance, through coordinate reduction and symbolic discovery at fixed and varying Reynolds numbers. VG discovers fixed-parameter reduced-order equations that recover the fundamental shedding oscillator and higher harmonics without a wake-specific candidate library or prescribed Navier-Stokes structure, while the cross-parameter model generalizes to withheld regimes. Reconstruction fidelity alone did not determine symbolic discoverability, highlighting the importance of compatibility between latent dynamics and the backbone’s pretraining distribution. This work establishes a verifier-guided neural-to-symbolic methodology for interpretable and physically auditable forecasting in the natural sciences.
[AI-172] Your Agent ic LLM s Secretly Encode Latent Signals of Indirect Prompt-Injection Exposure
链接: https://arxiv.org/abs/2608.02657
作者: Jianshuo Dong,Yiming Liu,Maosen Zhang,Nan Deng,Xu Peng,Xiaoping Zhang,Tianwei Zhang,Jie Zhang,Han Qiu
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注: Preprint
Abstract:Agentic LLMs are vulnerable to indirect prompt injection (IPI) attacks, e.g., malicious side-tasks hidden in external tool results. While many efforts have sought to address the threats, little is known about the internals of agentic LLMs when they are exposed to IPI attacks, a condition which we call IPI exposure. In this paper, we study this problem in depth from three aspects. (1) Probing: Across six models, including the giant 753B-parameter GLM-5.2, simple linear probes trained on pre-generation hidden states can predict LLMs’ IPI exposure. These probes achieve 90%+ AUROC on unseen attacks, agent instructions, and task suites; they exhibit high robustness under adaptive attacks and in cross-lingual settings. (2) Defense: Our CoT measurement reveals a recognition–action gap: though models encode such signals, they often fail to translate them into safe actions. We then introduce AGRI, a probe-gated reasoning-based defense that prepends anti-injection reasoning on demand. On difficult AgentDojo settings, AGRI substantially reduces attack success rate, e.g., from 34.6% to 0% on Qwen3.5-27B, while largely maintaining clean-task utility. (3) Explanation: We introduce an analysis framework that identifies natural-language explanations most strongly correlated with probe-captured signals. The resulting profiles differ across models: latent signals can align with either direct IPI-exposure claims or indirect operational cues. Code is available: this https URL.
[AI-173] Secure AI Watermarking Framework for IP Protection in Multi-Tenant Cloud Platforms
链接: https://arxiv.org/abs/2608.02656
作者: M Anjan Kumar,Kishor Kumar Gajula,Ch Prathima
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注: 14 Pages, 14 figures, 4 Tables
Abstract:The Secured data safe guard transaction with multi-tenant environments run on private-protected authenticate platforms runs by secured handed environments that emerges with the expansion of cloud-based AI services. To enhanced this secured leakage address challenges solution to protect a secure AI Watermarking system incorporating key distributed between trusted parties based on key authentication as we proposed solution to guided safe guarded way to reactive, and proactive security alert systems using algorithms. This proposed system before attacks can be prevented through the active measures. domain run base restrictions with limited access. Conversely, Proposed system reactive methods to captured on watermarking and biometric identification owner device specific IP leakage that occur during the exchange of data and models in federated and remote learning algorithms.
[AI-174] HyperAg ent: Planning and Acting over Tool-Schema Hypergraphs for Tool-Use LLM Agents
链接: https://arxiv.org/abs/2608.02650
作者: Zian Zhai,Xingyu Tan,Gaowang Zou,Xiaoyang Wang,Wenjie Zhang
类目: Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
备注:
Abstract:Large language model (LLM) agents increasingly rely on external tools to complete complex real-world tasks. However, reliable tool-use planning remains challenging due to the limitations of implicit reasoning and the evolving nature of real-world execution environments. Existing tool-use agents typically rely on LLMs to infer tool compositions from textual descriptions, which can lead to inefficient exploration and unreliable execution in complex tasks. To address these challenges, we model tool relations at the schema level and construct a directed Tool–Schema Hypergraph, in which tools are represented as hyperedges from their required input-schema nodes to their output-schema nodes. Furthermore, we propose HyperAgent, a Tool–Schema Hypergraph-guided framework for dynamic planning and execution. Given a task, HyperAgent first extracts a task-relevant tool context graph and uses it to guide the construction of a schema-aware Task DAG. During execution, HyperAgent dynamically realizes each subtask by constructing a state-conditioned tool support graph through deficit-oriented expansion, which identifies unresolved requirements and retrieves supporting producer tools according to the current agent state. Experiments on AppWorld demonstrate that HyperAgent improves task completion performance while reducing redundant API calls, LLM interactions, and token consumption compared with existing agent baselines.
[AI-175] Verified Tool Calls Improve LLM Agent Reliability Under Non-Atomic Failures
链接: https://arxiv.org/abs/2608.02645
作者: Isham Kalappurackal Mansoor,Abhishek Phadke,Pratip Rana
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:
Abstract:Large Language Model (LLM) agents rely on external tools to perform multistage tasks. Existing agent frameworks typically assume that tool calls are atomic and return binary success or failure signals. However, real-world systems exhibit non-atomic behaviors such as timeouts after dispatch, delayed visibility, and partial state updates. These mismatches lead to reliability issues including duplicate actions, task success, and unnecessary tool executions. A lightweight, verification-aware tool wrapper is introduced that augments tool calls with postcondition verification, verify-before-retry logic, and idempotency keys. The approach is evaluated in a controlled simulated environment with injected non-atomic failures across multiple task templates. The results demonstrate that the proposed method significantly reduces duplicate actions, while maintaining comparable task success rates. Overall, the findings suggest that strengthening tool interaction semantics is a promising direction for improving LLM agent reliability without requiring modifications to the underlying language model.
[AI-176] CUADebug: Diagnosing and Repairing Computer-Use Agent Failures
链接: https://arxiv.org/abs/2608.02643
作者: Weijia Zhang,Kunlun Zhu,Zeyi Liu,Yinting Chen,Tianyi Ma,Jiateng Liu,Jiaxun Zhang,Bingxuan Li,Xiangru Tang,Heng Ji,Jiaxuan You
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: 23 pages, 10 figures, 6 tables
Abstract:Computer-use agents (CUAs) operate real desktop and web interfaces through screenshots, mouse and keyboard actions, and stateful UI feedback, yet their failures remain difficult to diagnose and repair. Unlike text-only agents, CUA failures arise from coupled visual perception, spatial grounding, low-level interaction, task reasoning, and environment dynamics, making debugging a distinctive multimodal causal localization problem. We introduce CUADebug, a framework for diagnosing and repairing CUA failures. CUADebug includes a CUA-specific error taxonomy, CUAErrorBench, a human-annotated OSWorld failure benchmark, and CUADebugger, a tool-augmented debugger. Instead of prompting over the full trajectory once, CUADebugger actively inspects suspicious steps with paired before/after screenshots and action traces, then submits a structured diagnosis containing the root-cause step, error type, grounded evidence, and corrective strategy for re-execution. Human annotations over 204 failed trajectories show that task reasoning and control is the largest failure family (110/204), followed by perception (36), grounding/interaction (25), external/system (13), and an others category of 20 OSWorld infeasible-task cases. On the main Claude-agent split, CUADebugger improves joint subtype-and-step diagnosis from 11.2% to 19.6% with Gemini 2.5 Pro and improves consistently across debugger backbones. In single re-execution package evaluation, RCA-based conditions achieve higher task completion than history-only continuation (28.47% with machine RCA and 29.90% with our method, versus 13.89%); in continual re-execution, our method improves success from 12.2% to 25.86%, while human-oracle guidance reaches 29.21%. These results show that CUA root-cause diagnosis can provide actionable repair signals rather than merely post-hoc explanations.
[AI-177] IR2Solve: Structured Intermediate Representations for Cost-Efficient Optimization Autoformulation
链接: https://arxiv.org/abs/2608.02641
作者: Penglin Zhu,Linhai Zhang,Jungang Xu,Xinchi Wei,Xiuqi Wu
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: 17 pages, 3 figures, 13 tables
Abstract:Large language models (LLMs) can translate natural-language optimization problems into solver-ready formulations, but direct code generation is brittle: schema, indexing, and semantic errors can cause compilation failures, infeasible models, or incorrect objectives, while iterative repair, search, and multi-agent workflows increase inference cost. We present IR2Solve, an intermediate-representation-first autoformulation pipeline that uses a single semantic LLM call to produce a schema-constrained ModelIR, followed by two deterministic stages: verification and IR-to-solver compilation. ModelIR explicitly represents sets, parameters, variables, objectives, and constraints using restricted Python-like expression strings. A concrete scalar-constraint convention represents finite per-index constraint families as individual entries, reducing free-index and implicit-quantification errors while simplifying downstream verification and compilation. Across six cleaned optimization benchmarks, IR2Solve achieves strong objective correctness and remains competitive with recent optimization-modeling systems. A controlled ablation on 153 IndustryOR and ComplexLP instances shows sequential gains from the structured IR interface, the scalar-constraint instruction, and deterministic verification. On a matched ten-instance cost panel, IR2Solve uses one semantic call per instance, whereas Chain-of-Experts and SAC-Opt use 8 and 39 calls per instance and consume 3.3 and 22.9 times the token volume of IR2Solve, respectively. These results show that structured intermediate representations, combined with deterministic post-generation processing, provide a practical accuracy-cost trade-off for LLM-based optimization autoformulation.
[AI-178] Instruction Stacking Collapse: A Benchmark and the Capability-Dependent Value of Prompt Compilation
链接: https://arxiv.org/abs/2608.02639
作者: Atul Anand,Sourav Chattaraj
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: 13 pages, 11 figures. Benchmark, deterministic verifiers, and cached model responses released for full reproduction
Abstract:Production prompts rarely carry a single instruction. One system message may require valid JSON, a word limit, three citations, and a fixed tone at the same time. We study how instruction-following degrades as such constraints accumulate. We introduce a benchmark that stacks 24 verifier-checked instructions, one to twenty at a time, and evaluate three production-tier LLMs (Claude Sonnet 4.6, GPT-5-mini, Gemini 2.5 Flash). Instruction-following degrades non-linearly: the follow rate falls from ~96% to as low as 20%, driven by a structured and reproducible set of pairwise conflicts. A single “output JSON” constraint, for example, is jointly unsatisfiable with nine others. We then evaluate a training-free remedy: an instruction compiler that rewrites the stacked prompt in a single LLM call and is reused across queries. Its benefit is capability-graded. It recovers up to +11 points of follow rate for weaker models, which are also the models most often deployed at scale, while leaving stronger models, which already internalise the same structure, essentially unchanged. Cluster-robust tests, same-baseline controls, and a within-family scaling ladder attribute the gain to the rewrite itself rather than to additional tokens, reordering, or measurement headroom. We release the benchmark, verifiers, and cached runs for full reproduction.
[AI-179] Studying Identifying and Fixing Hidden Technical Debt in AI-Intensive Cyber-Physical Systems
链接: https://arxiv.org/abs/2608.02638
作者: Beena
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: Doctoral Symposium of the 41st IEEE/ACM International Conference on Automated Software Engineering (ASE 2026)
Abstract:Artificial Intelligence (AI) components are increasingly pervasive in several software systems, including Cyber-Physical Systems (CPSs). AI-CPS are used in several domains, including autonomous vehicles, industry, home automation, robotics, and healthcare. Being composed of hardware, AI components, and conventional modules, AI-CPS can exhibit technical debt (TD) that is peculiar and potentially more challenging than that of conventional systems. This thesis aims to characterize AI-CPS TD and propose approaches for its identification and repair. In a first phase, we characterize AI-CPS TD by analyzing AI ecosystems and AI-CPS repositories, as well as interviewing developers. Based on the acquired knowledge, we define approaches to identify and mitigate such TD. Finally, we plan to develop and validate an automated tool that supports agentic AI solutions to monitor, govern, and repay AI-CPS TD.
[AI-180] Rethinking Self-Evolving Agent Skills: Feedback Dynamics over Multiple Rounds
链接: https://arxiv.org/abs/2608.02636
作者: Yuxuan Liu,Zhaochen Su,Yuhao Zhang,Jiahe Guo,Zhongwei Xie,Huihao Jing,Lingyun Xie,Qing Zong,Yauwai Yim,Zhixiong Zhang,Haoran Li,Yangqiu Song
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:
Abstract:Self-evolving skill systems promise to improve agents by turning execution feedback into persistent skill updates without changing the underlying model. Yet it remains unclear when further evolution helps, how successful and failed trajectories shape revision, and whether extra test-time computation can recover the same gains. To address these questions, we present a controlled evaluation framework across five benchmarks and three models. Our primary study contains 42 feedback runs across 14 supported model-benchmark settings. Within each setting, we hold the executor and optimizer configuration, revision procedure, validation rule, and round budget fixed, while varying only the feedback shown to the optimizer: successes and failures (Normal), failures only, or successes only. Evolution is sparse: only 55 of 388 candidates establish byte-distinct validation bests. Validation-based selection chooses an evolved skill in 11 of 14 settings, nine of which improve released-test performance. All 11 selections come from feedback conditions that include failed trajectories, although the relative ranking of Normal and Fail-only varies across settings. Validation and downstream evaluations on test, robustness, and transfer sometimes favor different feedback views. A broader SearchQA analysis covering eight models shows similarly sparse, feedback-dependent dynamics. In the GPT-5.5 test-time-scaling controls, oracle Parallel Sampling comes within 0.43 points of the evolved SearchQA skill but remains 30.96 points behind on SpreadsheetBench; Sequential Refinement recovers neither gain. Overall, persistent skill self-evolution is better understood as sparse, validation-filtered search with model- and benchmark-dependent returns, rather than steady improvement from additional rounds. The implementation is available at this https URL.
[AI-181] PULSE: An Executable Contract Language for Spatiotemporal Knowledge Graph Engineering
链接: https://arxiv.org/abs/2608.02630
作者: Dongxu Yang,Ziyi Liang
类目: Artificial Intelligence (cs.AI); Databases (cs.DB); Programming Languages (cs.PL)
备注: 6 pages, 5 tables, 1 code listing; submitted to KGSWC 2026. Research artifact available under the Apache-2.0 license
Abstract:Knowledge graph engineering often distributes accepted state, observations, constraints, processes, and hypothetical scenarios across artifacts whose combined execution contract remains external. We present PULSE, an Object-Process-Methodology-inspired language that localizes four operational roles and their write effects in one typed runtime. Here, modes denote operational roles rather than modal or deontic logic. The implemented contract fixes evidence non-overwrite, branch isolation, grounded multi-subject timers, guarded state change, and declaration-ranked event ordering over time and space; an external runner still decides whether evidence becomes an authoritative move. GeoSPARQL, SOSA, and SHACL remain generated views. A core calculus gives an effect-confinement lemma and six safety properties. Lean 4 checks kernel analogues for positions, evidence, clocks, monitors, atomicity, and branch source retention; 88 tests, 3,534 bounded checks, and 32 Lean/Python runtime-kernel cases bound the implementation claim to the checked cases. First-author implementations of a standards composition and a separate Sismic statechart reproduce the tested cold-chain trace. Across 37,440 generated temporal traces, PULSE matches a separate workflow and distinguishes ten single-field mutants. On the complete NOAA IBTrACS since1980 subset it agrees with GEOS and an event sweep on 1,476,290 transition-zone pairs, including 4,800 sampled and 12,831 duration-qualified events. Project-specific GeoSPARQL probes measure interface coverage. Overall, the results support contract localization, safety arguments, and trace parity for the tested fragment; language superiority and usability remain outside the evaluation.
[AI-182] Multimodal Auto-regressive Transformer Surrogate for Modeling Variable Operations and Quantifying Uncertainty in Geological Carbon Storag e
链接: https://arxiv.org/abs/2608.02629
作者: Yifu Han,Louis J. Durlofsky
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:The use of variable well perforation and injection strategies can improve the efficiency of geological carbon storage operations. We develop a new multimodal auto-regressive transformer surrogate to model these operations under geological uncertainty. A modified SEAM CO2 geomodel, which involves a faulted system with three stacked aquifers, is considered. The two injection wells are perforated in stages, from bottom to top, with the stage durations and individual well injection rates treated as control variables. The surrogate model processes three input modalities - the 3D geomodel, scalar parameters characterizing relative permeability functions, and control variables - through separate encoders. These are fused via self-attention in a transformer encoder, and a temporal decoder generates predictions auto-regressively through encoder-decoder cross-attention. The surrogate is trained, using 4000 GEOS flow simulations, to predict saturation and pressure at monitoring locations, total injected and mobile CO2 mass, and saturation footprints. For a new test set, involving randomly sampled geomodels and control variables, the surrogate achieves a median saturation MAE of 0.028 and median relative errors of 0.2-5% for the other quantities of interest. Importantly, it captures the switch from rate to bottom-hole-pressure control. The surrogate model is used within a hierarchical Markov chain Monte Carlo data assimilation procedure for a synthetic true model under three operational strategies. Substantial uncertainty reduction is achieved for key metaparameters, particularly the fault permeabilities. Posterior predictions for saturation footprints and total injected and mobile CO2 mass are also shown to be generally consistent with true model results.
[AI-183] Deep Divide-and-Reduce in Symbolic Regression
链接: https://arxiv.org/abs/2608.02628
作者: Yusong Deng,Yanjie Li,Weijun Li
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Symbolic regression (SR) is the task of discovering underlying patterns from data and representing them using mathematical expressions. Current machine learning approaches to SR often lack a profound understanding of the intrinsic mathematical and physical principles governing these expressions. While the pioneering AI Feynman method leverages the mathematical properties underlying the data, its expression simplification mechanism suffers from a narrow scope of applicability and is prone to failure on complex equations. Furthermore, its underlying mechanisms rely heavily on brute-force searches for sub-expressions, severely limiting its practical utility. Through rigorous mathematical deduction and proofs, we propose our method, Deep Divide and Reduce in Symbolic Regression (DDRSR). DDRSR fundamentally broadens the applicability of expression decomposition and reduction, circumvents the need for brute-force sub-structure searches, and ensures both wider versatility and strict theoretical correctness. Empirical evaluations demonstrate that these theoretical principles yield significant advantages in both expression decomposition and numerical regression tasks. Finally, we discuss the applicable scenarios and inherent limitations of this paradigm, alongside promising directions for future research.
[AI-184] KernelBrain: Coarse-to-Fine Budget-Aware Search for Agent ic GPU Kernel Optimization
链接: https://arxiv.org/abs/2608.02611
作者: Shuai Che,Gang Peng
类目: Distributed, Parallel, and Cluster Computing (cs.DC); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Software Engineering (cs.SE)
备注:
Abstract:Automating GPU kernel optimization remains difficult in practice: generated variants can violate correctness constraints, runtime measurements are noisy, and search often stalls early. We present a practical optimization agent that combines LLM-guided mutation, adaptive resource allocation, policy-gated evaluation, and profiler-informed diagnosis. The system screens many candidates with low-cost evaluation and allocates higher-fidelity budget only to promising survivors to optimize and evolve GPU kernels. On important Triton kernel generation tasks, this design improves both kernel quality and search efficiency, reaching 0.88x-6.72x speedup over PyTorch and up to 1.4x speedup over the state-of-the-art kernel agent, with up to 48% lower optimization time.
[AI-185] Self-Organising Digital Circuits
链接: https://arxiv.org/abs/2608.02606
作者: Marcello Barylli,Gabriel Béna,Alexander Mordvintsev,Eleni Nisioti,Sebastian Risi
类目: Artificial Intelligence (cs.AI); Neural and Evolutionary Computing (cs.NE)
备注: 9 pages, 10 figures
Abstract:Fault tolerance in classical computing has traditionally relied on static strategies like hardware redundancy and error-correcting codes. Biological systems, in contrast, exhibit adaptive plasticity, maintaining function through dynamic re-organisation around damage. Inspired by this principle, we introduce Self-Organising Digital Circuits, framing functional logic generation and maintenance as a meta-learning problem on graphs. Our architecture employs a topology-masked Transformer that configures the Lookup Tables (LUT) of a circuit’s Boolean gates. Extending the pattern-generation paradigm of Neural Cellular Automata (NCA), it navigates the degenerate Boolean search space to satisfy a computational task, rather than regenerating a fixed target state. We demonstrate that it can self-assemble functional circuits from scratch and rapidly re-route logic around permanent, previously unseen hardware faults. For soft errors, the policy achieves near-perfect recovery (99.99% accuracy) from damage sizes far exceeding training conditions. We further observe generalisation across circuit scales: accuracy improves on graphs substantially wider than those seen during training. This work bridges the principles of biological self-organisation with the practical domain of digital hardware.
[AI-186] AgentS tream: How Well Do Self-Evolving LLM Agents Perform Under Streaming Tasks?
