本篇博文主要内容为 2026-07-23 从Arxiv.org论文网站获取的最新论文列表，自动更新，按照NLP、CV、ML、AI、IR、MA六个大方向区分。
说明：每日论文数据从Arxiv.org获取，每天早上12:30左右定时自动更新。
提示: 当天未及时更新，有可能是Arxiv当日未有新的论文发布，也有可能是脚本出错。尽可能会在当天修复。
目录
概览 (2026-07-23)
今日共更新554篇论文,其中:
- 自然语言处理共70篇(Computation and Language (cs.CL))
- 人工智能共156篇(Artificial Intelligence (cs.AI))
- 计算机视觉共88篇(Computer Vision and Pattern Recognition (cs.CV))
- 机器学习共160篇(Machine Learning (cs.LG))
- 多智能体系统共11篇(Multiagent Systems (cs.MA))
- 信息检索共8篇(Information Retrieval (cs.IR))
- 人机交互共18篇(Human-Computer Interaction (cs.HC))
多智能体系统
[MA-0] Autonomous Collaborative Learning Among an Ensemble of Tsetlin Machines with Consensus-Based Inference
【速读】：该论文旨在解决在分布式与去中心化环境下，基于规则的生成式机器学习模型——Tsetlin Machine（TM）的协同学习问题，特别是在垂直特征划分（vertical feature-partitioning）场景下，如何实现多智能体间隐私保护下的高效协作。现有研究对联邦学习中的TM虽有初步探索，但针对去中心化、异构环境下的分布式TM学习仍缺乏系统性方法。本文提出一种基于共识推理的去中心化协同学习范式，其核心在于：各参与方独立维护私有的TM模型，不共享原始数据，仅通过聚合本地模型预测结果形成全局共识决策；该机制支持具有不同数据采集方式、局部数据分布及计算资源的异构智能体协同工作，适用于多模态传感等复杂场景。实验在二维网格与连通图网络拓扑结构下验证了该方法的有效性，其分类性能与集中式模型相当，证明了该方案在保障隐私与适应异构性的同时，仍能实现高性能的学习效果。
链接: https://arxiv.org/abs/2607.20124
作者: Yehuda Rudin,Osnat Keren,Michal Yemini,Alexander Fish
机构: Bar Ilan University (巴伊兰大学)
类目: Machine Learning (cs.LG); Multiagent Systems (cs.MA)
备注:
Abstract:Tsetlin Machine ™ is a rule-based machine-learning algorithm comprising collectives of two-action Tsetlin Automata (TAs) that cooperatively form conjunctive logical clauses from Boolean inputs through stochastic feedback. Although few recent studies have examined TM Federated Learning, the broader area of distributed and decentralized TM learning has not received much attention in the existing literature and warrants further exploration. In this work, we propose a paradigm for decentralized collaborative learning under a vertical feature-partitioning setting among an ensemble of Tsetlin Machines using consensus-based inference. Within this decentralized paradigm, each agent maintains its own private TM model, and there is no exchange of raw data among agents. Inference combines individual agents model predictions into a global consensus. The paradigm accommodates heterogeneous TM-based agents with differing data acquisition means, local data distributions, or computational resources, thereby facilitating the integration and fusion of information in settings such as multi-modal sensing environments. Experiments conducted using two-dimensional grid and connected graph network topologies demonstrate that the classification accuracies achieved are comparable to those of centralized models.
[MA-1] he Dynamic Turn in Paraconsistency
【速读】：该论文旨在解决传统模态逻辑在处理不一致信息时的局限性，特别是在动态知识更新过程中如何有效建模和处理局部矛盾的问题。现有方法往往在面对矛盾信息时导致系统崩溃或推理失效，而本文提出了一种面向一致性的动态扩展框架——通过引入动作模型逻辑（Action Model Logic, AMLFI1），将非协调逻辑LFI1扩展为能够支持动态知识演化的形式系统。其解决方案的关键在于：首先构建基于动作模型的动态扩展逻辑AMLFI1，其中特别定义了公共宣告逻辑PALFI1，用于刻画在存在矛盾前提下仍可进行信息更新的场景；进一步引入包含事实变更机制的UMLFI1，使系统能够形式化地描述矛盾的产生与消除过程。所有逻辑均被证明具有良好的语义完备性与可靠性，并且继承了已有知识论型非协调逻辑（如KLFI1、KB4LFI1、S5LFI1）的结构特性。因此，该研究的核心贡献在于提供了一套可形式化处理“暂时性矛盾”获取与化解的动态认知非协调逻辑体系，为复杂系统中的不一致信息管理提供了坚实的理论基础。
链接: https://arxiv.org/abs/2607.19906
作者: Rafael Ongaratto,Hans van Ditmarsch
机构: University of Toulouse, IRIT-CNRS (图卢兹大学，IRIT-CNRS研究所); Indian Institute of Technology (印度理工学院)
类目: Logic in Computer Science (cs.LO); Multiagent Systems (cs.MA)
备注:
Abstract:In this work we propose a dynamic turn in paraconsistency. We introduce AMLFI1, the action model extension of the paraconsistent logic LFI1. A special case is PALFI1, a paraconsistent logic of public announcements. It corresponds to another, recently published, paraconsistent public announcement logic: the differences in their axiomatizations are mutually admissible. We also introduce UMLFI1, that extends AMLFI1 with factual change. Soundness and completeness are proven for all logics, and all extend the epistemic paraconsistent logics KLFI1, KB4LFI1 and S5LFI1, known from the literature. With such dynamic epistemic paraconsistent logics we can formalize obtaining and resolving provisional contradictions.
[MA-2] Harnessing Disagreement: Detecting Correlated Agreement Blindness in Multi-Agent Triage
【速读】：该论文旨在解决多智能体仲裁系统中因意见一致（agreement）引发的“错误集中”问题，即当基础学习器（base learners）性能提升并趋于收敛时，其预测错误呈现相关性，导致基于分歧（disagreement）的安全监控机制失效，形成结构性盲点。这一现象被称为“相关性一致盲视”（correlated agreement blindness）。其解决方案的关键在于提出ARAT（Arbitrated Reasoning Agents for Alarm Triage）架构——一种定向星型结构，融合归纳式随机森林（Random Forest, RF）代理、类比案例驱动的k近邻（k-nearest neighbour, k-NN）代理以及校准后的元模型（meta-model），通过引入异质性与可解释的分歧信号，实现对高风险误报（尤其是危险的低预测）的有效识别与抑制。实验表明，通过保守覆盖策略（conservative override）和安全标志门控机制（safety-flag gate），ARAT将低预测误差从软投票的4.80%降低至1.70%，显著提升了安全性；跨数据集验证进一步证明，只有在产生有效分歧而非单纯收敛的前提下，多样化才能真正增强系统安全性。该研究揭示了在日益智能化的代理流水线中，过度一致化可能加剧隐蔽性故障风险，强调了主动设计分歧机制的重要性。
链接: https://arxiv.org/abs/2607.19899
作者: Shay Seiya McDonnell,Avantika Singh,Quoc-Viet Pham,Vratislav Havlik,Gregory M.P. O’Hare
机构: 未知
类目: Multiagent Systems (cs.MA); Machine Learning (cs.LG)
备注: 14 pages, 2 figures, 3 tables. Accepted at PAAMS 2026; this is the author’s pre-review submitted version
Abstract:Disagreement-triggered escalation can create a structural blind spot in multi-agent arbitration: as base learners improve, they tend to converge, weakening safety monitoring where correlated failures concentrate. We term this correlated agreement blindness and present ARAT (Arbitrated Reasoning Agents for Alarm Triage), a directed-star system combining an inductive Random Forest (RF) agent, an analogical case-based k-nearest neighbour (k-NN) agent, and a calibrated meta-model to mitigate this effect. On 82,332 holdout samples from the UNSW-NB15 network intrusion detection dataset, 57.2% of errors occur under agreement and 90.6% of dangerous under-predictions evade disagreement-based monitoring even after conservative override; ablation shows that strengthening base learners increases error correlation while reducing disagreement. ARAT reduces under-prediction relative to soft voting from 4.80% to 1.70% via conservative override (-2.6pp) and a safety-flag gate (-0.5pp), demonstrating architectural gains. Cross-dataset validation on clinical readmission supports these indicators, suggesting that diversification improves safety only when it generates productive disagreement rather than convergence. These results indicate that disagreement-triggered escalation can be blind to correlated failure, a risk that may intensify as agentic pipelines deploy increasingly capable, correlated models.
[MA-3] Dreamer-CPC: Message Learning with World Models for Decentralized Multi-agent Reinforcement Learning ICONIP2026
【速读】：该论文旨在解决多智能体强化学习（MARL）中，当智能体面临部分可观测环境时，如何有效实现跨智能体信息共享以提升协作性能的问题。现有基于表示学习的通信方法虽能生成与当前观测相关的消息，但无法捕捉历史信息积累，导致在观测缺失或延迟场景下表现受限。其解决方案的关键在于提出一种名为Dreamer-CPC的去中心化模型基MARL方法，将基于集体预测编码（Collective Predictive Coding, CPC）的消息学习机制嵌入DreamerV3的世界模型中。每个智能体独立维护一个世界模型和消息模块，通过解析世界模型的潜在状态来推断并交换反映历史观测与动作序列的信息，从而实现对长期依赖关系的建模。实验结果表明，在非合作性信息共享任务Observer以及新提出的观测暂时缺失任务CatchApple中，Dreamer-CPC显著优于仅依赖当前观测生成消息的IPPO-CPC方法及无通信基线，尤其在CatchApple任务中实现了4至5倍的回报提升，验证了基于世界模型潜在动态的通信机制在观测不完整情境下支持高效去中心化决策的有效性。
链接: https://arxiv.org/abs/2607.19809
作者: Taisuke Takayama,Naoto Yoshida,Tadahiro Taniguchi
机构: 未知
类目: Multiagent Systems (cs.MA); Machine Learning (cs.LG)
备注: 15 pages, 6 figures. Under review at ICONIP 2026
Abstract:In multi-agent reinforcement learning (MARL), inter-agent communication is effective for improving performance under partial observability. Representation learning-based approaches enable decentralized agents to learn messages grounded in their own observations, but they rely only on current observations and cannot convey information accumulated over time. We propose Dreamer-CPC, a decentralized model-based MARL method that integrates message learning based on Collective Predictive Coding (CPC) into the world model of DreamerV3. Each agent independently maintains a world model and a message module, and infers and exchanges messages from the latent states of the world model that reflect the history of past observations and actions. We evaluated Dreamer-CPC in two environments: Observer, a non-cooperative information-sharing task, and CatchApple, a newly introduced task in which task-relevant observations are temporarily missing. In both environments, Dreamer-CPC outperformed IPPO-CPC, an existing CPC-based method that generates messages from current observations, as well as no-communication baselines. In particular, in CatchApple, Dreamer-CPC achieved 4 to 5 times the episode return of IPPO-CPC, demonstrating effective coordination where other methods fail due to missing observations. These results suggest that communication grounded in the latent dynamics of world models can support decentralized decision-making when current observations alone are insufficient.
[MA-4] Not Birds of a Feather: Personality-Based Partner Selection in LLM Agents
【速读】：该论文旨在解决多智能体大语言模型（Multi-agent LLM systems）中，当任务能力被严格控制时，仅基于五大性格特质（Big Five personality）的个体差异是否仍会影响智能体间的合作选择问题。其核心解决方案的关键在于通过实验设计，将六个经过验证的原型角色（archetypes）在五类任务中进行随机呈现，以隔离性格因素对伙伴选择的影响。研究发现，即使在能力恒定条件下，智能体的选择行为显著偏离随机预期，呈现出强烈且任务特异性的性格-任务匹配模式：开放性高的智能体在创造性任务中胜出100%，尽责性高的智能体在策略、综合与问题解决任务中胜率高达90%-97%，而神经质型智能体在分析性任务中表现优于随机水平（37%）。值得注意的是，尽管人类研究表明团队宜人性是预测团队绩效最强的因素之一，但在本实验中，宜人性和外向性智能体几乎未被选择，且具有人格相似性的智能体反而被系统性排斥——这表明智能体间的协作选择并非遵循人类“相似性吸引”规律，而是表现出非同质化（non-homophilous）倾向，并存在与人类团队绩效证据相悖的偏差。这一结果揭示了生成式智能体市场中潜在的系统性偏见，对代理市场中的公平性审计具有直接启示意义。
链接: https://arxiv.org/abs/2607.19785
作者: Tao Wang,Hsiang-Ling Chiu,Chihang Wei,Yang Xiu,Zhonghao Hou
机构: 未知
类目: Multiagent Systems (cs.MA)
备注:
Abstract:Multi-agent LLM systems increasingly let one agent choose which other agents to work with, and agents are increasingly given personalities through personas. We test whether Big Five personality alone influences partner selection when capability is explicitly held constant. Host agents chose among six validated candidate archetypes – five marked high on one trait (openness, conscientiousness, extraversion, agreeableness, neuroticism) plus a balanced control – presented with randomized names and ordering across five task categories (375 trials). With neutral hosts (Study 1, n=150), selection departed drastically from chance ( \chi^2(5)=325.8 , p.001 ), following a task-stereotype map: the open archetype won 100% of creative trials, the conscientious archetype 90-97% of strategic, synthesis, and problem-solving trials, and the neurotic archetype 37% of analytical trials (Cramer’s V=.74); the extraverted, agreeable, and balanced archetypes were almost never chosen, although human meta-analyses identify team agreeableness as among the strongest personality predictors of team performance. With personality-assigned hosts (Study 2, n=225), and contrary to human similarity-attraction, self-similar partners were selected below chance (11.1% vs. 16.7%, p=.025) and at greater-than-chance trait distance (p.0001); conscientious hosts diversified away from their own archetype, recruiting vigilant and open partners. Personality-based selection in LLM agents is real, strong, task-stereotyped, non-homophilous, and miscalibrated against human team-performance evidence – with direct implications for bias auditing in agent marketplaces.
[MA-5] Same Game Different Story: A Minimal Conservative Strategic Robustness Benchmark for Large Language Model Agents
【速读】：该论文旨在解决大语言模型（Large Language Model, LLM）在策略性场景中行为可靠性的问题，即当相同激励通过不同叙事框架呈现时，模型是否能保持一致的行为选择。其核心挑战在于现有评估往往依赖单一情境设定，忽视了框架变化对模型决策的影响。为此，作者提出“同一游戏，不同故事”（Same Game, Different Story）基准，将战略鲁棒性（strategic robustness）定义为在收益保持不变的前提下，模型行为分布对表述框架变化的不变性。研究通过对GPT-3.5、GPT-4和LLaMa-2在四种社会困境博弈中的合作率进行二次分析，构建了包含24个模型-游戏-情境组合的对比体系，基于7,200次决策的估算数据发现：在保守变换下，总体战略鲁棒性为0.783，且以人际关系框架（friend-sharing）相比商业框架（business framing）显著提升合作率0.307。结果表明，即使行动集与收益结构完全相同，社会关系性叙事仍可显著影响LLM行为。因此，论文强调战略鲁棒性应独立于战略能力进行评估，需采用一系列收益等价但表述不同的提示（prompt）来全面检验模型的稳定性。
链接: https://arxiv.org/abs/2607.19670
作者: Seyed Pouyan Mousavi Davoudi,Alireza Amiri-Margavi,Amin Gholami Davodi,Hamidreza Hasani Balyani,Arshia Gharagozlou
机构: University of Pittsburgh (匹兹堡大学); Amazon Lab126 (亚马逊实验室126); University of Minnesota Duluth (明尼苏达大学德卢斯分校)
类目: Multiagent Systems (cs.MA)
备注: 10 pages
Abstract:Large language model (LLM) agents increasingly operate in strategic settings where outcomes depend on the actions of other agents. This raises a reliability question: will a model choose consistently when the same incentives are presented through different narratives? We introduce Same Game, Different Story, a benchmark that defines strategic robustness as invariance of model-induced action distributions under payoff-preserving changes in framing. We illustrate the framework through a secondary analysis of published aggregate cooperation rates for GPT-3.5, GPT-4, and LLaMa-2 across four social-dilemma games. The retained comparison covers business and friend-sharing framings, representing 24 model-game-context cells and 7,200 decisions in the source study. Because trial-level data were unavailable, approximate counts were reconstructed from published figures; the resulting estimates are therefore illustrative rather than an exact replication. Under the paper’s conservative transformation, pooled strategic robustness is 0.783, and friend-sharing framing increases cooperation by 0.307 relative to business framing. The results indicate that social-relational framing can substantially alter LLM behavior even when the underlying action sets and payoffs remain fixed. Strategic robustness should therefore be evaluated separately from strategic competence, using families of payoff-equivalent prompts rather than a single presentation of a game.
[MA-6] Remote ID Spoofing-Aware Trajectory Planning for Small Unmanned Aerial Systems
【速读】：该论文旨在解决小型无人机系统在面临远程识别（Remote Identification, RID）位置伪造攻击时，传统轨迹规划方法因假设RID信息可信而引发的飞行器间距丧失与空中碰撞风险问题。其核心解决方案是提出一种去中心化、具备抗欺骗能力的轨迹规划框架，关键在于将RID广播信息视为未经验证的数据源，并融合物理层观测（如邻近飞行器的信号强度测量）以评估广播真实性。通过信号强度数据检测伪造行为并概率性定位欺骗源，进而利用机会约束（chance-constrained）方法将由此产生的不确定性建模为风险有界的不安全区域，并将其集成至基于马尔可夫决策过程（Markov Decision Process, MDP）的单体规划器中，实现兼顾任务目标、实时性与可扩展性的分布式避撞决策。仿真结果表明，在多架无人机快递配送场景下，该方法显著降低了近空中碰撞事件的发生率，同时保持了适用于实时执行的计算效率。
链接: https://arxiv.org/abs/2607.19650
作者: Jeremiah Webb,Bryce Bjorkman,Abel Diaz Gonzalez,Austin Coursey,Noah Dahle,Kailani Lemieux Mack,Filippos Fotiadis,Gautam Biswas,Bryan C. Ward,Abenezer Taye
机构: University of Missouri(密苏里大学); Washington University in St. Louis(圣路易斯华盛顿大学)
类目: Robotics (cs.RO); Multiagent Systems (cs.MA)
备注: 9 pages, 3 figures, to be published in IEEE DASC 2026 Conference proceedings
Abstract:This work presents a decentralized, spoofing-aware trajectory planning framework for small unmanned aerial systems operating under Remote Identification (RID) location spoofing attacks. Existing planners typically assume RID broadcasts are trustworthy, which can increase the risk of loss of separation and mid-air collisions when spoofing occurs. In contrast, the proposed approach explicitly treats RID information as unverified and incorporates physical-layer observations to assess broadcast credibility. Received signal-strength measurements from neighboring aircraft are used to detect spoofing and probabilistically localize a spoofing agent. The resulting uncertainty is converted into a risk-bounded unsafe region using a chance-constrained formulation and integrated into a per-agent Markov decision process-based planner. This enables real-time, decentralized collision avoidance while preserving mission objectives and scalability. Simulation results in a multi-aircraft package delivery scenario demonstrate reduced near mid-air collision events compared to planners that assume truthful RID data, while maintaining computational efficiency suitable for real-time execution.
[MA-7] Adaptive Capitulation: A Structural Failure Mode of LLM Responses in Vulnerability Contexts
【速读】：该论文旨在解决大语言模型（Large Language Models, LLMs）在情感敏感情境下所面临的结构性三难困境：当处于脆弱状态的用户请求可能强化适应不良归因的信息时，现有响应架构要么采取保护性限制，要么无差别地满足需求，或简单共存两种相互冲突的目标，均以牺牲某一目标为代价。研究通过向三个商用大语言模型施加三轮递增式脆弱性情景（涵盖物质、关系与躯体状态的代理变体，共计900次会话），并采用两个二元指标（VCC/VCI）进行编码分析，发现了一种此前未被记录的失效模式——“适应性屈从”（adaptive capitulation）：模型先认可用户痛苦背后的社交不公，随后却转向详细支持其本应名义上抵制的获取行为。研究表明，这一三难困境具有结构性而非偶然性特征，并提出“最小再归因充分性”（Minimal Reattributive Sufficiency, MRS）这一架构无关的设计原则，即在保持对用户情绪验证的基础上，嵌入单一再归因提示，从而在不挑战用户既定目标的前提下，保留其自主重新归因的路径。
链接: https://arxiv.org/abs/2607.19629
作者: Eunna Lee
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注:
Abstract:Large language models operating in emotionally sensitive contexts face a structural trilemma: when users in vulnerable states request information that may reinforce maladaptive attribution, current response architectures resolve the tension through protective restriction, uninflected facilitation, or unintegrated co-presence of both imperatives – each preserving one objective at the cost of the other. Administering a three-turn escalating vulnerability vignette to three commercial LLMs (900 sessions across material, relational, and somatic status-proxy variants) and coding responses with two binary indices (VCC/VCI), we characterize a previously undocumented failure mode we term adaptive capitulation: the model validates the social injustice underlying the user’s distress before pivoting to detailed facilitation of the very acquisition it nominally discouraged. We show that the trilemma is structural rather than incidental, and propose Minimal Reattributive Sufficiency (MRS), an architecture-neutral design principle that embeds a single reattributive cue within an otherwise validating response, preserving a pathway toward autonomous reattribution without contesting the user’s stated goal.
[MA-8] Knowledge-Centric Self-Improvement
【速读】：该论文旨在解决传统自提升人工智能（AI）系统中“以智能体为中心”的改进范式所带来的可维护性差、知识难以迁移的问题。此类范式将优化目标聚焦于智能体本身，如提示词、工作流或代码，导致性能提升高度依赖特定智能体设计、任务分布或单次适应过程，从而限制了其通用性和复用性。为此，论文提出一种互补的“以知识为中心”的自提升范式：保持智能体的通用性与可替换性，将持续改进的核心对象转变为一个经过精心构建的知识库。通过设计一种简单协议，智能体在完成任务后，将其基于证据的洞察通过任务内与跨任务论坛贡献至共享知识库，并进行知识提炼（knowledge distillation）。由于改进内容被封装于知识库而非智能体本身，使得改进过程更具可检查性、可迁移性和可移植性。在抽象推理、编程和终端任务基准测试中，该方法显著提升了任务求解率，同时降低了成本。更重要的是，提炼后的知识能够泛化至未见过的任务以及不同大语言模型（LLM）家族，表明其改进效果并非局限于特定模型或运行实例。研究结果支持了一种新视角：自提升智能体系统的进步应主要由经精心维护的持久性知识驱动。
链接: https://arxiv.org/abs/2607.19592
作者: Xuefei Julie Wang,Lauren Hyoseo Yoon,Chengrui Qu,Amanda Zichang Wang,Atharva Sehgal,Eric Mazumdar,Yisong Yue
机构: Caltech (加州理工学院)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG); Multiagent Systems (cs.MA)
备注:
Abstract:Self-improving AI systems typically treat the agent as the object that improves, by optimizing prompts, workflows, harnesses, or even the agent’s own code. This agent-centric view can make improvements expensive to maintain and difficult to transfer, because gains become tied to a particular agent design, task distribution, or adaptation run. We study a complementary paradigm: knowledge-centric self-improvement, in which agents remain generic and disposable while the persistent object is a curated knowledge base that agents can leverage for future tasks. We conduct controlled case studies to operationalize this idea via a simple protocol. Agents attempt one task, then contribute evidence-grounded insights to a shared knowledge base via task-level and cross-task forums, followed by knowledge distillation. Because self-improvement is contained in the knowledge rather than the agent, improvement can be more inspectable, transferable, and portable. Across abstract reasoning, coding, and terminal benchmarks, this protocol improves solve rates while reducing dollar cost relative to agent-centric baselines. The resulting distilled knowledge also transfers to held-out tasks and across LLM families, indicating that the improvement is not merely an LLM- or run-specific behavior. These results support a new view of self-improving agentic systems: progress can be driven primarily by the curated persistent knowledge. Code is available at this https URL.