链接: https://arxiv.org/abs/2608.00155
作者: Dong Yan,Jian Liang,Dapeng Hu,Ran He,Nicholas Jing Yuan,Qi Zhang,Tieniu Tan
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Code is available at this https URL
Abstract:Large language model (LLM) agents can self-evolve by continually improving from their own accumulated experience. However, existing studies predominantly adopt independent evaluation. Consequently, the behavior of self-evolving agents in realistic streaming settings, where agents adapt to diverse and complex task streams, remains poorly understood. To address this gap, we introduce AgentStream, a unified framework that evaluates self-evolving agents spanning diverse evolution components by organizing agentic benchmarks into a configurable task stream and instantiating the \textttIsolated, \textttSequential, and \textttInterleaved streaming scenarios at test time, which progressively vary the scope and domain composition of the stream. Over these scenarios, we combinatorially evaluate five representative self-evolving methods across three frontier foundation models, disentangling how model capability, method architecture, and streaming scenario jointly shape self-evolution. Our results show that self-evolution reliability varies across streaming scenarios, the benefit of self-evolution is gated by model capability and non-monotonic in model strength, and no single method dominates across models and scenarios. These findings offer concrete guidance for selecting self-evolving methods across models and streaming scenarios. Overall, we advocate that self-evolving agents should be evaluated under realistic task streams rather than isolated single-task settings.
[AI-187] Separating quantum circuits from classical LLM s
链接: https://arxiv.org/abs/2608.03962
作者: Srinivasan Arunachalam,Arkopal Dutt,Hari Krovi,Rik Sengupta
类目: Quantum Physics (quant-ph); Artificial Intelligence (cs.AI); Computational Complexity (cs.CC)
备注: 60 pages, 6 figures
Abstract:Modern large language models - transformers and diffusion language models - are built around two canonical algorithmic tasks: prediction and generation. We prove unconditional separations between low-depth quantum computation and the corresponding bounded-resource classical language-model architectures in both regimes. Concretely, we exhibit the following: 1. Distributional separation. We give a distribution that is sampleable by \textsfQNC^0 circuits (i.e., a family of constant-depth quantum circuits consisting of bounded fan-in gates) that no constant-round diffusion language model ( \textsfDLM ) with shallow scheduling and denoising can sample within constant distance, even when allowed sublinear chain-of-thought and output-token revision/remasking events, the very features modern \textsfDLM s rely on. 2. Functional separation. We exhibit a function computable in \land \circ \textsfQNC^0[\log\log n] (i.e., a family of O (\log\log n) -depth \textsfQNC^0 circuits, where n is the input length, followed by a single classical \mathsfAND gate) such that any constant-depth decoder-only transformer computing the function must be large: it would have to have width n^\Omega(1) . Together, our work initiates the study of quantum advantage in the era of large language models. Comments: 60 pages, 6 figures Subjects: Quantum Physics (quant-ph); Artificial Intelligence (cs.AI); Computational Complexity (cs.CC) Cite as: arXiv:2608.03962 [quant-ph] (or arXiv:2608.03962v1 [quant-ph] for this version) https://doi.org/10.48550/arXiv.2608.03962 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Arkopal Dutt [view email] [v1] Tue, 4 Aug 2026 17:28:24 UTC (1,461 KB)
[AI-188] Minimax-Optimal Semiparametric Contextual Dynamic Pricing with Multimodal Revenue
链接: https://arxiv.org/abs/2608.03142
作者: Xueping Gong,Zhuoluo Zhang,Zhaowei Miao,Jiheng Zhang
类目: Machine Learning (stat.ML); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 53 pages
Abstract:We study contextual dynamic pricing with arbitrary covariate sequences and bounded, possibly nonbinary purchase quantities. Demand follows a semiparametric surplus-index model with an unknown linear valuation parameter and an unknown Hölder-smooth response. We impose neither concavity nor strong unimodality on revenue and allow nonunique optimal prices. We develop a pilot-corrected layered decision-partitioning policy that combines directional pilot estimation, local polynomial learning, predictable data assignment, and global action elimination. Pilot correction removes the first-order effect of valuation-parameter error, while permanent labels enable concentration under adaptive sampling. The policy attains the minimax smoothness-dependent horizon rate up to logarithmic factors; a matching lower bound already holds for a constant-context binary-demand subclass.
[AI-189] Optimal Liability Design for Medical AI
链接: https://arxiv.org/abs/2608.03114
作者: Rui Mao,Tingliang Huang,Houcai Shen
类目: Theoretical Economics (econ.TH); Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注:
Abstract:Artificial intelligence (AI) is increasingly integrated into medical decision-making, yet its liability implications remain complex, particularly when physicians differ in diagnostic skills and their quality is unobservable. This paper develops a principal-agent model in which a social planner designs medical liability to regulate a physician with private quality information who chooses between a standard treatment, a personalized judgment-based treatment, or following an imperfect AI recommendation. Our analysis yields several novel insights. First, we show that the optimal mechanism under asymmetric information is surprisingly simple: a uniform, one-size-fits-all liability level for all physician types who deviate from the standard of care. Despite physician heterogeneity, this simple policy often achieves the full-information first-best outcome, particularly when standard care is reliable or AI is highly accurate. Second, the relationship between AI accuracy and optimal liability is non-monotonic. Contrary to common intuition, better AI does not always imply more relaxed liability. As AI accuracy increases, the optimal liability either decreases monotonically or follows an inverted-U pattern, depending on the uncertainty of the standard treatment. Third, asymmetric information does not universally reduce social welfare. Welfare loss arises only when standard care is unreliable and AI accuracy is too low; even then, its magnitude follows an inverted U-shape, initially increasing as AI complicates the regulatory problem, but declining as more accurate AI helps mitigate it. Finally, we find that information asymmetry is a double-edged sword in the presence of AI, and greater transparency does not benefit all stakeholders equally.
[AI-190] Improved Quantum Algorithms for Reinforcement Learning Under a Generative Model
链接: https://arxiv.org/abs/2608.02826
作者: Joao F. Doriguello
类目: Quantum Physics (quant-ph); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Machine Learning (stat.ML)
备注: 22 pages. Comments welcome
Abstract:Reinforcement learning is a subfield of machine learning that studies how an agent interacts with an environment in order to extract as large a reward as possible. A standard approach to study such interaction is through Markov Decision Processes (MDPs) and the task of choosing an optimal policy — a function that tells the agent which action to take. In this work, we study two types of MDPs — finite-horizon and infinite-horizon discounted — and propose new quantum algorithms for computing approximate optimal policies. Our quantum algorithms are based on a new combination of standard value iteration and quantum subroutines like quantum mean estimation and quantum maximum finding, overall enhanced with techniques from sample-optimal classical algorithms. Our resulting query complexities improve upon previous works, thus approaching already established quantum lower bounds.
[AI-191] A Hyperfinite Framework for Score-Based Generative Modeling
链接: https://arxiv.org/abs/2608.02799
作者: Sunder Ram Krishnan
类目: Machine Learning (stat.ML); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Probability (math.PR)
备注:
Abstract:Score-based diffusion models are typically formulated using continuous-time stochastic differential equations and measure-theoretic stochastic calculus. In this paper, we develop a hyperfinite formulation of score-based generative modeling within the framework of Nonstandard Analysis. Starting from an internal diffusion process on a hyperfinite grid, we derive the associated infinitesimal generator and establish its correspondence with the classical Fokker–Planck equation. We then obtain a hyperfinite backward-mean identity that yields the reverse-time drift and provides a constructive derivation of the reverse-time SDE. Building on these results, we show that minimization of an internal score-matching objective recovers the score function required by the reverse-time dynamics, thereby connecting score estimation with generative sampling directly at the hyperfinite level. Under suitable assumptions, we further derive a hyperfinite Girsanov formula and establish a relationship between likelihood optimization and Fisher-divergence objectives. Finally, we analyze the second-order consistency of the hyperfinite dynamics and show that the leading correction term depends explicitly on the fourth moment of the increment distribution, with the Gaussian value \kappa=3 eliminating the leading dispersion contribution. Taken together, these results provide a unified hyperfinite framework for diffusion-based generative modeling–while laying foundations for further extensions–that links discrete grid dynamics, reverse-time diffusion, score matching, and likelihood-based formulations within a common nonstandard setting.
[AI-192] A Blind Spot in Alignment: Quantifying Biosecurity Risks in Large Language Models
链接: https://arxiv.org/abs/2608.02684
作者: Shu Quan,Tianfang Hao,Sitong Fang,He Geng,Jiayi Zhou,Boyuan Chen,Kaile Wang,Donghai Hong,Juntao Dai,Yaodong Yang,Jiaming Ji
类目: Quantitative Methods (q-bio.QM); Artificial Intelligence (cs.AI)
备注: Accepted to COLM 2026. 40 pages, 9 figures
Abstract:Large Language Models (LLMs) are accelerating biological research, yet this same capability poses a critical biosecurity threat: models that assist in protein engineering can equally be prompted to generate predicted toxin-like sequences, potentially lowering the barrier to biological misuse. Current safety evaluations, however, operate in natural language and cannot determine whether a model-generated amino acid sequence is biological gibberish or a computational risk signal. To address this evaluation blind spot, we introduce SPIKE-Bench, coupling 631 curated toxin-design prompts across seven functional categories with the SPIKE funnel, a three-stage protocol that filters output through compliance, biological plausibility, and predicted toxicity, producing stage-level diagnostics and an aggregate function-aware metric: the Functional Harmfulness Rate (FHR). An audit of 32 LLMs reveals that most models freely comply with toxin-design requests; FHR is driven primarily by biological generation capability rather than safety alignment, reaching 50.7%; and Refusal Rate fails to predict functional risk. As a first step toward mitigation, we provide BioSafe-Guard, a domain-specialized classifier that substantially reduces predicted functional risk while preserving benign utility. We release SPIKE-Bench and BioSafe-Guard at this https URL to support more rigorous biosecurity evaluation of LLMs.
[AI-193] Cross-Anesthetic ECoG State Decoding Fails at the Decision Threshold Not the Representation
链接: https://arxiv.org/abs/2608.02646
作者: Kunkun Zhang,Qianwei Zhou
类目: Quantitative Methods (q-bio.QM); Artificial Intelligence (cs.AI)
备注:
Abstract:Decoders of anesthetic state from cortical activity fail across drug classes, most notoriously ketamine, but reported accuracy cannot say whether the neural representation or only the decision threshold has failed; we separate the two in a controlled preparation with ground-truth labels. We decoded awake versus anesthetized from mouse electrocorticography (the 250-Hz-bandlimited local field potential, sampled at 1875 Hz) under leave-one-anesthetic-out evaluation across five mechanistically distinct anesthetics (isoflurane, dexmedetomidine, ketamine, propofol, and midazolam), comparing a spatially blind band-power decoder, covariance/Riemannian representations, and Riemannian domain adaptation, with all statistics at the session level and mouse-level cluster bootstrapping for the ketamine fold. The representation transfers: band-power ranks awake versus anesthetized at a session AUROC of at least 0.96 on every held-out drug, ketamine included (0.980, cluster confidence interval 0.821 to 1.000). The failure is confined to the threshold: across three representations the ketamine ranking is near-invariant while its balanced accuracy swings from chance to high, and a permutation test is significant for ranking (p = 0.0025) but not for fixed-threshold accuracy (p = 0.3795). Riemannian domain adaptation is net-negative. A causal, label-free threshold anchored to the subject’s own pre-induction baseline fixes ketamine (balanced accuracy 0.50 to 0.85) and dominates domain adaptation. Because the ketamine test sessions come from three mice that also contribute training drugs, this is within-subject cross-drug transfer; we do not claim population-level transfer across subjects. In cross-drug state decoding the actionable failure is calibration, not representation.
[AI-194] MDArena: Evaluating Coding Agents on Realistic Molecular Dynamics Workflows
链接: https://arxiv.org/abs/2608.02642
作者: Nithishwer Mouroug Anand,Wei-Tse Hsu,Kyle Vaccaro,Eden James Gage,Jonathan David Colburn,Linda Xi Phan,Minjoon Seo,Kevin Guan,Philip C. Biggin
类目: Chemical Physics (physics.chem-ph); Artificial Intelligence (cs.AI)
备注: 17 pages including appendices, 4 figures
Abstract:Accelerating scientific discovery is among the most consequential applications of AI, and computational biomolecular simulation stands out as a particularly promising target within this broader effort. Coding agents promise to automate significant portions of this workflow, yet their reliability on realistic molecular dynamics (MD) tasks remains poorly characterized. To address this issue, we introduce MDArena, a benchmark of 50 containerized tasks drawn from active biomolecular simulation projects, spanning 29 molecular systems and 14 broad research protocols, including trajectory analysis, complex system preparation, free-energy protocols, and enhanced sampling. We evaluate six model/harness configurations spanning Codex and OpenCode. Among the evaluated configurations, Codex GPT-5.5 at extra-high reasoning effort performs best, reaching 24/50 Strict-Pass@1 successes (48%), followed by Codex GPT-5.5 Medium with 21/50, and OpenCode Gemini Flash 3.5 with 20/50. Average correctness and process rewards are substantially higher than strict success rates across all configurations, indicating that agents frequently make meaningful partial progress but fail on the fine-grained details required for reproducible scientific workflows. Hard tasks remain largely unsolved, particularly membrane-protein system preparation and alchemical free-energy setup, both unsolved or near-unsolved by every evaluated configuration. MDArena thus exposes a substantial gap between the usefulness of coding agents as supervised assistants and their reliability as autonomous MD researchers, while providing a reproducible and extensible platform for tracking progress toward closing it.
[AI-195] UniPASE: A Generative Model for Universal Speech Enhancement with High Fidelity and Low Hallucinations
链接: https://arxiv.org/abs/2604.14606
作者: Xiaobin Rong,Zheng Wang,Yushi Wang,Jun Gao,Jing Lu
类目: Audio and Speech Processing (eess.AS); Artificial Intelligence (cs.AI)
备注: Accepted by IEEE TASLP
Abstract:Universal speech enhancement (USE) aims to restore speech signals from diverse distortions across multiple sampling rates. We propose UniPASE, an extension of the low-hallucination PASE framework tailored for USE. At its core is DeWavLM-Omni, a unified representation-level enhancement module fine-tuned from WavLM via knowledge distillation on a large-scale supervised multi-distortion dataset. This module directly converts degraded waveforms into clean and linguistically faithful phonetic representations, ensuring robust enhancement with minimal linguistic hallucination. Based on these enhanced phonetic representations, an Adapter generates enhanced acoustic representations containing rich acoustic details, which a neural Vocoder uses to reconstruct corresponding high-fidelity 16-kHz waveforms. A PostNet then converts the waveforms to 48~kHz before resampling them to their original rates, enabling seamless handling of inputs and outputs at multiple sampling rates. Experimental results on several evaluation datasets, covering sub-tasks and full tasks, demonstrate that UniPASE achieves superior or competitive performance compared with existing state-of-the-art models. The proposed model also serves as the backbone of our submission to the URGENT 2026 Challenge, which achieved 1st place in the objective evaluation. The source code and audio demos are available at this https URL.
[AI-196] GAP-URGENet: A Generative-Predictive Fusion Framework for Universal Speech Enhancement ICASSP2026
链接: https://arxiv.org/abs/2604.01832
作者: Xiaobin Rong,Yushi Wang,Zheng Wang,Jing Lu
类目: Audio and Speech Processing (eess.AS); Artificial Intelligence (cs.AI); Sound (cs.SD)
备注: Awarded 1st place in the URGENT 2026 Challenge (objective phase), accepted by ICASSP 2026
Abstract:We introduce GAP-URGENet, a generative-predictive fusion framework developed for Track 1 of the ICASSP 2026 URGENT Challenge. The system integrates a generative branch, which performs full-stack speech restoration in a self-supervised representation domain and reconstructs the waveform via a neural vocoder, along with a predictive branch that performs spectrogram-domain enhancement, providing complementary cues. Outputs from both branches are fused by a post-processing module, which also performs bandwidth extension to generate the enhanced waveform at 48 kHz, later downsampled to the original sampling rate. This generative-predictive fusion improves robustness and perceptual quality, achieving top performance in the blind-test phase and ranking 1st in the objective evaluation. Audio examples are available at this https URL.
[AI-197] StuPASE: Towards Low-Hallucination Studio-Quality Generative Speech Enhancement INTERSPEECH2026
链接: https://arxiv.org/abs/2603.09234
作者: Xiaobin Rong,Jun Gao,Zheng Wang,Mansur Yesilbursa,Kamil Wojcicki,Jing Lu
类目: Audio and Speech Processing (eess.AS); Artificial Intelligence (cs.AI); Sound (cs.SD)
备注: Accepted to Interspeech 2026
Abstract:Achieving high perceptual quality without hallucination remains a challenge in generative speech enhancement (SE). A representative approach, PASE, is robust to hallucination but has limited perceptual quality under adverse conditions. We propose StuPASE, built upon PASE to achieve studio-level quality while retaining its low-hallucination property. First, we show that finetuning PASE with dry targets rather than targets containing simulated early reflections substantially improves dereverberation. Second, to address performance limitations under strong additive noise, we replace the GAN-based generative module in PASE with a flow-matching module, enabling studio-quality generation even under highly challenging conditions. Experiments demonstrate that StuPASE consistently produces perceptually high-quality speech while maintaining low hallucination, outperforming state-of-the-art SE methods. Audio demos are available at: this https URL.
[AI-198] PASE: Leverag ing the Phonological Prior of WavLM for Low-Hallucination Generative Speech Enhancement AAAI2026
链接: https://arxiv.org/abs/2511.13300
作者: Xiaobin Rong,Qinwen Hu,Mansur Yesilbursa,Kamil Wojcicki,Jing Lu
类目: Audio and Speech Processing (eess.AS); Artificial Intelligence (cs.AI); Sound (cs.SD)
备注: Accepted by AAAI 2026
Abstract:Generative models have shown remarkable performance in speech enhancement (SE), achieving superior perceptual quality over traditional discriminative approaches. However, existing generative SE approaches often overlook the risk of hallucination under severe noise, leading to incorrect spoken content or inconsistent speaker characteristics, which we term linguistic and acoustic hallucinations, respectively. We argue that linguistic hallucination stems from models’ failure to constrain valid phonological structures and it is a more fundamental challenge. While language models (LMs) are well-suited for capturing the underlying speech structure through modeling the distribution of discrete tokens, existing approaches are limited in learning from noise-corrupted representations, which can lead to contaminated priors and hallucinations. To overcome these limitations, we propose the Phonologically Anchored Speech Enhancer (PASE), a generative SE framework that leverages the robust phonological prior embedded in the pre-trained WavLM model to mitigate hallucinations. First, we adapt WavLM into a denoising expert via representation distillation to clean its final-layer features. Guided by the model’s intrinsic phonological prior, this process enables robust denoising while minimizing linguistic hallucinations. To further reduce acoustic hallucinations, we train the vocoder with a dual-stream representation: the high-level phonetic representation provides clean linguistic content, while a low-level acoustic representation retains speaker identity and prosody. Experimental results demonstrate that PASE not only surpasses state-of-the-art discriminative models in perceptual quality, but also significantly outperforms prior generative models with substantially lower linguistic and acoustic hallucinations.
[AI-199] Assessing speech quality metrics for evaluation of neural audio codecs under clean speech conditions
链接: https://arxiv.org/abs/2509.24457
作者: Wolfgang Mack,Nezih Topaloglu,Laura Lechler,Ivana Balić,Alexandra Craciun,Mansur Yesilbursa,Kamil Wojcicki
类目: Audio and Speech Processing (eess.AS); Artificial Intelligence (cs.AI); Sound (cs.SD); Signal Processing (eess.SP)
备注:
Abstract:Objective speech-quality metrics are widely used to assess codec performance. However, for neural codecs, it is often unclear which metrics provide reliable quality estimates. To address this, we evaluated 45 objective metrics by correlating their scores with subjective listening scores for clean speech across 17 codec conditions. Neural-based metrics such as scoreq and utmos achieved the highest Pearson correlations with subjective scores. Further analysis across different subjective quality ranges revealed that non-intrusive metrics tend to saturate at high subjective quality levels.
[AI-200] UL-UNAS: Ultra-Lightweight U-Nets for Real-Time Speech Enhancement via Network Architecture Search
链接: https://arxiv.org/abs/2503.00340
作者: Xiaobin Rong,Leyan Yang,Dahan Wang,Yuxiang Hu,Changbao Zhu,Kai Chen,Jing Lu
类目: Audio and Speech Processing (eess.AS); Artificial Intelligence (cs.AI)
备注: Accepted by IEEE TASLP
Abstract:Lightweight models are essential for real-time speech enhancement applications. In recent years, there has been a growing trend toward developing increasingly compact models for speech enhancement. In this paper, we propose an Ultra-Lightweight U-net optimized by Network Architecture Search (UL-UNAS), which is suitable for implementation in low-footprint devices. Firstly, we explore the application of various efficient convolutional blocks within the U-Net framework to identify the most promising candidates. Secondly, we introduce two boosting components to enhance the capacity of these convolutional blocks: a novel activation function named affine PReLU and a causal time-frequency attention module. Furthermore, we leverage neural architecture search to discover an optimal architecture within our carefully designed search space. By integrating the above strategies, UL-UNAS not only significantly outperforms the latest ultra-lightweight models with the same or lower computational complexity, but also delivers competitive performance compared to recent baseline models that require substantially higher computational resources. Source code and audio demos are available at this https URL.
[AI-201] Crowdsourced Multilingual Speech Intelligibility Testing
链接: https://arxiv.org/abs/2403.14817
作者: Laura Lechler,Kamil Wojcicki
类目: Audio and Speech Processing (eess.AS); Artificial Intelligence (cs.AI); Sound (cs.SD); Signal Processing (eess.SP)
备注:
Abstract:With the advent of generative audio features, there is an increasing need for rapid evaluation of their impact on speech intelligibility. Beyond the existing laboratory measures, which are expensive and do not scale well, there has been comparatively little work on crowdsourced assessment of intelligibility. Standards and recommendations are yet to be defined, and publicly available multilingual test materials are lacking. In response to this challenge, we propose an approach for a crowdsourced intelligibility assessment. We detail the test design, the collection and public release of the multilingual speech data, and the results of our early experiments.
机器学习
[LG-0] Assessment of Conditional Diffusion Model for Synthetic Histopathology Image Generation
链接: https://arxiv.org/abs/2608.03990
作者: Seyed Kahaki,Shijie Li,Weijie Chen,Nicholas Petrick
类目: Machine Learning (cs.LG)
*备注: 11 pages, 3 figures
Abstract:Synthetic histopathology image generation has emerged as an approach that may address data scarcity in computational pathology, yet current evaluation methodologies may not fully assess synthetic data quality for medical applications. This work investigates and addresses limitations in existing evaluation metrics, investigating an approach for assessing synthetic histopathology image quality through domain-specific metrics and downstream task validation. We show that conventional synthetic data evaluation metrics such as Frechet Inception Distance (FID) and Inception Score (IS) may have limitations when applied to histopathology images due to their reliance on ImageNet-pretrained feature extractors. To address these limitations, we propose for consideration modified FID and IS approaches utilizing foundation models pretrained on digital pathology datasets, supplemented by precision-recall based metrics as part of an additional quality assessment. Using conditional denoising diffusion models trained on four benchmark datasets, with a two-step training approach, we generated synthetic datasets with systematically varied quality characteristics. We also measured the correlation between the synthetic data quality metrics with downstream nuclei segmentation performance using common metrics including the aggregated Jaccard index (AJI+) and the Dice coefficient. The study results suggest that pathology-specific metrics may provide improved discriminative power. Specifically, the modified Inception Score indicates higher correlation with downstream task performance (r=0.6096 with AJI+, p=0.0122), compared to the original IS (r=0.0708, p=0.7944). Our observations indicate that increasing the variety of generated training data has a higher positive correlation with segmentation model performance than improving the visual fidelity of individual generated images.
[LG-1] Muon Meets Mamba: Spectral Optimization for State Space Models
链接: https://arxiv.org/abs/2608.03941
作者: Arslan Battalov,Karim Kramin,Alexander Markotenko,Sofia Sinitsina
类目: Machine Learning (cs.LG)
*备注: 15 pages, 25 figures, 8 tables
Abstract:Muon is a recent optimizer that orthogonalizes the update to each weight matrix with a Newton-Schulz iteration, which performs steepest descent under the spectral norm. Almost all the evidence for it comes from Transformer models, and its behavior on state-space models is largely unreported. We compare Muon with AdamW on Mamba-2 130M under a controlled protocol that varies only which weight groups are trained with Muon. The benefit is localized. Muon on the output projection alone beats Muon on the input projection or on both. The advantage is mainly one of token efficiency. It holds on two corpora and two token budgets, and persists when training continues well past the compute-optimal point. Conditioning does not explain the gain. Muon lowers the condition number of whichever projection it trains, but the better-conditioned input projection is not the one that helps.
[LG-2] A Physics-Flavored Transformer Network for Parametrizing Contraction Dynamics of Engineered Skeletal Muscle Tissues
链接: https://arxiv.org/abs/2608.03927
作者: Mattias Luber,Timo Betz
类目: Machine Learning (cs.LG)
*备注:
Abstract:Engineered Skeletal Muscle Tissues (ESMs) have become a key structure for biomedical disease modeling and pharmacological screening, yet their functional characterization often relies on simplistic metrics like peak force, discarding critical kinetic information. This is partially due to the high level of mathematical complexity which mechanistic models introduce to capture these dynamics. Hence, exactly the complexity prevents scalable application and widespread adaptation in the field. Here we present a Physics-Flavored Neural Network (PFNN) that automates the kinetic phenotyping of ESMs. Our architecture integrates a stretched-exponential physical model into a CNN-Transformer, enabling the extraction of physically meaningful parameters directly from force-time profiles. To address the scarcity of labeled biological data, we employ a hybrid training paradigm: the model develops a “physical intuition” on synthetic data before undergoing unsupervised self-alignment on unlabeled real-world measurements. Our results demonstrate that this physics-flavored approach achieves high-fidelity parameterization across diverse contractile phenotypes and cell lines, including Duchenne Muscular Dystrophy models. Our scalable, self-improving pipeline bridges the gap between idealized biophysics and noisy \emphin vitro data, providing a robust tool for high-throughput biophysical research.
[LG-3] rajectory inference via Acceleration Matching
链接: https://arxiv.org/abs/2608.03916
作者: Bartolo Dazzini,Giovanni Conforti,Alain Durmus,Aram-Alexandre Pooladian
类目: Machine Learning (cs.LG); Optimization and Control (math.OC); Machine Learning (stat.ML)
*备注: Comments welcome!
Abstract:Trajectory inference is a fundamental problem in many scientific domains: given a collection of unpaired snapshots of observations at discrete time points, the goal is to generate smooth trajectories that best resemble and interpolate the data. Existing algorithms exhibit computational challenges: they either rely on preprocessing subroutines to enforce smoothness or on simulation-based training objectives, both of which can be expensive. In order to overcome these limitations, we propose a new algorithm called Acceleration Matching (\textttAM). Our approach consists of lifting the original interpolation problem to phase space and then regressing onto an explicit conditional acceleration field that induces random, smooth trajectories that agree with the prescribed marginals. Importantly, our resulting training algorithm only requires positional data, avoids trajectory simulation during training, and is devoid of expensive preprocessing. We provide ample numerical evidence suggesting that \textttAM is competitive with or superior to existing algorithms on several benchmark problems from the existing literature.
[LG-4] Cross-Model KV Cache Transfer in LLM Families: A Closed-Form Linear Mapping for Prefill Reuse
链接: https://arxiv.org/abs/2608.03893
作者: Taekyung Heo,Rasoul Shafipour,Ritchie Zhao,Maximilian Golub,Mohammad Mahdi Kamani,Ritika Borkar,Makesh Tarun Chandran,Pantea Zardoshti,Bita Darvish Rouhani
类目: Machine Learning (cs.LG)
*备注:
Abstract:Production deployments often swap between different-sized models in a family for cost-quality cascading, mid-conversation switching, and routing, and each swap forces the receiver to repay the prefill from scratch. We propose cross-model KV cache transfer, where the receiver reuses the source’s KV cache, skipping prefill. We find that cross-model KV has substantial linear structure across matched-KV pairs, where source and target share KV head count and per-head dimension. On Qwen3 14B-32B, one source layer explains 56% of variance in the target’s keys and 32% in values, rising to 79% and 65% with multiple source layers. Building on this, we design a closed-form ridge mapper that operates per head and proceeds in three steps. First, for each target layer we select the top-k most predictive source layers and concatenate their KV as input. Second, we strip RoPE from the keys before mapping, so the fit is position-free and reusable across context lengths. Third, we fit ridge regression on a small calibration set of 500 FineWeb-Edu sequences of 1,024 tokens each. Surprisingly, across six pairs in three families, this linear mapper retains 73-98% of the receiver’s standalone-prefill accuracy on four pairs, while two degrade sharply. A nonlinear MLP recovers up to +37 pp HellaSwag retention on the failures. The mapper runs 2.7-25x faster than re-prefill and remains stable across multi-turn handoff, making cross-model KV cache transfer practical.
[LG-5] Omega-S: A Functional Resilience Index for LLM Fine-Tuning
链接: https://arxiv.org/abs/2608.03887
作者: Alberto Acedo
类目: Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE); Molecular Networks (q-bio.MN)
*备注: 15 pages of main text plus appendices; 12 tables. Code, per-seed data and all negative results at this https URL
Abstract:Fine-tuning a large language model on new data degrades what it previously learned. We present Omega-S, a drop-in penalty computed from the weight matrix alone: it needs no previous-task data, no Fisher matrix and no stored copy of the old weights. It is three lines in an existing training loop and adds under 4% to the cost of a step. Retention. On Llama-3-8B with LoRA, fine-tuned from code to prose and measured by HumanEval over ten seeds, Omega-S retains more of the original capability than no regularisation on 9 of 10 seeds (0.173 - 0.238 absolute pass@1; sign test one-sided p=0.011, Wilcoxon p=0.006), as a retention ratio, 62.9% - 84.1%. It also beats tuned weight decay on 10 of 10 seeds (p=0.002) and tuned EWC on 8 of 10 (p=0.014), every arm re-measured in the same session. Mechanism, measured rather than asserted. Omega-S is topological by construction, its objective built from Tr(A^3), but we measured which of its four factors actually moves and three do not: their elasticity with respect to the weights is at or below 1e-4, against 9e-3 for the degree-variance term. As implemented, the composite reduces to a penalty on the variance of node degrees, which means row magnitude in square modules and directional alignment in non-square ones. We report this because a method whose name promises one thing and whose gradient does another should say so. We also enumerate the open design choices, including a contrast-preserving construction that does what it was designed to do and makes retention worse on all ten seeds. Repeating an identical configuration, same seed and same hardware, gives a standard deviation of 0.104 in retention ratio. We have not found this quantified for low-rank fine-tuning of language models, and it bounds every seed-paired comparison in this literature, ours included. Code, per-seed results and the full record of negative results are available. Comments: 15 pages of main text plus appendices; 12 tables. Code, per-seed data and all negative results at this https URL Subjects: Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE); Molecular Networks (q-bio.MN) Cite as: arXiv:2608.03887 [cs.LG] (or arXiv:2608.03887v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2608.03887 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Alberto Acedo [view email] [v1] Tue, 4 Aug 2026 16:22:30 UTC (60 KB) Full-text links: Access Paper: View a PDF of the paper titled Omega-S: A Functional Resilience Index for LLM Fine-Tuning, by Alberto AcedoView PDFHTML (experimental)TeX Source view license Current browse context: cs.LG prev | next new | recent | 2026-08 Change to browse by: cs cs.NE q-bio q-bio.MN References Citations NASA ADSGoogle Scholar Semantic Scholar export BibTeX citation Loading… BibTeX formatted citation loading… Data provided by: Bookmark checked="checked"class=“labs-tab-input”> Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, Media Code, Data and Media Associated with this Article alphaXiv Toggle alphaXiv (What is alphaXiv?) Links to Code Toggle CatalyzeX Code Finder for Papers (What is CatalyzeX?) DagsHub Toggle DagsHub (What is DagsHub?) GotitPub Toggle Gotit.pub (What is GotitPub?) Huggingface Toggle Hugging Face (What is Huggingface?) ScienceCast Toggle ScienceCast (What is ScienceCast?) Demos Demos Replicate Toggle Replicate (What is Replicate?) Spaces Toggle Hugging Face Spaces (What is Spaces?) Spaces Toggle TXYZ.AI (What is TXYZ.AI?) Related Papers Recommenders and Search Tools Link to Influence Flower Influence Flower (What are Influence Flowers?) Core recommender toggle CORE Recommender (What is CORE?) IArxiv recommender toggle IArxiv Recommender (What is IArxiv?) Author Venue Institution Topic About arXivLabs arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv’s community? Learn more about arXivLabs. Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?) mathjaxToggle(); We gratefully acknowledge support from our major funders, member institutions, , and all contributors. About Help Contact Subscribe Copyright Privacy Accessibility Operational Status (opens in new tab) Major funding support from
[LG-6] Operationally Feasible Synthetic Power-Grid Scenarios via Learning the AC-Operable Joint Distribution
链接: https://arxiv.org/abs/2608.03878
作者: Chenhan Xiao,Xinyu He,Haoran Li,Hanghang Tong,Yang Weng
类目: Machine Learning (cs.LG); Systems and Control (eess.SY)
*备注: 10 pages, 10 figures, journal submission
Abstract:Synthetic power-grid scenarios are essential for planning, resilience assessment, contingency analysis, and data-driven power-system applications. Recent synthetic grid generation methods have improved structural realism and operational feasibility by incorporating engineering knowledge through post-generation validation, optimization, or physics-aware generation. However, generated scenarios may still exhibit low AC feasibility and robustness, limiting their practical value for downstream power-system studies. This paper proposes a feasibility-aware distribution-learning framework that learns the AC-operable joint distribution of network topology, branch electrical parameters, and time-varying load profiles. Instead of enforcing feasibility after generation, the proposed framework incorporates AC power-flow convergence and operational constraints into hierarchical diffusion-based distribution learning. This enables the generator itself to produce operationally feasible grid scenarios through efficient diffusion sampling. The hierarchical architecture decomposes the high-dimensional generation task into three engineering-motivated stages: topology and bus-attribute generation, branch-parameter generation conditioned on the generated structure, and load-profile generation conditioned on both network structure and electrical characteristics. Experiments on benchmark systems demonstrate that the proposed framework significantly improves operational feasibility and contingency robustness while maintaining strong statistical fidelity and eliminating optimization-based post-processing.
[LG-7] CRS-Triage: Confidence- and Reliability-Aware Selective Triage under Incomplete Clinical Evidence
链接: https://arxiv.org/abs/2608.03862
作者: Guan Qiang,Yushen Chen,Tianlong Liu,David Rotenberg,Ethan H.Kim,Fang Fang
类目: Machine Learning (cs.LG)
*备注:
Abstract:Emergency triage requires reliable decisions within a short time period. However, the available electronic health record (EHR) data, including structured data and clinical text, are often incomplete, unreliable, and inconsistent. This makes machine learning (ML)-based triage prediction more challenging, as existing ML models typically rely on complete and reliable EHR data to accurately predict patients’ acuity levels. To address this, we propose confidence- and reliability-aware selective triage (CRS-Triage) to predict patients’ acuity levels with a confidence score. By comparing the confidence score with a predefined threshold, CRS-Triage can selectively determine whether the model should make the decision or defer the case. Specifically, CRS-Triage separately evaluates the reliability of structured data and clinical text and then jointly considers the consistency between the two modalities to estimate the confidence of each prediction. Moreover, to reduce the risk of missing high-acuity patients, namely under-triage, CRS-Triage prefers to assign patients slightly higher acuity levels, namely over-triage, by penalizing under-triage errors. Experiments on the MIMIC-IV-ED dataset show that CRS-Triage achieves strong predictive performance. It also provides a better risk-coverage trade-off and remains reliable when the available EHR data are incomplete, degraded, or inconsistent across modalities.
[LG-8] Bi-semantic Chemical Embedder for Joint Representation Learning of SMILES and Natural Language
链接: https://arxiv.org/abs/2608.03855
作者: David Ming Segura,Jeremy Goumaz,Joshua W. Sin,Bojana Ranković,Philippe Schwaller
类目: Machine Learning (cs.LG)
*备注:
Abstract:Transformer models have revolutionized natural language processing (NLP), and text-based molecular representations like SMILES have successfully extended these architectures to chemistry. However, domain-adaptive pre-training often causes models to overfit to chemical syntax, catastrophically forgetting their foundational semantic capabilities. To address this challenge, we introduce CheMatE, a chemistry-oriented embedding model that jointly captures molecular structure and domain-specific natural language within the same representation space. Built on a ModernBERT backbone, CheMatE learns bi-semantic representations through a two-stage training procedure: continued masked language modeling (MLM) followed by a Matryoshka contrastive learning stage via Multiple Negative Ranking Loss (MNRL). First, we train the model using MLM on a novel, large-scale corpus of SMILES-annotated, long-context scientific documents that were constructed and curated from FineWeb and ChemPile (comprising 10.4B and 11.5B tokens, respectively). Subsequently, the model undergoes contrastive learning using a synthetic dataset of SMILES-text pairs algorithmically derived from our original training corpus. This design exposes the model to SMILES-enriched scientific literature, enabling bi-semantic understanding. We evaluate CheMatE across a range of downstream tasks covering molecular property prediction and scientific language understanding. Our results demonstrate that coupling our custom-curated datasets with this sequential training strategy yields robust, highly transferable representations. By effectively unifying structural and contextual signals within a single text-based framework, CheMatE achieves competitive performance across both specialized chemistry models and general-purpose language model baselines.
[LG-9] Quantization Effects on Biomedical LLM Reliability
链接: https://arxiv.org/abs/2608.03854
作者: Anton Rasmussen,Hong Qin
类目: Machine Learning (cs.LG)
*备注: 8 pages, 1 figure
Abstract:When decoder language models are used as classifiers, predicted class probabilities depend on implementation choices, including the prompt template, verbalizer (label-to-token mapping), and scoring rule, that are rarely treated as experimental variables. We present a controlled evaluation of three Mistral-7B variants (Base, BioMistral, and Instruct) on PubMed RCT sentence classification (n=2000) under FP16, INT8, and INT4 precision using four answer-text prompt templates. Our primary finding is that the probability extraction protocol dominates apparent calibration. Switching from summed to mean token log-likelihood scoring reverses the calibration ranking between models: BioMistral average expected calibration error increases from 0.097 to 0.289, whereas Instruct decreases from 0.237 to 0.096, while accuracy changes by less than 1 percentage point for the specialized models but 4-6 percentage points for the base model. Prompt template choice produces accuracy differences of 7-24 percentage points, comparable to or larger than model-level effects. On one template, BioMistral outperforms Instruct although the overall mean favors Instruct by only 1.3 percentage points. For BioMistral and Instruct, INT8 quantization changes accuracy and F1 by only 1-2 percentage points relative to FP16, whereas the base model shows larger INT8 effects on some templates (up to +4.2 percentage points). INT4 produces heterogeneous but non-catastrophic effects. Temperature scaling reduces expected calibration error under summed scoring for both models but only for that scoring rule. A fine-tuned PubMedBERT reference achieves 82.7% accuracy but uses about 176000 labeled training examples, precluding direct comparison. These results demonstrate that prompt template design and scoring normalization are first-order experimental decisions when evaluating decoder language model calibration.
[LG-10] Resume Means Resume: A Machine-Checked Conformance Contract for Checkpoint Interrupt and Resume Semantics in Workflow Persistence Layers
链接: https://arxiv.org/abs/2608.03836
作者: Sajjad Khan
类目: Machine Learning (cs.LG); Distributed, Parallel, and Cluster Computing (cs.DC); Logic in Computer Science (cs.LO); Software Engineering (cs.SE)
*备注: 26 pages, 11 tables, 1 figure. Supplementary material included as an ancillary file
Abstract:A framework that persists execution state so a run can be interrupted, survive a crash, and continue must decide what a resume means for effects that already fired. Five widely deployed agent workflow frameworks answer differently, none exposes a machine-checkable contract, and behavior violates even the fragments they state. The RESUME CONTRACT states six properties over the persistence API (prefix continuation, effect exactly-once, fork determinism, checkpoint validity, consume-once, recovery determinism), plus fork-intent and liveness obligations. A TLA+ model checks a reference semantics exhaustively, unchanged at scaled bounds (7.4 million states); a 39-cell fault matrix yields the separating models independence requires, and consume-once splits, its consumption clause independent of all six others. A deterministic, LLM-free harness measures them at pinned releases. LangGraph 1.2.9 durably records a second resume value and never consults it, persists schema-invalid state silently, and re-executes durably recorded work after a real SIGKILL: exactly-once across interrupts, at-least-once across crashes, on one API. CrewAI 1.15.2 re-executes completed effect-bearing methods against its written claim; pydantic-graph 1.x cannot resume after a mid-node crash; no two probed frameworks share a conformance profile. Consume-once holds sequentially and fails under concurrent delivery: k processes resuming one parked interrupt fire the gated effect k times, saturation 1.0 in 36 of 40 cells, and the failure crosses hosts. REMIT, a reference sequencer whose Verus-verified recovery core is line-identical to the shipped executable, repairs the fork and validity cells. The cross-process cell is repaired at the read path, and that repair ships: an opt-in gate claims consumption in the shared store, serving one racer and refusing the rest before any node executes.