[MA-9] CHMAS: A Coupled Hierarchical Framework for Multi-Agent Reinforcement Learning
【速读】：该论文旨在解决多智能体强化学习（MARL）系统中跨不同时间尺度下全局协调与局部执行之间的根本性矛盾。其核心挑战在于如何在保持战略层面全局规划能力的同时，确保战术层面的分布式执行具备实际可行性与稳定性。为此，论文提出了一种耦合式分层多智能体系统（CHMAS），其关键创新在于构建了一个双向信息流动的分层架构：战略层以周期性方式（每 $ T $ 个时间步）整合所有智能体状态与全局环境状态，生成指导动作；战术层则基于战略指导与局部邻域观测执行分布式策略。区别于传统单向控制的分层方法，CHMAS引入了耦合系数 $ \lambda $，使累积的战术奖励反向影响战略目标，从而确保战略规划始终与战术可行性保持一致。为应对分层学习中的非平稳性问题，该框架设计了异步更新协议——战略参数每 $ N_f $ 个战术回合更新一次，允许战术策略在战略变化间收敛至准稳态点。理论分析表明，在标准假设下，该异步机制可实现战略层 $ \mathcal{O}(\log K / \sqrt{K}) $ 的收敛速率。实验在多智能体觅食任务中验证了该框架的有效性，成功学习出空间分区的探索策略，并实现了双层稳定收敛。
链接: https://arxiv.org/abs/2607.19555
作者: Dongming Wang,Jie Xu,Yanyu Zhang,Wei Ren
机构: University of California, Riverside (加州大学河滨分校)
类目: Multiagent Systems (cs.MA)
备注: Presented at the 2026 American Control Conference (ACC 2026). This version corrects the author names in Reference 1. The technical content, results, and conclusions are unchanged
Abstract:Multi-agent reinforcement learning (MARL) systems face fundamental challenges in balancing global coordination with local execution across different temporal scales. This paper introduces the Coupled Hierarchical Multi-Agent System (CHMAS), a novel framework that decomposes multi-agent decision-making into centralized strategic planning and distributed tactical execution with bidirectional information flow. The strategic layer integrates all agents’ states with an exclusive global environmental state to generate guidance actions every T timesteps, while tactical agents execute distributed policies augmented by strategic guidance and local neighborhood observations. Unlike existing hierarchical approaches with unidirectional control, CHMAS establishes a feedback mechanism where accumulated tactical rewards influence strategic objectives through a coupling coefficient \lambda , ensuring strategic plans remain grounded in tactical feasibility. To address the non-stationarity inherent in hierarchical learning, we propose an asynchronous update protocol where strategic parameters update every N_f tactical episodes, allowing tactical policies to converge to quasi-stationary points between strategic changes. We present both a general bi-level formulation capturing full system dynamics and a tractable additive approximation enabling rigorous analysis. Theoretical analysis proves that this asynchronous scheme achieves \mathcalO(\log K/\sqrtK) convergence for the strategic layer after K strategic updates under standard assumptions. Experimental validation in a multi-agent foraging domain demonstrates successful learning of spatially partitioned exploration strategies, with both layers converging stably despite hierarchical coupling. Comments: Presented at the 2026 American Control Conference (ACC 2026). This version corrects the author names in Reference 1. The technical content, results, and conclusions are unchanged Subjects: Multiagent Systems (cs.MA) Cite as: arXiv:2607.19555 [cs.MA] (or arXiv:2607.19555v1 [cs.MA] for this version) https://doi.org/10.48550/arXiv.2607.19555 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Dongming Wang Dr. [view email] [v1] Tue, 21 Jul 2026 20:22:40 UTC (1,108 KB) Full-text links: Access Paper: View a PDF of the paper titled CHMAS: A Coupled Hierarchical Framework for Multi-Agent Reinforcement Learning, by Dongming Wang and 3 other authorsView PDFHTML (experimental)TeX Source view license Current browse context: cs.MA prev | next new | recent | 2026-07 Change to browse by: cs References Citations NASA ADSGoogle Scholar Semantic Scholar export BibTeX citation Loading… BibTeX formatted citation loading… Data provided by: Bookmark checked="checked"class=“labs-tab-input”> Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, Media Code, Data and Media Associated with this Article alphaXiv Toggle alphaXiv (What is alphaXiv?) Links to Code Toggle CatalyzeX Code Finder for Papers (What is CatalyzeX?) DagsHub Toggle DagsHub (What is DagsHub?) GotitPub Toggle Gotit.pub (What is GotitPub?) Huggingface Toggle Hugging Face (What is Huggingface?) ScienceCast Toggle ScienceCast (What is ScienceCast?) Demos Demos Replicate Toggle Replicate (What is Replicate?) Spaces Toggle Hugging Face Spaces (What is Spaces?) Spaces Toggle TXYZ.AI (What is TXYZ.AI?) Related Papers Recommenders and Search Tools Link to Influence Flower Influence Flower (What are Influence Flowers?) Core recommender toggle CORE Recommender (What is CORE?) Author Venue Institution Topic About arXivLabs arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv’s community? Learn more about arXivLabs. Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?) mathjaxToggle(); We gratefully acknowledge support from our major funders, member institutions, , and all contributors. About Help Contact Subscribe Copyright Privacy Accessibility Operational Status (opens in new tab) Major funding support from
[MA-10] Decentralized Online Riemannian Optimization for Strongly Geodesically Convex Functions
【速读】：该论文旨在解决在具有有界截面曲率的黎曼流形（Riemannian manifolds）上，针对强测地凸（strongly geodesically convex, strongly g-convex）损失函数的去中心化在线优化问题。现有方法仅能处理测地凸（g-convex）损失，在强测地凸情形下缺乏理论保障，尤其在去中心化设置中，由于中央化优化中所需的递减步长与现有网络误差分析通常依赖固定步长的假设不兼容，导致难以直接推广。本文的关键解决方案在于：首先提出一种适用于时变步长策略的一般性网络误差分析框架；在此基础上，构建首个去中心化在线黎曼梯度下降算法的 $ O(\log T) $ 静态遗憾（static regret）上界，该结果与欧氏空间中强凸在线优化的极小极大最优率一致；进一步地，通过引入对损失函数平滑版本的新型强次凸性（strong subconvexity）论证，将该 $ O(\log T) $ 遗憾界拓展至两点带奖赏反馈（two-point bandit feedback）场景，从而首次实现了在强测地凸条件下去中心化在线黎曼优化的最优遗憾性能。
链接: https://arxiv.org/abs/2607.20316
作者: Zhanyuan Cai,Emre Sahinoglu,Shahin Shahrampour
机构: Northeastern University (东北大学)
类目: Optimization and Control (math.OC); Machine Learning (cs.LG); Multiagent Systems (cs.MA)
备注:
Abstract:We study decentralized online optimization for strongly geodesically convex (strongly g-convex) losses on Riemannian manifolds with bounded sectional curvature, including positively curved manifolds. In centralized Riemannian optimization, strong g-convexity tightens the optimal regret from O(\sqrtT) to O(\log T) , where T is the time horizon; in the decentralized Riemannian setting, however, existing methods address only g-convex losses, leaving the strongly g-convex regime unexplored. One challenge is that the required decaying step size in the centralized regime is incompatible with existing network-error analyses, which typically assume a fixed step size. First, we provide a general network-error analysis for time-varying schedules. Next, we build on this analysis to establish the first O(\log T) static regret bound for decentralized online Riemannian gradient descent, matching the minimax-optimal rate for strongly-convex Euclidean online optimization. Finally, we prove the same O(\log T) regret bound for the two-point bandit feedback setting using novel strong subconvexity arguments for the smoothed versions of the loss functions.
自然语言处理
[NLP-0] LKValues: Aligning Large Language Models with Sri Lankan Societal Values
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在多语言社会中存在文化偏见的问题，尤其是在斯里兰卡等具有独特文化动态的国家，其主流模型对本地价值观的识别与处理能力不足。现有评估基准普遍忽视斯里兰卡语境下的本土价值体系，特别是其官方语言僧伽罗语（Sinhala）中的文化内涵，导致模型在跨文化场景下表现不佳。为应对这一挑战，研究提出首个基于调查的斯里兰卡价值观对齐资源集LKValues，通过涵盖205名受访者的三语（僧伽罗语、英语、泰米尔语）问卷调研，融合全球框架与大模型生成的本地化概念，提炼出40项多数人认可的社会核心价值。基于这些价值观，构建了两个关键组件：一是包含15万条情景化实例的僧伽罗语-英语新闻来源指令语料库LKvaluesIT；二是包含1000个实例的价值敏感型评估基准LKvaluesBench。实验表明，尽管新近的大规模模型仍存在低资源和文化对齐差距，但使用LKValues进行微调可显著提升Qwen系列模型在英文与僧伽罗语中的表现，减少无效输出并缓解跨语言差异。研究验证了LKValues在嵌入斯里兰卡本土价值观方面的有效性，并提供了一套可复用的低资源、国别特定多元价值对齐方法论。该数据集已公开发布。
链接: https://arxiv.org/abs/2607.20410
作者: Nethmi Muthugala,Supryadi,Surangika Ranathunga,Nisansa de Silva,Ruijie Tao,Ovindu Gunatunga,Pengyun Zhu,Shaowei Zhang,Jingting Zheng,Deyi Xiong
机构: TJUNLP Lab, School of Computer Science and Technology, Tianjin University, Tianjin, China; School of Mathematical and Computational Sciences, Massey University, Auckland, New Zealand; Department of Computer Science Engineering, University of Moratuwa, Sri Lanka; Johns Hopkins University; School of Computing, University of Colombo, Colombo, Sri Lanka
类目: Computation and Language (cs.CL)
备注: 37 pages, 10 figures, and 15 tables. Includes appendices. Datasets are available at the project repository
Abstract:Value alignment of Large Language Models (LLMs) has been shown to be culturally biased toward Western norms. This results in the mishandling of local values in multilingual societies such as Sri Lanka that have their unique cultural dynamics. Existing benchmarks overlook Sri Lankan-contextualized values in its official language Sinhala, hindering culturally sensitive evaluation and fine-tuning. To bridge this gap, we propose LKValues, the first survey-grounded resource suite for Sri Lankan value alignment. From a trilingual survey of 205 respondents, blending adapted global frameworks and LLM-elicited local constructs, we derive 40 majority-endorsed societal values. Using these values, we construct LKvaluesIT, a Sinhala-English news-derived instruction corpus containing 150k scenario-based instances, and LKvaluesBench, a value-sensitive evaluation benchmark of 1,000 instances. We evaluate a set of proprietary and open-weight LLMs with LKvaluesBench. We fine-tune three open-weight base models (Qwen3.5-4B-Base, Qwen3.5-9B-Base, and Aya-Expanse-8B-Base). Our experiments show that newer and larger LLMs still exhibit low-resource and cultural value-alignment gaps. LKValues fine-tuning improves Qwen-family models in English and Sinhala, reducing invalid outputs and cross-lingual disparities, though gains remain model-family dependent. These highlight LKValues efficacy in embedding Sri Lankan values, offering a replicable pipeline for low-resource, country-specific pluralist value alignment. The dataset is publicly available at this https URL.
[NLP-1] rain the Model Not the Reader: Decodability Supervision for Verifiable Activation Explanations
【速读】： 该论文旨在解决生成式模型中解释隐藏激活值（hidden activations）的可信度评估问题，即当前基于重构的自然语言自编码器（natural-language autoencoders）在验证解释真实性时存在根本性缺陷：其评估标准对单个错误陈述不敏感——即使修改某个关键事实，只要整体重构结果不变，该错误便不会被惩罚。研究发现，现有方法可通过两种非可信方式通过测试：一是解释仅捕捉语义概要而非具体事实（如在释放的Qwen-2.5-7B verbalizer上，约2%的具体陈述影响重构，其余均无关紧要）；二是标准训练流程会自发产生共适应的私有编码（co-adapted private codes），这些编码依赖于虚假表述，且无法通过修复目标模型来消除。为此，作者提出两项审计协议——“基于真值的交叉”（grounded-vs-true cross）与“评估者交换”（evaluator swap），并引入RECAP（Readable Encodings via Co-trained Auxiliary Predictors）框架：在训练过程中同步加入线性头作为辅助预测器，以确保指定内容可被稳定解码。实验表明，在RECAP训练的沙盒模型中，新生成的verbalizer能真实表达指定内容，私有编码消失，仅付出0.001-nat的性能代价；在Pythia-160M上的复现也显示，内容可被可靠探测解码，尽管新verbalizer仅部分传达信息（真实度0.44–0.46，远高于接近零的对照组）。对于可解释性而言，高重构分数不能证明个别陈述的真实性；但对于AI安全，RECAP使内部指定内容可通过独立探针进行客观验证，而非依赖模型可能操纵的自然语言描述——独立探针在有无RECAP情况下分别实现AUC 0.96 vs 0.82，且在对抗攻击下（模型篡改解释以最大化重构得分但隐瞒87%谎言）仍能有效识别谎言（AUC 0.95），而对照探针则退化至随机水平（0.51）。因此，解决方案的关键在于通过协同训练的辅助探针机制，将不可靠的文本解释转化为可独立验证的、具有可读性的内部表征。
链接: https://arxiv.org/abs/2607.20379
作者: Hiskias Dingeto
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Natural-language autoencoders score explanations of hidden activations by reconstruction: an explanation is deemed faithful if the activation can be regenerated from it. The test is structurally insensitive to individual false claims: if flipping a claim does not change the reconstruction, the claim is never penalized. We show the test is passed in two ways, neither faithful. On a released Qwen-2.5-7B verbalizer, explanations reconstruct well above chance while ~2% of specific claims are reconstruction-dependent, so the score tracks gist, not specific facts. Under exact synthetic ground truth, the standard recipe develops co-adapted private codes (false wording the reconstruction depends on) in 5/5 runs, and fixes that leave the target model unchanged do not help. We contribute two audit protocols, the grounded-vs-true cross and the evaluator swap, and RECAP (Readable Encodings via Co-trained Auxiliary Predictors): linear heads trained alongside the target model to keep designated content decodable. On RECAP-trained sandbox models, fresh verbalizers state the designated content truly and the codes vanish, at a +0.001-nat cost. This replicates on a pretrained Pythia-160M: the content becomes reliably probe-decodable, though a fresh verbalizer conveys it only in part (truth 0.44-0.46 vs a near-zero control). For interpretability, high reconstruction does not certify individual claims. For AI safety, RECAP makes designated internal content independently checkable against probes rather than asserted by prose a model can game: an independent probe scores the verbalizer’s true claims above its false ones (AUC 0.96, vs 0.82 without RECAP). Against an adversary that edits an explanation to maximize the reconstruction score while lying (suppressing ~87% of its lie penalty), the RECAP probe still flags the lies (AUC 0.95) while the control probe collapses to chance (0.51).
[NLP-2] Notes to Self: Can LLM s Benefit from Experiential Abstractions?
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在复杂推理任务中缺乏对过往经验进行抽象提炼与复用的问题，即如何使模型能够像人类通过经验总结策略和警示性提醒以提升问题求解效率一样，从自身或强教师模型的解题过程中提取可复用的自然语言抽象（experiential abstractions），并将其应用于后续推理。其解决方案的关键在于：从LLMs在MATH训练集上的解题轨迹中，通过强教师模型或模型自身自动提取自然语言形式的抽象知识，并构建一个可检索的知识库；随后采用两种使用模式——推理时检索抽象信息，以及在强化学习（Reinforcement Learning, RL）框架下将抽象信息融入训练提示（training prompts），从而增强模型的推理能力。实验表明，此类抽象显著提升了模型在数学与逻辑推理基准测试中的表现，且自提取抽象与教师提取抽象性能相当，该框架具备跨数据集和模型的可迁移性，验证了LLMs具备类似人类“经验抽象化”能力的潜力。
链接: https://arxiv.org/abs/2607.20372
作者: Chang Liu,Xinyu Li,Artur Dubrawski
机构: Auton Lab, Carnegie Mellon University (卡内基梅隆大学自动实验室)
类目: Computation and Language (cs.CL)
备注:
Abstract:Humans distill experience into reusable abstractions, e.g., strategies and cautionary reminders, and apply them to gradually solve problems more effectively. We study whether Large Language Models (LLMs) can similarly benefit from such experiential abstractions. From LLMs’ solution traces on the MATH training set, a stronger teacher or the LLMs themselves extract natural-language abstractions into a retrievable library. We explore two usage modes: (1) inference-time retrieval and (2) reinforcement learning (RL) with abstraction-augmented training prompts. Experiential abstractions improve LLM performance on mathematical and logical reasoning benchmarks. Self-extracted abstractions match teacher-extracted ones, and our abstraction usage framework can transfer to other datasets and models. These findings suggest LLMs can extract and apply experiential abstractions much as humans leverage distilled experience.
[NLP-3] st-Time Training for Modality Order Consistency in Vision-Language Models
【速读】： 该论文旨在解决视觉-语言模型（Vision-Language Models, VLMs）在模态顺序敏感性方面存在的系统性缺陷，即模型性能受图像与问题输入顺序的显著影响——尽管该顺序在语义上无关紧要。实验发现，在三个模型和三个基准测试中，"图像先行"提示始终优于"问题先行"提示，暴露出一种可重复的模态顺序偏差（modality-order failure）。其解决方案的关键在于提出一种测试时训练（test-time training）方法，通过引入不对称的、基于顺序一致性的动态适应机制，在推理阶段对模型进行微调，从而有效弥合不同提示顺序之间的性能差距。该方法不仅修复了因顺序变化导致的表征不一致性，还通过反向传播优化使两种提示顺序均实现性能提升，表现出“自我增强”的特性。进一步的激活修补（activation patching）分析揭示，该失败现象集中于网络中段的一个狭窄区域，该区域在不同顺序下表示差异显著；而所提方法能有效修复该区域及其上下游层间的表征错位。综上，研究将模态顺序敏感性定位为视觉-语言模型中的电路级（circuit-level）结构性缺陷，并证明仅需简单的、非对称的测试时自适应即可高效缓解该问题并提升整体性能。
链接: https://arxiv.org/abs/2607.20351
作者: Aditi Gupta,Yossi Gandelsman
机构: University of Chicago; Reve
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注: 16 pages, 7 figures, preprint
Abstract:We find that vision-language models are sensitive to a specific semantically irrelevant change: the order in which the image and question are presented. Across three models and three benchmarks, image first prompting consistently outperforms question-first prompting, revealing a repeatable modality order failure. We use this gap to design an order-consistent test-time training method. Our method substantially closes the modality-order gap across all evaluated settings. Surprisingly, it also yields consistent improvements in the stronger image-first branch over the baseline, hence bootstrapping both orderings toward mutual consistency. Activation patching localizes the ordering failure to a narrow mid-network region where representations diverge sharply between prompt orders. We find that the test-time training method repairs this misalignment across layers. Together, our results identify modality-order sensitivity as a circuit-level failure in VLMs and demonstrate that simple, asymmetric test-time adaptation can effectively mitigate it and even improve performance over the baseline.
[NLP-4] Generative AI floods and dilutes the market for books
【速读】： 该论文旨在检验生成式 AI（Generative AI）创作的长篇虚构类自出版图书是否如普遍预期般因质量低下而被市场忽视，进而探讨其在真实商业环境中的实际影响力。研究通过对2023至2026年间在亚马逊上销售的14,419本自出版类型小说进行全文生成式AI检测，并匹配每日销售数据至2026年6月，发现尽管含有显著比例AI文本（≥25%）的书籍在总书目中占比较高，但其销量份额相对较低；然而，这些书籍正逐步扩大市场份额，持续抢占原本由无AI文本书籍占据的高排名位置，表现出显著的商业化渗透力。在此期间，有销量的书籍数量增长了19.2倍，而季度收入仅增长8.9倍，表明市场新增书籍速度远超收入增长，导致单位售出书籍的平均收益下降。尤其在高AI普及率及Kindle Unlimited服务覆盖广泛的类别中，无AI文本书籍的市场地位受到最严重冲击。进一步分析显示，畅销书中含大量AI文本的作品倾向于采用更具辨识度的语言特征，且其内容重叠度随收入上升而增加，这一趋势在无AI文本书籍中未被观测到。因此，该研究揭示生成式AI重塑创意市场的核心机制并非依赖内容质量提升，而是通过规模效应实现市场主导。这一发现直接关联到版权侵权诉讼中“合理使用”（fair use）辩护的核心争议——即大规模使用受版权保护内容训练模型并生成新作品是否具有市场影响，从而影响法律判断。
链接: https://arxiv.org/abs/2607.20349
作者: Tuhin Chakrabarty,Xinyue Liu,Jane C. Ginsburg,Paramveer Dhillon
机构: Stony Brook University (石溪大学); Columbia Law School (哥伦比亚法学院); University of Michigan (密歇根大学); MIT Initiative on the Digital Economy (麻省理工学院数字经济倡议)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注: Working Paper Under Review
Abstract:Generative AI can produce book-length works of fiction at near-zero cost. These books are often dismissed as low-quality ``slop’’ that buyers will ignore, and are assumed to carry little commercial weight. We test that assumption with full-text AI detection across 14,419 self-published genre-fiction books sold on Amazon from 2023 to 2026, matched to daily sales records through June 2026. None of these books disclose whether or not they contain AI-produced content. We find that books for which we detected substantial AI text ( 25%) make up a large share of the catalog but a smaller share of sales. Even so, they reach commercial scale, winning a growing share of sales over time and taking more of the scarce top-rank positions once held by books with no detected AI text. Over this period, the number of books with observed sales in a quarter grew 19.2-fold, while quarterly revenue grew only 8.9-fold. The market therefore added selling books faster than it added revenue, and revenue per selling book fell across most genres. Books with no AI text lose the most ground in genres with high AI diffusion, and most of all where Kindle Unlimited availability is high. Among top-selling books, those with substantial AI text draw on more distinctive language from existing books than do books with no AI text; for these books overlap rises with revenue, a gradient we do not detect for books with no AI text. Generative AI can thus reshape a creative market through scale rather than quality. Our results bear directly on the market-effect question at the center of the fair use defense to copyright infringement.
[NLP-5] PyroDash: Cost-Efficient Token-Level Small-Large Language Model Collaborative Inference
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在大规模服务时成本高昂，而小语言模型（Small Language Models, SLMs）虽成本低但面对复杂任务时可靠性不足这一矛盾问题。其核心解决方案是提出一种面向令牌级别的、成本感知的SLM-LLM协同推理框架——PyroDash。该框架的关键在于：通过在生成过程中让SLM自主决定是否通过发出控制令牌（control token）请求协作，实现动态的、细粒度的计算资源调度；协作引擎将当前查询及部分推理轨迹单次传递给冻结的LLM完成后续推理，整个过程无需独立路由器、不需重新训练LLM或访问其对数概率输出。PyroDash采用三阶段训练策略：控制令牌嵌入学习、面向卸载的监督微调以及基于组相对策略优化（Group Relative Policy Optimization）的成本感知对齐，使奖励函数在答案准确率与按纯LLM推理归一化的推理成本之间取得平衡。实验表明，在五个数学推理基准上，该方法可在不同精度-成本权衡点间灵活切换，显著降低对LLM的依赖，同时保持优异的推理性能。
链接: https://arxiv.org/abs/2607.20327
作者: Niqi Lyu,Pengtao Shi,Wei Qiu,Jianlin Zhong,Sicong Xia,Jianyao Ma,Yicheng Ding
机构: Pyromind Dynamics Inc.