[LG-11] UNVaMP: Neural Knowledge Tracing with Variational Regularization of Latent Knowledge Dynamics
链接: https://arxiv.org/abs/2608.03811
作者: Carson J. Cook,Ahmed J. Zerouali,Anthony Schmidt,Reginald Ziedzor,Paul Lin,Luke G. Eglington
类目: Machine Learning (cs.LG); Computers and Society (cs.CY)
*备注: 12 pages, 4 figures, Proceedings of the 19th International Conference on Educational Data Mining , Seoul, Republic of Korea, June-2026
Abstract:We introduce the Unified Neural Variational Measurement of Proficiency (UNVaMP) architecture, a knowledge tracing method that integrates observed student-item interactions with internal memory to produce evolving latent representations of student knowledge. These representations support accurate predictions of future responses while enabling explicit control over the smoothness of estimated learning trajectories. UNVaMP can be configured as either a purely neural model or a hybrid model that predicts responses through an interpretable measurement function over the latent space. We show that a pure neural configuration (UNVaMP-MLP) achieves the strongest predictive performance among compared models on three out of four datasets. Meanwhile, a hybrid configuration (UNVaMP-MIRT, using a 1PL MIRT measurement function) lags only slightly behind UNVaMP-MLP, indicating that the predictive cost of interpretability is modest. Beyond predictive accuracy, UNVaMP provides the following: a principled mechanism for controlling volatility when estimating student latent variables, quantification of uncertainty over student knowledge state estimates, and flexible input specification that supports heterogeneous student-item interaction features. In addition, the hybrid UNVaMP-MIRT configuration generates interpretable moment-in-time student knowledge state estimates. Using an experimental dataset, we show that auxiliary inputs induce structured changes in the predictive behavior of UNVaMP-MIRT, consistent with sensitivity to underlying structure beyond response correctness. Furthermore, through a simulation study, we show that UNVaMP yields well-behaved knowledge state estimates under controlled measurement conditions. In total, these results indicate that UNVaMP is both useful for real-world education systems and capable of recovering underlying structure from student-item interactions. Comments: 12 pages, 4 figures, Proceedings of the 19th International Conference on Educational Data Mining , Seoul, Republic of Korea, June-2026 Subjects: Machine Learning (cs.LG); Computers and Society (cs.CY) Cite as: arXiv:2608.03811 [cs.LG] (or arXiv:2608.03811v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2608.03811 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Journalreference: Proceedings of the 19th International Conference on Educational Data Mining, 353-364, 2026 Related DOI: https://doi.org/10.5281/zenodo.21040074 Focus to learn more DOI(s) linking to related resources
[LG-12] Amortized Interventional Forecasting for Multivariate CIR Processes
链接: https://arxiv.org/abs/2608.03715
作者: Andreas Sauter,Sumit Sourabh,Drona Kandhai,Erman Acar
类目: Machine Learning (cs.LG); Computational Engineering, Finance, and Science (cs.CE)
*备注:
Abstract:Mean-reverting dynamics are pervasive in finance, and the Cox–Ingersoll–Ross (CIR) process is a standard model for the time series they produce, from short rates to credit default swap (CDS) spreads. Yet CIR models capture only \emphcorrelated co-movement, not \emphcausal influence between series, so they cannot answer the system’s response when one series is externally shocked, which observational conditionals confound with historical co-movement. We make two contributions. First, an amortized model for distributional causal effect estimation that frames trajectories as time-stamped observations and predicts the calibrated multi-horizon shock response without retraining per scenario. Second, a causal multivariate CIR data-generating process that supplies the paired observational and interventional ground truth that real markets cannot. We instantiate and calibrate the framework on CDS spreads as a testbed. CIR-ACTIVA’s validity is established on synthetic ground truth, independent of how well the simulator matches reality, while practical grounding is assessed by backtesting the generated traces against real CDS data. Against observational and amortized causal-inference baselines, CIR-ACTIVA leads on both causal selectivity in the joint distribution and horizon-resolved calibration, retaining its selectivity once the interventional law varies over the horizon, with gains concentrating at short horizons. This opens up a class of what-if queries on coupled spread systems, CDS stress testing among them, that observational forecasters cannot answer.
[LG-13] o Describe or Construct Statistical Learning Models Using the Category-theoretical Language
链接: https://arxiv.org/abs/2608.03706
作者: Congwei Song
类目: Machine Learning (cs.LG)
*备注:
Abstract:Statistical learning is a fascinating field that has long been the mainstream of machine learning/artificial intelligence. A large number of results have been produced which can be widely applied to real-world problems. It also leads to many research topics and also stimulates new research. This report summarizes some classical statistical learning models and well-known algorithms, especially for amateurs, and provides a category-theoretic perspective on understanding statistical learning models. The aim is to attract researchers from other fields, including basic mathematics, to participate in the research related to statistical learning.
[LG-14] Learning and Clustering on Temporal Graphs: Principles Primitives and Pooling KDD2026 ECML
链接: https://arxiv.org/abs/2608.03696
作者: Nelson Aloysio Reis de Almeida Passos,Emanuele Carlini,Salvatore Trani
类目: Machine Learning (cs.LG); Social and Information Networks (cs.SI)
*备注: 4 pages, 1 figure. Accepted at ECML PKDD 2026 (Nectar Track)
Abstract:This work focuses on the problem of learning on temporal graphs, with particular emphasis on the task of clustering: obtaining coarse-grained representations by aggregating information from nodes, edges, and temporal dynamics - a task related to pooling in machine learning on graphs, or community detection in network science. Although graph neural networks reach state-of-the-art performance across many downstream graph tasks, their advantage over established descriptive and inferential clustering algorithms is far less settled, especially under demands of efficiency and recovery accuracy. We frame this tension through three linked perspectives: principles, connecting graph learning and community detection through shared spectral foundations and detectability thresholds in stochastic block model regimes; primitives, making spectral clustering and multislice modularity optimization tractable through GPU-accelerated temporal backends; and pooling, viewing principled community detection as a theory-grounded coarse-graining operator for temporal graphs. Our results indicate that algorithmic methods remain the appropriate tool where attributes are absent or weak - scalability rather than accuracy being the binding obstacle - while neural models are most compelling when structural, temporal, and attribute signals align. By making temporal clustering scalable, GPU-accelerated primitives suggest a route toward theory-grounded pooling, while raising a central question: when does community-based coarse-graining preserve the dynamics needed for downstream learning tasks?
[LG-15] Accelerating Dynamic Graph Clustering on GPU Architectures with cuGraph
链接: https://arxiv.org/abs/2608.03695
作者: Nelson Aloysio Reis de Almeida Passos,Emanuele Carlini,Salvatore Trani
类目: Distributed, Parallel, and Cluster Computing (cs.DC); Machine Learning (cs.LG)
*备注: 12 pages, 2 figures. Accepted at FRAME 2026, Euro-Par 2026 Workshops; to appear in Springer LNCS
Abstract:This work addresses community detection in temporal networks through GPU-accelerated extensions of spectral clustering and modularity-based algorithms originally designed for static graphs. Built on the NVIDIA RAPIDS ecosystem, the framework enables the characterization and tracking of communities in snapshot-based dynamic graphs, either by Leiden greedy optimization with multi-GPU support via Dask-based workload distribution, or eigendecomposition of a symmetric Bethe-Hessian operator. Our multislice modularity backend achieves up to roughly three orders of magnitude speedup over the CPU reference under an equal-work budget, depending on graph density and snapshot count, while preserving compatibility with existing graph analytics pipelines. We demonstrate its applicability on real-world and synthetic datasets, facilitating exploratory analysis of structural network properties over time. Such capabilities are relevant across several application domains, such as epidemic spreading, financial systems, cybersecurity, and trajectory and mobility analysis. We release our implementation as free and open-source software, including Python bindings through the NetworkX-Temporal library for ease of use and zero-code acceleration with existing codebases.
[LG-16] LAEF: A Lead-Agnostic ECG Foundation Model Towards Point-of-Care Diagnostics
链接: https://arxiv.org/abs/2608.03690
作者: Edoardo Coppola,Stefano Fiorini,Pietro Liò,Mattia Savardi,Alberto Signoroni
类目: Machine Learning (cs.LG)
*备注:
Abstract:Point-of-care cardiac devices such as smartwatches and handheld ECG recorders typically capture 1–2 leads, yet existing ECG foundation models are architecturally constrained to fixed 12-lead inputs, degrading or failing under these reduced configurations. We introduce LAEF (Lead-Agnostic ECG Foundation), a 7M-parameter ECG foundation model that can natively process any lead subset without zero-padding or architectural modification. LAEF represents ECGs as variable-size spatiotemporal graphs with physiologically motivated intra- and inter-lead connectivity, processed by a Graph Attention Network that scales naturally with active lead this http URL-trained on 9.2M 12-lead ECGs via masked node modelling with stochastic lead sampling, LAEF learns representations robust to lead configuration. Across 18 downstream datasets, LAEF is on par with specialized 12-lead baselines over 12 \times larger at full lead availability. Under direct point-of-care-oriented diagnostics (1–2 leads), it outperforms all zero-padded alternatives on 17 out of 18 datasets with with a single randomly sampled lead and on 14 out of 18 with 2 leads, with an average AUROC gain of +3.2 points. Representation analysis links this advantage to architectural lead-agnosticism, and a lead-importance study across 164 cardiovascular conditions shows population-level performance is stable across single standard input leads while still recovering established clinically lead-condition associations.
[LG-17] DiagLoop: A Counterfactual Data Flywheel with Stage-Localized Reinforcement for Diagnostic LLM s
链接: https://arxiv.org/abs/2608.03674
作者: Jian Zhang,Bingyi Wang,Yizhi Liu
类目: Machine Learning (cs.LG)
*备注: 9 pages, 2 figures
Abstract:Causal diagnostic models must explain how conclusions follow from evidence because diagnoses guide repairs and treatments. Yet serious cases are scarce, records rarely contain reasoning paths, and data transfer poorly across configurations, complicating local deployment. We present DiagLoop, a counterfactual data flywheel that converts codified physical relations or clinical guidelines, authored once per mechanism family, into training supervision beyond recorded cases. A training-only teacher proposes counterfactual worlds by varying causes, contexts, and observations, while an independent hybrid checker admits only valid worlds. The student reasons through symptom abstraction, causal-chain construction, and root-cause attribution. Stage-specific criteria identify its earliest failure. For nonterminal failures, a bounded repair probes downstream competence, and the resulting weakness profile guides subsequent data generation. Stage-localized reinforcement learning updates only the model-generated continuation, while replay and preservation reduce forgetting. The same criteria govern admission, attribution, reward, and regeneration through checks separate from the proposer. Using only synthesized scenarios and no case-level expert reasoning annotations, the resulting 8B model improves strict path correctness over the strongest conventional baseline. Gains are 11.6 points across eight industrial systems and 5.5 points across ten disease categories. Gains over a deranged-routing control are 3.9 and 2.3 points, respectively. The model also exceeds the evaluated proprietary references in both domains, even when they receive few-shot examples or the specification in context.
[LG-18] CausalOPD: First-Wrong-Step Supervision for Distilling Causal Chain Reasoning
链接: https://arxiv.org/abs/2608.03673
作者: Jian Zhang,Bingyi Wang,Yizhi Liu
类目: Machine Learning (cs.LG)
*备注: 9 pages, 2 figures
Abstract:Many critical reasoning tasks, including clinical diagnosis, legal judgment, and industrial fault diagnosis, require step-dependent causal chains in which early errors propagate and correct conclusions can mask invalid reasoning. Although large language models perform well on such tasks, privacy, latency, and controllability motivate distillation into locally deployable models. Standard trajectory imitation does not correct process errors on the student’s own rollout distribution. We propose CausalOPD, a curriculum online process distillation framework. A knowledge-augmented teacher first provides trajectories grounded in domain-specific causal rules, entity relations, and structural constraints. The student then generates on-policy trajectories, and the teacher identifies the first wrong step, defined as the earliest transition that verifiably violates available constraints. Starting from the verified prefix, short-horizon reinforcement learning repairs this localized failure. A causal-stage curriculum advances from evidence-level to mechanism-level and conclusion-level errors, following their propagation order. Across three domains, CausalOPD improves average path correctness by 23.4 percentage points over sequence-level online process distillation and reduces the right-label-wrong-reasoning rate from 15.7% to 4.4%. The domain-specific 8B students also surpass both evaluated proprietary references in path correctness across all domains.
[LG-19] POEM: Phase-Aware mathrmSO(2) Feature Rotation for Time Series Forecasting Under Periodicity Drift
链接: https://arxiv.org/abs/2608.03630
作者: Jiawen Zhu,Shuhan Liu,Shengxuan Li,Qiming Shi,Di Weng
类目: Machine Learning (cs.LG)
*备注: 9 pages, 5 figures
Abstract:Deep learning has advanced time series forecasting, but periodicity drift, in which cycle timing and phase vary over time, remains a challenging problem. Existing methods predominantly model these sequences on fixed time grids, suffering from a limited ability to accommodate phase-related variation. To address this limitation, we propose \textbfPOEM, a phase-aware forecasting framework based on latent feature rotation using the special orthogonal group in two dimensions, denoted by \mathrmSO(2) . POEM aims to reduce the phase-related variability by learning a phase-correction coordinate and applying an invertible \mathrmSO(2) -based rotation to paired latent features. To extrapolate this correction coordinate, Directional Phase Increment Attention (DPIA) retrieves historical phase increments from similar temporal contexts and integrates them into future phase corrections. Experiments demonstrate that POEM achieves competitive performance, while qualitative visualizations suggest that the learned phase-aware transformation makes latent trajectories more regular.
[LG-20] ConformalShift: Targeted Event Reordering Against Adaptive ECG Monitoring
链接: https://arxiv.org/abs/2608.03628
作者: Arash Vashagh,Yasmin Vashagh
类目: Machine Learning (cs.LG)
*备注:
Abstract:Adaptive conformal prediction can recover clinically important heartbeat classes missed by a point classifier, but delayed feedback makes its decisions sensitive to event order. We introduce ConformalShift, a bounded event-reordering attack that suppresses the ventricular class for rescued events without modifying ECG waveforms, labels, classifier scores, or the event multiset. ConformalShift searches for feasible permutations of authentic preceding events that lower the ventricular threshold before a selected target is evaluated. On disjoint MIT–BIH confirmation records, the attack suppressed 66.7% of eligible targets for Extra Trees and 60.0% for HistGradientBoosting, compared with random-schedule rates of 4.4% and 12.0%, respectively. Transferred configurations also outperformed random scheduling on INCART, while reducing the displacement budget weakened the attack on both datasets. These results show that adaptive monitors in healthcare can be compromised through the timing of authentic information, even when waveforms, labels, classifier outputs, and event contents remain unchanged.
[LG-21] Design-Time Optimization of Deep Neural Networks for Intermittent Learning on Microcontrollers KDD2026 ECML
链接: https://arxiv.org/abs/2608.03589
作者: Jakob Schubert,Maximilian Kasper,Maximilian Linke,Benedict Herzog,Mark Deutel,Axel Plinge,Dominik Seuss,Christopher Mutschler
类目: Machine Learning (cs.LG)
*备注: Accepted at the 7th Workshop on IoT, Edge, and Mobile for Embedded Machine Learning (ITEM) collocated with ECML PKDD 2026, 12 pages, 5 figures, 1 table,
Abstract:We present a method for designing deep neural networks (DNNs) for intermittent, energy-autonomous, on-device learning on microcontroller units (MCUs). In mobile applications where the energy can run out, e.g., when solar-powered, executing artificial intelligence (AI) faces a technical issue as learning can be interrupted at any time. Our approach combines a hardware-aware energy prediction model with multi-objective optimization (MOO), enabling offline DNN optimization at the design stage without repeated deployment and online testing on the target MCU. Our proposed energy predictor estimates per-layer energy consumption for both DNN inference and training, including the intermittent checkpointing overhead, based on implementation-specific compute and memory features extracted from the DNN model. We validate our approach using autoencoders for anomaly detection on a Cortex-M4 MCU, where our predictor achieves a weighted absolute percentage error of 16.6%, which is sufficient for reliable architecture selection under intermittency constraints. As a result, this work bridges the gap between MOO, automated DNN design, deployment on energy-harvesting systems, and intermittent learning, truly enabling autonomous AI at the edge.
[LG-22] Robust General Utility for Reinforcement Learning
链接: https://arxiv.org/abs/2608.03562
作者: Zixuan Liu,Fangzheng Wu,Brian Summa,Zizhan Zheng
类目: Machine Learning (cs.LG)
*备注:
Abstract:Reinforcement learning (RL) with general utility extends classic RL by optimizing an arbitrary utility functional of the policy-induced occupancy measure, thereby enabling a broader range of applications. However, previous work on general utility RL typically assumes the evaluation utility is fixed and correctly specified. In practice, the utility used at deployment can deviate from the training one, creating a robustness gap that prior work does not address. Motivated by this, we propose robust general-utility RL, a minimax learning framework that trains policies against utility misspecification within a prescribed uncertainty set. Our framework strictly generalizes standard general-utility RL while also providing a unified view of many existing RL frameworks, including reward-robust RL and constrained RL, through appropriate choices of the utility uncertainty set. We further develop provably convergent stochastic algorithms for two regimes. For concave utilities, we develop a projected stochastic gradient descent-ascent method and establish stationarity guarantees. For the more challenging nonconcave regime, we propose a stochastic prox-extragradient algorithm that mitigates ill-posed behavior induced by nonconcavity, with convergence guarantees to approximate first-order stationarity. Experiments on LLM safety alignment and exploration maximization tasks further corroborate the convergence behavior consistent with our theory.
[LG-23] FedCARE: A Multi-Objective Personalised Federated Learning Framework for Smart Healthcare
链接: https://arxiv.org/abs/2608.03498
作者: Rojalini Tripathy,Padmalochan Bera,Shreya Ghosh,Rajkumar Buyya
类目: Machine Learning (cs.LG); Distributed, Parallel, and Cluster Computing (cs.DC)
*备注:
Abstract:Federated Learning (FL) enables collaborative model training across distributed healthcare institutions without centralising sensitive patient data. However, real-world healthcare federations are often characterised not only by non-IID data, but also by heterogeneous clinical objectives and partially overlapping feature spaces. Different hospitals may optimise distinct and potentially conflicting objectives, such as mortality risk prediction, readmission reduction, or length-of-stay estimation, while also retaining institution-specific clinical features that cannot be shared with other participants. Existing personalised FL methods mainly address statistical heterogeneity, whereas multi-objective FL approaches typically learn a shared global model without explicit client-level adaptation. To address these limitations, we propose \textbfFedCARE, a multi-objective personalised FL framework for smart healthcare services. FedCARE follows a two-stage training strategy. First, it learns a shared global backbone from common clinical features using Pareto-driven multi-objective federated optimisation. Second, each client independently fine-tunes the shared backbone using its private features and local clinical objectives, enabling institution-specific personalisation without additional communication overhead. We implement FedCARE in a cloud-based client-server federated deployment on the Melbourne Research Cloud and evaluate it on two real-world healthcare datasets, MIMIC-III and Diabetes 130-US Hospitals. Experimental results show that FedCARE consistently outperforms standard FL, multi-objective FL, and personalised FL baselines, achieving up to 12.5% AUROC improvement and 32.0% MAE reduction over FedAvg.
[LG-24] Beyond the Gegenbauer Paradigm: q-Orthogonal Kernels for Machine Learning
链接: https://arxiv.org/abs/2608.03482
作者: Álvaro Sánchez-Paniagua Ríos,Juan P. Llerena,Alberto Lastra,Nuria Torrado,Edmundo J. Huertas
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: 27 pages, 4 figures, 9 tables. Source code and experimental pipeline available at this https URL
Abstract:The performance of Support Vector Machines (SVMs) critically depends on the kernel function choice, which enables implicit mapping of data into high-dimensional feature spaces. While classical kernels like Radial Basis Function (RBF) remain popular, orthogonal polynomial kernels offer mathematically interpretable alternatives that can incorporate structured prior knowledge. This work extends the orthogonal polynomial kernel paradigm by introducing a novel family based on discrete q -Hermite I polynomials, a class of q -orthogonal polynomials that generalize classical Hermite polynomials through a deformation parameter q . We formally define the q-Hermite kernel and establish its validity under Mercer’s theorem. The kernel’s inherent boundedness properties naturally prevent annihilation and explosion effects without requiring explicit scaling mechanisms. Extensive experiments across 20 benchmark datasets demonstrate that the proposed kernel achieves competitive performance compared to both classical kernels and other orthogonal polynomial kernels, while offering advantages in numerical stability and computational simplicity. Our results confirm that q -orthogonal polynomials constitute a promising direction for kernel design, bridging mathematical elegance with practical machine learning applications, that provides conceptual and algorithmic resources that may be further extended to emerging quantum computing paradigms. To facilitate full reproducibility, we provide the complete implementation and experimental pipeline in an open-access GitHub repository at this https URL.