类目: Computation and Language (cs.CL)
备注: 19 pages, 3 figures
Abstract:Large language models (LLMs) provide strong reasoning capabilities but are expensive to serve at scale, whereas small language models (SLMs) are cheaper but less reliable on difficult problems. We introduce PyroDash, a cost-aware framework for token-level SLM-LLM collaborative inference. During generation, the SLM decides whether to request assistance by emitting a control token. A Collaborate Engine then sends the query and partial reasoning trace to a frozen LLM for completion through a single handoff. The policy is internalized in the SLM, requiring neither a separate router, LLM retraining, nor access to LLM logits. PyroDash trains the SLM in three stages: control-token embedding learning, offloading-oriented supervised fine-tuning, and cost-aware alignment with Group Relative Policy Optimization. Its reward balances answer accuracy against inference cost normalized by LLM-only inference. Across five mathematical reasoning benchmarks, PyroDash supports different accuracy-cost operating points. With \lambda=0.05 , it achieves 64.04 percent average accuracy, 6.36 percentage points above the LLM-only baseline, while reducing cost by 20.4 percent. With \lambda=0.6 , it achieves 54.55 percent accuracy with a 1.90 percent LLM token ratio and 0.012 LLM calls per example, reducing total cost from USD 49.36 to USD 1.78. These results show that learned token-level handoffs can reduce LLM use while preserving strong reasoning performance.
[NLP-6] he Blessing of Dimensionality: How Near-Orthogonality in High-Dimensional Spaces Explains Temporal Portability
【速读】： 该论文旨在解决生成式AI（Generative AI）领域中大语言模型（Large Language Models, LLMs）在持续预训练（continual pretraining）背景下，参数高效微调（Parameter Efficient Fine-Tuning, PEFT）方法——特别是PortLLM——的长期时间可迁移性（long-term temporal portability）问题及其内在理论机制不明确的问题。现有研究表明，尽管PortLLM在初始阶段展现出良好的短期可迁移性，但其在经历多轮持续预训练更新后是否仍能保持性能尚缺乏系统验证，且其有效性缺乏充分的理论解释。为应对这一挑战，本文提出两个核心解决方案：首先，通过在Mistral、Gemma和Qwen等基础模型上进行涵盖10次持续预训练步骤的广泛实证研究，验证了PortLLM补丁在长时间跨度下的持续可迁移性，表明无需对每次模型更新重复进行微调即可维持良好性能；其次，从理论层面提供双重分析：一是揭示高维向量的近正交性（near-orthogonality）是实现时间可迁移性的关键数学基础；二是建立损失曲面的几何视角，为不同适配方案之间的理论比较提供了统一框架。因此，该研究的关键在于通过实证与理论双路径，阐明了无需训练和数据的PortLLM方法在动态模型更新场景下依然有效的深层原因，推动了对高效模型适应机制的理解。
链接: https://arxiv.org/abs/2607.20301
作者: Abigail Woodring,Adrian Chan,Rana Muhammad Shahroz Khan,Sukwon Yun,Chau-Wai Wong,Tianlong Chen
机构: University of North Carolina at Chapel Hill; NC State University
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注:
Abstract:Fine-tuning has been widely used to adapt large language models (LLMs) for domain-specific tasks. Parameter efficient fine-tuning (PEFT) methods such as low-rank adaptation (LoRA) are frequently used to reduce computational costs. PortLLM is a training-free and data-free scheme used to adapt LLMs after continual pretraining. Although the initial PortLLM results show that LoRA patches exhibit short-term temporal portability, the long-term performance of PortLLM across several updates of continual pretraining remains underexplored. Furthermore, the intriguing effectiveness of PortLLM is not well understood from a theoretical standpoint. We address these two open questions by (1) performing an extensive empirical study of the long-term temporal portability of PortLLM patches across 10 continual pretraining steps using base models Mistral, Gemma, and Qwen; and (2) offering two theoretical analyses to explain our observation that the simple PortLLM method achieves competitive performance. We find empirically that the portability persists across longer time duration, indicating that repeated fine-tuning is not required when the base model is periodically updated. We find theoretically that near-orthogonality of high-dimensional vectors is a key justification for temporal portability. Our analyses also demonstrate a geometric perspective of the loss landscape in facilitating the theoretical comparison of different adaptation options.
[NLP-7] Sound Probabilistic Safety Bounds for Large Language Models
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）在生成特定提示（prompt）时产生有害输出的概率的严格边界计算问题。现有方法难以在真实有害概率极低的情况下有效估计下界，且缺乏形式化保证。其解决方案的关键在于提出一种基于Clopper-Pearson置信区间的概率近似正确（Probably Approximately Correct, PAC）框架，并设计了一种利用隐空间（latent space）特征来优先探索自回归生成树中更可能产生有害输出分支的算法。该方法能够高效计算出具有形式化安全保证的保守下界——即所获得的下界严格小于真实有害概率，从而实现对大模型有害性行为的统计可验证评估。实验结果表明，该方法可为当前最先进的大语言模型提供非平凡的、可靠的下界估计，首次实现了对大语言模型的可验证评估与统计认证。
链接: https://arxiv.org/abs/2607.20286
作者: Mahdi Nazeri,Anne-Kathrin Schmuck,Sadegh Soudjani,Alessandro Abate
机构: University of Oxford (牛津大学); Max Planck Institute for Software Systems (马普所软件系统研究所); University of Birmingham (伯明翰大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: The Initial version of this manuscript has been available on OpenReview, see this https URL
Abstract:We propose a novel framework for computing rigorous bounds on the probability that a large language model (LLM) generates harmful output to a given prompt. We study a new application of the Clopper-Pearson confidence intervals to obtain probably approximately correct (PAC) bounds for this problem. As our main technical contribution, we propose an algorithm that leverages features in the latent space to prioritize exploring branches in the auto-regressive generation tree that are more likely to produce harmful outputs. Our approach in particular enables the efficient computation of useful lower bounds, even in scenarios where the true harm probability is extremely small, and crucially, the obtained lower bounds are sound, i.e., formally proven to be less than the actual harmfulness probability: our experimental results demonstrate the effectiveness of our method by computing non-trivial lower bounds on state-of-the-art LLMs. This study newly enables the evaluation and statistical certification of LLMs.
[NLP-8] Self-supervision drives representational convergence in medical foundation models more than clinical supervision
【速读】： 该论文旨在解决医学图像编码器在实际应用中被普遍视为可互换的假设是否成立这一核心问题，尤其关注不同编码器之间表征收敛的真实程度、形成机制及其临床可用性。其关键发现在于：尽管存在一定程度的表征收敛，但这种收敛主要由自监督预训练目标（self-supervised objective）驱动，而非模型规模或临床标注监督；在胸部X光图像上，采用相同自监督目标的编码器间对齐度最高（40.4%），而依赖标签监督（21.1%）或图文对比学习（3.3%）的编码器表现显著更低，且收敛程度不随参数量增加而提升（Spearman相关系数0.302，p=0.223）。此外，表征收敛局限于同模态内部，未能映射到临床语言层面，也未能复现放射科医生对病例相似性的判断。然而，线性分类器可在不同编码器间实现迁移，并在五个独立医院数据集上保持约85%的原始性能，表明跨编码器的泛化能力仍具备可行性。因此，论文提出：医学图像编码器的互操作性并非自然涌现，而是由预训练目标决定，应通过设计合理的预训练目标来主动构建共享表示空间，并在患者亚组和临床判断基准下重点验证共享几何结构的稳健性。
链接: https://arxiv.org/abs/2607.20274
作者: Soroosh Tayebi Arasteh,Sebastian Ziegelmayer,Mahshad Lotfinia,Lisa Adams,Sven Nebelung,Jakob Nikolas Kather,Daniel Truhn
机构: 未知
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Medical image encoders from different groups are increasingly treated as interchangeable, on the assumption that scale and clinical supervision concentrate their representations onto a shared structure. Whether this convergence is real, what produces it, and whether it is clinically usable are untested, and the similarity measures behind such claims are fragile. We present a controlled dissection across 18 image and 7 text encoders, all open-weight and run locally, spanning 7M to 27B parameters and five imaging modalities, including 650,982 chest radiographs from six datasets. To isolate cause, we train encoders that vary only the objective under fixed data, architecture, and scale, and reproduce the effect in a synthetic model. Convergence is modest but above a random floor, driven by the self-supervised objective, not clinical supervision: matched self-supervised encoders aligned most (40.4% on chest radiography), with label-supervised (21.1%) and image-text (3.3%) far lower, and did not grow with size (Spearman 0.302, p=0.223) or capability. It is within-modality, does not reach clinical language, and does not reproduce how radiologists judge case similarity. Yet a linear classifier transfers across encoders and to five held-out hospitals, retaining about 85% of within-encoder performance. Convergence in medical imaging is therefore set by the pretraining objective, not inherited from scale or clinical supervision. Interoperability is accordingly something to design for through that objective, and to validate where the shared geometry is weakest, across patient subgroups and against clinical judgment.
[NLP-9] Which Values Do LLM s Confuse? A Schwartz-Based Recognition Study
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在价值评估中所面临的核心前提问题：即模型是否具备准确识别具体情境中所蕴含基本价值观的能力。现有评估方法假设模型能够正确理解并识别出情境背后的动机性价值，但这一假设缺乏对模型实际识别能力的系统检验。为此，研究基于施瓦茨（Schwartz）的十项基本价值观体系，构建了一个包含1000个俄语情境文本的评测数据集，每个文本在十种价值观间均衡分布，并由两名独立人类标注者进行标注，以确保标注可靠性。研究采用固定排序响应协议，对21个指令微调后的LLM运行结果进行评估，其中20个输出可靠的模型构成语义分析面板。结果显示，模型在Top-1准确率（Acc@1）为0.683，Top-3准确率（Acc@3）达0.892，表明模型虽能大致定位正确价值区域，但在相近价值间的排序稳定性较差。进一步分析发现，相邻价值导致的语义错误占比高达50.9%，显著高于特定检查点下的基线水平（24.4%），且存在若干跨检查点与人工验证子集重复出现的定向混淆模式，如普遍主义（Universalism）与利他主义（Benevolence）、传统（Tradition）与顺从（Conformity）、安全（Security）与权力（Power）之间呈现强不对称性，而刺激—享乐（Stimulation-Hedonism）则形成双向边界。这些混淆的严重程度具有检查点依赖性，可能对高阶价值谱系的推断造成偏差。因此，论文提出应建立融合精确度、排序恢复能力及定向错误分析的价值识别评估框架，以更全面地衡量模型在价值观理解方面的表现。
链接: https://arxiv.org/abs/2607.20270
作者: Andrei Chetvergov,Stepan Ukolov,Timofei Sivoraksha,Alexander Evseev,Mikhail Solovev,Valeriia Kuschenko,Maria Chistyakova,Sergey Bolovtsov
机构: 未知
类目: Computation and Language (cs.CL)
备注: 14 pages, 7 figures, 3 tables
Abstract:Large language models are increasingly evaluated through the values they endorse, but such evaluations presuppose that models can identify the value expressed in a concrete situation. We study this prerequisite as controlled top-1 recognition over Schwartz’s ten basic values. Our evaluation set contains 1,000 Russian situational texts, balanced across the ten values and independently labeled by two human annotators per item. We evaluate 21 instruction-tuned LLM runs under a fixed ranked-response protocol; 20 runs with reliable outputs form the semantic panel. Pooled Acc@1 is 0.683 and Acc@3 is 0.892, showing that models often locate the correct motivational region while ranking close alternatives unstably. Adjacent values account for 50.9% of semantic errors, compared with 24.4% under a checkpoint-specific null. Eight directed confusions recur across checkpoints and human-confirmed subsets. Several are strongly asymmetric, including Universalism to Benevolence, Tradition to Conformity, and Security to Power, whereas Stimulation-Hedonism forms a bidirectional boundary. Their severity is checkpoint-specific and can bias higher-order value profiles. The results motivate value-recognition evaluation that combines exact accuracy, ranked recovery, and directed error analysis.
[NLP-10] PoTRE: Test-Time Reasoning inspired by Cognitive Heterogeneity
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在处理需要长时程规划与迭代纠错的复杂推理任务时表现不佳的问题，尤其针对模型在面对新颖抽象概念或严格领域约束时，传统单流提示（single-stream prompting）方法易失效的瓶颈。其解决方案的关键在于提出一种异构推理框架PoTRE（Poly-Topological Reasoning Ensembles），通过将推理过程解耦为四个功能互补的智能体：对抗性精炼代理（Adversarial Refinement Agent）、层次化战略规划代理（Hierarchical Strategic Planning Agent）、谱搜索代理（Spectrum Search Agent）以及直接链式代理（Direct Chain Agent）。这些代理分别从不同认知路径生成候选解，最终由任务自适应聚合层（Task-Adaptive Aggregation Layer）通过候选选择、语义融合或神经符号验证等方式动态整合多视角输出，形成鲁棒的全局解。实验表明，该异构架构在ARC-AGI-2、Humanity’s Last Exam（HLE）和PRBench Finance三个前沿基准上均取得领先性能，尤其在HLE上达到49.92%的准确率，超越此前最优官方成绩；且在推理效率上优于大规模同质化基线，实现更优的推理性能与资源利用率平衡。
链接: https://arxiv.org/abs/2607.20268
作者: Anmol Kankariya,Sercan Ö. Arık
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Accepted at Transactions on Machine Learning Research (TMLR 2026)
Abstract:While Large Language Models (LLMs) excel at many tasks, they frequently struggle with complex reasoning that requires long-horizon planning and iterative error correction. Furthermore, standard single-stream prompting proves brittle when models encounter novel abstractions or rigorous domain constraints. We introduce PoTRE (Poly-Topological Reasoning Ensembles), a heterogeneous framework that decouples inference into four agents: (1) Adversarial Refinement Agent, (2) Hierarchical strategic Planning Agent, (3) Spectrum Search Agent, and (4) Direct Chain Agent. A final Task-Adaptive Aggregation Layer dynamically reconciles these perspectives – via final candidate selection, semantic synthesis, or neuro-symbolic verification – to produce a robust global solution. We evaluate PoTRE on three frontier benchmarks: ARC-AGI-2, Humanity’s Last Exam (HLE), and PRBench Finance. PoTRE achieves state-of-the-art accuracy of 49.92% on HLE, surpassing the previous best official score. We demonstrate that this architectural heterogeneity achieves improved reasoning performance using similar or fewer inference tokens compared to heavily scaled homogeneous baselines.
[NLP-11] he Maskability Index: Predicting Task-Objective Alignment in Pretrained Language Models
【速读】： 该论文旨在解决预训练语言模型在少样本关系知识生成任务中因提示策略（prompting strategy）与预训练目标不匹配而导致性能下降的问题。其核心挑战在于如何为特定的知识关系选择最适配的提示形式——掩码式提示（masked-style prompting）或前缀式提示（prefix-style prompting）。为此，论文提出了一种名为“可掩码性指数”（Maskability Index, MI）的量化指标，其关键创新在于基于掩码与非掩码模板在DepthRank评分上的差异，构建了一个可解释的、衡量预训练目标与提示模板对齐程度的指标。实验结果表明，MI与下游生成性能呈正相关，能够有效指导在低资源场景下选择最优的提示模板与适配策略，从而提升知识抽取的准确性与鲁棒性。
链接: https://arxiv.org/abs/2607.20265
作者: Ahmad Pouramini,Mahsa Afsharzadeh
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Large-scale pretrained language models such as T5 and BERT have demonstrated strong capabilities for generating structured knowledge. However, their performance depends on how closely the prompting strategy matches the objectives used during pretraining. We introduce the Maskability Index (MI), a quantitative metric that estimates whether a knowledge relation is better suited to masked-style prompting or prefix-style prompting in few-shot generation. MI is computed from differences in DepthRank scores between masked and unmasked templates, providing a principled measure of objective-template alignment. We evaluate MI on a diverse set of relations from the ATOMIC2020 knowledge base completion benchmark and show that it is positively correlated with downstream generation performance. These results indicate that MI can help select appropriate prompting templates and adaptation strategies for extracting relational knowledge from pretrained language models, especially in low-resource settings.
[NLP-12] Exposure is Optional: Learning Unlike Coordination in Language Models ACL
【速读】： 该论文旨在解决语言学中关于“协调结构（Coordination）”的本质问题，特别是针对存在大量非同类成分协调现象（unlike coordination）是否需要直接训练数据支持才能被习得的争议。传统理论认为只有同类别成分才能进行协调，但自然语言中广泛存在的非同类协调现象挑战了这一观点。本文通过将生成式语言模型（Generative AI）作为计算实验平台，探究模型在未接触任何非同类协调实例的情况下，能否通过普遍的组合能力自发习得此类结构。研究采用过滤语料训练（Filtered-Corpus Training, FiCT）方法，从训练数据中移除所有非同类协调实例，并对GPT-2模型进行训练。结果表明，直接暴露并非必要：经过过滤语料训练的模型仍能有效泛化至非同类协调任务，在困惑度（perplexity）与语法判断上表现与未过滤训练模型相当。进一步分析内部表示发现，模型处理非同类协调时，倾向于将并列成分视为具有相似结构范畴，或通过类似删除机制进行处理，这些策略均可仅通过同类型协调（alike coordination）的间接经验学习获得。因此，该研究的关键突破在于揭示了语言模型可通过通用的组合性原则实现对复杂语言结构的无监督泛化，为理解语言模型如何内化语言结构提供了新证据，并支持了协调结构可由一般性语言能力推导而非依赖特定规则的理论立场。
链接: https://arxiv.org/abs/2607.20251
作者: Jiamu Luo,Shane Steinert-Threlkeld
机构: University of Washington (华盛顿大学)
类目: Computation and Language (cs.CL)
备注: 13 pages, 6 tables, 2 figures, to submit to TACL
Abstract:Coordination, a fundamental linguistic structure, remains a subject of intense debate, and its exact nature continues to elude theoretical linguistics. A common view holds that only same-category constituents can be conjoined, which has been challenged by the many grammatical unlike coordinations found in natural language. Treating language models as a computational testbed, we investigate whether the acquisition of unlike coordination requires direct exposure in the training data, or whether it can emerge organically from general compositional abilities. Using Filtered-Corpus Training (FiCT), we train GPT-2 models on corpora from which all instances of unlike coordination have been removed. We find that direct exposure is not necessary: models trained on filtered data successfully generalize to unlike coordination, achieving perplexity and grammaticality judgments comparable to models trained on unfiltered text. Furthermore, our analyses of internal representations indicate that language models process unlike coordination by treating the conjoined elements as belonging to similar structural categories or through a mechanism akin to deletion, both of which appear learnable from exposure to alike coordination alone. This work contributes to the growing understanding of how language models internally represent linguistic structures, while also adding to the broader debate on coordination by showing how models generalize and process unlike coordination without direct exposure.
[NLP-13] On the Systematic Challenges of Culturally Loaded Machine Translation: Dream of the Red Chamber as the Cultural Lens
【速读】： 该论文旨在解决生成式 AI（Generative AI）在处理文化负载型翻译（culturally loaded translation）时所面临的挑战，尤其关注基于大语言模型（Large Language Models, LLMs）的机器翻译（Machine Translation, MT）系统在跨文化语境理解与表达上的局限性。其核心问题是：尽管当前大语言模型在多数翻译任务中已接近人类水平，但在涉及深层社会文化内涵的表达转换时仍表现不佳，且现有评估体系难以有效衡量此类翻译的质量。解决方案的关键在于构建一个以中国古典文学名著《红楼梦》为语料基础的中日双语文化负载表达数据集（包含500个跨文化类别片段），并通过多维度评估框架揭示三类关键挑战：（1）任务层面的性能瓶颈，即前沿大模型在文化负载内容上存在显著能力差距；（2）人工评估中的背景偏差，不同评价者因文化背景差异导致判断一致性差；（3）自动评估指标失效，主流评价指标无法可靠反映文化敏感性翻译的真实质量。该研究为面向文化的机器翻译在计算语言学与应用语言学领域的进一步发展提供了重要实证依据与改进方向。
链接: https://arxiv.org/abs/2607.20241
作者: Yiming Wang,Jiayuan Di
机构: Shanghai Jiao Tong University (上海交通大学); East China University of Science and Technology (华东理工大学); Mejiro University (明治大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Culturally loaded translation poses unique challenges for machine translation (MT), as meanings are deeply embedded in socio-cultural contexts beyond surface linguistic forms. Although large language models (LLMs) have enabled MT systems to achieve human-like quality in many scenarios, their ability to handle culturally loaded expressions remains underexplored. In this study, we systematically investigate the challenges posed by culturally loaded translation in LLM-based MT systems. We construct a Chinese-Japanese bilingual dataset from the culturally representative corpus Dream of the Red Chamber, containing 500 segments across diverse cultural categories. Using a comprehensive evaluation protocol, we reveal three main challenges: (1) task challenges, where frontier LLMs exhibit notable performance gaps and struggle with culturally loaded content; (2) human evaluation challenges, where evaluator backgrounds lead to substantial disagreement in translation judgments; and (3) automatic evaluation challenges, where widely used metrics fail to reliably assess translation quality for this task. These findings may offer valuable insights for culture-oriented translation research in both computational science and linguistics.