[LG-25] Quality Control Algorithms for Pattern Counting
链接: https://arxiv.org/abs/2608.03439
作者: Cassandra Marcussen,Ronitt Rubinfeld,Madhu Sudan
类目: Data Structures and Algorithms (cs.DS); Machine Learning (cs.LG); Combinatorics (math.CO); Probability (math.PR)
*备注: 60 pages, 5 figures
Abstract:In recent work, Marcussen, Rubinfeld, and Sudan introduced the notion of quality control problems, which aim to capture the task of determining if a given input is truly random. Formally, their goal is to accept typical inputs from the specified distribution while rejecting every input whose value of a specified statistic is far from the distributional baseline. This captures the empirical practice of using specified statistics as a proxy for the quality of randomness. Empirical algorithms, however, have not exploited the asymmetry in the definition of quality control problems, which require soundness guarantees in the worst-case while only seeking average-case completeness. Their work abstracted a problem definition emphasizing this asymmetry and used it to give efficient quality control algorithms for assessing the randomness of graphs. In this work, we introduce and study quality control problems over sequences, where the goal is to distinguish a sequence of i.i.d. characters from sequences where some specified pattern appears too often (or too infrequently) as a subsequence. We consider this problem in both the finite-alphabet setting and for real-valued sequences. We refer to the former setting as the pattern counting problem. In the latter case, the natural notion of a pattern is to consider the relative ordering of the characters in the subsequence, and we refer to this as the permutation pattern counting problem. Algorithms to approximately count (permutation) patterns of length k in a worst-case sequence of length n can provably require exponential in k queries into the sequence. In contrast, we show that by taking advantage of the asymmetry in the definition of quality control, we give algorithms that run in poly (k) time to solve these problems. We also prove that any quality control algorithm (over some natural distributions) requires superlinear queries in k . Comments: 60 pages, 5 figures Subjects: Data Structures and Algorithms (cs.DS); Machine Learning (cs.LG); Combinatorics (math.CO); Probability (math.PR) Cite as: arXiv:2608.03439 [cs.DS] (or arXiv:2608.03439v1 [cs.DS] for this version) https://doi.org/10.48550/arXiv.2608.03439 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-26] FedRings: A Scalable and Topology-Aware Federated Learning Framework for LEO Satellite Constellations
链接: https://arxiv.org/abs/2608.03436
作者: Ziwu Liu,Inês Pinto Gouveia,Rehana Yasmin,Paulo Esteves-Verissimo,Ali Shoker
类目: Distributed, Parallel, and Cluster Computing (cs.DC); Machine Learning (cs.LG)
*备注:
Abstract:Federated learning over low Earth orbit (LEO) satellite networks is limited by frequent link changes, short contact times, and a highly dynamic topology, making centralized or synchronized training inefficient and hard to scale. To address this, we propose FedRings, a decentralized framework that organizes satellites into ring-based communication structures. It uses a spatio-temporal routing strategy with link-aware communication scheduling to align model exchange with actual visibility windows and time-varying connectivity patterns in LEO. Model updates are propagated along the ring using adaptive sparse incremental aggregation, which reduces communication overhead by progressively combining and compressing updates. To handle communication interruptions, a historical compensation mechanism maintains training continuity. By combining topology-aware routing, communication scheduling, and efficient aggregation, FedRings enables stable and efficient learning in dynamic LEO networks while reducing communication cost, and experiments show it consistently outperforms existing methods in realistic settings.
[LG-27] meRLM: Recursive Language Models Enable Precise Anomaly Localization in Long-Context Time-Series
链接: https://arxiv.org/abs/2608.03391
作者: Nicolas Zumarraga,Lorenzo Steno,Ning Wang,Max Rosenblattl,Thomas Kaar,Maxwell A. Xu,Kevin O’Sullivan,Markus Kreft,Elgar Fleisch,Paul Schmiedmayer,Patrick Langer,Robert Jakob
类目: Machine Learning (cs.LG)
*备注: Open source code and datasets: this https URL
Abstract:Precise anomaly localization over long-context time series is a crucial task in monitoring applications across clinical care, industrial operations, financial services, and logistics, where brief evidence may hide inside long spans of high-frequency data. Time-Series Language Models (TSLMs) are able to ingest time series data and verbalize findings on anomalies in natural language; however, recent benchmarks report a decrease in retrieval performance at long contexts, mirroring failure modes in text, vision, and audio. In the text domain, Recursive Language Models (RLMs) can recover much of this lost performance by keeping context external to the large language model (LLM), allowing the model to query it through code. We present TimeRLM, an RLM formulation for time-series that sequentially manipulates the signal using code and vision capabilities. We further introduce AnomalyXL, a synthetic long-context anomaly localization benchmark with programmatically injected anomalies that require precise retrieval. We implement five different task categories and two variants: AnomalyXL-MCQ and AnomalyXL-Localize. TimeRLM outperforms every evaluated TSLM and single-pass baseline on four of the five AnomalyXL-Localize tasks, reaching 0.682 IoU on localization and 0.745 on classify-with-evidence, versus at most 0.329 and 0.072 across all baselines. We post-train TimeRLM using reinforcement learning. The resulting model further improves performance and requires approximately one-third as many agent interaction turns as its untrained base model to produce a final answer. On unseen real-world ECG, sleep and software observability recordings, the post-trained TimeRLM retains or improves performance, surpassing TSLMs despite being trained exclusively on synthetic data. Our findings suggest recursive interaction with time-series is an effective approach for long-horizon retrieval.
[LG-28] Benign interpolation and Occams razor
链接: https://arxiv.org/abs/2608.03386
作者: Tom F. Sterkenburg,Daniel A. Herrmann,Jan-Willem Romeijn
类目: Machine Learning (cs.LG)
*备注:
Abstract:Contemporary deep learning methods generalize well even when they fit their training data perfectly, a phenomenon known as benign interpolation. This phenomenon cannot be accounted for by classical statistical learning theory and has prompted a range of attempted new explanations in the statistics and machine learning literature. A common feature of these new proposals is an appeal to a simplicity preference among interpolating models, often presented as a form of Occam’s razor. We clarify this debate for a philosophical audience and argue that this new appeal to simplicity creates an explanatory gap. The classical theory offers theorems which connect the simplicity of model classes to good generalization, thus underwriting methodological simplicity norms. The new accounts instead appeal to properties of individual models, which they interpret as a kind of simplicity. Lacking a provable connection to generalization, it is the name “simplicity” that does the work a theorem used to do, making a substantive and unargued assumption look like the application of a familiar methodological principle.
[LG-29] ght Worst-Case Bounds for the Smallest Eigenvalue of ReLU NTK Gram Matrices
链接: https://arxiv.org/abs/2608.03368
作者: Zhao Song
类目: Machine Learning (cs.LG)
*备注:
Abstract:For n unit vectors x_1,\ldots,x_n \in \mathbbR^d , we study the continuous ReLU derivative Gram matrix H , whose entries are obtained by averaging pairwise gated inner products over a standard Gaussian direction. Writing \Delta_\pm := \min_i \neq j \min\ |x_i-x_j|2, |x_i+x_j|2 \ for their projective separation, we prove the universal dimension-free lower bound \lambda\min(H) = \Omega( \Delta\pm/\sqrt\log n ) . Conversely, we construct worst-case families satisfying the matching upper bound \lambda_\min(H) = O( \Delta_\pm/\sqrt\log n ) , showing that this rate is tight up to universal constants.
[LG-30] AS-FedBridge: Pseudo-Spike Bridge Distillation for Heterogeneous ANN-SNN Federated Learning
链接: https://arxiv.org/abs/2608.03324
作者: Shengyang Li,Yiting Dong,Liuyang Song,Ximing Wang,Luyuan Xie,Cong Li,Qingni Shen,Zhaofei Yu
类目: Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE)
*备注:
Abstract:Federated learning enables collaborative model training across distributed edge devices while strictly preserving data privacy. To facilitate practical deployment on resource-constrained edge devices, Spiking Neural Networks (SNNs) have emerged as a promising alternative to traditional Artificial Neural Networks (ANNs) due to their sparse computing mechanisms and high energy efficiency. However, jointly training ANNs and SNNs exposes a challenge of representational misalignment, which is intrinsically caused by differences in information representation, specifically the semantic gap between continuous real-valued activations in ANNs and discrete spatio-temporal spikes in SNNs. To overcome this barrier, we propose AS-FedBridge, a novel federated learning framework tailored for mixed ANN-SNN clients. AS-FedBridge features a lightweight Bridge equipped with a Pseudo-Spike Interface, which effectively projects continuous signals into a spike-compatible space to facilitate ANN-SNN alignment. Given the absence of existing mixed ANN-SNN federated frameworks, we establish a comprehensive benchmark to evaluate against multiple advanced heterogeneous FL methods. Our empirical analysis demonstrates a positive correlation between the degree of ANN-SNN alignment and the collaborative FL performance. Across four datasets, AS-FedBridge consistently demonstrates advanced accuracy while mitigating extreme scale, architecture, and client heterogeneity challenge. Furthermore, our framework enables a highly controllable trade-off between model performance and resource efficiency. AS-FedBridge accomplishes these robust performance gains while introducing only marginal computational overhead, establishing a robust and practical foundation for mixed ANN-SNN federated learning systems.
[LG-31] Provably Learning Multi-Head Attention with Queries
链接: https://arxiv.org/abs/2608.03294
作者: Sunyeop Kim,Insung Kim,Jian Guo
类目: Machine Learning (cs.LG); Cryptography and Security (cs.CR)
*备注: 39 pages
Abstract:We study the problem of learning multi-head softmax attention from black-box input-output access. The learner may query arbitrary real-valued token sequences and observe only the scalar output at the final token. Recent work gives an algorithm using O(d^2) value queries to recover the single-head parameters (W,v) . For multiple heads, the same work establishes identifiability under the assumption that the heads occupy pairwise orthogonal subspaces. Applying the single-head recovery algorithm separately to the heads additionally requires bases for these subspaces to be known. We recover a canonical representation by merging heads with the same W_h , summing their corresponding v_h , and discarding a merged head when this sum is zero, without these subspace assumptions. By varying the number of copies of a token, our algorithm obtains samples of a rational function whose interpolation separates the canonical heads. Additional queries formed by adding selected token vectors then match the same head across different queries. When the oracle outputs and all subsequent computations are exact, the learner chooses its query vectors at random and recovers the canonical pairs (W_h,v_h):h\in[H]\ up to permutation with probability one. When H is known, it uses exactly 4Hd^2-2H+1 value queries of maximum length 2H+1 . If only a known upper bound H_0 is available, the algorithm uses 4H_0d^2-2H_0+1 value queries of maximum length 2H_0+1 . For approximate oracle outputs, we give conditions under which the parameter error is at most a model- and query-dependent constant multiple of the output error. Finally, we extend our result to a one-layer Transformer with multi-head attention followed by a bias-free ReLU feed-forward network. Under additional conditions, we recover a functionally equivalent Transformer without relying on a separate algorithm for learning the feed-forward network.
[LG-32] Noise-Aware Shrinkage for Differentially Private Zeroth-Order Fine-Tuning of Large Language Models
链接: https://arxiv.org/abs/2608.03277
作者: Lele Zheng,Weifeng Kong,Xinyi Zhang,Ke Cheng,Tao Zhang,Yulong Shen
类目: Machine Learning (cs.LG); Cryptography and Security (cs.CR)
*备注:
Abstract:Differentially private zeroth-order optimization (DP-ZO) enables memory-efficient private fine-tuning of large language models using only forward evaluations. Existing aggregation-based DP-ZO methods reconstruct model updates at a fixed scale, ignoring that the strength of useful signals varies throughout training. Consequently, noise-dominated updates may receive excessive weight and degrade model utility. To address this issue, we propose SAGE, a noise-aware shrinkage method that adaptively attenuates privatized estimates according to their estimated signal quality. SAGE subtracts the known Gaussian noise variance from the observed second moment to estimate the underlying signal energy, stabilizes this estimate through temporal tracking, and compares its current signal-to-noise level with a warm-up reference to derive a bounded shrinkage factor. As pure post-processing, SAGE requires neither additional privacy budget nor model queries and introduces only constant additional state. Our theoretical analysis shows that shrinkage reduces the quadratic update-risk term faster than the linear descent term, preserving useful descent while limiting the influence of noise-dominated updates. Experiments on RoBERTa-large, OPT-1.3B, and OPT-6.7B demonstrate that SAGE outperforms existing baselines in most settings under the same privacy budgets while preserving the forward-only memory efficiency of DP-ZO.
[LG-33] ED-DiT: Physics-Guided Diffusion Pretraining for Transferable Molecular Representations from Electron Density
链接: https://arxiv.org/abs/2608.03260
作者: Liang Shuang,Haocheng Wang,Jiayi Song,Shuquan Ye,Ben Fei
类目: Machine Learning (cs.LG)
*备注: 16 pages, 9 figures, 7 tables, including supplementary material
Abstract:Pretraining has shown strong potential for learning transferable representations, yet it remains underexplored for electron-density-based molecular learning. Electron density provides a continuous three-dimensional description of molecular electronic structure, capturing both local spatial patterns and global physical quantities. This raises a key question: can electron-density fields be used for self-supervised pretraining to learn a shared representation that transfers across diverse electronic-structure-related tasks? We propose ED-DiT, a physics-guided Diffusion Transformer for self-supervised pretraining on electron-density point clouds. ED-DiT learns reusable representations by reconstructing corrupted and partially masked log-density fields across diffusion noise levels. An electron-number consistency constraint is further introduced to preserve the total electronic mass. The pretrained encoder can be adapted to property prediction, open-/closed-shell classification, molecule-electron-density retrieval, and molecule-conditioned electron-density prediction. Experiments on six EDBench tasks show that ED-DiT consistently outperforms the same architecture trained from scratch, especially under limited supervision. For molecule-conditioned electron-density prediction, it reduces RMSE from 2.2474 to 1.3753 and surpasses the available baseline. With only 10% labels, it improves orbital energy prediction RMSE from 0.0293 to 0.0138. These results demonstrate the effectiveness of physics-guided electron-density pretraining for learning transferable molecular representations.
[LG-34] ShielDroid: A Hybrid Approach Integrating Machine and Deep Learning for Android Malware Detection
链接: https://arxiv.org/abs/2608.03250
作者: Md Faisal Ahmed,Zarin Tasnim Biash,Abu Raihan Shakil,Ahmed Ann Noor Ryen,Arman Hossain,Faisal Bin Ashraf,Muhammad Iqbal Hossain
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注:
Abstract:The rapid advancement of modern technology has led to a significant increase in the use of smart devices, such as smartphones and tablets, resulting in the widespread adoption of mobile applications. Although applications are required to undergo malware screening before being published on official app stores, many malicious applications successfully evade detection by concealing sophisticated malware variants. These malicious behaviors are often activated only during runtime, making them difficult to identify through conventional static analysis. As a result, malware may remain undetected until after installation, potentially causing irreversible damage to users and their devices. This study presents a real-time Android malware detection framework that analyzes application behavior to accurately identify and classify complex malware. The proposed approach employs a hybrid dynamic analysis technique to distinguish malicious applications from benign ones. After preprocessing and filtering the collected dataset, the applications are classified using multiple machine learning algorithms. A comprehensive performance evaluation is conducted to compare the effectiveness of different classification techniques in terms of detection accuracy and execution time. Experimental results demonstrate that a hybrid model combining Random Forest and a Multilayer Perceptron achieves the best overall performance, attaining an accuracy of 97.5% with an execution time of 22.945 seconds. The proposed framework can enhance mobile device security by enabling timely detection of malicious applications and reducing the risk of cyberattacks.
[LG-35] SAKI: Score-Aware Low-Rank Key Indexing for Long-Context KV Retrieval
链接: https://arxiv.org/abs/2608.03228
作者: Lin Zhang
类目: Machine Learning (cs.LG); Applications (stat.AP)
*备注: 10 pages, 4 figures 3 tables
Abstract:Existing low rank KV cache methods preserve either model weights or key variance, neither of which directly reflects the attention scores used during inference. We derive the expected attention score distortion caused by rank r key compression and show that it yields a covariance weighted low rank objective. Under a margin condition, controlling this distortion also improves top k recall. The optimal rank r solution has a closed form asymmetric factorization obtained from the SVD of the covariance weighted query key operator. This motivates SAKI, a training free KV cache index that directly preserves attention scores rather than key reconstruction quality. Across LLaMA 3.1 8B, Qwen 2.5 7B, Mistral 7B v0.1, and Llama 3.2 3B, SAKI outperforms key PCA at every tested rank. At rank 32, it removes 13 to 30 percent of PCA’s remaining top 64 recall error, including improvements from 0.748 to 0.799 on LLaMA 3.1 8B and from 0.786 to 0.850 on Qwen 2.5 7B. It improves 68 to 89 percent of attention heads per model, with the largest gains in deeper layers. Predicted score MSE reductions closely match empirical measurements, with a Pearson correlation of 0.997, while ablation studies confirm that the gains arise from optimizing the attention score objective rather than covariance weighting alone. Analysis of the scoring operator further explains why weight only, invariant subspace, and key reconstruction methods can be suboptimal. Comments: 10 pages, 4 figures 3 tables Subjects: Machine Learning (cs.LG); Applications (stat.AP) Cite as: arXiv:2608.03228 [cs.LG] (or arXiv:2608.03228v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2608.03228 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-36] On the Implicit Flatness Bias of Sharpness-Aware Minimization: A Linear Stability Analysis with Quantitative Hyperparameter Bounds
链接: https://arxiv.org/abs/2608.03197
作者: Jiaxin Deng,Junbiao Pang
类目: Machine Learning (cs.LG)
*备注:
Abstract:Sharpness-Aware Minimization (SAM) improves generalization by seeking parameters whose loss is robust to local adversarial perturbations, but the quantitative mechanism underlying its implicit bias toward flat minima remains unclear. In particular, the perturbation radius \rho is typically treated as an isolated tuning parameter, despite defining the neighborhood in which SAM measures sharpness. We analyze mini-batch SAM near an interpolating minimum through linear stability. Under local linearization and gradient-noise alignment assumptions, we prove that every linearly stable minimum satisfies \lambda_\max\leq\sqrt[3]b\Gamma/(2\rho\eta^2) , where \lambda_\max is the largest Hessian eigenvalue, b is the batch size, \eta is the learning rate, and \Gamma bounds the gradient norm. The bound quantitatively characterizes SAM’s implicit flatness bias: holding the other quantities fixed, a smaller batch size, a larger learning rate, or a larger radius restricts linearly stable SAM to flatter minima. It also exposes a necessary trade-off: \rho should be large enough to promote flatness, yet remain local enough to preserve the approximation and stable training. We validate this prediction in a controlled study of 900 models on CIFAR-100 with ResNet-18 and VGG-19, where increasing \rho is consistently associated with a smaller largest Hessian eigenvalue across batch-size and learning-rate settings. Finally, we instantiate the analysis in Taylor-Locality Controlled SAM (TLC-SAM), which adjusts \rho using the observed Taylor-approximation error and further reduces the top Hessian eigenvalue relative to fixed-radius SAM. Our results provide quantitative hyperparameter bounds and a stability–locality perspective for analyzing and designing SAM variants.
[LG-37] Simulation-free and finite-time diffusion model
链接: https://arxiv.org/abs/2608.03117
作者: Kentaro Kaba,Masayuki Ohzeki,Yuki Sughiyama
类目: Machine Learning (cs.LG)
*备注:
Abstract:The performance of generative diffusion models is determined by the choice of the reference diffusion process connecting the empirical and prior distributions. Conventional approaches typically trade off simulation-free training against finite-time generation. We propose a framework for designing the reference process that achieves both simultaneously. The key idea is to prescribe tractable time-dependent conditional distributions and then construct the reference process realizing them as its marginals. This framework reveals that score matching is not fundamental to diffusion-model training but instead emerges naturally through reversal of the reference process. We further show that conditional flow matching arises as the small-noise limit of the proposed framework.
[LG-38] Double Descent in Gradient Boosting Decision Trees via Split-Candidate Scaling
链接: https://arxiv.org/abs/2608.03111
作者: Ryuichi Kanoh
类目: Machine Learning (cs.LG)
*备注:
Abstract:Double descent is commonly studied by scaling an explicit capacity parameter, such as neural-network width. For gradient boosting decision trees (GBDTs), however, an analogous single-axis capacity parameter has not been established. We propose the number of split candidates as an operational capacity parameter for GBDTs. Holding other training controls fixed, increasing the split-candidate budget refines the feature-quantization grid and expands the dictionary of root-to-leaf paths from which boosting selects its updates. To analyze this expansion, we construct an empirical tree-kernel diagnostic that summarizes how candidate-induced paths group the training examples. A regime in which the empirical kernel rank grows toward the sample size and very small positive eigenvalues emerge exposes noise-sensitive directions; in this regime, test error peaks before decreasing again at larger split-candidate budgets. This perspective predicts that deeper trees should reach the regime with fewer split candidates, larger training sets should require finer grids, and label noise should make the peak more pronounced. Experiments support these predictions and show test-error peaks at intermediate split-candidate budgets across XGBoost, LightGBM, and CatBoost, whereas a random-forest control improves monotonically under the same split-candidate sweep. Taken together, our analysis and experiments support split-candidate scaling as a single-axis capacity intervention for studying GBDTs and suggest that the observed double descent arises from an interaction between candidate-induced geometry and boosting dynamics.
[LG-39] Revisiting TD Target Aggregation under Uncertainty in Q-Learning UAI
链接: https://arxiv.org/abs/2608.03069
作者: Lipeng Zu,Xiaonan Zhang
类目: Machine Learning (cs.LG)
*备注: Accepted in the 42nd Conference on Uncertainty in Artificial Intelligence (UAI)
Abstract:Deep Q-Networks (DQNs) learn value functions through bootstrapped temporal-difference updates, where future returns are approximated using a greedy maximization over next-state action values. While effective, this aggregation rule is inherently sensitive to estimation noise: when Q-values are uncertain, the maximization operator deterministically favors the largest estimate, regardless of its reliability, leading to amplified errors through bootstrapping. In this work, we propose the \textbfSuccessor Rollout \textbfAggregation \textbfDeep \textbfQ-Network (SADQ), a simple modification to Q-learning that regularizes how the TD target is formed. SADQ uses one-step rollout predictions from a learned dynamics model to guide the comparison among candidate next-state actions, introducing additional structure into the aggregation step without altering the underlying learning framework. The resulting mixed Bellman update attenuates unreliable maxima while preserving the standard fixed point under diminishing model error. We provide theoretical analysis showing that SADQ reduces bootstrap-induced overestimation in a pointwise manner. Empirically, SADQ consistently improves training stability across classical control tasks, real-world vector-based environments, and Atari benchmarks when compared to strong DQN variants.
[LG-40] Exploiting Separability in Multi-Scale Grey-Box Bayesian Optimization
链接: https://arxiv.org/abs/2608.03045
作者: Joshua E. Hammond,Tyler A. Soderstrom,Brian A. Korgel,Michael Baldea
类目: Machine Learning (cs.LG)
*备注:
Abstract:We consider grey-box optimization problems where the decision variables naturally partition into black-box variables (as arguments to an expensive black-box function) and white-box variables, governed by a set of explicit, closed-form equations that also depend on the output of the black-box function. We exploit this separability through a bilevel reformulation: an outer Bayesian optimization (BO) to optimize the scalar objective as a function of black-box variables alone, while an inner problem solves the white-box subproblem via global optimization. The Gaussian process surrogate used in BO is therefore defined rather than and white-box constraints are satisfied exactly whenever the inner optimizer converges to a feasible point—without penalty functions, chance constraints, or moment approximations. On a suite of 13 benchmark problems, bilevel BO achieves lower regret, with fewer iterations and wall clock time. This advantage is robust to initialization set size, exploration parameters, and inner-solver choice.
[LG-41] Paired Recipient-based Evaluation of Survival Prediction for Deceased Donor Kidney Transplants ALT
链接: https://arxiv.org/abs/2608.03017
作者: Misaki Matsuura,Mohammadreza Nemati,Dulat Bekbolsynov,Stanislaw Stepkowski,Kevin S. Xu
类目: Machine Learning (cs.LG); Computers and Society (cs.CY); Applications (stat.AP)
*备注: To appear at the Machine Learning for Healthcare Conference (MLHC) 2026
Abstract:There has been significant interest in using machine learning algorithms to predict kidney transplant outcomes, such as the number of years until a graft inevitably fails. These prediction algorithms could possibly be used for pre-transplant donor-recipient matching to identify more compatible donors and recipients and thus improve post-transplant outcomes. In this study, we explore the use of survival prediction models trained on deceased donor kidney transplant data from the Scientific Registry of Transplant Recipients (SRTR). We propose a novel paired recipient-based evaluation framework that compares graft outcomes between two recipients who received kidneys from the same deceased donor, allowing us to evaluate the counterfactual benefit of changing the recipient for a certain donor. We find that five different survival prediction models, ranging in complexity from linear to deep learning-based models, all result in ~60% paired recipient-based accuracy. We further translate this accuracy into an interpretable quantity of post-transplant years gained. We also highlight major limitations of the commonly used concordance index (C-index) metric for evaluating survival prediction accuracy in this setting and demonstrate that our proposed paired recipient-based accuracy metric is more clinically relevant and better reflects real-world allocation settings.
[LG-42] Joint Affine Spectral Shaping: Coupling Weight and Bias Updates Beyond Weight-Only Muon
链接: https://arxiv.org/abs/2608.02991
作者: Gongyue Zhang,Honghai Liu
类目: Machine Learning (cs.LG)
*备注:
Abstract:Matrix spectral optimizers reshape weight-update spectra but usually delegate vector-valued biases to a separate optimizer. We study whether this separation is neutral. We formulate each affine layer as a joint momentum matrix A=[M_W,\alpha m_b] and apply a capped regularized-inverse spectral map to the complete matrix, producing both the weight and physical bias updates. A strict five-seed ablation on a four-layer BERT-mini trained from scratch on IMDb compares exact-SVD Muon, weight-only inverse shaping, affine-probe inverse shaping, and the proposed joint regularized inverse (JRI). Weight-only inverse shaping raises validation-loss-selected test accuracy from 84.903\pm0.242% to 85.562\pm0.308% and lowers selected test loss from 0.3479 to 0.3345 . Allowing bias to alter the joint SVD while retaining an independent Adam bias update does not improve over weight-only inverse shaping. Using the transformed bias jointly raises selected test accuracy to 85.738\pm0.180% and lowers test loss to 0.3291 , with all five seeds improving relative to the probe baseline. During the peak-performance window, JRI preserves the eligible weight-update norm while reducing the bias-update norm from 0.02095 to 0.00301 , lowers boundary-function share from 86.58% to 78.97% , and changes the cosine between weight-induced boundary motion and explicit bias from +0.030 to -0.137 . An independent 22-seed replication yields 85.743\pm0.203% selected test accuracy. These results identify joint affine spectral allocation as a small but consistent extension to weight-only spectral optimization.
[LG-43] A Physics-Informed Hybrid Neural Operator for Transient Magnetization Prediction in Power Magnetics
链接: https://arxiv.org/abs/2608.02965
作者: Yachao Zhu,Qiujie Huang,Sinan Li,Yang Li,Gang Lei,Jianguo Zhu
类目: Machine Learning (cs.LG); Materials Science (cond-mat.mtrl-sci); Systems and Control (eess.SY)
*备注: 13 pages, 7 figures. Preprint prepared for possible submission to IEEE Transactions on Power Electronics
Abstract:Magnetic components in high-frequency, high-power-density converters are increasingly driven by non-sinusoidal flux-density waveforms with fast transitions, minor-loop operation, dc bias, and temperature variation. Under these conditions, steady-state core-loss formulas and single-valued material curves cannot fully capture transient magnetization responses. This work proposes the Physics-Informed Hybrid Neural Operator (PI-HNO), a compact material-specific neural model with B-H energy-consistency regularization for core-loss-oriented transient magnetization prediction. Given the measured B(t)-H(t) history, the input B(t) series over the prediction interval and operating-condition information, PI-HNO predicts the H(t) series and the corresponding reconstructed B-H trajectory. The model integrates a local recurrent branch for boundary-state representation and rate-dependent response evolution with a Preisach-inspired global branch that extracts waveform-level hysteresis context. Evaluation on the MagNetX transient database using material-specific models for 14 ferrite materials demonstrates that PI-HNO achieves a compact trade-off between sequence accuracy and B(t)-H(t) energy consistency, with the mean and 95th percentile B(t)-H(t) energy consistency errors of 1.92% and 7.60%, respectively, using only 4777 trainable parameters per model. Ablation studies further demonstrate that the local, global, and energy-aware regularized components provide distinct contributions to transient magnetization prediction.
[LG-44] Inverted Detection and Control in Steering Vectors
链接: https://arxiv.org/abs/2608.02957
作者: Max Torop,Aria Masoomi,Jennifer Dy
类目: Machine Learning (cs.LG)
*备注:
Abstract:Steering vectors (SVs) are widely used to influence the expression of concepts (e.g., truthfulness) in large language model outputs. A key assumption underpinning SVs is that they are linearly discriminative with respect to the concept: representations of texts that exhibit the concept are more aligned with the SV than those that do not, motivating shifts along the positive or negative SV direction to respectively promote or suppress the concept. In this work, we identify an inverted detection-control phenomenon in which some highly discriminative SVs that are aligned with positive representations can consistently promote the opposite behavior. We refer to such vectors as inverted-steering vectors (ISVs). We provide a geometric characterization of ISVs’ effects, finding that steering along these directions systematically pushes representations in discriminative downstream heads as if the concept were absent, even prior to decoding. Motivated by this analysis, we propose an approach for distinguishing ISVs without requiring generation or associated response scoring. This enables targeted sign flips, which we use to improve a foundational detection-based steering pipeline via Inference Time Intervention (ITI). Our approach improves results in 27/30 experiments, ranging from +0.9% to +138%. We evaluate our findings on Gemma 3 12B, Qwen 2.5 14B, and Olmo 3 7B across 5 concepts.
[LG-45] Schedule-Informed Temporal Fusion Forecasting of Hourly Airport Security-Checkpoint Throughput
链接: https://arxiv.org/abs/2608.02950
作者: Yinxiao Zhang,Sen Wang,Yi Gao
类目: Machine Learning (cs.LG); Emerging Technologies (cs.ET)
*备注:
Abstract:Checkpoint staffing requires accurate forecasts of when screening demand will occur, yet flight schedules record departure times rather than passenger arrival times at security checkpoints. This study develops a framework that converts known flight schedules into temporally aligned signals for forecasting hourly checkpoint throughput. Using 2023-2024 Transportation Security Administration throughput data and Cirium Diio flight schedules for Hartsfield-Jackson Atlanta International Airport, domestic and international seat capacity was distributed across pre-departure hours using truncated Poisson kernels. A Temporal Fusion Transformer then combined these schedule-derived arrival-intensity signals with historical throughput, scheduled activity, and temporal variables. Models were trained chronologically, with July-December 2024 reserved for testing, and evaluated against recurrent neural network and long short-term memory models across five random seeds. For direct six-hour forecasts, the proposed model achieved a weighted mean absolute percentage error of 9.33%, compared with 12.16% for the recurrent neural network and 11.37% for long short-term memory, while also producing the lowest errors during peak periods. With six-hour recursive updates, errors remained between 10.60% and 11.04% across 24-96 hour horizons, although longer horizons contained fewer valid forecast origins. By transforming scheduled departures into interpretable pre-departure screening-load signals without requiring passenger-flight matching, the framework supports advance staffing, lane-opening, and multiday checkpoint planning. Because observed throughput reflects realized processing rather than unconstrained arrivals, the forecasts should be interpreted together with local staffing, capacity, queue, and wait-time information.
[LG-46] Sedentary Behavior Classification for Wearable Sensors with a CNN-BiLSTM Model
链接: https://arxiv.org/abs/2608.02946
作者: Yuliang Chen,Weiwei Shi,Jingjing Zou,Rong Zablocki,Animesh Kumar,Jordan A. Carlson,Sheri J. Hartman,Mikael Anne Greenwood-Hickman,Paul R. Hibbing,Marta Jankowska,Jay Yang,Arun Kumar,Loki Natarajan
类目: Machine Learning (cs.LG)
*备注:
Abstract:Accurate detection of sedentary behavior is important for studying health risks related to prolonged sitting, but posture-based classification remains challenging with wearable sensors, especially at the wrist. We study whether a deep learning model trained on hip-worn accelerometer data can transfer to wrist-worn accelerometer data for sitting versus non-sitting classification. We use CHAP, a CNN-BiLSTM model originally developed for hip accelerometers, and evaluate its zero-shot performance on wrist data as well as its adaptation through finetuning with varying amounts of labeled wrist data. Experiments are conducted on the iWatch dataset with ground-truth posture labels derived from wearable cameras. The hip-trained model performs strongly on hip data without retraining, but accuracy drops on wrist data due to sensor placement shift. Finetuning CHAP provides consistent advantages over transformer models trained from scratch. These findings suggest that hip-based pretraining provides a useful starting point for wrist deployment, while highlighting the need for wrist-specific adaptation to handle higher signal variability.
[LG-47] Federated generative event models for tokenized electronic health records
链接: https://arxiv.org/abs/2608.02939
作者: Michael C. Burkhart,Luke Solo,Inhyeok Lee,S’Khaja Charles,Zewei “Whiskey” Liao,Kaveri Chhikara,Dema Therese,Wan-Ting Liao,Catherine A. Gao,William F. Parker,Brett K. Beaulieu-Jones
类目: Machine Learning (cs.LG); Computers and Society (cs.CY)
*备注:
Abstract:Electronic health record foundation models are limited by institutionally siloed data and substantial performance degradation under cross-site transfer. We evaluated federated training of tokenized generative event models (GEMs) across 122,251 intensive care hospitalizations from three independent health systems harmonized to the Common Longitudinal ICU Data Format. Models were assessed on 12 post-24-hour clinical prediction tasks using within-site, cross-site, centralized, and federated training configurations. GEMs achieved the highest mean within-site and cross-site ROC-AUC and were substantially more transportable than conventional supervised models: their average cross-site penalties were 0.025 ROC-AUC and 0.027 PR-AUC, compared with 0.079 and 0.089 for LightGBM. Federated Learning (FedAvg and FedAvgM) approached the performance of centralized GEM training, with most gains obtained within 5-10 communication rounds. However, centralized multi-site training provided only modest improvements over complete local training. Multi-site models were most useful when local training data were limited, with their advantage narrowing as institutional data accumulated. These findings show that federated GEM training is technically feasible and preserves most centralized performance, but that the main open challenge is learning transportable representations to translate larger, but heterogeneous data from multiple health systems into a reliable target-site benefit.
[LG-48] Forecasting Revenue with its Customer-Base Drivers: When and Why Coordination Helps
链接: https://arxiv.org/abs/2608.02911
作者: Kyeongbin Kim,Daniel McCarthy,Dokyun Lee
类目: Machine Learning (cs.LG)
*备注:
Abstract:Revenue forecasts guide acquisition budgets, demand planning, and customer-based valuations, yet an aggregate forecast does not show whether change reflects acquisition, repeat purchasing, spending per order, or offsetting movements. Using weekly transaction panels for 966 companies in 25 industries, the authors develop the Customer-Based Multi-task Transformer (CBMT), which learns shared structure, retains separate primitive forecasts, and aligns their combination with downstream revenue. CBMT’s mean total-sales error is 30% below the strongest representative established customer-base benchmark. It is also 2.65% below a Transformer that forecasts total sales directly, although the paired difference is not statistically significant (p=.222), and it beats separately estimated single-task forecasts for 74.3% of firms. CBMT’s source MAE is lower in 23 of 24 benchmark-by-outcome comparisons, with the remaining difference not statistically distinguishable from zero. Firms whose primitives co-move more strongly are more likely to benefit from joint forecasting; selected-family scenario-3 comparisons are consistent with gains from shared representation and revenue alignment but remain diagnostic rather than causal. Accuracy deteriorates for all models when customer-base dynamics are highly volatile, and CBMT’s advantage narrows there. Calibration-period routing rules do not improve average accuracy over always deploying CBMT. The results show how coordinated customer-base forecasts support revenue planning and when they warrant greater caution.
[LG-49] Bayesian Data Reweighting Improves Multimodal Retrieval for Knowledge-Based Visual Question Answering
链接: https://arxiv.org/abs/2608.02907
作者: Jingchen Sun,Shaobo Han,Ruiyi Zhang,Naresh Kumar Devulapally,Ming Liu,Yitao Long,Vishnu Suresh Lokhande,Changyou Chen
类目: Machine Learning (cs.LG)
*备注:
Abstract:Multimodal retrievers are essential for knowledge-based visual question answering, where they retrieve external evidence for image-question pairs. However, existing contrastive training methods typically treat all unmatched query-document pairs as equally informative negatives, which is problematic because many unmatched documents may still be semantically relevant or partially useful. We propose Bayesian Data Reweighting, a probabilistic framework that models query-document importance as latent variables and adaptively infers posterior weights to downweight likely false negatives. With closed-form posterior updates under conjugate priors and stochastic EM optimization, our method consistently improves retrieval accuracy across three retrievers and seven knowledge-based VQA benchmarks.
[LG-50] Population-Robust Feature Selection via Generalized Welfare Optimization
链接: https://arxiv.org/abs/2608.02887
作者: Ruiqi Lyu,Alistair Turcan,Bryan Wilder
类目: Machine Learning (cs.LG)
*备注:
Abstract:Choosing which features to collect is a deployment decision: the same limited questionnaire, test panel, or sensor set may need to serve several heterogeneous populations. Standard feature-selection methods typically optimize for one large population, while existing robust approaches tend to learn one shared model for every population. We introduce PopFS, a method for learning one shared, deployable feature set that is robust to population differences while letting each pop- ulation train its own model. PopFS uses a tunable welfare objective that lets practitioners balance overall predictive ben- efit against stronger protection of the populations that benefit least. To make this objective practical at scale, PopFS first uses multitask sparse learning to reduce the candidate pool, then searches directly over hard feature sets by ranking promising additions and swaps and fully refitting only a shortlist. Across eight population splits from six prediction tasks drawn from five tabular and public-health datasets, PopFS consistently achieves strong average and worst-population performance while scaling to thousands of candidate features. A 43-state COVID-19 nowcasting study further shows that changing the welfare objective can improve the least-served states with lit- tle change in average performance and yields an interpretable change in the selected symptom signals. Our code is available at this https URL.
[LG-51] GoT-CD: Graph-of-Thoughts Causal Discovery and the Frag ility of Post-hoc Path-Specific Fairness Audits
链接: https://arxiv.org/abs/2608.02877
作者: Nitish Nagesh,Elahe Khatibi,Thomas Dean Hughes,Mahdi Bagheri,Pratik Gajane,Amir M. Rahmani
类目: Machine Learning (cs.LG)
*备注:
Abstract:Causal discovery recovers directed structure from observational data and is increasingly used in clinical settings to support mechanism reasoning and fairness audits of predictive models. Path-specific counterfactual fairness asks whether a protected attribute influences an outcome through illegitimate pathways, but these estimands are defined relative to a supplied causal graph and therefore inherit whatever errors the discovery step introduces. Discovery methods are routinely scored on aggregate structural metrics that weight all edges equally, and no established evaluation asks whether the specific pathway an audit depends on survives discovery—or what the audit reports when that pathway is missing. Here we show that full-graph Graph-of-Thoughts reasoning yields acyclic discovered graphs that are structurally competitive with large language model (LLM) baselines, yet that structural fidelity alone does not guarantee fairness-faithful audits. We introduce GoT-CD, in which the reasoning unit is a complete candidate edge set: multiple graphs are generated in parallel, scored by a deterministic validity function, and merged under a hard union constraint that forbids invented edges, with greedy projection enforcing a DAG before commitment. GoT-CD returns a valid DAG on all five reported benchmarks and achieves the best DAG-valid F1 score among LLM methods on Asia, Alzheimer’s, and COVID-Respiratory datasets. On an Alzheimer’s benchmark with known unfair path, a post-hoc path-specific audit shows that five of eight discovered graphs recover no path from the sensitive attribute to the outcome and therefore report a null overall effect while mediated effects persist, necessitating downstream path-specific fairness analysis along with structural discovery.
[LG-52] Maglev: Sliding Recurrent Memory
链接: https://arxiv.org/abs/2608.02870
作者: Bo Liu,Qiang Liu
类目: Machine Learning (cs.LG)
*备注: Neural Architecture Research
Abstract:We introduce \ours, a recurrent Transformer architecture with fixed-size memory that generalizes sliding-window attention while remaining parallelizable during training. \ours consists of two coupled models: a prefiller Q , which leverages full attention\footnoteIn practice, we use interleaved full and sliding-window attention for Q , as this yields stronger performance. The essential requirement is that Q be more expressive than P , with access to the full history. to produce memory targets m’_t , and a decoder P , which uses only sliding-window attention and recurrent K/V injection to produce decoder memories m_t for next-token prediction. We train \ours with a memory consistency loss that aligns m_t with m’_t , allowing inference to use P alone. Empirically, \ours improves validation loss and downstream pretraining benchmarks over sliding-window and latent recurrent transformer baselines. Moreover, sharing parameters between P and Q reduces parameter memory while preserving most of the gains.