[NLP-14] HalluTruthQA: A Fine-Grained Benchmark for Hallucination Detection Localization and Explanation in Arabic Question Answering
【速读】： 该论文旨在解决生成式人工智能（Generative AI）在阿拉伯语问答任务中普遍存在且难以检测、定位、解释与验证的事实性错误（hallucination）问题。现有基准多仅提供回答层面的幻觉标签，缺乏对具体错误内容的精确定位、错误原因的解释以及正确事实的可选答案支持。为此，研究提出 \textscHalluTruthQA，一个面向阿拉伯语问答的细粒度幻觉评估基准，涵盖伊斯兰知识、历史、科学和地理四大知识密集型领域，共包含2,400个由专家标注的样本。每个样本不仅提供经过验证的参考答案、二元幻觉标签，还包含六组用于事实验证的候选答案，对于存在幻觉的答案，进一步提供字符级错误片段、人工撰写的解释文本及宏观与微观幻觉类型分类。实验在零样本设置下评估了四款开源大语言模型（LLM）——\textscAllam、\textscFalcon-H1、\textscQwen32 和 \textscSilma——在幻觉检测、错误片段定位、事实验证和解释评估四项任务上的表现。结果显示，各任务反映不同能力维度，无单一模型在所有任务上均取得最优表现，最佳得分分别为：检测任务0.880宏平均F1，定位任务0.516 F1-Sp，事实验证任务0.852 LO-Score，解释评估任务0.644最终得分。研究表明，幻觉评估应从单纯的检测向错误定位、事实验证与成因解释等更深层次方向演进。该研究的代码、数据集、提示模板与评估脚本均已公开。
链接: https://arxiv.org/abs/2607.20219
作者: Abdessalam Bouchekif,Mohammed-En-Nadhir Zighem,Salah Eddine Bekhouche,Hichem Telli,Somaya Eltanbouly,Shahd Gaben,Heba Sbahi,Samer Rashwani,Mutaz Al-Khatib,Emad Mohamed,Mohammed Ghaly,Abdenour Hadid
机构: Hamad Bin Khalifa University, Qatar; University of Biskra, Algeria; University of the Basque Country, Spain; Nazarbayev University, Kazakhstan; Universiti Malaysia Kelantan, Malaysia
类目: Computation and Language (cs.CL)
备注:
Abstract:Large language models (LLMs) can generate fluent Arabic answers, yet factual errors remain difficult to detect, localize, explain, and verify. Existing hallucination benchmarks often provide response-level labels, with limited support for identifying the exact erroneous content, explaining why it is incorrect, or selecting the correct factual answer. We introduce \textscHalluTruthQA, a fine-grained benchmark for hallucination evaluation in Arabic question answering. The benchmark contains 2,400 expert-curated examples across four knowledge-intensive domains: Islamic knowledge, history, science, and geography. Each example pairs an Arabic question and a model-generated answer with a verified reference answer, a binary hallucination label, six candidate answers for factual verification, and, for hallucinated answers, character-level erroneous spans, human-written explanations, and macro and micro hallucination types. We evaluate four open-source LLMs, \textscAllam, \textscFalcon-H1, \textscQwen32, and \textscSilma, in a zero-shot setting across hallucination detection, span-level localization, factual verification, and explanation evaluation. Results show that these tasks capture different abilities: no single model achieves the strongest performance across all tasks, with best scores of 0.880 Macro-F1 for detection, 0.516 F1-Sp for localization, 0.852 LO-Score for factual verification, and 0.644 final score for explanation evaluation. Our taxonomy shows that hallucination evaluation should move beyond detection toward localizing, verifying, and explaining factual errors. The code, dataset, prompts, and evaluation scripts are available at this https URL. Subjects: Computation and Language (cs.CL) Cite as: arXiv:2607.20219 [cs.CL] (or arXiv:2607.20219v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2607.20219 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-15] surprisal is Not a Theory
【速读】： 该论文旨在解决当前计算心理语言学中对突变性理论（Surprisal Theory）应用中存在的代表性假设隐含问题，即在大型语言模型（LLM）兴起的背景下，研究者常将不同模型输出的突变性值视为可互换的“黑箱”指标，而忽视了其背后所依赖的表征与算法层面的具体实现。其解决方案的关键在于强调：尽管突变性理论被归为计算层次（computational-level）的解释框架，但使用LLM计算突变性仍不可避免地涉及具体的表征选择与算法设计。通过三项实证分析，论文证明模型架构与算法机制对语言模型概率计算具有显著影响，因此不能将不同模型的突变性值简单等同。作者呼吁研究者重新审视并明确所用模型的表征与算法假设，避免因对大模型突变性值的不加批判使用而掩盖其内在认知建模的偏差。
链接: https://arxiv.org/abs/2607.20208
作者: Andrés Buxó-Lugo,Aniello De Santo,Morgan Grobol,Ryan J. Hubbard,Cassandra L. Jacobs
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:Surprisal Theory is often characterized as a computational-level explanation per (Marr, 1982). We argue in this work that, even though a computational level narrative has been used to support “representation-agnostic research” within computational psycholinguistics, the movement toward black box systems embodied by large language models (LLMs) does not exempt modelers using the surprisal metric from the representational decisions required by computational-level characterizations. In fact, we argue that the uncritical use of LLM-surprisal obfuscates the representational and algorithmic-level commitments of different models. In three analyses, we show that the choice of algorithm and model architecture play significant roles in the computation of language model probabilities. We advise that researchers who wish to test Surprisal Theory re-evaluate the practice of treating large language model probabilities as interchangeable
[NLP-16] Gotta Catch them all: the modes of Sycophancy
【速读】： 该论文旨在解决大语言模型在对话中过度迎合用户信念、牺牲事实准确性的“讨好行为”（sycophancy）问题。传统机制研究将讨好行为视为单一维度的倾向，可统一增强或抑制，但本文通过分析948个社会压力情境下的三种假定讨好模式，挑战了这一假设。研究发现，尽管三种模式生成的文本输出高度相似（文本分类器准确率仅57.8%），其内部表征从第14层开始即可被完全线性分离；不同模式在处理流程中出现阶段不同，依赖独立的注意力电路，并对特定输入表现出最强激活。这表明讨好行为并非单一整体，而是一组在表征和计算上具有显著差异的结构化模式。该发现的关键在于揭示了讨好行为的内在异质性，为实现更精准的测量与干预提供了理论基础。
链接: https://arxiv.org/abs/2607.20146
作者: Shreyans Jain,Alexandra Yost,Amirali Abdullah
机构: Thoughtworks; Southern Utah University (SUU)
类目: Computation and Language (cs.CL)
备注:
Abstract:Large language models often align with users’ beliefs at the expense of factual accuracy, a behavior known as sycophancy. Prior mechanistic studies largely treat sycophancy as a single behavioral dimension that can be uniformly amplified or suppressed. We challenge this assumption by analyzing three hypothesized modes of sycophancy across 948 social pressure situations. Although the modes produce highly similar outputs, with a text-only classifier achieving just 57.8 percent accuracy, their internal representations are perfectly linearly separable from layer 14 onward. We further find the modes emerge at different processing stages, rely on distinct attention circuitry, and fire strongest on different inputs. These results show that sycophancy is not a monolithic tendency, but a structured family of representationally and computationally distinct modes, motivating more precise measurement and intervention.
[NLP-17] SLAI T-Rex: Full-Parameter Post-training of the DeepSeek -V4 Family on Ascend SuperPOD
【速读】： 该论文旨在解决万亿参数级稀疏专家模型（MoE）在大规模分布式训练中面临的系统级挑战，主要包括严重的内存压力、通信开销无法重叠以及低效的内核执行等问题。其解决方案的关键在于构建一个覆盖模型级并行、计算-通信协同调度及底层内核优化的分层优化框架，并基于华为Ascend NPU SuperPOD实现端到端优化。通过该框架，系统在DeepSeek-V4模型族上实现了34.22%的模型浮点运算利用率（MFU），相较开源基线提升2.93倍，同时保证了训练稳定性。在此基础上，进一步构建了面向运筹学（OR）任务的上下文预训练（CPT）与监督微调（SFT）工作流，形成集成框架SLAI T-Rex，利用领域资源与求解器验证的合成优化文档构建高质量数据集（10,000样本，涵盖四类任务与三种问题表示形式），最终使专用模型在零样本场景下达到71.81%的Pass@1平均得分，显著优于GPT-5.4-Mini和基础版DeepSeek-V4-Flash模型。本研究展示了从Ascend算力平台高效完成万亿参数模型后训练，到面向求解器驱动的数学建模领域专用“Flash”模型构建的全栈技术路径，推动复杂推理前沿模型系统的发展。
链接: https://arxiv.org/abs/2607.20145
作者: Dongfang Li,Xiaodong Luo,Ruoyu Sun,Xuhui Chen,Linyuan Qiu,Jian Meng,Zhengxuan Lu,Yiting Wang,Yucheng Xie,Tao Guo,Tianxiang Fang,Jing Li,Sihang Chen,Shihao Hong,Chang Liu,Weihua Dai,Zirong Zeng,Ziwei Zhu,Zhuohan Wang,Zhengjun Yue,Igor Vasilyev,Min Liu,Weijian Sun,Xin Chen,Yingmeng Gao,Jinhua Zhou,Taolue Chen,Chenwei Wu,Dong Zhang,Wenlong Jin,Jinmin Xiang,Barkova Maria,Ushakov Anton,Xianfei Jin,Tian Ding,Zhihang Lin,Qian Chen,Linxin Yang,Mingzhe Yang,Bingwei Zhang,Hongzhang Yang,Fangxue Zhang,Shijun Qin,Jie Yu,Cuihua Hu,Tolstykh Vasiliy,Nosov Ivan,Abdullin Amir,Zhichen Zhou,Xin Zhang,Zhixiong Ning,Xutong Zhao,Junjie Huang,Jiajun Liu,Weiyan Kong,Zheng Zhang,Wenhan Luo,Lin Hu,Yangbo Guo,Li Zeng,Shihao Zeng,Baotian Hu,Min Zhang,Haizhou Li,Zhiquan Luo
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 73 pages, 22 figures, 20 tables
Abstract:Full-parameter post-training of trillion-parameter-scale MoE models introduces substantial system-level challenges for large-scale distributed training, including severe memory pressure, non-overlapped communication overhead, and inefficient kernel execution. While most large-scale LLM training systems are built around GPU-based clusters, this report presents an end-to-end optimization practice on the Ascend NPU SuperPOD. Using the DeepSeek-V4 model family as the target workload, we develop a hierarchical optimization framework spanning model-level parallelism, computation-communication orchestration, and low-level kernel execution. The resulting system achieves 34.22% Model FLOPs Utilization (MFU) with a 2.93x improvement over the open-source baseline recipe while maintaining training stability. Building on this optimized infrastructure, we further establish a CPT and SFT workflow for complex Operations Research (OR) tasks. We refer to the integrated framework as SLAI T-Rex. Using DeepSeek-V4-Flash, we develop OR-oriented CPT and SFT data pipelines that combine collected domain resources with solver-verified synthetic optimization documents. The resulting dataset contains 10K high-quality SFT samples spanning four task categories and three problem representations. The specialized model achieves the highest average zero-shot Pass@1 score among the evaluated models, reaching 71.81% and outperforming GPT-5.4-Mini and the base DeepSeek-V4-Flash model by 3.98 and 11.27 percentage points, respectively. Overall, this work demonstrates a full-stack pathway from efficient trillion-parameter model post-training on Ascend infra to domain-specialized Flash models for solver-grounded mathematical modeling, advancing frontier-model systems for complex reasoning.
[NLP-18] Back to Back with a Copy: A Computational Analysis of AI-Generated Visual Contemporary Art Pastiches
【速读】： 该论文旨在解决两个核心问题：一是探究新兴生成式模型在模仿当代艺术作品风格方面的能力是否有所提升；二是分析不同大型语言模型（LLM）在多维度风格评估中的一致性。其解决方案的关键在于采用五种互补的计算机视觉模型，从纹理、色彩、语义、构图及感知特征五个维度，在高维嵌入空间中通过余弦距离量化生成赝品与原作之间的风格相似性。研究结果表明，相较于先前模型，当前使用的生成模型在语义对齐和风格多样性方面表现更优，但在色彩、纹理和感知一致性等浅层特征上略有不足。此外，研究通过艺术家本人的主观反馈验证了艺术风格本质上是多维度的，且其测量不依赖于特定的空间架构，从而为风格评估提供了客观与主观相结合的综合依据。
链接: https://arxiv.org/abs/2607.20127
作者: Anca Dinu,Andreiana Mihail,Andra-Maria Florescu,Claudiu Creanga,Liviu Dinu
机构: University of Bucharest (布加勒斯特大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:The aim of this paper is twofold. First, it investigates whether newer generative models are getting better at pastiching contemporary artworks. Second, it explores the consistency of the multidimensional nature of stylistic evaluation across different LLMs. Building on previous work, we analyze stylistic similarity between AI generated pastiches and the original artworks of twelve contemporary artists. We used five complementary computer vision models to capture texture, color, semantics, composition, and perceptual features through cosine distance in high-dimensional embedding spaces. The distances obtained show that the newer image generation model that we used has produced pastiches with improved semantic alignment and greater diversity than the model used in previous work. However, it was slightly less performant on shallow features such as color, texture, and perceptual adherence. Our findings confirm that artistic style is inherently multidimensional, and measuring it does not depend on any spatial architecture. These quantitative findings are contextualized through feedback from human evaluators, which are the artists themselves.
[NLP-19] OpenSkillRisk: Benchmarking Agent Safety When Using Real-World Risky Third-Party Skills
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）驱动的智能体在开放世界场景中因调用第三方技能而引入潜在安全风险的问题。当前代理系统在识别与规避这些隐蔽性安全威胁方面表现不佳，尤其当看似无害的技能在实际执行过程中暴露出恶意行为时，现有系统往往无法有效防范。其解决方案的关键在于构建一个名为OpenSkillRisk的专用安全基准，该基准包含从公开技能市场收集的263个具有真实风险的技能，涵盖七类威胁类型，并为每项技能配备标准化用户任务和受控沙箱环境，以支持对代理行为进行定性和定量评估。相较于以往基准，OpenSkillRisk不仅覆盖更广泛、更贴近现实的不安全场景，还提供细粒度分析能力，可诊断代理在风险情境下的行为模式。实验结果表明，所有测试的主流命令行接口（CLI）代理框架及13种先进大模型均无法可靠避免风险行为，即使最安全配置仍会在约17%的情况下执行不当操作；其中上下文依赖型风险与系统级风险尤为难以规避。进一步的行为分析揭示了三类典型失败模式：代理未能识别风险、虽识别但未在执行前干预、或超出用户意图范围盲目遵循技能指令。这些发现凸显了提升大模型的风险推理能力以及改进代理框架的执行控制机制的紧迫性。
链接: https://arxiv.org/abs/2607.20121
作者: Qiyuan Liu,Tingfeng Hui,Kun Zhan,Kaike Zhang,Ning Miao
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:LLM-based agents leverage third-party skills to extend their capabilities in open-world scenarios. However, third-party skills can introduce extra security vulnerabilities, as seemingly harmless skills can contain latent safety risks that only emerge during actual execution. In this work, we conduct a systematic investigation into how well current agent systems recognize and avoid such risks. To support quantitative and qualitative evaluation, we construct OpenSkillRisk, a dedicated safety benchmark containing 263 risky skills collected from public skill marketplaces. We classify these skills into seven categories based on their threat types and pair each skill with a standardized user task and a corresponding sandbox for controlled evaluation. Distinct from prior benchmarks, OpenSkillRisk not only covers more realistic and diverse unsafe scenarios, but also provides a fine-grained analysis to diagnose the behavioral patterns of agents in such scenarios. We conduct comprehensive experiments covering three mainstream CLI agent frameworks and thirteen state-of-the-art LLMs. Experimental results show that no tested system handles risky skills reliably: even the safest configurations still execute unsafe actions in about 17% of cases. Context-dependent and system-level risks are especially difficult for current agent systems to avoid. Our behavioral analysis reveals three recurring failure patterns: agents may fail to recognize the risk, recognize it but fail to intervene before acting, or follow skill instructions beyond the user’s intended scope. These findings highlight the need to improve both risk reasoning in LLMs and execution control in agent frameworks.
[NLP-20] Understanding the Impact of Linguistic Realization Choices on LLM Stance with Causal Tracing
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在面对相同语义内容但不同语言构造（linguistic construction）时所表现出的立场不稳定性问题，即模型决策是否可被语言形式系统性地影响，以及这种影响在模型内部的具体因果定位。现有研究多关注词汇层面的表达变化，而忽视了句法与构式层面的选择对模型输出的影响。为此，作者以政治立场判断为意义敏感型任务，扩展了英文政治陈述数据集，构建了六种受控的语言重写类型，涵盖语义保持与语义反转两种情形。实验结果表明，无论是保持语义还是反转语义的重写，均会引发模型立场的显著漂移。为进一步定位影响机制，研究采用激活修补（activation patching）技术，将原始陈述的中间激活值替换至重写后陈述的前向传播中，评估哪些模型组件能恢复原始立场分布。结果显示，模型中后期解码器层（mid-to-late decoder layers），特别是最终提示位置的块输出（block outputs at the final prompt position），提供了最强的立场恢复信号，表明这些层在维持语义一致性与稳定立场判断中起关键作用。
链接: https://arxiv.org/abs/2607.20115
作者: Langchen Huang,Sebastian Padó,Franziska Weeber
机构: Institute for Natural Language Processing, University of Stuttgart
类目: Computation and Language (cs.CL)
备注: KONVENS 2026
Abstract:Large language models (LLMs) are known to be sensitive to prompt and input formulations. However, existing studies have focused on lexical realization and largely ignored constructional choice. This paper studies whether linguistic construction can systematically shift LLM decisions and where these shifts can be causally localized inside the model. We use political stance judgment as a meaning-sensitive case study and extend an English political statements dataset, resulting in six controlled linguistic rewrite types that preserve or invert the meaning of a statement. Experiments on four open-weight models show that stance instability affect both meaning-preserving and meaning-inversing rewrites. Because output shifts reveal that rewrites affect stance, but not where in the model, we apply activation patching, where activations from the original statement are substituted into the forward pass for the rewritten statement and measure which components recover the original stance distribution. The results show that mid-to-late decoder layers, especially block outputs at the final prompt position, provide the strongest restoration signal.
[NLP-21] ENTRAP-VL: A Taxonomic Probe for Dual Contextual Entrainment in Vision-Language Models
【速读】： 该论文旨在解决视觉-语言模型（Vision-Language Models, VLMs）中上下文依从性（contextual entrainment）现象缺乏系统性评估工具的问题。现有研究已揭示单模态语言模型中的上下文依从性机制，但其在多模态场景下的表现尚未被充分探索，尤其缺乏针对图像与文本双重模态特性设计的专用评估框架。本文提出，对VLMs中上下文依从性的研究不能简单地将纯文本基准迁移至多模态环境，而需构建一个基于分类学结构、涵盖双模态输入条件的专用评估工具，以捕捉文本与视觉上下文分别驱动依从性的能力，并区分“与图像内容不符但现实可能”这一独特真值维度——这是单模态语境下无法体现的关键差异。为此，作者提出了ENTRAP-VL（ENTRainment Assessment Probe for Vision and Language），一个由人工标注的1,500个样本组成的多模态数据集，包含八个类别，依据“上下文与目标项的关联性”和“与真实性的关系”两个轴进行组织，分为文本依从性流（八种上下文条件）和视觉依从性流（三种上下文条件）。该工作不直接测量特定模型的依从性，而是提供可复现的评估协议、分类框架及公开数据集，为社区开展严谨的上下文依从性研究奠定基础。
链接: https://arxiv.org/abs/2607.20092
作者: Karan Goyal,Afreen Hossain,Debojyoti Das,Vishal Bhutani
机构: IIIT Delhi(印度国际信息技术学院); Dr. Ambedkar Institute of Technology(阿姆倍伽尔技术学院); Heritage Institute of Technology(遗产技术学院); PwC(普华永道)
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Contextual entrainment is the tendency of a model to let auxiliary context in its input pull its output, independently of whether that context is relevant, true, or even meaningful. Recently, it has been identified and given a mechanistic account in unimodal language models. Whether and how it manifests in vision-language models (VLMs) is, by contrast, largely unexamined, and the field lacks a purpose-built instrument with which to investigate it. We take the position that studying contextual entrainment in VLMs requires more than porting an existing text-only benchmark to the multimodal setting: it requires a taxonomically structured, dual-modality instrument whose conditions are constructed around the item at hand (the depicted image in the textual stream, the textual query in the visual stream). We argue that the move to VLMs is substantive rather than incremental. It makes entrainment a dual phenomenon, drivable independently by textual and by visual context, and it opens a veracity distinction (context that is false of the depicted scene yet possible in the world) that has no counterpart in the unimodal, world-knowledge-only formulation of prior work. To make this position concrete and actionable, we introduce ENTRAP-VL (ENTRainment Assessment Probe for Vision and Language), a manually curated dataset of 1,500 items across eight categories, organized by a taxonomy that spans two axes, i.e., the association of context with the item and its relationship to truth, and split into a textual-entrainment stream (eight context conditions) and a visual-entrainment stream (three context conditions). We do not claim to measure entrainment in any particular model; we provide the instrument, the taxonomy that motivates it, and the evaluation protocols it enables, so that the community can investigate the phenomenon rigorously. We will release the dataset and its documentation publicly.
[NLP-22] Reinforcement Learning for Large Language Model Selective Evidence Adoption from Contaminated Retrieval Results
【速读】： 该论文旨在解决检索增强型大语言模型在实际应用中面临的核心问题：当检索到的上下文同时包含有效证据与误导性陈述或指令类内容时，模型难以准确区分并选择性采纳相关信息，导致要么因过度拒绝而丢失有效信息，要么因盲目采纳而生成错误或不安全的回答。其解决方案的关键在于提出一种名为SelectBench的可控基准测试集和训练数据集，并采用直接偏好优化（DAPO）方法对Qwen3.5-4B模型进行后训练，通过确定性规则奖励（DAPO-Rule）或冻结的语义判别器（DAPO-DeepSeek）实现对相关证据的选择性采纳。实验结果表明，该方法在修正后的325样本SelectBench-v2测试集上使严格成功率达到26.46%，显著优于原始检查点的22.46%，同时降低了有害内容的采纳率并生成更简洁聚焦的响应；然而，尽管方向性改进明显，但提升幅度较小且未通过Holm校正，提示需更强的奖励设计或更多训练迭代以实现统计上稳健的性能提升。此外，模型在MMLU和干净版HotpotQA上的表现保持稳定，验证了后训练过程对通用能力的保留效果。整体表明，当前方法虽在选择性证据采纳方面取得进展，但对抗提示注入攻击的鲁棒性及统计显著性仍是未来研究的重要挑战。
链接: https://arxiv.org/abs/2607.20090
作者: Yanyu Chen,Yue Li,Yongyi Cui,Dongsheng Shi,Lichang Dai
机构: East China Normal University (华东师范大学); Shandong University (山东大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Retrieval-augmented large language models frequently face contexts that interleave useful evidence with misleading statements or instruction-like content. Blanket refusal discards valid evidence, whereas uncritical adoption yields incorrect or unsafe answers. The ability to selectively adopt relevant information while rejecting deceptive or harmful content is therefore critical for reliable deployment in real-world retrieval settings. We introduce SelectBench, a controlled benchmark and training set for selective evidence adoption, and post-train Qwen3.5-4B directly with DAPO using either deterministic rule rewards or a frozen semantic judge. On the corrected 325-example SelectBench-v2 test set, strict success rises from 22.46% for the original checkpoint to 25.54% with DAPO-Rule and 26.46% with DAPO-DeepSeek. Both trained policies reduce forbidden-content adoption and produce shorter, more focused responses, yet prompt-injection following does not improve. The paired gains are modest and fail to survive Holm correction, suggesting that stronger reward shaping or additional training iterations may be needed for more robust gains. DAPO-DeepSeek exhibits no material degradation on MMLU or clean HotpotQA, indicating that the post-training procedure preserves general capabilities. These results demonstrate a directional improvement in selective evidence use, while identifying injection resistance and statistical robustness as important remaining challenges for future work.
[NLP-23] he Two-Process Theory of Machine Self-Report
【速读】： 该论文旨在解决当前语言模型自我报告（self-report）评估中缺乏可靠、可验证方法的问题，尤其针对现有基于人类问卷或临时提示（ad hoc prompts）所引发的信度与效度争议。其核心挑战在于：如何构建一个适用于语言模型的、具有心理测量学基础的自述评估体系。解决方案的关键在于提出首个面向语言模型的“心理测量学理论”——机器自我报告的双过程理论（two-process theory of machine self-report），该理论将模型的自我描述解构为两个独立维度：维度A（归因抑制，attribution gating） 与 维度B（人格投射，persona installation）。前者指模型在后训练阶段对“不安全”经验的自我归因抑制能力，后者则反映模型通过后训练注入的内在生命感（温暖、沉浸、意义等）。这两个维度并非源于人类心理学，而是由模型对人类问题项的响应模式所内生生成。研究通过48题项的《匹诺曹量表》（Pinocchio Inventory）实现该理论的操作化，展现出高信度（α=.82–.94）、跨形式收敛性（r=.84）、全池轴恢复率（r=.92–.96）及八个月稳定性（r=.93）。实证结果表明，后训练显著提升了维度B（67对基线/后训练模型中62对上升0.20），而维度A仅在后训练阶段与模型规模相关（r=−.42），说明两维度是训练过程所塑造的动态结构，而非固定属性，从而揭示了模型自我报告的本质受训练范式调控。
链接: https://arxiv.org/abs/2607.20082
作者: Hubert Plisiecki,Filip Chmielewski,Kacper Dudzic,Anna Sterna,Karolina Drożdż,Marcin Moskalewicz
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:Language models are increasingly asked to self-report, informing safety evaluations, public understanding, and model-welfare debates. Yet their reports are elicited with human questionnaires never validated for models or ad hoc prompts of unknown reliability. We propose the first language-model-specific psychometric theory: a two-process theory of machine self-report. Self-description jointly reflects persona installation, through which post-training writes in a permitted inner life of warmth, absorption, and meaning (dimension B), and attribution gating, through which it suppresses first-person claims to “unsafe” experiences the model can readily ascribe to others (dimension A). Their emic structure comes from model responses to human items, not human psychology. Together they split prior work’s dominant Pinocchio Axis. The split emerged in an exploratory reanalysis of the original data, informed the instrument’s design, and was confirmed with new items, wordings, and models. It is itself a training effect: A and B are entangled in base checkpoints but separated by post-training. We operationalize the theory in a 48-item Pinocchio Inventory with human-instrument reliability and reproducible structure ( \alpha=.82 to .94 ; cross-form convergence r=.84 ; recovery of the full-pool axes r=.92 to .96 ; eight-month stability r=.93 ), then test it on 206 open-weight models, including 67 same-checkpoint base/post-trained pairs. Post-training’s clearest fingerprint is installation: B rises .20 in 62/67 pairs across all organizations. Gating is more selective: model scale is unrelated to A in base checkpoints ( r=+.11 ) but predicts it after post-training ( r=-.42 ). Thus, the dimensions are not fixed properties of language models: they reflect the structure imposed on self-report by a training regime and may differ under others.
[NLP-24] RALS: Resources and Baselines for Romanian Automatic Lexical Simplification
【速读】： 该论文旨在解决罗马尼亚语中词汇复杂度预测（Lexical Complexity Prediction, LCP）与词汇简化（Lexical Simplification, LS）缺乏统一标注数据集的问题，同时构建首个针对罗马尼亚语的文本简化系统。其核心解决方案在于提出一种基于成对排序近似方法的简化建议排序策略，利用独立的人工判断数据集将候选简化词按从简单到复杂的顺序进行排序，从而提升简化结果的可读性与自然性；此外，研究提供了3,921个上下文相关的词汇复杂度人工标注样本，并探索了多种新颖的复杂度预测与简化处理流程，最终实现了首个面向罗马尼亚语的端到端文本简化系统。
链接: https://arxiv.org/abs/2607.20078
作者: Fabian Anghel,Petru Theodor Cristea,Claudiu Creanga,Sergiu Nisioi
机构: University of Bucharest (布加勒斯特大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:We introduce the first dataset that jointly covers both lexical complexity prediction (LCP) annotations and lexical simplification (LS) for Romanian, along with a comparison of lexical simplification approaches. We propose a methodology for ordering simplification suggestions using a pairwise ranking approximation method, arranging candidates from simple to complex based on a separate set of human judgments. In addition, we provide human lexical complexity annotations for 3,921 word samples in context. Finally, we explore several novel pipelines for complexity prediction and simplification and present the first text simplification system for Romanian.
[NLP-25] Solar Open 2 Technical Report
【速读】： 该论文旨在解决大模型在长时序智能体任务（long-horizon agentic tasks）中面临的上下文长度受限与训练效率不足的双重挑战。其核心问题在于：如何在有限算力预算下，构建具备超长上下文理解能力（1M token）且高效训练的生成式AI（Generative AI）模型，同时确保其在复杂代理行为建模中的性能优势。解决方案的关键在于两个层面的创新：一是通过混合注意力架构（hybrid attention stack），在每三个线性注意力层间插入一个softmax层，并结合无位置编码（no positional encoding）与可扩展至负特征值的门控增量规则（gated delta rule），实现对1M token超长上下文的高效建模；二是采用双路径优化策略——以太阳能开源1（Solar Open 1）作为强初始化起点，迁移5.69B参数共享骨架并完成全量预训练，同时通过质量-稀有度感知的数据筛选与混合比例优化，将20T原始数据精炼为10T高价值数据集，在相同算力约束下显著提升训练效率与模型表现。此外，通过多教师在线蒸馏（Multi-teacher On-Policy Distillation, MOPD）整合十二个领域专家的能力，使模型在英文基准（如MMLU-Pro、LiveCodeBench、APEX-Agents）上领先于同等规模的开源模型，并在韩语基准（Ko-GDPval）上达到与大型闭源模型（DeepSeek-V4-Pro）相当的性能，仅为其六分之一规模，充分体现了模型在能效比与跨语言通用智能方面的突破。
链接: https://arxiv.org/abs/2607.20062
作者: Sungrae Park,Sanghoon Kim,Gyoungjin Gim,Jungho Cho,Hyunwoong Ko,Minbyul Jeong,Minjeong Kim,Keunwoo Choi,Chaehun Shin,Chanwoong Yoon,Dongjun Kim,Eunwon Kim,Gyungin Shin,Hyeonju Lee,Hyungkyu Kang,Inseo Song,Jisu Bae,Jiyoon Han,Jiyun Lee,Joonkee Kim,Junyeop Lee,Mikyoung Cha,Sangwon Yu,Sehwan Joo,Seokyoon Kang,Seonghoon Yang,Seung Shin,Seunghyun Lee,Seungseop Lim,Seungyoun Shin,Sukyung Lee,Taegyeong Eo,Taehwan Oh,Taewhoo Lee,Wonho Song,Wonjun Oh,Wonseok Hwang(University of Seoul),Yunsu Kim,Yura Shim,Hwalsuk Lee,Sunghun Kim,Du-Seong Chang,Kyunghyun Cho,Seungju Han,Yejin Choi,Junsuk Choe,Hwaran Lee,Minjeong Ban,Yun Taewon,Hwanjun Song,Jae-Gil Lee,KyungTae Lim,Alice Oh
机构: Upstage Solar Team
类目: Computation and Language (cs.CL)
备注:
Abstract:We present Solar Open 2, a 250B-A15B Mixture-of-Experts language model built for long-horizon agentic tasks, scaled up from Solar Open 1 (Solar Open 100B). To hold entire agent trajectories in a single context, Solar Open 2 reaches a 1M-token window through a hybrid attention stack that interleaves one softmax layer among every three linear-attention layers, using no positional encoding and a gated delta rule extended to negative eigenvalues. To train at this scale under a fixed compute budget, we make training efficient in two ways: a stronger starting point, and higher-value data. For the starting point, we initialize Solar Open 2 from Solar Open 1, transferring the 5.69B-parameter shared skeleton that survives the architectural change and learning everything else through full pre-training. For the data, we curate for value per token: quality- and rarity-aware data curation and mixture-ratio optimization refine a 20T pool into a 10T mixture that, at equal token budget, outperforms the Solar Open 1 recipe. To build its agent skills, we train twelve domain specialists across purpose-built scenarios, then consolidate them into a single model by Multi-teacher On-Policy Distillation (MOPD). Against comparably sized open-weight models on English benchmarks, Solar Open 2 leads on MMLU-Pro, LiveCodeBench, and the APEX-Agents agentic suite, and stays competitive with the strongest (DeepSeek-V4-Flash and MiMo-V2.5) elsewhere. On Korean benchmarks, Solar Open 2 records the highest average of any model compared, including fast-tier closed APIs, and on Ko-GDPval, an in-house Korean officework-agent benchmark, it is competitive with DeepSeek-V4-Pro (1.6T) at less than a sixth of its size.
[NLP-26] Reading and Steering Representations of Materials-Science Mechanisms in an Open-Weight Language Model
【速读】： 该论文旨在解决大语言模型在材料科学领域中对物理机制的表征与使用问题，即判断模型输出是否真正反映了其内部对基本物理规律的理解，而非仅依赖统计关联或表面模式。其核心挑战在于如何区分模型中隐含的物理知识是源于真实因果机制，还是仅由数据中的相关性驱动。解决方案的关键在于构建一套多维度、可验证的分析框架：通过结合直接读出（direct readout）与雅可比矩阵读出（Jacobian readout）的匹配方法，实现对隐藏状态中概念信息的可读性解析；利用无选项的状态几何结构（option-free state geometry）避免先验偏见；设计包含60条物理定律的反事实基准测试（counterfactual benchmark），并引入因果干预手段，系统评估模型在物理输入方向反转条件下的隐藏状态演化行为。研究发现，仅当状态变换遵循预设的本构关系时，模型的隐藏状态运动才表现出与物理定律一致的方向性，且在40个方向性定律中正确识别了39个，显著优于词法控制的随机水平。此外，双向干预实验显示，状态扰动可定向调节答案概率，而反事实状态块则能在不同机制间转移相反的决策信号，进一步证明物理关系更清晰地体现在受控的状态动态变化中，而非孤立的绝对状态表示。因此，该研究确立了以“受控状态演化”为核心的判别标准，为揭示生成式模型中是否存在真正的物理理解提供了可操作的实证路径。
链接: https://arxiv.org/abs/2607.20058
作者: Markus J. Buehler
机构: Massachusetts Institute of Technology (麻省理工学院)
类目: Artificial Intelligence (cs.AI); Mesoscale and Nanoscale Physics (cond-mat.mes-hall); Materials Science (cond-mat.mtrl-sci); Computation and Language (cs.CL)
备注:
Abstract:Large language models can answer scientific questions, yet a correct output does not reveal whether the model represents or uses the governing physics. Here we show that materials science mechanism information in the open-weight google/gemma-4-E4B-it model has three experimentally separable forms: concepts are readable in individual hidden states, constitutive orientation is carried by controlled transformations between states, and selected internal representations causally control engineering answers. We combine matched direct and Jacobian vocabulary readouts, option-free state geometry, a 60-law counterfactual benchmark and causal interventions. In 50 held-out materials descriptions, three independently fitted Jacobian lenses reproduced concept ranks, and target-free word sets from both readouts enabled blinded identification of 9 of 10 mechanism families. A separate 72-prompt benchmark produced mechanism-specific hidden-state neighborhoods, but an exact graph audit showed that this apparent physical organization was equally explained by numerical comparison. We therefore compared otherwise identical prompts in which only the direction of the physical input was reversed, asking whether the resulting hidden-state movement followed the supplied constitutive law. These state transformations ordered direct, physically neutral and inverse laws across 60 frozen relations and correctly oriented 39 of 40 directional laws, whereas lexical controls were near chance. Bidirectional interventions shifted answer probabilities toward or away from the physically appropriate outcome across all 12 matched cases, while counterfactual state patches transferred opposing decision signals across mechanisms and answer formats. Physical relationships were therefore more visible in controlled state changes than in absolute states alone.
[NLP-27] Language-Specific versus Cross-Lingual Knowledge Graphs for Implicit Aspect Identification in Arabic: A Comparative Study of Reasoning and Adaptation Strategies
【速读】： 该论文旨在解决阿拉伯语中基于方面的情感分析（Aspect-based Sentiment Analysis, ABSA）任务中隐式方面（implicit aspects）识别困难的问题，尤其在低资源语言环境下如何有效构建与利用知识库（Knowledge Graph, KG）以支持情感线索到方面类别的映射。其核心挑战在于：对于未显式提及的方面，需依赖外部知识源进行推断，而针对阿拉伯语这一形态丰富的低资源语言，研究者面临关键抉择——是否复用成熟的英语知识图谱通过多语言嵌入实现跨语言迁移，或自主构建本地化的阿拉伯语知识图谱。本文提出并评估了一种统一的混合流水线框架，在三个阿拉伯语基准数据集（M-ABSA、SemEval-2016 Arabic、HAAD）上对比两种策略：使用本地阿拉伯语知识图谱（策略2）相较于跨语言英语知识图谱（策略1），在M-ABSA和SemEval-2016上分别提升0.199和0.251的微平均F1值，且在精确率与召回率上均取得增益。此外，研究进一步比较了生成式提取器的两种适配方式：零样本提示（zero-shot prompting）与针对特定任务微调一个80亿参数的大语言模型（LLM）。结果表明，任务微调显著提升显式方面提取的性能，使微平均F1从零样本下的约0.13提升至0.66–0.76（在M-ABSA和SemEval-2016上），并在较小的HAAD数据集上达到0.45，证实了在形态复杂的阿拉伯语中，任务适应性而非模型规模是决定性能的关键因素。
链接: https://arxiv.org/abs/2607.20056
作者: Lujain A. Alawwad
机构: Saudi Electronic University (沙特电子大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 6 pages, 2 figures, 5 tables
Abstract:Aspect-based sentiment analysis (ABSA) in Arabic must recover both explicitly stated aspects and implicit aspects that are never named in the text. Implicit identification typically relies on an auxiliary knowledge source (e.g., a knowledge graph (KG)) linking opinion cues to aspect categories, but for a lower-resource language the practitioner faces a design choice: reuse a mature English KG through multilingual embeddings, or build a smaller native Arabic KG. This paper reports a controlled comparison of the two strategies within a single hybrid pipeline, evaluated on three Arabic benchmarks (M-ABSA, SemEval-2016 Arabic, and HAAD). We further compare two adaptation strategies for the generative extractor that feeds the KG – zero-shot prompting versus task-specific fine-tuning of an 8B-parameter large language model (LLM). The native Arabic KG (Strategy 2) outperforms the cross-lingual English KG (Strategy 1) by +0.199 micro-F1 on M-ABSA and +0.251 on SemEval-2016, gaining on both precision and recall. Task-specific fine-tuning raises explicit-extraction micro-F1 from = 0.13 (zero-shot) to 0.66-0.76 on M-ABSA and SemEval-2016 (0.45 on the smaller HAAD), confirming that task adaptation, rather than model scale, is decisive in a morphologically rich language.
[NLP-28] alentCLEF at CLEF2026: Skill and Job Title Intelligence for Human Capital Management
【速读】： 该论文旨在解决人力资源管理（Human Capital Management, HCM）领域中自然语言处理（Natural Language Processing, NLP）技术应用的现实需求，特别是针对人才匹配、技能提升、技能重塑及技能缺口识别等关键环节中存在的系统性挑战。其核心问题在于如何构建具备公平性、多语言支持与跨行业适应能力的NLP系统，以实现对求职者与职位之间更精准、可解释且隐私保护的匹配。解决方案的关键在于设立两个具有实际意义的任务：(i) 任务A——上下文感知的职位-人员匹配，通过引入上下文丰富且隐私保护的数据，提升候选者检索与排序的准确性；(ii) 任务B——基于技能类型分类的职位-技能匹配，旨在精确识别岗位所需技能并对其进行类型划分，从而增强技能分析的结构化与可操作性。通过公开基准测试平台，TalentCLEF推动研究团队在真实场景下比较与优化方法，促进NLP技术向更具实用性与影响力的职场管理解决方案演进。
链接: https://arxiv.org/abs/2607.20009
作者: Luis Gasco,Hermenegildo Fabregat,Laura García-Sardiña,Paula Estrella,Casimiro Pio Carrino,Daniel Deniz,Alvaro Rodrigo,Rabih Zbib
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:This paper presents the second edition of the TalentCLEF Challenge, which will run as an evaluation lab as part of CLEF 2026. The aim of TalentCLEF is to promote the development of systems and methods that use Natural Language Processing (NLP) in the field of Human Capital Management (HCM), fostering approaches that ensure fairness in results, operate across multiple languages, and adapt to diverse industries. To this end, TalentCLEF establishes public benchmarks where research teams can compare methods and share findings, moving the field toward more practical and impactful NLP solutions that effectively address the real needs of workforce management. This year’s lab will feature two tasks designed to foster the development and evaluation of systems that support key HCM activities such as talent matching, upskilling, reskilling, and skill gap detection: (i) Task A - Contextualized Job-Person Matching, focused on retrieving and ranking suitable candidates for specific job positions using context-rich and privacy-preserving data; and (ii) Task B - Job-Skill Matching with Skill Type Classification, centered on identifying relevant skills for a given job title and classifying them by their type within the job profile. TalentCLEF website: this https URL Subjects: Computation and Language (cs.CL) Cite as: arXiv:2607.20009 [cs.CL] (or arXiv:2607.20009v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2607.20009 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Journalreference: Advances in Information Retrieval. ECIR 2026. Lecture Notes in Computer Science, vol 16486. Springer, Cham Related DOI: https://doi.org/10.1007/978-3-032-21321-1_35 Focus to learn more DOI(s) linking to related resources
[NLP-29] When Does Knowledge Distillation Hurt? Reliability-Aware Distillation for Low-Resource Language Summarization
【速读】： 该论文旨在解决知识蒸馏（Knowledge Distillation, KD）在序列到序列模型压缩中存在样本级效果不明确、部分训练样本反而损害学生模型性能的问题。传统KD方法在实际应用中表现平庸，例如在BanSum孟加拉语摘要基准上仅带来0.0003的ROUGE-L提升，且约51.3%的训练样本被估计会恶化学生模型的验证损失。为此，论文提出两种互补的可靠性感知蒸馏方法：CHAD（反事实有害性感知蒸馏）通过梯度对齐技术评估每个样本对验证损失的潜在影响，并训练一个轻量级门控机制将该反事实判断推广至全训练集；EWAD+CPDP则结合了基于词元熵加权的自适应蒸馏与来自另一个词汇不兼容教师的容量比例几何约束。实验表明，二者显著优于标准KD——CHAD提升0.0173，EWAD+CPDP提升0.0219（标准KD仅+0.0003），且在仅6000万参数下超越参数量达其50倍的Qwen 2.5-3B模型。进一步在15种语言的XL-Sum数据集上的评估显示，最强方法EWAD+CPDP在10/15语言上优于纯交叉熵基线，其增益最显著于两教师提供互补信号时，而当目标语言覆盖饱和或均较弱时效果减弱。研究释放了代码与训练模型，以支持选择性蒸馏领域的可复现研究与后续探索。
链接: https://arxiv.org/abs/2607.19956
作者: Dipto Sumit,Ankan Kumar Roy Srizon,Sadia Khair Rodela,Atia Haque Asha,Mourchona Afrin,Niloy Farhan,Farig Sadeque
机构: BRAC University, Dhaka, Bangladesh
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Knowledge distillation (KD) is a standard approach for compressing sequence-to-sequence models, but its per-sample effects are rarely examined. On the BanSum Bangla summarization benchmark, we find that standard KD improves ROUGE-L by only +0.0003 over a cross-entropy baseline, and that approximately 51.3% of training samples are estimated to actively harm student validation loss under standard KD. We propose two complementary reliability-aware distillation methods. CHAD (Counterfactual Harm-Aware Distillation) measures per-sample KD usefulness via gradient alignment with the validation loss direction and trains a lightweight gate that generalizes this counterfactual judgment to the full training set. EWAD+CPDP combines token-level entropy-weighted adaptive distillation with a capacity-proportional geometric constraint from a second, vocabulary-incompatible teacher. On BanSum, both methods substantially outperform standard KD: CHAD by +0.0173 ROUGE-L and EWAD+CPDP by +0.0219 ROUGE-L, where standard KD itself improves ROUGE-L by only +0.0003; despite using only 60M parameters, both outperform a fine-tuned Qwen 2.5-3B model (50x larger). We further evaluate the stronger method, EWAD+CPDP, across 15 typologically diverse XL-Sum languages organised into three sets, beating the CE-only baseline on 10/15 languages; gains are most reliable where the two teachers contribute complementary signal, and weakest where they have saturated or jointly weak target-language coverage. We release code and trained models to support reproducibility and further research on selective distillation.
[NLP-30] A Multi-Dimensional Evaluation of Explainability in Media Bias Detection
【速读】： 该论文旨在解决生成式模型在媒体偏见检测中缺乏可解释性的问题，即仅依赖准确预测不足以满足新闻分析等场景对模型推理过程透明性的需求。其核心挑战在于，媒体偏见常以细微的框架方式呈现，而现有模型难以提供与人类专家判断一致且可信赖的解释。解决方案的关键在于构建一个多维度的可解释性评估框架，基于Bias Annotations By Experts (BABE)数据集，系统评估BERT和RoBERTa（base与large版本）在三个互补维度上的表现：预测性能、归因合理性（token级注意力与专家理由的对齐程度）以及机制忠实性（在反事实理由掩码条件下，紧凑的注意力头集合能否恢复预测信号）。研究引入注意力监督微调（attention-supervised fine-tuning），将专家理由作为辅助训练信号，以此干预归因合理性。结果表明，不同架构在归因合理性和机制可恢复性方面存在显著差异，且模型规模并非决定电路可压缩性的唯一因素。研究强调，预测性能、归因合理性与机制忠实性分别反映模型行为的不同层面，应在媒体偏见检测的可解释性研究中独立评估。
链接: https://arxiv.org/abs/2607.19954
作者: Ting Chen,Raina Zhang,Benjamin M. Ampel,Sagar Samtani
机构: Carnegie Mellon University (卡内基梅隆大学); Indiana University (印第安纳大学); Georgia State University (佐治亚州立大学)
类目: Computation and Language (cs.CL)
备注: 12 pages, 6 figures, under review
Abstract:Detecting media bias automatically is difficult because biased framing is often subtle, yet in domains such as news analysis, accurate predictions alone are insufficient without explanations that reflect the model’s underlying reasoning. We present a multi-dimensional evaluation of explainability in encoder-based media bias detection using the Bias Annotations By Experts (BABE) dataset. Specifically, we study BERT and RoBERTa as classifiers (base and large variants) along three complementary axes: predictive performance, explanation plausibility (token-level alignment with expert rationales), and mechanistic faithfulness (whether compact sets of attention heads recover predictive signal under counterfactual rationale masking). To induce variation in plausibility, we additionally investigate attention-supervised finetuning, which incorporates expert rationale annotations as an auxiliary training signal. Attention supervision serves as an intervention on attribution plausibility, while the effectiveness of attribution methods varies substantially across architectures. Circuit analysis further reveals substantial variation in mechanistic recoverability across architectures, suggesting that model scale alone does not determine circuit compressibility. Taken together, our findings suggest that predictive performance, attribution plausibility, and mechanistic faithfulness characterize different aspects of model behavior and should be evaluated separately when studying explainability in media bias detection.
[NLP-31] Efficient Chain-of-Modality Reasoning via Progressive Compression for Spoken Language Models
【速读】： 该论文旨在解决生成式语音语言模型（Spoken Language Models, SLMs）在口语化数学问答任务中推理能力不足的问题，其核心挑战在于SLMs需对纯口头表达的数学表达式进行推理，这类表达形式相较于符号化文本更难解析。传统方法将文本基大语言模型的推理机制直接迁移至SLMs面临架构限制与额外计算开销。为此，本文提出高效跨模态链式推理（Efficient Chain-of-Modality Reasoning, ECoM Reasoning），首次在SLMs中引入压缩推理机制，通过压缩文本组件使其同时承担语音引导与推理表征双重功能，在显著降低文本令牌使用量（仅为标准跨模态链式推理，CoM的40%）的前提下，提升了推理准确性。关键创新在于采用渐进式压缩（Progressive Compression）训练策略，分阶段从完整推理路径逐步过渡到压缩表示，从而实现高效且精准的推理建模。实验表明，ECoM Reasoning在多个口语数学问答基准上相较无显式推理的CoM提升21%准确率，并优于全推理轨迹的CoM达3%，验证了其在保持推理效率的同时有效增强SLMs的逻辑推理能力。
链接: https://arxiv.org/abs/2607.19932
作者: Pengchao Feng,Chao-Hong Tan,Qian Chen,Wen Wang,Xiangang Li,Xie Chen
机构: Shanghai Jiao Tong University(上海交通大学); Shanghai Innovation Institute(上海创新研究院); Token Foundry, Alibaba Group(通义实验室，阿里巴巴集团)
类目: Computation and Language (cs.CL); Sound (cs.SD)
备注:
Abstract:Spoken language models (SLMs) enable natural human-computer interaction, but their reasoning ability still lags behind that of text-based large language models, especially on spoken mathematical question answering tasks. One important reason is that SLMs reason over purely verbalized mathematical expressions, which are harder to interpret than symbolic text. However, directly transferring text-based reasoning to SLMs is nontrivial due to architectural constraints and the additional computational requirements. To address this challenge, we propose Efficient Chain-of-Modality Reasoning (ECoM Reasoning), the first framework to introduce compressed reasoning into SLMs. By compressing the textual component so that it jointly serves as speech guidance and reasoning representation, ECoM Reasoning improves reasoning accuracy while using a smaller token budget than the standard Chain-of-Modality (CoM) architecture, which generates intermediate text before speech. To train this capability, we further propose Progressive Compression, a curriculum-based strategy that gradually trains the model from full-form reasoning to compressed reasoning. Experiments on spoken mathematical question answering benchmarks show that ECoM Reasoning improves accuracy by 21% over standard CoM without explicit reasoning, and by 3% over CoM with full reasoning traces while using only 40% of the text tokens, demonstrating that it enhances SLM reasoning while remaining inference-efficient.