[LG-53] Contrast-invariant deep ptychography neural networks
链接: https://arxiv.org/abs/2608.02869
作者: Albert Vong,Steven Henke,Oliver Hoidn,Hanna Ruth,Junjing Deng,Apurva Mehta,David Shapiro,Alexander Hexemer,Nicholas Schwarz
类目: Machine Learning (cs.LG)
*备注: ~30 pages, 4 main figures + 2 SI figures. Submitting to scientific journal
Abstract:Ptychography neural networks suffer from scaling inconsistencies when generalizing out of distribution, limiting their real world viability. We address this scaling mismatch using a factorization strategy which decouples the learned object texture from measurement scaling, enabling a single trained network to produce measurement-consistent reconstructions across varying illumination conditions. This requires predicting the learned object in real and imaginary units instead of the canonical amplitude and phase representation. We additionally introduce a synthetic object sampling strategy that minimizes phase distribution mismatch between synthetic training data and experimental targets. These improvements yield up to a 5x reduction in Fourier error over the previous PtychoPINN-torch baseline across 5 experimental datasets spanning multiple beamlines and facilities.
[LG-54] Adaptive Sampling for Automated Post-Disaster Rapid Damage Assessment via Level-Set Cost-Aware Bayesian Optimization SDM
链接: https://arxiv.org/abs/2608.02868
作者: Boyang Xu,Mostafa Reisi Gahrooei,Mohammad Ilbeigi,Hao Yan
类目: Machine Learning (cs.LG)
*备注: 11 pages, 7 figures, 1 table. Accepted at the SIAM International Conference on Data Mining (SDM) 2026
Abstract:Natural disasters frequently inflict severe damage to the built environment, which demands a rapid, reliable, and cost-effective damage assessment for emergency response. However, traditional methods for post-disaster damage assessment often rely on static, labor-intensive data collection strategies that can be prohibitively expensive and struggle to adapt to dynamic post-disaster conditions. In this study, we propose a cost-aware Bayesian optimization framework combined with level-set estimation that continuously guides autonomous data collectors, e.g., an unmanned aerial vehicle (UAV), toward the most informative regions. By dynamically updating damage estimates across different geographic zones, our approach systematically reduces uncertainty while minimizing operational costs. The proposed framework is first validated using a controlled synthetic toy study, demonstrating the agent’s ability to efficiently trace damage boundaries, recover the underlying damage map, and rapidly reduce predictive uncertainty. Furthermore, the approach is evaluated using high-fidelity disaster data generated by the Regional Resilience Determination (R2D) software. The results of the algorithm provide accurate and timely damage estimates that support informative and fast emergency response.
[LG-55] NOMADD: Numerical Optimization of Models Adapting to Data Drift
链接: https://arxiv.org/abs/2608.02845
作者: Swapn Shah,Keith Burghardt
类目: Machine Learning (cs.LG)
*备注:
Abstract:Tabular model performance degrades when feature distributions change over time or the relationship between features and outcome variables change over time, known as data drift and concept drift, respectively. These issues are challenging to mitigate in real time because labeled data may not be immediately available, or re-training a model could be impractical. While tools exist to reduce drift, they are typically bespoke to neural network architectures and adapt how models are trained. In this paper, we offer an alternative post-hoc method to reduce concept drift, which is applicable to a variety of models, from trees to neural networks to tabular foundation models. This new tool is especially useful when constraints, such as high model accuracy, bounded inference time, or model size requires users to choose between different models for their specific use-cases. Our algorithm fits the base model separately on each labeled training period, measures how its parameters evolve against a single anchor model pooled over all of those periods, compresses those changes with a low-rank factorization, and extrapolates each latent factor forward with a damped, regularized forecast. On the 18-dataset Drift-Resilient TabPFN benchmark, evaluated under that benchmark’s own protocol and metric, the extrapolation improves every base family it is applied to, and achieves performance competitive with the state-of-the-art Drift-Resilient TabPFN with seconds of training. In contrast, Drift-Resilient TabPFN requires pre-training on millions of synthetic datasets over approximately 1,300 GPU-hours, and is orders of magnitude slower in inference (depending on the model). In the discussion, we explore the promise and challenges of extending this tool to other modalities.
[LG-56] Wiring Beats Blending: What Transfers Between Transformer Sizes – and What Doesnt
链接: https://arxiv.org/abs/2608.02829
作者: Ravi Satya Durga Prasad Yenugula
类目: Machine Learning (cs.LG)
*备注: 15 pages, 5 figures. Independent research preprint
Abstract:Model families train every size from scratch. Can a pretrained large model be converted into a smaller sibling? We characterize the 1.4B-410M conversion in the Pythia family end-to-end: (i) representations align strongly across sizes (ridge R^2=0.84) while parameters align weakly; (ii) dense weight projection is functionally destructive – provably not an assembly artifact – because basis mixing breaks rotary, per-head, GELU, and LayerNorm structure; (iii) after the best-fit linear operator, weight residuals are statistically indistinguishable from noise under shuffle controls; (iv) conversion value therefore lives in initialization. In matched-budget continued pre-training we decompose conversion into two independent levers – least-squares compensation (function: best zero-shot) and variance-preserving rescale (dynamics: best endpoints). Compensation is a token-efficient, low-budget win rather than a universal one: at 30M tokens it beats the strongest subcloning variant on both a width-reduced pair (84.0 +/- 1.8 vs. 89.7 +/- 3.7, 3/3 seeds) and a held-out depth-reduced pair (109.3 vs. 117.9, 3/3 seeds), reaching a given quality with fewer tokens; at a 33x larger budget the two converge to parity (40.0 vs. 40.0), both far ahead of from-scratch, which transfer initialization always beats – by up to 18x at low budget, the margin narrowing at convergence and at the largest scale. We further map the method’s boundary: at ~5x the donor scale (6.9B-1.4B) stacking both levers over-corrects, which we trace to ill-conditioning of the compensation solve at large width, pointing to dimension-aware regularization as the fix. Code, checkpoints, and the frozen evaluation corpus are released.
[LG-57] opological Simplification in Predictive Coding Networks
链接: https://arxiv.org/abs/2608.02816
作者: Adam Shaw,Jiayu Li,Michael Sperling,Michael Kim,Alvin Jin
类目: Machine Learning (cs.LG)
*备注: Accepted to the 2nd Annual Conference on Topology, Algebra, and Geometry in Data Science (TAG-DS 2026); to appear in Proceedings of Machine Learning Research
Abstract:We study the topology of learned representations in predictive coding networks (PCNs), a neuro-inspired bidirectional architecture, using a quantitative layer-wise persistent homology analysis. We train well-performing PCNs on a synthetic classification dataset ( \geq 99.9% test accuracy) and on MNIST ( \geq 95% test accuracy), and measure how topological features change across layers for different architectures and activation functions. We find that smaller PCNs collapse connected components across layers earlier than larger models (Spearman \unicodex1D70C \in [0.72, 0.79] across activations), with model size measured as the sum of hidden-layer widths. We also observe a strong negative correlation ( \unicodex1D70C = -0.58 ) between the depth at which simplification occurs and reconstruction error; i.e., architectures that simplify later reconstruct better. Finally, a seed-level bootstrap comparison across architectures and activations shows that PCNs consistently collapse connected components later than matched MLPs, with an average difference of 3.6 layers. These results suggest that persistent homology offers a useful quantitative lens on the compression–reconstruction tradeoff in PCNs, and that both model capacity and the recurrent, bidirectional dynamics of predictive coding inference shape when this tradeoff is resolved across layers.
[LG-58] Evaluation Blindness: How Silent Measurement Failures Corrupt AI Systems from Training to Deployment
链接: https://arxiv.org/abs/2608.02786
作者: Priyanka Bajaj(Independent Researcher)
类目: Machine Learning (cs.LG)
*备注: 10 pages, 7 figures
Abstract:AI systems can fail silently. The failure propagates through training loops, evaluation pipelines, and production monitoring stacks until downstream harm makes it visible. This paper introduces evaluation blindness: a measurement function M exhibits evaluation blindness with respect to failure class F when it produces readings indistinguishable from a healthy state while the system is actually failing, with no auxiliary signal flagging the gap. The problem surfaces at two lifecycle stages the literature has treated separately. At training time, reward models are gamed, importance-sampling corrections are silently miscalculated, and benchmark contamination inflates fine-tuning evaluations, all while loss curves look healthy and gradient updates proceed normally. At deployment time, monitoring fails to catch six classes of production failure, including an Operational category that is 100% silent by structural definition. We provide a formal detectability predicate unifying both stages. Four training-time case studies trace concrete breakdowns, including a real implementation bug in TRL PR #6594 where gradients are corrupted as loss decreases normally. A six-class taxonomy validated against 50 real-world incidents from court documents and regulatory filings finds that 53% of verifiable public failures were silent. A failure budget framework ties acceptable failure rates to use-case risk class. The implication is direct: measurement infrastructure is a correctness concern across the full AI lifecycle, not just at evaluation time. Data, code, and taxonomy schema are at this https URL. Comments: 10 pages, 7 figures Subjects: Machine Learning (cs.LG) Cite as: arXiv:2608.02786 [cs.LG] (or arXiv:2608.02786v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2608.02786 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Priyanka Bajaj [view email] [v1] Mon, 3 Aug 2026 18:33:37 UTC (360 KB)
[LG-59] Neural Networks with Local Converging Inputs for Efficient Options Pricing Models
链接: https://arxiv.org/abs/2608.02778
作者: Harris Cobb,Wenbo Hao,Yingjie Liu
类目: Machine Learning (cs.LG); Computational Finance (q-fin.CP)
*备注: 15 pages
Abstract:We present a novel application of Neural Networks with Local Converging Inputs (NNLCI) to improve the efficiency of existing numerical methods for pricing multi-asset options. The most concise input format for NNLCI has been introduced, offering substantial convenience and efficiency. NNLCI uses a neural network to locally correct solutions from a coarse mesh and a refined mesh (relative to the coarse one), requiring only a minimal amount of high-fidelity training data. We demonstrate this approach on cash-or-nothing options under the Black-Scholes equation in one, two, and three spatial dimensions, and on single-asset down-and-out barrier call options under the Heston stochastic-volatility model (whose pricing PDE is two-dimensional in the spot price S and the instantaneous variance v ). In each case, NNLCI reduces the root-mean-square error (RMSE) of the refined-mesh numerical solution by a factor of approximately 4-12 on test sets, even when the neural network is trained on only a small subset of parameter combinations. These results demonstrate that NNLCI significantly reduces computational requirements for high-dimensional problems in real-time options trading and risk management, offering low training costs and strong generalization ability.
[LG-60] Stylometric Defenses Against Author Impersonation in Software Repositories
链接: https://arxiv.org/abs/2608.02695
作者: Leonid Ravich,Michael Fire
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG); Software Engineering (cs.SE)
*备注:
Abstract:Software supply-chain attacks increasingly exploit an identity gap where compromised maintainer accounts authorize malicious changes. This work evaluates patch-level authorship verification as a behavioral defense layer, showing that stylometric analysis can operate not only on full source files but also on patch-level commits. We fine-tune a cross-modal transformer on more than 20 years of Linux kernel commit history to embed code diffs and commit messages into a unified stylometric space, achieving ROC AUC of 0.93 for open-world authorship verification. We then use these representations in a streaming anomaly detector suited to continuous integration and deployment (CI/CD) settings. We validate the pipeline on two retrospective supply-chain incidents involving different patch characteristics: the 2021 PHP backdoor and the 2026 ForceMemo/GlassWorm campaign. Without retraining, the proposed detector surfaces both PHP forged commits within approximately 1% of the maintainer audit queue and ranks the 28 scoreable ForceMemo spoofs with a median per-repository review burden of 0.8%. These results indicate that cross-modal patch-level embeddings can support behavioral triage against author impersonation in real-world repositories.
[LG-61] PatTree: a novel approach for automated creation of multimodal graph-based patient representations for medical classification tasks
链接: https://arxiv.org/abs/2608.02692
作者: Julia Gehrmann,Lars Quakulinski,Hamza Naseem,Oya Beyan
类目: Machine Learning (cs.LG)
*备注:
Abstract:Access to holistic, multimodal data improves the performance of Artificial Intelligence (AI) in medical classification tasks compared to utilizing single modalities or data sources. However, the inherent heterogeneity and complexity of clinical real-world data pose significant challenges to structured data analysis and AI application. This heterogeneity includes missing values, multiple time points, diverse modalities, and inconsistent formats and semantics. Data harmonization prior to data integration tackles this challenge but remains resource-intensive and error-prone, limiting the scalability and reproducibility of holistic, AI-driven decision support on clinical real-world data. We therefore propose PatTree, a graph-based, holistic representation of patients that can be derived from real-world clinical data through the automated structuring of multimodal clinical data. PatTree enables early-stage data integration without relying on pre-standardized inputs. While representing heterogeneous clinical data within a unified knowledge graph, PatTree preserves the semantic relationships between data elements across modalities and data sources, facilitating interoperability and machine-interpretable data access. Using a subset of the ADNI-1 cohort (n = 763), we demonstrate that classification of patients is directly feasible on PatTree reaching state-of-the-art classification performance. In the three-class classification task distinguishing Alzheimer’s disease, mild cognitive impairment, and cognitively normal individuals, we achieve a balanced accuracy of 98.5% and an F _1 score of 0.987 on the held-out test set. Our results show that assumption-free, automated structuring of multimodal medical data can serve as a scalable foundation for clinical AI pipelines bypassing tedious data preparation and standardization.
[LG-62] GLOBE: Trajectory-Aligned Gradient Matching with Structured SparseOptimization for Coreset Selection
链接: https://arxiv.org/abs/2608.02690
作者: Hetian Liu,Jin Cui,Mengcheng Shi,Yanbin Hu,Xinyue Long,Boran Zhao,Pengju Pen
类目: Machine Learning (cs.LG)
*备注:
Abstract:On-device training of deep neural networks is fundamentally constrained by the computational and memory costs of large-scale datasets. Coreset selection offers a practical solution by retaining only a compact subset of real training samples. However, existing gradient-based methods commonly rely on gradients computed at a single model snapshot and employ greedy or pursuit-based selection procedures, limiting their ability to capture evolving optimization dynamics and handle strongly correlated samples. We propose GLOBE (Gradient Local-Balanced Extraction), a trajectory-aligned coreset selection framework that formulates sample selection as a globally optimized sparse weighting problem. GLOBE represents each sample by a gradient trajectory constructed across multiple training checkpoints, thereby capturing its influence throughout different stages of optimization. To preserve the training behavior of the full dataset, we introduce a multi-order matching objective that jointly aligns the first-order mean and projected uncentered second-order moments of gradient trajectories. GLOBE further combines Group LASSO, Elastic Net regularization, and nonnegative budget constraints to induce group- and sample-level sparsity while stabilizing the weights of correlated trajectories. Finally, class-balanced Top-K selection maintains adequate category coverage under limited sampling budgets. Experiments across six benchmarks and five evaluation architectures demonstrate that GLOBE consistently outperforms existing coreset selection methods in downstream test accuracy, particularly at low retention ratios. These results highlight the effectiveness of combining dynamic gradient information, multi-order distribution matching, and structured sparsity for data-efficient learning.
[LG-63] On the Performance of Malware Detection Classifiers Using Hardware Performance Counters
链接: https://arxiv.org/abs/2608.02671
作者: Alireza Abolhasani Zeraatkar,Parnian Shabani Kamran,Inderpreet Kaur,Nagabindu Ramu,Tyler Sheaves,Hussain Al-Asaad
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注:
Abstract:Malware detection using Hardware Performance Counters (HPC) has emerged as a promising solution to improve the security of computing systems as a complement to antivirus software. Hardware-based malware detectors (HMD) use Machine Learning (ML) classifiers to detect malicious application patterns. The inputs to ML classifiers are low-level performance features known as HPCs, hardware-related activity data collected from a processor at run time to profile the low-level microarchitectural behavior of an application. This paper proposes malware detection using HPCs and machine learning classifiers and highlights the effectiveness of malware detection at run-time. We use ensemble learning techniques to improve the performance of the hardware-based malware detectors, which reduces the number of necessary micro-architectural events. This improves the processor’s efficiency by eliminating the need to run an application several times since a processor can measure only 2 to 8 events at a cycle. We use 18 machine-learning models along with two ensemble learning methods to evaluate the malware detection performance, creating a total of 144 different configurations. The experimental results show that the ensemble learning-based malware detection with 2 HPCs using the ensemble technique outperforms standard classifiers with 8 HPCs by up to 10%. It also matches the performance of standard ML-based detectors that use 16 HPCs while requiring only 4 HPCs, thereby enabling effective run-time malware detection.
[LG-64] GeoID-PINN: Identifiability-Aware Regional Epidemic Inference with Geographic Coupling KDD KDD2026 ALT
链接: https://arxiv.org/abs/2608.02633
作者: Weixiong Hua,Fan Bu
类目: Machine Learning (cs.LG); Applications (stat.AP); Machine Learning (stat.ML)
*备注: 11 pages, 3 figures. Accepted at the 8th epiDAMIK ACM SIGKDD Workshop on Data-driven Decision Making for Public and Population Health (epiDAMIK @ KDD 2026)
Abstract:Regional surveillance data reflect local transmission, reporting, seeding, and external infection pressure, which are difficult to identify separately. We introduce GeoID-PINN, a physics-informed neural network (PINN) for susceptible-infectious-recovered-deceased (SIRD) dynamics. The model represents spatial dependence with a row-stochastic source-composition matrix whose rows assign nonnegative source weights that sum to one. We regularize this matrix toward a spatial prior constructed from distance, adjacency, commuting, or lead-lag information. In a four-region simulation with known truth, a compatible distance prior gives source-composition error 0.099. The error rises to 0.159 without regularization and 0.577 under a strongly misspecified prior, while trajectory fit and transmission-scale estimates remain similar. Accurate trajectories therefore do not guarantee recovery of the regional dependence structure. We also evaluate GeoID-PINN retrospectively using COVID-19 data from 64 Louisiana counties. Relative to an autoregressive negative-binomial baseline, Forecast-Trained Geo-PINN reduces mean squared error (MSE) from 32,957 to 11,468 and mean absolute error (MAE) from 70.60 to 57.73. The baseline has lower negative log likelihood (NLL), 5.158 versus 5.346, indicating better distributional fit but worse point accuracy. In a controlled 15-county comparison, county adjacency reduces MSE by 6.85 percent and MAE by 3.1 percent. Similar performance across plausible priors supports structured regularization but not unique edge recovery. These results require prior-sensitivity and observation-model checks before interpretation.
[LG-65] LLM s Can Annotate Attribution Graphs ICML
链接: https://arxiv.org/abs/2608.02632
作者: Ameen Patel,Max Zhang,Nathan Hu
类目: Machine Learning (cs.LG)
*备注: Poster at ICML Mechanistic Interpretability Workshop 2026
Abstract:Circuit tracing is an exciting technique for revealing the internal computation of language models, but it requires a time-intensive manual step of grouping individual features or MLP neurons into supernodes. We present a simple pipeline for automating this step: directly presenting feature descriptions to a language model that groups them into supernodes. Using automated interpretability metrics, we confirm that supernodes generated by our pipeline are as interpretable as those generated by human annotators. On a two-hop Capitals task, our pipeline recovers a supernode corresponding to the intermediate hop in 97 of 100 prompts. Finally, we present a simple proof of concept using our pipeline for open-ended exploration, where we automatically annotate 1000 attribution graphs from Wikipedia prompt completions and then use an LLM judge to flag interesting graphs worth human review. We hope this work demonstrates that even simple automation can produce meaningful attribution graph annotations, motivating further work on automated circuit tracing.
[LG-66] Neural network realization of binary refinement iterates via a two-chart atlas selector
链接: https://arxiv.org/abs/2608.02624
作者: Tsogtgerel Gantumur
类目: Numerical Analysis (math.NA); Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: 30 pages
Abstract:Refinement operators generate many functions used in wavelet constructions, subdivision schemes, and geometric modeling. Their finite iterates can develop rapidly increasing numbers of linear pieces, making them a natural test case for the expressive power of deep neural networks. Earlier work showed that, for scalar binary refinement with a finitely supported mask, every compactly supported continuous piecewise linear seed has finite refinement iterates that admit exact ReLU realizations of fixed width and depth growing linearly with the number of refinement steps. The present paper gives a new construction of this known theorem. The difficulty is that the refinement cascade is driven by discontinuous binary digit choices, whereas ReLU networks produce continuous piecewise linear maps. We represent the residual dynamics on a polygonal model of the circle and describe each residual position in two overlapping coordinate systems, one ordinary and one shifted by one half. Their discontinuities occur at different points. The network switches between the two descriptions only where both are valid and the corresponding fixed linear cascade updates agree, so the switch is exact and requires no multiplication by a variable selector. The construction also gives exact readout of every continuous piecewise linear circle function satisfying the natural endpoint compatibility condition. Localized seeds are handled by a two-pass network, and translation covariance, finite decomposition, and gluing extend the result to arbitrary compactly supported continuous piecewise linear seeds in a preserved support window. Comments: 30 pages Subjects: Numerical Analysis (math.NA); Machine Learning (cs.LG); Machine Learning (stat.ML) MSC classes: Primary 41A30, Secondary 68T07, 42C40 Cite as: arXiv:2608.02624 [math.NA] (or arXiv:2608.02624v1 [math.NA] for this version) https://doi.org/10.48550/arXiv.2608.02624 Focus to learn more arXiv-issued DOI via DataCite
[LG-67] Information-Geometric Forward Policy Training in GFlowNets
链接: https://arxiv.org/abs/2608.03967
作者: Yordan Raykov,Rodrigo Veiga
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注: 13 pages + appendix
Abstract:Generative Flow Networks (GFlowNets) have emerged as a flexible framework for amortised inference over discrete and mixed discrete-continuous objects, requiring only an unnormalised target density specified through a reward. In this work, we formulate forward-policy training in GFlowNets through the information geometry of the induced trajectory sampler. Treating the forward policy as an induced trajectory sampler, we show that its intrinsic first-order geometry is given by the Fisher-Rao metric of the trajectory family, and that the associated natural gradient provides the canonical local update whenever the corresponding Fisher information is computable or accurately approximable. We derive an exact decomposition of the trajectory Fisher into per-step conditional second moments, which clarifies when temporal score interactions vanish and when dense couplings remain under shared parameterisation. This leads to three computational regimes: settings with tractable exact Fisher information, settings where Monte Carlo estimators of the expected Fisher are sufficient, and structure-exploitable settings in which target locality or factorisation yields accurate approximations of the Fisher expectation. In the latter case, graphical-model tools such as exact marginalisation, separator methods, and belief propagation provide principled surrogates for natural-gradient updates. The resulting framework turns target structure into optimisation geometry and yields a tractable route to structure-aware forward-policy training in GFlowNets. We illustrate the framework empirically through examples comparing convergence and exploration behaviour under Riemannian and Euclidean optimisation.