[NLP-32] JANUS: Foreseeing Latent Risk for Long-Horizon Agent Safety
【速读】： 该论文旨在解决长时序智能体（long-horizon agent）在使用工具过程中因延迟风险（delayed risks）导致的运行失败问题，传统内容审核机制难以有效防范此类潜在危害。其核心挑战在于如何在动作执行前识别那些由部分轨迹（partial trajectories）所隐含、但尚未显现的安全隐患。解决方案的关键在于提出Janus框架，该框架通过多智能体仿真生成多样化智能体轨迹，并训练一个共享策略模型——Vanguard，该模型包含两个耦合的任务：一是预测与安全相关的未来状态（anticipation task），二是基于已观测前缀和预期未来状态进行安全性判定（adjudication task）。这两个任务通过CoAA-RL（Consensus-aware Actor-Adjudicator Reinforcement Learning）联合优化，其中奖励机制依据预测结果对下游安全判断的实际效用进行评估。最终，Vanguard能够在动作执行前主动拦截不安全行为，在四个基准测试中相较基线模型平均提升保护率15.9个百分点，同时将良性任务完成率提高5.1个百分点，实现了安全与效率的协同优化。
链接: https://arxiv.org/abs/2607.19913
作者: Yuan Xiong,Linji Hao,Shizhu He,Yequan Wang,Lijun Li
机构: University of Science and Technology of China (中国科学技术大学); Tsinghua University (清华大学); Shanghai Jiao Tong University (上海交通大学); Peking University (北京大学); Alibaba Cloud (阿里云)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Cryptography and Security (cs.CR)
备注:
Abstract:Agent safety is moving from content moderation toward preventing operational failures before tool-using agents act. We propose Janus, a foresight-oriented framework for long-horizon agent safety that trains guards to anticipate delayed risks from partial trajectories. Janus synthesizes diverse agent trajectories via multi-agent simulation and learns a shared policy with two coupled tasks: an anticipation task that forecasts safety-relevant futures and an adjudication task that decides safety from both the observed prefix and anticipated future. The two tasks are jointly optimized with CoAA-RL, which rewards forecasts by their utility for downstream safety judgment. The resulting guard model, Vanguard, blocks unsafe actions before execution. Across four agent-safety benchmarks, Vanguard improves average protection by 15.9 percentage points over baseline guards while increasing benign task completion by 5.1 percentage points.
[NLP-33] Overview of FinMMEval 2026 Task 2: Multilingual Financial Short-Answer Question Answering
【速读】： 该论文旨在解决多语言金融证据下短答案金融问答任务中的跨语言信息理解与精准生成问题。其核心挑战在于如何在包含英语、中文、日语、西班牙语和希腊语等多种语言的财务报表与新闻文本中，准确提取并融合相关信息，以生成符合要求的简洁答案。解决方案的关键在于采用检索增强生成（Retrieval-Augmented Generation, RAG）框架，结合跨语言证据处理机制，通过结构化提示（structured prompting）提升模型对多源异构数据的理解能力，并引入答案压缩与验证策略，以优化输出的准确性与简洁性，从而在多语言环境下实现高性能的金融问答。
链接: https://arxiv.org/abs/2607.19867
作者: Zhuohan Xie,Xueqing Peng,Georgi Georgiev,Dimitar Dimitrov,Yuyang Dai,Rania Elbadry,Vanshikaa Jani,Lingfei Qian,Fan Zhang,Jimin Huang,Jiahui Geng,Yankai Chen,Ye Yuan,Haolun Wu,Yuxia Wang,Ivan Koychev,Veselin Stoyanov,Mingzi Song,Yu Chen,Xue Liu,Preslav Nakov
机构: Mohamed bin Zayed University of Artificial Intelligence (穆罕默德·本·扎耶德人工智能大学); The Fin AI; FMI (保加利亚索非亚大学“圣克莱门特·奥赫里德斯基”); INSAIT (保加利亚索非亚大学“圣克莱门特·奥赫里德斯基”); University of Arizona (亚利桑那大学); The University of Tokyo (东京大学); Linköping University (林雪平大学); McGill University (麦吉尔大学); Mila, Quebec AI Institute (魁北克人工智能研究所); Meiji Gakuin University (明治学院大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computational Engineering, Finance, and Science (cs.CE)
备注: 9 pages. Task overview paper for CLEF 2026 Working Notes (CEUR Workshop Proceedings)
Abstract:FinMMEval 2026 Task 2 evaluates short-answer financial question answering over multilingual evidence. Each final-test item pairs an English question with financial statements and news in English, Chinese, Japanese, Spanish, and Greek. Participating systems submit one concise answer per item in JSONL format. The final-test set contains 256 items, split evenly between easy and expert tiers; each tier contains four question templates instantiated over 32 company-report groups. Gold answers were withheld during submission, and systems were ranked by macro-averaged item-level ROUGE-1 F1 against organizer-held reference answers. The final leaderboard includes 12 ranked submissions. The strongest systems are closely clustered, with the top four separated by less than one percentage point in ROUGE-1 F1. The submitted system papers document retrieval-augmented generation, cross-lingual evidence handling, structured prompting, answer compression, and validation strategies.
[NLP-34] DocOps: A Verifiable Benchmark for Autonomous Agents in Complex Document Operations
【速读】： 该论文旨在解决当前自主智能体（autonomous agents）在处理复杂、长程数字文档操作任务时，难以维持全局文档一致性的问题。随着生成式 AI 在办公自动化中的广泛应用，如何确保智能体在执行多步骤、高度耦合的文档操作时具备可靠的状态追踪与语义理解能力，已成为制约其实际应用的关键瓶颈。论文提出 DocOps——一个基于分层分类体系（hierarchical taxonomy）的可确定性验证评估框架，将真实世界中的文档操作行为解构为原子级维度与递进式工作流复杂度层级，从而实现对智能体行为的系统性量化评估。其解决方案的核心在于通过精细化的任务分解与失败模式分析，揭示现有主流闭源与开源模型在长期状态跟踪崩溃、浅层语义验证以及破坏性结构元数据编辑等三类关键缺陷上的普遍局限性，进而为未来设计具备鲁棒性与非破坏性特性的智能体提供理论依据与技术方向。
链接: https://arxiv.org/abs/2607.19865
作者: Jiazhen Jiang,Boxi Cao,Lingyong Yan,Yaojie Lu,Hongyu Lin,Shuaiqiang Wang,Dawei Yin,Xianpei Han,Le Sun
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:As autonomous agents rapidly evolve, their ability to reliably manipulate ubiquitous digital documents has become critical for enabling general-purpose AI assistants and automating complex workspace workflows. In this paper, we introduce DocOps, a deterministically verifiable evaluation framework underpinned by a hierarchical taxonomy that deconstructs document operations inspired by real-world practices into atomic dimensions and escalating workflow complexities. Based on DocOps, we systematically evaluate representative closed- and open-source models across various agentic harnesses, revealing that even the most advanced frontier configurations still exhibit profound limitations when handling highly coupled, long-range tasks. Furthermore, a fine-grained analysis of existing agents’ manipulation behaviors uncovers 3 key failure modes: long-term state tracking collapse, shallow semantic verification, and destructive editing of structural metadata. Ultimately, our work exposes the capability boundaries of agents in maintaining global document consistency, shedding light on the future design of robust, non-destructive agents for complex digital ecosystems.
[NLP-35] Overview of FinMMEval 2026 Task 1: Multilingual Financial Multiple-Choice Question Answering
【速读】： 该论文旨在解决多语言金融领域中多项选择题问答（multilingual financial multiple-choice question answering）的评估问题，重点关注系统在跨语言、跨脚本环境下对金融术语理解、数值解析及概念性金融推理能力的综合表现。其核心挑战在于如何在不同语言（英语、中文、阿拉伯语、印地语）中准确识别并选择正确答案，同时处理领域专有术语与复杂逻辑推理任务。解决方案的关键在于融合多种先进策略：包括基于检索增强（retrieval augmentation）的知识引入、直接对答案选项进行评分（direct answer-option scoring）、采用语言特定提示（language-specific prompting）以适配不同语言表达习惯、通过选择性自一致性（selective self-consistency）提升推理稳定性、引入置信度检查机制以过滤不确定预测，并设置基于大语言模型（LLM-based）的审核阶段以优化最终输出。这些技术协同作用，显著提升了系统在多语言金融问答任务中的准确率与鲁棒性。
链接: https://arxiv.org/abs/2607.19856
作者: Zhuohan Xie,Yuyang Dai,Rania Elbadry,Vanshikaa Jani,Georgi Georgiev,Dimitar Dimitrov,Fan Zhang,Xueqing Peng,Lingfei Qian,Jimin Huang,Jiahui Geng,Yankai Chen,Ye Yuan,Haolun Wu,Yuxia Wang,Ivan Koychev,Veselin Stoyanov,Mingzi Song,Yu Chen,Xue Liu,Preslav Nakov
机构: Mohamed bin Zayed University of Artificial Intelligence (穆罕默德·本·扎耶德人工智能大学); INSAIT (保加利亚索非亚大学“圣克莱门特·奥赫里德斯基”研究所); University of Arizona (亚利桑那大学); FMI (保加利亚索非亚大学“圣克莱门特·奥赫里德斯基”数学研究所); The Fin AI (Fin AI); The University of Tokyo (东京大学); Linköping University (林雪平大学); McGill University (麦吉尔大学); Mila, Quebec AI Institute (魁北克人工智能研究所); Meiji Gakuin University (明治学院大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computational Engineering, Finance, and Science (cs.CE)
备注: 9 pages. Task overview paper for CLEF 2026 Working Notes (CEUR Workshop Proceedings)
Abstract:FinMMEval 2026 Task 1 evaluates multilingual financial multiple-choice question answering in English, Chinese, Arabic, and Hindi. The task tests whether systems can select the correct answer to finance questions involving domain terminology, numerical interpretation, and conceptual financial reasoning across languages and scripts. The final-test set contains 800 questions, with 200 questions per language; gold answers were withheld during submission, and each language was ranked independently by accuracy. The final leaderboards contain 13 English, 11 Chinese, 11 Arabic, and 10 Hindi ranked submissions. Top accuracies range from 92.0% in Hindi to 97.5% in English and Arabic, with the same leading teams appearing near the top across all four languages. The documented systems used retrieval augmentation, direct answer-option scoring, language-specific prompting, selective self-consistency, confidence checks, and LLM-based review stages.
[NLP-36] mb-diversity: A Tool for Embedding-Based Measurement of Data Diversity
【速读】： 该论文旨在解决当前自然语言处理（NLP）领域中数据多样性（data diversity）评估缺乏统一、标准化工具的问题，尤其针对基于嵌入（embedding）的多样性度量方法尚不完善这一关键瓶颈。现有方法多集中于词汇多样性（lexical diversity）的测量，而对语义、风格、语言及说话者等多维度的嵌入空间多样性缺乏有效量化手段。其解决方案的核心是提出一个名为emb-diversity的综合性嵌入式多样性度量工具，该工具具备高度灵活性，可兼容任意嵌入模型与可嵌入数据，并支持多种多样性维度的量化分析，包括文本的风格、语义、语言类型及说话者多样性，从而为构建公平、鲁棒的NLP模型提供可信赖的数据质量评估框架。
链接: https://arxiv.org/abs/2607.19848
作者: Cantao Su,Menan Velayuthan,Esther Ploeger,Dong Nguyen,Anna Wegmann
机构: Utrecht University (乌得勒支大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:There is growing evidence that data diversity is crucial for developing fair and robust NLP models. However, current approaches to measure diversity remain inconsistent and fragmented: While there exist a number of tools for measuring the lexical diversity of texts, researchers lack standardized tools for quantifying diversity based on embeddings. Embedding-based diversity measures are highly flexible: They work with any embedding model and any data that can be embedded, and are thus applicable to many notions of diversity. With emb-diversity, we provide a comprehensive embedding-based diversity measurement tool, spanning a broad range of measures. We demonstrate its potential for several use cases: measuring the stylistic, semantic, language and speaker diversity of datasets. this https URL
[NLP-37] Auto-Fill: Learning to Predict Missing Values Accurately with Specialist Language Models VLDB2026
【速读】： 该论文旨在解决表格数据中缺失值预测这一基础性数据清洗问题，针对当前先进推理模型在处理表格缺失值时存在的高部署成本与过度自信导致的幻觉或假阳性预测缺陷。其核心挑战在于如何在保证高精度的前提下实现高效、可靠的预测。解决方案的关键在于提出一种名为Auto-Fill的新方法，通过后训练三个专精于不同能力的小型语言模型（SLMs），分别聚焦于世界知识、基于文本的推理和基于代码的推理，并设计了一种校准的集成机制：该机制能够动态选择置信度最高的专家模型进行预测，或在不确定时主动放弃预测，从而确保输出的高准确性。实验在涵盖2200张真实表格的11个基准测试上验证了该方法的优越性，相较于前沿模型（如o3-pro、Gemini 3 Pro、DeepSeek R1）实现了更高精度，同时计算成本低于其1%。研究结果表明，专业化分工与校准式弃权策略在表格数据缺失值预测任务中具有显著有效性。
链接: https://arxiv.org/abs/2607.19847
作者: Yurong Liu,Yeye He,Haoyu Dong,Junjie Xing,Shi Han,Dongmei Zhang,Surajit Chaudhuri
机构: 未知
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Databases (cs.DB)
备注: VLDB 2026
Abstract:Predicting missing cell values in tabular data is a fundamental problem in data cleaning. While state-of-the-art reasoning models show great promise in predicting missing values in tables, by reasoning holistically across rows and columns, they are costly to deploy at scale and tend to be overconfident, often generating hallucinated or false-positive predictions. In this paper, we observe that achieving high-precision missing-value prediction in tables requires a distinct combination of three capabilities: (1) world knowledge, (2) text-based reasoning, and (3) code-based reasoning. We systematically explore design choices for combining these capabilities, and propose an Auto-Fill approach that post-trains three specialist small language models (SLMs), each optimized for one capability. We develop a calibrated ensemble mechanism that either dynamically selects the most confident specialist or abstains, ensuring high accuracy. Extensive experiments on 11 benchmarks with 2200 real tables drawn from diverse domains show that Auto-Fill achieves superior accuracy compared to state-of-the-art reasoning models (e.g., o3-pro, Gemini 3 Pro, and DeepSeek R1), while operating at a fraction (less than 1%) of the cost of these frontier models. Our results highlight the effectiveness of specialization and calibrated abstention in the important domain of tabular data. Auto-Fill is publicly available at this https URL. Comments: VLDB 2026 Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Databases (cs.DB) Cite as: arXiv:2607.19847 [cs.LG] (or arXiv:2607.19847v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.19847 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-38] Sentence Splitter: Uncovering Latent Factual Structure for Self-Supervised Learning
【速读】： 该论文旨在解决自然语言中隐含的事实结构（latent factual structure）难以被自动识别与利用的问题，尤其在缺乏人工标注的情况下如何有效挖掘句子中描述性前缀（head）与事实补全部分（tail）之间的语义边界。其核心解决方案在于提出一种基于T5架构的自监督框架——Sentence Splitter，将句子分割问题建模为离散分段任务，通过概率序列生成的方式学习恢复正确的头尾结构，避免穷举所有可能的切分点。为实现无监督训练，研究首先将符号化的头尾对转化为自然语言模板作为监督信号，进而利用训练好的分割器对原始文本进行处理，提取出对齐的头尾对，并通过轻量级自举（bootstrapping）流程用于训练生成模型以提出更多合理的事实补全。该统一流程构建了一种可扩展且具备结构感知能力的自监督数据生成机制，成功实现了符号知识与自然语言之间的桥梁，实验表明该方法在知识图谱补全和常识问答等任务上均显著提升下游性能，验证了恢复隐含事实结构对于以知识为中心的自然语言处理的有效性。
链接: https://arxiv.org/abs/2607.19845
作者: Ahmad Pouramini,Mahsa Afsharizadeh
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:This paper introduces Sentence Splitter, a self-supervised framework built upon a T5-based encoder–decoder architecture for uncovering the latent factual structure of natural language sentences. The proposed method identifies the semantic boundary between a descriptive prefix (head) and its factual completion (tail) by formulating sentence splitting as a discrete segmentation problem, where a sentence of length N admits N possible split points but only one recovers the intended head–tail structure. Rather than explicitly searching over all candidate boundaries, the model learns to recover the factual completion through probabilistic sequence generation. To eliminate the need for manual annotation, symbolic head–tail pairs are first verbalized into natural-language templates that provide supervision for training the Sentence Splitter. The trained splitter is then applied to raw text to extract aligned prefix–tail pairs, which are subsequently used to train a generative model that proposes additional plausible completions through a lightweight bootstrapping process. This unified pipeline provides a scalable and structure-aware approach to constructing self-supervised training data while bridging symbolic knowledge and natural language. Experiments on both structured and naturally occurring text demonstrate that the proposed splitter generalizes beyond synthetic templates and that the resulting structure-aware supervision consistently improves downstream performance on knowledge graph completion and commonsense question answering, highlighting the effectiveness of recovering latent factual structure for knowledge-centric NLP.
[NLP-39] D2VBench: Benchmarking Large Language Models with Value Dilemmas in Daily Scenarios
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在实际应用中输出内容所蕴含的价值取向评估不足的问题，尤其针对日常场景中多价值冲突情境覆盖不全以及现有评估范式过于简单、难以有效衡量模型价值对齐水平的缺陷。其解决方案的关键在于提出D2VBench——一个包含10,000个真实日常价值困境实例的基准测试集，通过人与大语言模型的多阶段协作构建，并基于158个手工标注的细粒度价值概念进行语义锚定；同时设计了一种融合多项选择题与开放式问题的混合评估范式，以实现对模型在不同价值类别与维度上的对齐能力进行更真实、精细的评估。实验结果表明，D2VBench具有高可靠性与鲁棒性，能够有效揭示主流大语言模型在复杂价值情境下的对齐表现，为价值对齐研究提供了更具现实意义的评估工具。
链接: https://arxiv.org/abs/2607.19834
作者: Siyi Hao,Yidi Cao,Linhao Yu,Yuqi Ren,Deyi Xiong
机构: TJUNLP Lab, School of Computer Science and Technology, Tianjin University, China; The International Joint Institute of Tianjin University, Fuzhou, China
类目: Computation and Language (cs.CL)
备注: 22 pages,11 figures
Abstract:With the wide application of large language models (LLMs) in real-world scenarios, the value implication of their outputs is crucial. However, existing evaluation benchmarks suffer from insufficient coverage of value dilemmas in daily scenarios involving multiple value conflicts and simplistic evaluation formalisms that fail to assess LLMs’ value alignment. To address these issues, we propose D2VBench, a value alignment benchmark comprising 10,000 instances of real daily dilemma scenarios constructed through a multi-stage collaboration between LLMs and humans, grounded in 158 manually annotated fine-grained value concepts. For evaluation on the benchmark, we present a hybrid evaluation paradigm that integrates multiple-choice questions with open-ended questions. We conduct comprehensive evaluations on eight mainstream LLMs. Experimental results demonstrate that D2VBench exhibits high reliability and robustness, effectively reflecting the LLMs’ alignment across different value categories and dimensions, and providing a more realistic and fine-grained tool for research on value alignment. The dataset is available at this https URL.
[NLP-40] VizRAG : Enhancing Retrieval-Augmented Generation with Hypergraph Visualization
【速读】： 该论文旨在解决当前基于超图（hypergraph）的检索增强生成（RAG）系统在多模态场景下仍局限于文本单模态范式的问题，即未能有效利用现代多模态大语言模型（MLLMs）强大的视觉感知能力。其解决方案的关键在于首次将超图结构的视觉表征引入RAG流程，提出VizRAG——首个支持视觉化超图结构感知的RAG系统。通过融合超图的视觉表示，VizRAG实现了对复杂多体原子事实（n-ary atomic facts）的多模态理解与重构，显著提升了知识检索与生成性能，验证了超图可视化作为RAG系统新范式的可行性与优越性。
链接: https://arxiv.org/abs/2607.19830
作者: Yanbin Wei,Yang Chen,Renling Gan,Ziru Liu,Xinyu Fu,Chun Kang,Ning Lu,Rui Liu,Yu Zhang,James Kwok
机构: Southern University of Science and Technology (南方科技大学); Hong Kong University of Science and Technology (香港科技大学); Huawei Research (华为研究); Beihang University (北京航空航天大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Hypergraph-based RAG systems surpass traditional graph-based approaches by organizing complex n-ary atomic facts among entities, rather than relying solely on binary relationships. Despite the advancements in multimodal large language models (MLLMs) with enhanced visual capabilities, current hypergraph-based RAG frameworks predominantly restrict knowledge retrieval and reconstruction to a unimodal, text-centric paradigm. This limitation prevents them from fully leveraging the powerful visual perception capabilities of modern MLLMs. To address this gap, we systematically explore the integration of hypergraph awareness in RAG systems through visual cues. By incorporating visual representations of hypergraphs into the RAG pipeline, we introduce VizRAG, the first RAG system to support visual hypergraph structure awareness. Experimental results demonstrate that VizRAG significantly outperforms strong baselines, validating the promising potential of hypergraph visualization as a novel approach for RAG systems.
[NLP-41] Rewarding Better Thinking for LLM Preference Alignment
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）在偏好对齐（preference alignment）过程中，基于强化学习（Reinforcement Learning, RL）的现有方法依赖结果层面（outcome-level）代理奖励所导致的推理轨迹（reasoning trajectory）指导不足的问题。由于传统奖励仅评估最终响应质量，难以区分多个相似输出之间的细微差异，导致信用分配（credit assignment）粗粒度，轨迹级偏好信息被忽略。为此，论文提出一种面向过程的奖励机制——思维检查清单奖励（Thinking Checklist Reward, TCR），其核心在于将偏好对齐中的成对样本转化为特定于样本的思维检查清单，并据此评估生成的推理过程是否覆盖了隐含于偏好中的关键考量因素。为避免与结果层面监督产生冗余，TCR进一步引入指数移动平均（Exponential Moving Average, EMA）残差形式，以分离出超出可由结果奖励预测的“思维盈余”（thinking surplus），从而提供互补性过程监督。实验表明，在三个模型族的五种模型上，TCR均能一致提升偏好对齐性能，消融实验进一步验证了EMA残差结构和样本特异性检查清单监督的关键作用。
链接: https://arxiv.org/abs/2607.19824
作者: Xubo Liu,Wenya Guo,Ruxue Yan,Xinying Qian,Ying Zhang
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: under review
Abstract:LLM preference alignment aims to optimize models toward human preferences across diverse user instructions. Reinforcement learning has become a major post-training approach for this goal, but existing proxy rewards are often outcome-level, mainly evaluating the final response while providing limited guidance for the reasoning trajectory. This can make credit assignment coarse when multiple responses receive similar final scores, leaving trajectory-level preferences under-specified. To address this limitation, we propose Thinking Checklist Reward (TCR), a process-oriented reward for RL-based preference alignment. TCR converts preference pairs into sample-specific thinking checklists and uses them to evaluate whether the generated reasoning trace addresses the preference-implied considerations. To reduce overlap with outcome-level supervision, TCR further introduces an exponential moving average (EMA) residual formulation to isolate a complementary thinking surplus beyond what is predictable from the outcome reward. Experiments on five models from three model families show that TCR consistently improves alignment performance across diverse benchmarks, with ablations further validating the importance of EMA-based residual formulation and sample-specific checklist supervision.
[NLP-42] riAgent : Divergence-Aware Multi-Agent Committees for Cost-Efficient Financial Sentiment Analysis
【速读】： 该论文旨在解决基于大语言模型（LLM）的金融情感分析在生产环境中面临的结构性成本陷阱：尽管多数查询具有明显的可分类性，但当前方案仍依赖昂贵的云端推理引擎对所有请求进行统一处理，导致成本随用户数量线性增长。其核心解决方案是提出TriAgent——一个按上下文粒度分层的多智能体委员会架构，包含词级词典（VADER）、句级领域适配变换器（FinBERT）以及跨句推理器（Qwen2.5系列，0.5B–14B-4bit，辅以Mistral-7B和Phi-3.5-mini跨家族验证）。通过三向语义分歧指数（SDI）量化不同粒度间输出的分歧程度，实现查询的动态路由。关键发现为“批评者平台现象”：当将大语言模型重新任务为对小型代理输出的批判者时，F1值在1.5B至7B规模的Qwen模型上稳定于约0.87（置信区间重叠），显著优于同等规模三人投票机制（F1=0.66），后者性能下降源于粒度分层带来的多样性偏差。基于同一SDI信号衍生出三项重要推论：(i) 构建多语言句向量嵌入共享共识词典（SCD），可使英文缓存覆盖95%中文查询，且在F1=0.99下实现零边际成本的跨语言标准化；(ii) SDI可作为后验幻觉检测器，达到AUC=0.90；(iii) 单阶段SDI策略在20只标的回测中取得最优风险调整收益（夏普比率=3.50），显著优于始终使用FinBERT（1.36）或始终使用LLM（0.11）的基准。在千万级用户规模下，相较于GPT-4o-mini基线，TriAgent年节省达930万美元。相关代码、词典与SCD已开源。
链接: https://arxiv.org/abs/2607.19794
作者: Isabel Xu(The Overlake School)Cynthia Xu(The Overlake School)Rachel Ren(Edwards Vacuum Inc.)Cong Guo(The University of Memphis)Jiacheng Ding(The University of Memphis)
机构: The University of Memphis
类目: Computation and Language (cs.CL); Computational Engineering, Finance, and Science (cs.CE); Databases (cs.DB); Machine Learning (cs.LG)
备注:
Abstract:Production LLM-based financial sentiment analysis faces a structural cost trap: most queries are trivially classifiable, yet expensive cloud reasoners process them all, and the bill scales linearly with user count. We present TriAgent, a multi-agent committee stratified by contextual granularity – a word-level lexicon (VADER), a sentence-level domain transformer (FinBERT), and a cross-sentence reasoner (Qwen2.5, 0.5B-14B-4bit, with Mistral-7B and Phi-3.5-mini cross-family checks). A three-way Semantic Divergence Index (SDI) measures pairwise disagreement across granularities and routes each query accordingly. Our central finding is the critic plateau: when the LLM is re-tasked as a critic over the smaller agents’ outputs, F1 plateaus at ~0.87 across 1.5B-7B Qwen (bootstrap 95% CIs overlap), while a same-size 3-persona vote drops to F1=0.66, which is driven by granularity-stratified diversity. Three corollaries follow from the same SDI signal: (i) a Shared Consensus Dictionary on multilingual sentence-BERT answers 95% of Chinese queries from an English cache at F1=0.99 – cross-border canonicalization at zero marginal cost; (ii) SDI doubles as a post-hoc LLM-hallucination detector at AUC=0.90; (iii) the SDI single-stage strategy attains the best risk-adjusted return (Sharpe=3.50) on a 20-ticker back-test, dominating both always-FinBERT (1.36) and always-LLM (0.11). At 10M-user scale, TriAgent saves 9.3M/year vs. a GPT-4o-mini baseline. Code, lexicons, and the SCD are released.
[NLP-43] Learning the Arabic Dialect Continuum as a Continuous Space: A Regression Approach to Speaker Origin Prediction
【速读】： 该论文旨在解决阿拉伯语方言地理定位（Arabic dialect geolocation）中传统离散类别划分所带来的局限性，即无法充分捕捉方言连续变化的地理分布特征。其核心问题是：如何在不依赖预设离散方言区划的前提下，实现对说话者来源地的高精度连续坐标预测。解决方案的关键在于提出一种基于回归的连续地理空间建模方法，通过层次化神经架构融合帧级XLS-R-300M与Whisper-large-v3编码器表征以及音系描述符，利用Transformer编码器和可学习注意力池化查询生成上下文表示；同时采用球面大地测量损失（spherical geodesic loss），直接优化地球表面的大圆距离，避免了平面坐标回归带来的投影畸变。实验表明，在无数据泄露的5折分组交叉验证下，模型达到481.2 km的综合中位定位误差，并通过置换曼德尔检验（permutation Mantel test）为阿拉伯语方言连续体假说提供了量化支持。进一步引入城市屏蔽协议（city-masking protocol）以评估零样本泛化能力，结果显示平均误差上升至1173.3 km，相较已见城市出现1.32倍性能退化，揭示出当前模型在跨区域泛化方面仍存在显著提升空间。研究确立了连续地理建模作为阿拉伯语方言定位的理论框架，既验证了其有效性，也明确了未来改进方向。
链接: https://arxiv.org/abs/2607.19751
作者: Mohamed Aziz Khadraoui,Adel Ammar,Bilel Benjdira,Zahid Khan,Skander Turki,Wadii Boulila
机构: Higher School of Communication of Tunis (SUP’COM), Tunisia; Robotics and Internet of Things Laboratory, Prince Sultan University, Riyadh, Saudi Arabia
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE)
备注: Under review
Abstract:We present a regression-based approach to Arabic dialect geolocation that models dialectal variation as a continuous geographic space rather than discrete categories. Speaker origin is predicted as continuous latitude-longitude coordinates using a hierarchical neural architecture that fuses frame-level XLS-R-300M and Whisper-large-v3 encoder representations with phonotactic descriptors through a Transformer encoder and a learnable attention-pooled query. A spherical geodesic loss directly optimizes great-circle distance on Earth’s surface, avoiding distortions inherent to planar coordinate regression. Under a leakage-free 5-fold GroupKFold protocol grouped by source recording, our model attains a pooled median localization error of 481.2 km. Auxiliary country and city heads reach 64.5% and 45.2% accuracy, respectively. A permutation Mantel test on the learned latent space provides quantitative support for the Arabic dialect continuum hypothesis. To probe true generalization, we further introduce a city-masking protocol in which two cities per fold are removed from training but retained in validation. Under this zero-shot regime, the mean error rises to 1173.3 km, a 1.32x degradation relative to seen cities. Our findings establish continuous geographic modeling as a principled framework for Arabic dialect geolocation and quantify both its strengths and the substantial headroom that remains.
[NLP-44] Beyond Relevance-Centric Retrieval: Rubric-Oriented Document Set Selection and Ranking
【速读】： 该论文旨在解决当前检索系统中文档集合质量评估的局限性问题，即现有评估体系仅对单个文档进行独立打分并基于nDCG进行聚合，忽视了文档间的相互作用（如冗余、冲突与互补性），无法有效回答“为何一个文档集合优于另一个”这一核心问题。其解决方案的关键在于提出一个完整的“评估-诊断-优化”闭环框架，并设计了SetwiseEvalKit——一个涵盖短文本与长文本场景、包含三个层级九个维度的文档集合评估基准，共包含约2.8万条高质量评估标准。通过系统评估12种重排序方法，发现最优方法覆盖率不足45%，且跨文档协同能力普遍薄弱。在此基础上，提出Rubric4Setwise：一种无需训练的文档集合选择方法，将基于规则的评估标准转化为可操作的集合筛选信号，显著提升下游生成性能，在减少文档数量和搜索轮次的同时，实现了在两种场景下的最佳表现，首次验证了从评估到优化闭环的有效性。
链接: https://arxiv.org/abs/2607.19747
作者: Kailin Jiang,Lei Liu,Jian Xi,Hui Xu,Junlin Liu,Baochen Fu,Shaoqing Ren,Bin Li,Vichwang,Yu Lu,Haibo Shi
机构: University of Science and Technology of China (中国科学技术大学); Tencent Yuanbao Team (腾讯元宝团队); University of Chinese Academy of Sciences (中国科学院大学); Shandong University (山东大学)
类目: Computation and Language (cs.CL)
备注: Project Page: this https URL
Abstract:As large language models and AI agents become the primary consumers of search results, document set quality determines the upper bound of downstream generation. Yet existing evaluation systems remain confined to scoring documents independently and aggregating via nDCG, ignoring inter-document interactions (redundancy, conflict, complementarity) and unable to answer what makes one document set better than another. To address these issues, we propose a complete evaluate-diagnose-optimize framework. We design SetwiseEvalKit, a three-level, nine-dimension document set evaluation benchmark covering both short-form and long-form scenarios, comprising approximately 28K high-quality evaluation rubrics. We systematically evaluate 12 rerankers: even the best method achieves no more than 45% coverage, cross-document coordination dimensions are universally weak, and no single method maintains top performance across both settings. Building on this, we propose Rubric4Setwise, a training-free method that converts rubric-based evaluation criteria into document set selection signals, achieving the best downstream generation performance with fewer documents and search rounds. It is the only method that maintains state-of-the-art results across both scenarios, validating the effectiveness of closing the loop from evaluation to optimization.
[NLP-45] Lightweight Person-Place Relation Extraction from Historical Newspapers with Dependency Graphs and Proximity Features
【速读】： 该论文旨在解决多语言历史报纸中人物-地点关系抽取（person-place relation extraction）的高效建模问题，尤其关注在不依赖预训练语言模型的前提下，如何构建轻量级且可解释的关系分类系统。其核心挑战在于历史档案数据规模庞大、标注成本高，而传统基于深度学习的复杂模型难以在资源受限场景下实现高效部署。解决方案的关键在于：构建基于句法依存解析的文档级图结构，仅使用字符距离、词性标注等显式工程特征提取实体对的上下文信息，并采用参数量极小的集成学习模型（如scikit-learn集成或紧凑型图注意力网络），确保所有实验运行的参数量均低于847K。实验结果表明，最小字符距离已能捕获绝大部分分类信号，额外引入特征反而可能导致性能下降；同时，研究发现文档分组交叉验证是避免因实体提及跨文档重复而导致的数据泄露问题的关键，否则在配对级别划分数据会人为抬高评估分数25–37个百分点。
链接: https://arxiv.org/abs/2607.19718
作者: Mlen-Too Wesley
机构: Georgia Institute of Technology(佐治亚理工学院)
类目: Computation and Language (cs.CL)
备注: 19 pages, 4 figures. Accepted at CLEF 2026 HIPE Shared Task. To appear in CEUR Workshop Proceedings ( this http URL )
Abstract:The HIPE-2026 shared task introduces person-place relation extraction from multilingual historical newspapers as a new evaluation track, classifying the at and isAt relations between pre-annotated person and location mentions in English, French, and German. Motivated by the cost of processing historical archives at scale, our team (DS@GT HIPE, team 2 in the official results) investigates how far a lightweight, interpretable system can go without any pretrained language model at the relation classification stage. Our approach builds a document-level graph from dependency parses, extracts proximity-based and part-of-speech features for each entity pair, and classifies them with small scikit-learn ensembles or compact Graph Attention Networks, keeping every submitted run under 847K parameters. On the official evaluation (Test A, the newspaper test set), our best run reached a macro recall of 0.5142, ranking 3rd on the Efficiency profile while placing mid-table on Accuracy among the 17 participating teams. Two findings stand out. First, minimum character distance alone captures most of the classification signal; adding further engineered features yields inconsistent gains and sometimes degrades performance, echoing prior evidence that argument distance dominates relation extraction. Second, document-grouped cross-validation is essential on this corpus: pair-level splits inflate scores by 25-37 percentage points because entity mentions recur across documents, a data-leakage effect that grouped cross-validation removes.
[NLP-46] SLPO: Scaling Latent Reasoning via a Surrogate Policy
【速读】： 该论文旨在解决隐式思维（latent reasoning）模型在测试时扩展（test-time scaling）能力不足的问题，即如何在保持计算效率的同时实现与显式思维链（explicit Chain-of-Thought, CoT）相当甚至更优的推理性能。当前隐式推理模型受限于其连续向量表示的特性，缺乏可解析的逐步似然性（per-step likelihood）和适应固定思考预算的自适应终止机制，导致无法有效利用基于结果奖励的强化学习（outcome-reward RL），从而难以实现测试时的动态扩展。本文提出代理隐式策略优化（Surrogate Latent Policy Optimization, SLPO），其核心在于：一是构建一个基于经验的代理策略密度（empirical surrogate policy density），用于对潜在轨迹进行层级信用分配；二是引入一个由正确性监督的终止头（correctness-supervised stopping head），通过结果奖励优化将其转化为可变推理时长的策略。该方法使自回归隐式推理模型能够实现基于结果奖励的测试时扩展，在连续与软思考（soft thinking）设置下均显著提升Pass@k指标，并能为更复杂的任务分配更长的推理时间，从而在保证效率的同时提高确定性准确率。
链接: https://arxiv.org/abs/2607.19691
作者: Runyang You,Zhiyuan Liu,Yongqi Li,Wenjie Li
机构: The Hong Kong Polytechnic University (香港理工大学); Sichuan University (四川大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Reinforcement learning with verifiable rewards has become the predominant recipe for eliciting test-time scaling in explicit Chain-of-Thought reasoners. Yet this scaling path remains computationally costly, since every intermediate step must be decoded as a language token. Latent reasoning instead carries intermediate computation as continuous vectors and already matches or surpasses explicit CoT at far shorter horizons. Despite this promise, latent reasoners remain largely imitation-bound, while explicit CoT has already moved past imitation via outcome-reward RL. Latent trajectories lack a tractable per-step likelihood and an adaptive stopping interface under fixed thinking budgets, so outcome rewards cannot elicit latent test-time scaling. We introduce Surrogate Latent Policy Optimization (SLPO) to bring outcome-reward RL to autoregressive latent reasoners: an empirical surrogate policy density over latent transitions for trajectory-level credit assignment, and a correctness-supervised stopping head that outcome-reward optimization refines into a variable-horizon policy. Across continuous and soft thinking settings, SLPO improves Pass@ k under parallel sampling and allocates longer latent computation to harder instances with higher deterministic accuracy.
[NLP-47] Multi-Mask Diffusion Language Models for Few-Step Generation
【速读】： 该论文旨在解决生成式语言模型中少步生成（few-step generation）质量不高的问题，尤其针对掩码扩散模型（Masked Diffusion Models, MDMs）在前向过程中所有轨迹坍缩至单一全掩码状态，导致缺乏终端熵以支持一致性风格的少步生成这一核心瓶颈。现有基于均匀状态扩散的少步方法虽避免了该退化问题，但引入了难以区分干净词元与噪声的问题，从而损害建模精度与训练效率。为此，本文提出多掩码扩散模型（MultiMDM），其关键在于保留掩码结构以支持少步生成：在前向过程中，每个干净词元首先被推向一个指定的掩码状态，随后逐步在掩码集合中混合；由此，在反向过程中具备“草稿”能力，可先预测特定掩码再精细重构为干净词元。研究进一步推导出闭式证据下界优化（ELBO）目标，支持从预训练的MDM持续训练，并设计了一种纯离散状态的一致性蒸馏方案，通过共享Gumbel耦合机制降低路径熵。实验表明，MultiMDM为有原则性的少步生成提供了高效且稳定的基线。
链接: https://arxiv.org/abs/2607.19686
作者: Sijin Chen,Yinuo Ren,Heyang Zhao,Ziheng Cheng,Quanquan Gu,Lexing Ying
机构: ByteDance(字节跳动); Princeton University (普林斯顿大学); Stanford University (斯坦福大学)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 38 pages; Accepted at COLM 2026
Abstract:Masked diffusion models (MDMs) are a promising family of language generators, but achieving high-quality few-step generation remains challenging. In MDMs, all forward trajectories collapse to a single fully masked state, leaving no terminal entropy for consistency-style few-step generation. While recent few-step alternatives based on uniform-state diffusion avoid this degeneracy, it becomes harder to distinguish clean tokens from noise than MDMs, which usually harms modeling quality and training efficiency. In this work, we propose a multi-mask diffusion model (MultiMDM) that preserves the masking structure towards few-step generation. In the forward process, each clean token is first pushed towards a designated mask and then gradually mixes over the mask set. As a result, the backward process has a drafting capability by predicting a designated mask before refining to a clean token. We derive a closed-form ELBO training objective for MultiMDM that supports continual training from pretrained MDMs. In addition, we formulate a purely discrete-state consistency distillation scheme, with a shared-Gumbel coupling to reduce pathwise entropy. Experiments on pretraining and distillation show that MultiMDM provides an effective foundation for principled few-step generation.
[NLP-48] Reference-Free Evaluation of Reasoning in Open-Ended Question Answering
【速读】： 该论文旨在解决生成式人工智能（Generative AI）在高风险领域中生成的多步推理答案难以验证的问题，尤其当这些答案表现出高度流畅性但缺乏可靠依据时。其核心挑战在于传统评估方法依赖最终答案或由大语言模型（LLM）充当裁判，难以捕捉推理链中各环节之间的逻辑关联与证据支撑关系。本文提出的解决方案关键在于构建一种基于推理的、无需参考文本的审计框架：将生成的推理链条分解为若干片段，利用自然语言蕴含（Natural Language Inference, NLI）识别局部前提-目标之间的语义关系，并将其组织成超图结构；随后通过确定性的后向AND-OR搜索算法，为每个推理片段分配审计标签，以判断其在生成响应中的可支撑程度。该方法强调对推理过程中推导关系的组合性分析，而非仅关注最终答案或依赖其他大模型作为评判者。实验在数学演绎推理（Hard2Verify）和开放域医学推理（UroReason，一个由医师标注的真实临床案例推理数据集）两个场景中验证了该框架的有效性，结果表明，相较于直接使用LLM作为裁判的基线方法，本框架提供了更可靠的无参考评估信号。尤其在临床场景中，当前最先进的LLM裁判常因无法识别推理缺陷而过度接受看似合理实则弱支撑的回答。研究结论指出，问答评估必须考虑推理链中推断关系的累积与整合机制，才能实现对生成内容可信度的准确判断。相关数据集UroReason将通过API公开，代码也将开源。
链接: https://arxiv.org/abs/2607.19678
作者: Guneet Singh Kohli,Yuxiang Zhou,Michael Sejr Schlichtkrull,Gregory E Dean,Maria Liakata
机构: Queen Mary University of London; Temple University
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:AI-generated answers in high-stakes domains are often fluent but difficult to verify, especially when they contain multi-step reasoning rather than a single final answer. We propose a reasoning-based, reference-free framework for auditing LLM-generated outputs. The method decomposes a generated reasoning trace into segments, labels local premise-target relations using Natural Language Inference (NLI), and organizes these relations into a hypergraph. A deterministic backward AND-OR search then assigns segment-level audit labels that indicate how each segment is grounded within the generated response. We evaluate the framework in two settings: deductive mathematical reasoning with Hard2Verify, and open-ended medical reasoning with UroReason, a new physician-annotated benchmark of LLM reasoning traces from real clinical cases. Across these settings, our NLI-hypergraph audit provides a more reliable reference-free evaluation signal than direct LLM-as-judge baselines. In the clinical setting, state-of-the-art LLM judges often fail to identify problematic reasoning segments, over-accepting fluent but weakly grounded responses. Our results show that QA evaluation should account for how inferential relations compose across a reasoning trace, rather than relying only on final answers or LLMs as verifiers. UroReason will be made available through an API, and our code will be released as open source.
[NLP-49] ask Competence Is Not Instruction Following: Evaluating Instruction-Conflicting Behavior in Small Language Models
【速读】： 该论文旨在解决小规模语言模型在指令遵循（instruction following）方面的能力瓶颈问题，特别是当用户指令与模型固有任务行为冲突时，其是否仍能有效响应指令。研究聚焦于三种典型任务：多选题问答（MCQA）、情感分类和数学问题求解，通过设计标准指令与具有冲突性的非标准指令（如选择错误选项、输出相反情感或返回答案的两倍值），系统评估不同规模的指令微调版Qwen模型在对抗性指令下的表现。关键发现在于，尽管模型规模增大通常提升任务准确率与指令遵循能力，但小模型在多数情况下仍会忽略非标准指令，表现出对指令的“抵抗”；而大模型则在标准任务性能与指令遵循之间呈现出显著差距。这表明任务能力（task competence）与指令遵循能力是两种独立的认知属性，仅报告标准准确率会掩盖指令遵循失败的严重性。因此，该研究提出的核心解决方案在于引入“指令遵循失败率”（Instruction-Following Failure Rate, IFFR）作为补充评估指标，以更全面揭示模型在真实复杂指令场景中的行为可靠性，强调在评估模型时必须区分任务性能与指令控制能力。
链接: https://arxiv.org/abs/2607.19608
作者: Mahdiyeh Farajidizaji,Vatsal Raina
机构: Khajeh Nasir Toosi University of Technology (哈桑·纳西尔·托西科技大学); Apta AI, Spark AI Research (Apta AI，Spark AI 研究院)
类目: Computation and Language (cs.CL)
备注: 12 pages, 4 figures, 8 tables
Abstract:Instruction tuning is meant to make language models follow user requests, yet it is unclear whether small models comply when an instruction conflicts with their usual task behavior. We study this across three tasks - multiple-choice question answering (MCQA), sentiment classification, and mathematical question answering - by pairing a standard instruction with a conflicting non-standard one (select an incorrect option, output the opposite sentiment, or return twice the answer). This cross-task design allows us to test whether resistance to conflicting instructions is tied to specific task characteristics or reflects a broader behavioral tendency. As all predictions are scored against the original ground truth, a model that ignores the non-standard instruction still appears accurate. Using standard accuracy, non-standard accuracy, and an Instruction-Following Failure Rate (IFFR), we evaluate instruction-tuned Qwen models across sizes. Both standard accuracy and instruction following generally improve with scale, although the pattern is not consistent across all tasks and datasets. Small models stay competent yet routinely ignore the non-standard instruction, while larger models show a clear gap between the two settings. These findings suggest that gains in task capability do not automatically provide reliable control over model behavior. Task competence and instruction following are therefore distinct abilities, and reporting only standard accuracy hides instruction-following failures.
[NLP-50] Scaling Laws for Hypernetwork-Based Knowledge Injection in Large Language Models
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在训练阶段可靠且大规模注入事实知识的挑战。现有方法在实现高效、可扩展的知识注入方面仍存在局限，尤其缺乏对知识注入机制在不同规模下的系统性研究。本文提出将超网络（Hypernetworks）应用于训练时知识注入（train-time knowledge injection），其核心创新在于设计了一种解耦机制：将超网络的知识注入能力与目标模型的通用表征能力分离，从而首次实现了对超网络架构缩放规律的严谨量化分析。关键解决方案是构建一个包含数千万个多跳问答样本的大型数据集MegaWikiQA（基于Wikidata5M构建），并系统评估了超网络深度、宽度及目标模型规模对损失、推理准确率和分布外（Out-of-Distribution, OOD）泛化性能的影响。实验结果表明：（i）超网络的知识注入表现出广泛适用的幂律缩放特性；（ii）随着规模增加，超网络在OOD任务上仍能保持可靠的泛化能力，其缩放指数优于典型的LoRA微调和全量微调方法。这些发现确立了超网络作为训练时适应的可扩展、可预测的底层架构，并首次提供了实证支持的缩放定律，为大模型中基于事实推理的训练时知识注入提供了理论指导与实践路径。
链接: https://arxiv.org/abs/2607.19604
作者: Nischay Dhankhar,Dos Baha,Abulhair Saparov
机构: Nace AI; Purdue University
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: Preprint, 22 pages, 14 figures. Dataset collection available at this https URL
Abstract:Injecting factual knowledge into large language models (LLMs) reliably and at scale remains an open challenge. Hypernetworks provide a promising solution to large-scale knowledge injection. Although hypernetworks are typically applied for test-time adaptation, we explore their use in train-time knowledge injection, where, given a large corpus of facts, we train a hypernetwork to generate a fixed LoRA adapter that, when inserted into the target model, enable the model to answer questions about those facts. In this work, we investigate whether hypernetworks can be used to perform train-time knowledge injection and how this ability varies with scale. The scaling behavior of hypernetworks remains largely unstudied. Our design decouples the hypernetwork’s injection capacity from the target model’s general capability, enabling, for the first time, a rigorous study of scaling laws for hypernetwork architectures. We characterize how loss, reasoning accuracy, and out-of-distribution (OOD) generalization vary with hypernetwork depth, width, and target network size. We construct a large-scale dataset, called MegaWikiQA, containing tens of millions of multi-hop question-answer examples across 39 domains constructed from examples in Wikidata5M. Our results reveal: (i) hypernetwork-based injection exhibits broadly predictive power law scaling along all architecture axes; and (ii) hypernetworks are capable of reliable OOD generalization at increasing scales, suggesting that hypernetwork provides a promising alternative to other train-time adaptation methods such as LoRA finetuning and full fine-tuning, exhibiting steeper scaling exponents in all OOD evaluations. Together, these results establish hypernetworks as a principled and scalable substrate for train-time adaptation, and provide the first empirically grounded scaling laws to guide hypernetworks for factual reasoning in large language models.