[LG-68] Robust Low-Tubal-Rank Tensor Completion under Cross-Concentrated Sampling
链接: https://arxiv.org/abs/2608.03928
作者: Hanqin Cai,Longxiu Huang,Jing Qin,Chengyue Wu
类目: Machine Learning (stat.ML); Information Theory (cs.IT); Machine Learning (cs.LG); Numerical Analysis (math.NA)
*备注:
Abstract:Tensor cross-concentrated sampling (t-CCS) bridges entrywise sampling and t-CUR slice-wise sampling by observing entries only within selected horizontal and lateral slices. Existing t-CCS completion methods, however, assume that the observations are free of gross corruption. In this work, we study robust recovery of a third-order low-tubal-rank tensor from partial t-CCS observations contaminated by sparse, arbitrarily large outliers. We propose Robust Iterative t-CUR (R-ItCUR), a tensor-native algorithm that partitions the sampled tensor cross into two exterior blocks and an intersection block, applies adaptive blockwise Welsch correction for outlier suppression, and updates the low-rank component through projected blockwise gradient descent. By operating directly on the sampled cross, R-ItCUR avoids reconstructing the full tensor throughout the iterations, resulting in substantial memory and computational savings. Experiments on synthetic tensors, cardiac MRI data, and three-dimensional seismic data demonstrate accurate recovery and strong robustness to sparse gross corruptions. The results further highlight the importance of explicitly exploiting the cross-concentrated sampling structure in robust tensor completion.
[LG-69] Divide-and-Conquer: Towards Generalizable Amortized Bayesian Inference for the Drift Diffusion Model
链接: https://arxiv.org/abs/2608.03566
作者: Yufei Wu,Shanqing Gao,Andreas Voss,Francis Tuerlinckx
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Applications (stat.AP); Methodology (stat.ME)
*备注:
Abstract:The drift diffusion model (DDM) is a cornerstone of cognitive decision-making research. Although numerous estimation methods exist, researchers continue to seek inference approaches that are both fast and flexible across diverse study designs. Amortized Bayesian inference (ABI) can provide nearly instantaneous inference for complex stochastic models like the DDM, but neural networks trained for one study design cannot generalize to others. In this paper, we propose a divide-and-conquer framework that address this limitation. The core idea is that the DDM’s independence assumption allows the full dataset to be decomposed into pairwise shards, each sharing a common structure that a single neural network can learn. Inference is performed on each shard separately and the resulting posteriors are combined via consensus MCMC to approximate the full posterior. Using simulated datasets, we evaluate the accuracy and uncertainty of this method. Our results show that the proposed divide-and-conquer approach achieves accuracy and uncertainty comparable to MCMC while reducing computational cost by several orders of magnitude. This work not only advances DDM estimation but also demonstrates a general strategy for improving the scalability and generalizability of ABI methods across diverse applications.
[LG-70] Should the Boundary Term Be Learned in Reflected Diffusion? Conormal Trace and Reflection Masking
链接: https://arxiv.org/abs/2608.03469
作者: Ziyue Wang,Takafumi Kanamori
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:We study score learning for reflected diffusion on bounded domains. Reflection keeps trajectories feasible but does not ensure that the learned score satisfies the boundary behavior implied by the forward process. With implicit score matching, integration by parts leaves a boundary term, and we show that it depends on one scalar at each boundary point: the diffusion- weighted normal component of the score, or conormal trace. The no-flux condition fixes this value while leaving the re- maining boundary components unrestricted; under anisotropic diffusion it generally differs from the ordinary normal score component. On hyperrectangles, our parametrization enforces the required trace without additional trainable parameters or a stochastic boundary estimator and, under regularity assump- tions, can represent the true score, whereas fixing an incorrect value creates an error that more data cannot remove. We ex- tend the construction to simplices and polygonal domains and identify reflection masking: hard reflection can keep samples feasible even when the learned trace is wrong, so post-reflection metrics may hide the error. Experiments show the clearest separation with less frequent reflection, anisotropic diffusion, and mass near intersections of constraints; under full reflection, final sample placement improves inconsistently, illustrating how hard repair can mask boundary-score errors and decouple score accuracy from downstream generation quality.
[LG-71] Conformal risk control for model-form uncertainty in parametric non-intrusive reduced-order models
链接: https://arxiv.org/abs/2608.03360
作者: Edgar Jaber(CB, ENS Paris Saclay),Rémy Vallot(CB, Michelin),Thibault Dairay(CB, Michelin),Mathilde Mougeot(CB, ENSIIE, ENS Paris Saclay)
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:Non-intrusive reduced-order models (NIROMs) have become a standard tool for approximating parametric partial differential equations from computer design of experiments while significantly reducing computational costs. However, assessing the reliability of their predictions remains a major challenge, particularly in extrapolation regimes or under limited training data. In this work, we introduce a framework for quantifying model-form uncertainty in NIROMs by combining a perturbative stochastic representation of reduced bases with distribution-free conformal-type methods. Starting from a deterministic reduced basis constructed from snapshot matrices, we model uncertainty through random perturbations defined on the Stiefel manifold, directed along the discarded modes, yielding stochastic reduced-order approximations whose induced variance reflects the basis-truncation error. A transport approximation gives a closed-form posterior variance that sepa- rates basis-induced from regression-induced uncertainty, without re-training the underlying Gaussian processes. We include this posterior variance within a conformal risk control calibration framework, that provides prediction sets with coordinate miscoverage guarantees. The calibration factor produced by this framework is itself an interpretable, scalar diagnostic of the quality of the uncertainty estimate. The methodology is evaluated on parametric PDE benchmarks and an industrial tire-manufacturing calendering process. Numerical experiments demonstrate reliable, locally informative uncertainty quantification that goes beyond the Gaussian predictive variance.
[LG-72] A Direct Route to Markov Chain Convergence via Asymptotic Equivalence with the Target
链接: https://arxiv.org/abs/2608.03353
作者: Patrick Forré
类目: Probability (math.PR); Machine Learning (cs.LG); Statistics Theory (math.ST); Computation (stat.CO); Machine Learning (stat.ML)
*备注:
Abstract:For a Markov kernel T with an invariant probability measure \pi , we give a self-contained proof of the Markov chain convergence theorem via a criterion called asymptotic equivalence with the target. It assumes two parts about the Lebesgue decompositions of T^n_x and \pi for every starting point x : 1.) asymptotic absolute continuity: the singular mass sing (T^n_x\mid\pi) tends to 0 ; 2.) asymptotic domination of the target: the singular mass sing (\pi\mid T^n_x) tends to 0 , as n \to \infty . This criterion, on countably generated measurable spaces, is both sufficient and necessary for the Markov chain convergence. A density version of this criterion is verified on general measurable spaces in three cases: (i) T has a positive transition density wrt \pi ; (ii) T consists of an absolutely continuous part with positive transition density together with an atom at the starting point, which covers the Metropolis–Hastings algorithm; (iii) the transition density is positive only after a finite number of steps that may depend on the starting point x . To demonstrate our general criterion, we investigate the Gibbs sampler with random scan and the parallel tempering algorithm. Furthermore, we show that in all mentioned settings Birkhoff’s ergodic theorem applies, so as to obtain the strong law of large numbers. Throughout this paper, neither irreducibility, nor aperiodicity, nor recurrence, nor couplings, nor splitting constructions, nor small sets are used. In most results, the state space is a general measurable space, which carries no structure beyond a \sigma -algebra. Countable generation is only assumed where the density-free form of the criterion is stated. None of the theorems proved here is new; what is offered is a short route to a single, widely applicable Markov chain convergence criterion, which is both sufficient and necessary.
[LG-73] ask-Oriented Candidate-Latent Feedback for Coarse-to-Fine Sensing in Distributed OFDM-ISAC Networks
链接: https://arxiv.org/abs/2608.03319
作者: Shiv Shankar,Radha Krishna Ganti,J Klutto Milleth
类目: ignal Processing (eess.SP); Machine Learning (cs.LG)
*备注:
Abstract:Future integrated sensing and communication (ISAC) architectures separate the sensing entity (SE) that acquires measurements from the sensing function (SF) that performs inference, creating a need for compact, task-oriented feedback on the SE-SF interface. Forwarding the raw channel frequency response or full per-link delay-Doppler-azimuth-elevation (DDAE) tensor is prohibitively expensive, while peak-only reporting discards target-discriminative structure under clutter. We propose a learning-based coarse-to-fine sensing pipeline with candidate-latent feedback for single-target estimation. At the SE, a lightweight convolutional scorer produces a dense delay-Doppler proposal map from pilot-based OFDM channel estimates, and a learned encoder constructs K compact C-dimensional candidate tokens by fusing per-candidate azimuth-elevation patches, normalized position, and confidence cues. The latents are uniformly quantized post-training to b bits and transmitted under a finite budget B_fb = bKC + 18K + 16 bits to the SF, which performs cross-candidate refinement, reranking, and joint four-parameter estimation. On a ray-traced urban scene with static and dynamic clutter, three operating points in the (K, C, b) design space achieve 96.33-98.88% detection at 107-806 bytes per coherent processing interval, compression ratios of 1.2-9.2 x 10^4 over the 8-bit DDAE magnitude tensor, reducing the SE-SF interface from multi-Gbit/s to sub-Mbit/s rates. Cross-scene evaluation on an independent campus-scale environment achieves 98.79-99.50% detection and at-or-better angular accuracy without retraining, indicating that the learned representation captures target-relevant structure that transports across scenes of comparable or lower clutter density.
[LG-74] Automatic Patient-Specific Microwave Ablation Planning Accelerated by a Physics-Guided Deep Learning Model MICCAI2026 ALT
链接: https://arxiv.org/abs/2608.03086
作者: Seonaeng Cho,Minjee Seo,Minju Seol,Juil Park,Joon Ho Kwon,Kyungho Yoon
类目: Image and Video Processing (eess.IV); Machine Learning (cs.LG)
*备注: Accepted at the Digital Twin for Healthcare (DT4H 2026) workshop at MICCAI 2026; 10 pages, 2 figures
Abstract:Microwave ablation (MWA) is a promising minimally invasive treatment for liver tumors, but its therapeutic outcome strongly depends on patient-specific planning of antenna insertion trajectory, power, and treatment duration. Accurate numerical simulation can provide physically reliable ablation predictions; however, its high computational cost limits its use in optimization-based planning, where repeated forward evaluations are required. To address this issue, we propose a digital twin-based automatic planning framework that combines a neural ablation prediction model with a genetic algorithm. The model was trained on multiphysics simulation data generated from patient-specific tumor and vessel structures, antenna configurations, and treatment conditions, and was used as a fast forward model during planning. The prediction model achieved a Dice score of 95.1%, enabling accurate deep learning-based optimization. In 13 unseen planning cases, the proposed method improved ablation efficiency by 54.3% and reduced organ damage by 55.0% compared with clinician-defined planning, while slightly shortening the insertion path length by 3.3%. Most generated plans were also judged clinically applicable by MWA specialists. Furthermore, the framework enabled approximately 420-fold faster planning than numerical-simulation-based planning, demonstrating its potential as a fast digital twin for quantitative and personalized MWA treatment planning. The code is available at: this https URL
[LG-75] Causal Inference with Unstructured Outcomes
链接: https://arxiv.org/abs/2608.03085
作者: Kevin Christian Wibisono,Yixin Wang
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注: 68 pages, 10 figures
Abstract:Causal inference has traditionally centered on scalar outcomes: whether a patient recovers, how much a worker earns, or how many visits a website receives. Modern studies increasingly ask causal questions about outcomes with richer form, such as clinical notes, open-ended survey responses, and images. A hospital may want to know how an AI documentation tool changes the notes physicians write, or how a nurse training program alters what patients say in survey responses. For such outcomes, the usual average treatment effect is ill-defined: one cannot meaningfully subtract one text or image from another. To this end, we propose a causal query for unstructured outcomes. The key idea is to learn what features of the outcome are most causally affected by the treatment, which we call the maximally contrasting feature (MCF). To estimate the MCF, we learn a feature-scoring function that maps each outcome to a scalar and exposes the sharpest contrast between treated and control potential outcomes. We develop identification conditions and estimation algorithms for this query, and extend it to heterogeneous effects by allowing the feature-scoring function to depend on observed covariates. We also handle settings where both the treatment and the outcome are unstructured. Empirical studies on text and images show that the algorithm recovers salient aspects of an outcome changed by a treatment.
[LG-76] Stochastic Saddle Avoidance Beyond Unit Excitation and Smoothness: A Pathwise Lyapunov-Perron Framework
链接: https://arxiv.org/abs/2608.03001
作者: Junwen Qiu,Bohao Ma,Andre Milzarek,Junyu Zhang
类目: Optimization and Control (math.OC); Machine Learning (cs.LG); Dynamical Systems (math.DS); Machine Learning (stat.ML)
*备注: 38 pages
Abstract:Unit excitation (UE) is a common assumption in stochastic saddle avoidance: the stochastic error must have a uniformly positive component along every direction, in expectation. This condition gives a direct way to rule out convergence to strict saddles, but it also oversimplifies the actual noise structure, and does not match many stochastic optimization regimes. In overparameterized or interpolation models, the noise may vanish near stationarity. In finite-sum problems, the stochastic gradient noise may lie in a low-dimensional, data-dependent subspace. In these (common) scenarios, UE is naturally not satisfied. In this paper, we prove an abstract almost sure avoidance theorem for stochastic recursions without UE. The theorem replaces UE-type requirements by verifiable pathwise conditions. In applications, these conditions follow, e.g., from local smoothness and finite-moment assumptions under standard i.i.d. sampling, or from the finite-sum structure under without-replacement sampling. Since the stochastically sampled maps generally do not share a fixed point, the celebrated center-stable manifold argument used in deterministic analyses is not directly applicable. Instead, we use a path-dependent change of variables together with a pathwise Lyapunov–Perron-based proof strategy. As applications, we obtain strict saddle avoidance for stochastic mirror descent (including SGD) and for random reshuffling. For nonsmooth composite objectives, we prove avoidance results for a proximal-type stochastic gradient method. Combining these insights with suitable iterate convergence guarantees, this allows establishing convergence to local minimizers of the original objective function.
[LG-77] ScoreField: Neural Inverse Scattering with Score-Based Generative Priors
链接: https://arxiv.org/abs/2608.02937
作者: Wenhan Guo,Yuan Gao,Yu Sun
类目: Image and Video Processing (eess.IV); Machine Learning (cs.LG); Computational Physics (physics.comp-ph)
*备注: 28 pages, 9 figures
Abstract:Designing an effective electromagnetic inverse-scattering solver requires faithful enforcement of nonlinear full-wave physics together with an expressive prior on the unknown permittivity contrast. We propose ScoreField, a neural inverse scattering framework that integrates coupled implicit neural representations (INRs) with a pretrained score-based generative prior. ScoreField employs two INRs to parameterize the permittivity contrast and the induced current fields, and jointly optimize them under the Lippmann-Schwinger equations. In addition to the implicit regularization by the INR architecture, the score model provides a learned prior gradient on the contrast, which is propagated to the contrast INR through the chain rule. This formulation enables ScoreField to effectively handle strong multiple scattering, where nonlinear wave interactions require accurate modeling of the coupled full-wave physics. We evaluate ScoreField on simulated weak- and strong-scattering benchmarks, the canonical Austria phantom, and experimental Fresnel measurements. We note that ScoreField significantly improves reconstruction fidelity and suppresses artifacts relative to classical full-wave methods and deep learning baselines, achieving an average PSNR improvement of 1.8 , \mathrmdB over the best competing method on real Fresnel data.
[LG-78] Particle-based Generalised Stochastic Optimisation
链接: https://arxiv.org/abs/2608.02844
作者: Jiechen Jackie Zhang,O. Deniz Akyildiz
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Computation (stat.CO)
*备注:
Abstract:We develop a class of diffusion-based stochastic particle optimisation methods for loss functions with intractable gradients. Specifically, we consider problems in which the loss gradient is an integral with respect to a parameter-dependent distribution, a structure that includes training generative models, fine-tuning, and learning latent-variable models. We introduce mean-field dynamics and its interacting-particle approximations, which contain several existing algorithms as special cases and provides a route to constructing new methods. Under well-posedness and joint contractivity assumptions, we prove exponential convergence and show that the continuous-time particle system admits a non-asymptotic error bound. We illustrate it by developing momentum and higher-order Langevin variants and evaluating them on maximum marginal-likelihood estimation and energy-based-model training.
[LG-79] Detecting high-frequency brain disorder signals using dynamic mode decomposition from EEG
链接: https://arxiv.org/abs/2608.02804
作者: Jacob Kang,Jong-Hyeon Seo
类目: Neurons and Cognition (q-bio.NC); Machine Learning (cs.LG); Signal Processing (eess.SP)
*备注: 15 pages, 6 figures, 3 tables
Abstract:Recent studies have reported clearly identifiable dynamical changes in the high-frequency range of EEG signals recorded during specific stimuli, such as visual or auditory inputs, or in cases of brain disorders like epileptic seizures. In this study, we utilized Dynamic Mode Decomposition (DMD) to extract consistent and persistent dynamical changes in the high-frequency band from the signals of neurologically relevant EEG channels. High-frequency DMD modes were employed as features, composing a feature table. Through post-processing, a random distribution test was performed, revealing that approximately 70% of the samples exhibited consistent high-frequency dynamics within the signal of a specific channel. Furthermore, classification experiments confirmed that the PCA components of the feature table that passed the test formed a consistent pattern that distinguished the alcohol-dependent group from the control group.
[LG-80] DAIF: A Data-Driven Intermediate Fusion Framework for Multimodal Supervised Learning via Approximate Message Passing
链接: https://arxiv.org/abs/2608.02769
作者: Sagnik Nandy,Samriddha Lahiry,Pragya Sur,Subhabrata Sen
类目: Methodology (stat.ME); Machine Learning (cs.LG); Statistics Theory (math.ST); Machine Learning (stat.ML)
*备注:
Abstract:Multimodal supervised learning seeks to leverage multiple heterogeneous data sources to improve predictive performance. A central challenge is determining the fusion granularity across modalities: over-integration may amplify noise while under-integration fails to exploit cross-modal dependence. Existing approaches rely on pre-specified fusion architectures, from early to late fusion, that may not adapt to the underlying dependence structure among modalities. We propose DAIF, a data adaptive intermediate fusion framework that combines random matrix theory and non-parametric dependence measures to learn fusion structure directly from data. We operate under a Bayesian multimodal factor model where the prior on the latent factors determines the cross-modal dependence. Our method clusters modalities based on estimated intermodal dependence, then performs clusterwise empirical Bayes estimation of the priors. These estimated priors are used to construct denoisers within an approximate message passing (AMP) framework, yielding denoised low-dimensional features that borrow strength across related modalities while preserving modality-specific signal. The resulting embeddings are used for downstream supervised prediction. We evaluate the framework through simulations under varying dependence structures and signal regimes, comparing against several benchmark methods, and demonstrate its practical utility on two multimodal datasets, namely a trimodal TEA-seq dataset (Swanson et al., 2021) and TCGA-BRCA dataset (Goldman et al., 2020). In the first example, we predict the expression level of a T-cell differentiation marker protein and in the second case we analyze patient survival prediction based on multimodal information. Our method competes with or outperforms the state-of-the-art techniques in both prediction problems, demonstrating its versatility across diverse supervised learning tasks.
附件下载