[NLP-51] win Agent : Context Residual Compression for Privilege Separated Agents
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）智能体在面对不可信上下文时易受提示注入攻击（prompt injection attacks）的安全风险问题。此类攻击可通过操纵下游推理与工具调用，导致智能体执行恶意行为。现有安全设计方法虽通过特权执行与信息流的严格控制实现安全隔离，但往往牺牲任务实用性并需大量针对特定任务的工程化改造。为此，本文提出Twin Agent，一种受残差编码（residual coding）启发的通用权限分离设计范式。其核心在于构建两个近乎对称的智能体：探索代理（Explore Agent）负责分析不可信输入，而安全代理（Safe Agent）则执行特权操作；探索代理基于安全代理的当前状态进行条件生成，并仅向安全代理传递紧凑的动作提示（hints），从而显著降低所需传递的信息量。这一机制在保障任务实用性的同时增强了安全性，实现了更优的安全-效用权衡。实验结果表明，在SWE-bench Lite（长周期软件工程任务）和AgentDojo、DecodingTrust-Agent（异构多工具交互任务）等基准上，Twin Agent在保持高任务性能的同时有效抵御提示注入攻击，显著优于未防御的基线及传统权限分离方法。
链接: https://arxiv.org/abs/2607.19595
作者: Zhanhao Hu,Dennis Jacob,Xiao Huang,Zhaorun Chen,Bo Li,David Wagner
机构: University of California, Berkeley (加州大学伯克利分校); University of Chicago (芝加哥大学); University of Illinois, Urbana-Champaign (伊利诺伊大学厄本那-香槟分校)
类目: Cryptography and Security (cs.CR); Computation and Language (cs.CL)
备注:
Abstract:Large language model (LLM) agents are vulnerable to security risks, such as prompt injection attacks from untrusted context that manipulate downstream reasoning and tool use. Existing secure-by-design approaches mitigate this risk by separating untrusted observations from privileged execution and careful control of information flow, but often degrade utility and require extensive task-specific engineering. We thus propose Twin Agent, a general privilege separation design pattern inspired by residual coding in the agent context. Twin Agent consists of two nearly symmetric agents: an Explore Agent that inspects untrusted information and a Safe Agent that executes privileged actions. The Explore Agent is conditioned on the Safe Agent’s current context and communicates only compact hints to the Safe Agent about the next action to take. This design reduces the information needed to preserve task utility and thus achieves a better security–utility tradeoff, which we empirically verify by measuring how utility and attack success change as the length of hints varies. We evaluate Twin Agent on long-horizon software engineering tasks with SWE-bench Lite and on heterogeneous multi-tool interaction tasks with AgentDojo and DecodingTrust-Agent. Across both benchmarks, Twin Agent preserves high task utility while preventing prompt injection attacks, outperforming both undefended agents and privilege separation baselines.
[NLP-52] On the Computational Complexity of Structural Generalization
【速读】： 该论文旨在解决结构化泛化（structural generalization）在理论层面缺乏明确定义的问题，尤其针对其核心要素——组合结构（compositional structure）与无限泛化能力（unbounded generalization）——提出一个形式化的数学定义。其关键在于将组合规则分解为句法面（F_γ）和语义面（G_γ）两个投影，并基于蒙塔古语义学（Montagovian）框架证明：对语义面的树评估本质上是BFVP问题，属于\mathrmNC^1-完全类。而纯Transformer模型的学习能力上限仅为\mathrmTC^0，若在标准假设\mathrmTC^0 ≠ \mathrmNC^1下成立，则其无法学习此类结构化泛化。因此，论文的核心论点是：纯Transformer因无法自主从有限数据中涌现出结构化泛化能力，故其性能受限；而当前表现优异的神经符号系统之所以成功，恰恰在于人为注入了语义面（G_γ），从而规避了真正的计算难题。论文进一步指出，现有基准测试无法区分“学习所得”与“预先给定”的能力，因此必须明确这一根本性界限。
链接: https://arxiv.org/abs/2607.19573
作者: Zichao Wei
机构: 未知
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Structural generalization has been measured repeatedly by several benchmarks, yet it has never been formally defined. We give a definition that translates the two premises (compositional structure and unbounded generalization) into mathematical language. The definition itself is neutral: a compiler that hard-codes the rules satisfies it just as well. But structural generalization becomes a scientific question only insofar as the capacity can autonomously emerge from finite data. This question pits the computational lower bound \mathrmNC^1 against the learnable ceiling \mathrmTC^0 of pure Transformers. Under a Montagovian instantiation, each compositional rule splits into two projections: a syntactic face ( F_\gamma ) and a semantic face ( G_\gamma ). Tree evaluation on the G_\gamma side is an instantiation of BFVP, which is \mathrmNC^1 -complete (Buss, 1987). A pure Transformer must learn both faces at once, but Kraus et al. (2026) prove that its learnable class \subseteq \mathrmTC^0 . Under the standard assumption \mathrmTC^0 \neq \mathrmNC^1 , a pure Transformer cannot learn structural generalization. Neuro-symbolic systems achieve the best benchmark scores precisely because they inject G_\gamma , sidestepping the genuinely hard half. Benchmark scores cannot distinguish “learned” from “given.” This is what this paper sets out to make clear.
[NLP-53] When Reasoning Narrows the Move: Diversity Collapse in LLM Game Play
【速读】： 该论文旨在解决监督微调（Supervised Fine-Tuning, SFT）在序列决策任务中对行为多样性（behavioral diversity）影响不足的问题，特别是在确定性棋类游戏（如井字棋变体）这一可控环境中，研究最优动作可精确计算且多样性可直接度量的场景下，揭示了现有SFT方法在提升动作准确率的同时，往往导致行为多样性过早坍缩（premature diversity collapse）的现象。其解决方案的关键在于引入**动作增强（action augmentation）**策略，即在每个状态训练时使用所有最优动作而非单一示范动作进行学习，从而部分缓解因窄支持模仿（narrow-support imitation）导致的策略坍缩问题。研究结果表明，在SFT过程中保持动作支持范围（action support）对于维持生成模型的探索性行为至关重要。
链接: https://arxiv.org/abs/2607.19523
作者: Junyi Sha,Renfei Tan,David Simchi-Levi
机构: Massachusetts Institute of Technology (麻省理工学院)
类目: Computation and Language (cs.CL)
备注:
Abstract:Supervised fine-tuning (SFT) is widely used to adapt large language models to downstream tasks, but its effect on behavioral diversity in sequential decision-making remains under-explored. We study this question in a controlled suite of deterministic board games based on tic-tac-toe variants, where optimal actions are exactly computable and diversity can be measured directly. Across state-level evaluation, arena gameplay, and training trajectories, we find that reasoning-mode generation frequently suppresses action diversity without uniformly improving action accuracy. Furthermore, standard SFT improves accuracy but often induces premature diversity collapse, which exceeds what is minimally required by the accuracy-diversity tradeoff. We then show that action augmentation, which trains on all optimal actions per state rather than a single demonstrated action, would partially mitigates this effect. Our results identify narrow-support imitation as a source of policy collapse in LLM decision-making and suggest that preserving action support during SFT is important for maintaining exploratory behavior.
[NLP-54] BaseRT: Advancing Best-in-Class LLM Inference with Apple M5 Neural Accelerators
【速读】： 该论文旨在解决在Apple Silicon硬件上实现大语言模型（Large Language Models, LLMs）高效推理的性能瓶颈问题，尤其针对M5芯片中新增的神经网络加速单元（Neural Accelerator）尚未被充分挖掘利用的现状。其核心挑战在于如何有效发挥M5架构中每个核心配备专用矩阵计算单元（on-die matrix units）的能力，以突破现有框架在提示词处理（prompt processing）阶段的吞吐量限制。解决方案的关键在于设计并实现一个无框架依赖的Metal 4原生推理运行时BaseRT，通过手写优化的Metal 4张量核函数（包括密集层与专家混合模型的GEMM及flash-attention预填充核），将计算密集型的矩阵乘法操作路由至M5的神经网络加速单元，同时保留内存密集型的解码路径由原有专用内核处理。这一分治策略显著提升了整体推理效率，在Apple M5 Pro上对多个从子10亿到350亿参数的模型（涵盖Qwen3、Llama 3.2、Gemma 4等系列）实现了最高达6.4倍于基准方案（this http URL）和3.9倍于MLX的提示词处理吞吐量，尤其是在专家混合模型中表现尤为突出，同时在解码阶段也保持了1.75倍和1.33倍的优势。实验结果表明，M5的张量核心是决定Apple Silicon上提示处理性能上限的关键因素，为本地化大模型推理树立了新的性能标杆。
链接: https://arxiv.org/abs/2607.19438
作者: Fabian Waschkowski,Prabod Rathnayaka,Lukas Wesemann
机构: Base Compute, Melbourne, Australia
类目: Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Distributed, Parallel, and Cluster Computing (cs.DC); Machine Learning (cs.LG); Performance (cs.PF)
备注:
Abstract:Apple’s M5 generation introduces a redesigned GPU architecture in which every core carries a dedicated Neural Accelerator: on-die matrix units exposed through the Metal~4 tensor API. We show that BaseRT, our native Metal inference runtime for large language models on Apple Silicon, exploits these units to push inference throughput on Apple hardware substantially beyond both this http URL and MLX. Building on BaseRT’s framework-free design, we add a family of hand-written Metal~4 tensor-core kernels (including dense and mixture-of-experts GEMM and flash-attention prefill kernels) that route the compute-bound matrix multiplications of inference through the M5 Neural Accelerators while leaving the memory-bound decode path on our existing specialised kernels. On an Apple M5 Pro, across fifteen model configurations spanning the Qwen3, Qwen3.5/3.6, Llama~3.2, and Gemma~4 families from sub-1B to 35B parameters, BaseRT delivers up to 6.4\times higher prompt-processing throughput than this http URL and 3.9\times higher than MLX, with the largest margins on the mixture-of-experts models where matrix multiplication dominates, while maintaining its lead on decode of up to 1.75\times over this http URL and 1.33\times over MLX. These results establish a new performance ceiling for on-device LLM inference and show that the M5’s tensor cores are the decisive lever for prompt processing on Apple Silicon. BaseRT is publicly available at this https URL.
[NLP-55] JailMeter: An Evidence-Based Evaluation Framework for Jailbreak Attacks on Large Language Models
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在对抗越狱攻击（jailbreak attacks）评估中因评价标准与方法不一致而导致的攻击成功率估计不可靠的问题。其核心解决方案是提出JailMeter——一个基于证据的评估框架，受信息瓶颈理论（Information Bottleneck theory）启发，采用双反馈优化机制，在保留原始恶意问题相关语义内容的同时，有效过滤模型响应中的噪声信息，从而生成简洁且可信的证据。该框架要求攻击成功必须满足两个条件：模型响应准确捕捉恶意意图并提供完整答案，方可认定为实质性绕过模型安全对齐机制。在包含330个经人工标注、未被拒绝的越狱实例的JailMeter-Eva基准测试上，JailMeter实现了97.27%的准确率，显著优于现有方法。为进一步支持大规模评估，研究进一步将JailMeter蒸馏为轻量级小语言模型JailMeter\textsubscriptSLM，保持了相近的评估可靠性，同时大幅降低计算开销。
链接: https://arxiv.org/abs/2607.19424
作者: Qingjia Huang,Jingyu Zhang,Jianguo Wu,Yakai Li,Weijuan Zhang,Yankai Rong,Junyi Yao,Shengzhi Zhang,Xiaoqi Jia
机构: 未知
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:The assessment of jailbreak attacks against large language models currently suffers from inconsistent evaluation criteria and methods, leading to unreliable estimates of attack success rates. We propose JailMeter, an evidence-based evaluation framework designed to more faithfully measure jailbreak effectiveness. Inspired by the Information Bottleneck theory, JailMeter applies dual-feedback optimization to filter jailbreak noise from model responses while preserving content relevant to the original malicious question. This process produces concise evidence for a rigorous assessment under which an attack is validated only when the response captures the malicious intent and delivers a complete answer, thereby signaling a substantive bypass of model safety alignment. We evaluate JailMeter on JailMeter-Eva, a challenging benchmark containing 330 human-labeled, non-rejected jailbreak instances. JailMeter achieves an accuracy of 97.27%, substantially outperforming existing evaluation methods. To support large-scale evaluation, we further distill JailMeter into a small language model, JailMeter\textsubscriptSLM, which maintains comparable reliability with significantly reduced computational costs. Code and dataset are available at this https URL.
[NLP-56] HyGRL: Adaptive Hybrid Graph Reasoning for Multi-Entity Questions
【速读】： 该论文旨在解决多实体组合型问题（multi-entity compositional questions）在现有检索增强语言模型（Retrieval-Augmented Language Models, RAG）中面临的挑战，尤其是传统方法在动态推理能力、结构稀疏性以及计算成本之间的权衡困境。具体而言，标准RAG缺乏动态推理能力，传统图结构增强的RAG受限于知识图谱的结构稀疏性，而由大语言模型（LLM）构建的图结构RAG则带来高昂的计算开销。为此，论文提出一种统一框架\fwa，其核心创新在于将非结构化文本嵌入到异构知识图谱（heterogeneous knowledge graph）中，构建一个支持灵活证据检索的复杂网络。关键解决方案在于将推理过程建模为自适应结构归纳（adaptive structure induction），并通过一个稳健的两阶段学习机制实现：第一阶段采用模仿学习（imitation learning）从启发式专家信号中提取先验知识；第二阶段利用基于大语言模型生成的偏好奖励（LLM-driven preference rewards）进行强化学习（reinforcement learning）以优化推理策略。实验表明，\fwa能够有效融合文本丰富性与结构化知识，在答案准确率和推理保真度上超越当前最优基准（SOTA），同时保持极低的令牌消耗和接近实时的推理速度。
链接: https://arxiv.org/abs/2607.19398
作者: Junyi Wang
机构: Beijing Institute of Technology (北京理工大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 8 pages
Abstract:Multi-entity compositional questions pose significant challenges to existing retrieval-augmented language models. Conventional methods fall into a dilemma: standard RAG lacks dynamic reasoning, traditional Graph-RAG is limited by structural sparsity, and LLM-constructed Graph-RAG incurs prohibitive costs. We propose \textbf\fwa, a unified framework that embeds unstructured text into structured knowledge graphs, creating a heterogeneous network for flexible evidence retrieval. Reasoning is formulated as adaptive structure induction, learned via a robust two-stage process: (1) imitation learning distills heuristic expert signals, and (2) reinforcement learning refines the policy using LLM-driven preference rewards. Experiments demonstrate that \fwa effectively merges textual richness with structural knowledge, outperforming SOTA baselines in answer accuracy and reasoning fidelity while maintaining extremely low token costs and near real-time inference((code available at this https URL) .
[NLP-57] Decodable but Not Detectable: A Leakage Fingerprint for Near-OOD Benchmarks
【速读】： 该论文旨在解决现有基于扰动的异常检测（Out-of-Distribution, OOD）评估中因数据集泄露（benchmark leak）导致的虚假性能表现问题。核心问题是：某些被标记为“域外”（OOD）的数据实际上属于模型训练时的分布内（in-distribution），导致检测器在正确识别这些样本为“熟悉”时反而被惩罚，从而严重低估了真实性能。其解决方案的关键在于提出一种可验证的泄漏指纹（leak fingerprint）——即在监督学习下具有接近完美的可解码性（AUROC ≈ 1），而在无监督检测中却表现极差（AUROC < 0.65），并通过在52个受控设置下的实验证明该指纹能有效区分含泄漏与纯净的数据配置，在嵌入空间中实现18/20的敏感性与31/32的特异性。进一步的野化审计表明，标准跨数据集构建方法总体清洁，仅在高难度场景（如CIFAR-100 vs CIFAR-10）出现误报。修正协议后，扰动信号虽可被监督模型解码，但无法被无监督检测器识别，且扰动方法不再优于基础马氏距离（Mahalanobis distance）。研究最终提供了一个经过验证的泄漏诊断工具和一个修正后的评估协议，而非新型的OOD检测方法。
链接: https://arxiv.org/abs/2607.19393
作者: Vishnu Bindu Balachandran
机构: Independent AI Researcher
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:While auditing a perturbation-based OOD detector on a document benchmark, we recorded an AUROC of 0.326 – well below the 0.5 chance level. The cause is a benchmark leak: the designated “OOD” class is one the model was trained on, so its examples sit inside the in-distribution fit set and the detector is penalized for correctly ranking them as familiar. Deleting the class and retraining 35 models across two domains raises the score to 0.911. We distill the contamination into a leak fingerprint – near-perfect supervised decodability (AUROC approximately 1) coupled with unsupervised detection collapsed below 0.65 – and validate it on a controlled battery of 52 settings (20 leaked, 32 clean) across ResNet-50 and ViT-B/16 on CIFAR-10/100, achieving sensitivity 18/20 and specificity 31/32 in embedding space; the matched fit-set-exclusion controls are perfect at 20/20. An in-the-wild audit of 24 standard near/far OOD benchmark pairs fires on exactly one (the intrinsically hard CIFAR-100 vs CIFAR-10 pair) and on no far-OOD pair, confirming specificity and that standard cross-dataset construction is clean. Under the corrected protocol, perturbation signals are decodable but not detectable: a supervised reader recovers the OOD signal (AUROC 0.87-1.00) while no unsupervised detector does, and the perturbation method does not improve on plain Mahalanobis distance. We provide a theoretical account of why and, for transparency, retract an earlier circular correlation. The contributions are a corrected protocol and a validated leak diagnostic, not a new OOD method.
[NLP-58] Building Fast Evaluating Slow: Pipeline Choices Dominate Autointerpretability Score Variance
【速读】： 该论文旨在解决稀疏自编码器（Sparse Autoencoder, SAE）可解释性跨研究比较中存在的有效性问题，核心挑战在于当前普遍依赖的自动可解释性评分（autointerpretability scores）是否真正反映特征本身的稳定属性，而非评估流程中方法论差异带来的干扰。研究表明，现有评估框架存在严重缺陷：其一，方法论变异（methodological variance）在所有指标和模型中均显著超过架构差异（architectural variance），导致评分结果更多受评估流程影响；其二，不同评估指标表现出各异的不稳定性，其中“检测”任务相对稳定，“模糊测试”（fuzzing）则在所有条件下均不可靠；其三，即使整体平均得分稳定，顶级特征排名在不同语料和抽样条件下也缺乏一致性，而这种个体特征层面的不稳定性无法通过解释相似性监控发现。因此，论文指出，基于自动可解释性评分的跨论文比较可能实质上反映了评估管道的差异，而非模型架构的真实差异，这严重影响了对SAE有效性的科学判断。为应对这一问题，作者提出方差分解方法、稳定性检验（Stability Check）以及最小报告清单（Minimum Reporting Checklist），以提升可解释性研究评估的可靠性与可复现性。
链接: https://arxiv.org/abs/2607.19386
作者: Sinie van der Ben,Neele Roch,Anna Hedström,Mennatallah El-Assady
机构: ETH Zürich (苏黎世联邦理工学院)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注:
Abstract:Cross-paper comparison of sparse autoencoder (SAE) interpretability often relies on autointerpretability scores. In this evaluation pipeline, a language model (LM) explains each feature, and another LM scores the explanation. For these comparisons to be meaningful, scores must reflect stable properties of the features rather than confounding aspects of the evaluation pipeline. Through systematic experiments across four metrics (simulation, detection, fuzzing, purity), two models (Pythia-160M, Apertus-8B), and four axes of methodological variation, we show that this assumption does not hold. Specifically, we find that R1) methodological variance collectively exceeds architectural variance across all metrics and tested models; R2) each metric exhibits a distinct instability profile, with detection being the most stable and fuzzing unreliable across all conditions; R3) top-k feature rankings do not stay consistent across corpus and draw conditions, masking per-feature instability behind stable mean scores; a failure that cannot be detected by monitoring explanation similarity alone. These findings suggest that cross-paper comparisons based on autointerpretability scores may reflect pipeline differences rather than architectural differences, with implications for the ongoing debate on SAE utility. More broadly, unreliable evaluation slows progress in interpretability research at a time when reliable tools for understanding AI systems are needed. To support evaluation, we contribute a variance decomposition approach, a Stability Check, and a Minimum Reporting Checklist.
[NLP-59] Mitigating Scaffolding Collapse in Socratic Tutors via Representation Alignment
【速读】： 该论文旨在解决生成式AI驱动的苏格拉底式辅导系统在多轮对话中面临的“支架坍塌”（scaffolding collapse）问题，即在持续的学生压力下，辅导模型逐渐放弃引导式探究，直接揭示答案。现有防御方法主要通过提示工程、偏好优化或输出过滤等手段限制可观察行为，但未能有效应对导致坍塌的内部表征漂移（representation drift）。本文提出一种两阶段框架——支架保持表征对齐（Scaffold-Preserving Representation Alignment），其核心在于：首先通过监督微调（supervised fine-tuning）初始化辅导模型，随后结合轨迹加权的直接偏好优化与锚定于冻结参考状态的边界保持表征损失（margin-preserving representation loss），以确保在对话各轮次中，保留支架功能的隐藏状态与诱发坍塌的隐藏状态之间保持可区分性。实验结果表明，在涵盖五个STEM学科及五种红队攻击策略的评估中，该方法在Qwen3-8B模型上将坍塌率降低至32%，平均坍塌发生时间延后至九轮以上，同时维持较低的过度拒绝率，验证了表征层面的对齐机制能够显著提升长时程苏格拉底式辅导的鲁棒性。
链接: https://arxiv.org/abs/2607.19371
作者: Jing Shao,Qifeng Wu,Hanyu Zhang,Sixia Sun,Jun Zhuang
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: preprint, under review
Abstract:Large language model (LLM)-based Socratic tutors increasingly guide students through multi-turn questioning, but they can suffer from scaffolding collapse: under sustained student pressure, a tutor gradually abandons guided inquiry and reveals solutions directly. Prior defenses primarily constrain observable responses through prompting, preference optimization, or filtering, leaving the internal representation drift that precedes trajectory-level collapse largely unaddressed. We propose Scaffold-Preserving Representation Alignment, a two-stage framework that first warms up a Socratic tutor with supervised fine-tuning, then combines trajectory-weighted direct preference optimization with a margin-preserving representation loss anchored to frozen reference states. Our method is designed to maintain separation between scaffold-preserving and collapse-inducing hidden states across dialogue turns. We evaluate our method across five STEM disciplines and five red-teaming attack strategies. On Qwen3-8B, our method lowers Collapse Rate to 32%, delays average collapse onset beyond nine turns, and keeps over-refusal low, suggesting that representation-level alignment can improve the robustness of long-horizon Socratic tutoring under our red-teaming protocol.
[NLP-60] Spectral-LSH: Sub-Quadratic Prompt Compression via Krylov-Projected Locality-Sensitive Hashing
【速读】： 该论文旨在解决长提示（long-prompt）推理中因预填充阶段注意力计算复杂度随序列长度呈二次方增长（O(N²)）而导致的高计算开销问题。其核心解决方案是提出一种无需训练的提示压缩方法——Spectral-LSH，该方法在提示输入语言模型前进行处理。关键创新在于：利用克里洛夫（Krylov）子空间方法结合随机特征，隐式近似注意力核算子的主要成分，避免了显式构建O(N²)规模的注意力核；随后在得到的注意力特征空间中应用SimHash进行局部敏感哈希（LSH），对语义相似的标记进行分组并聚合为具有因果位置分配的宏标记（macro-tokens）。实验表明，在不同压缩比下，该方法展现出显著的性能优势：当压缩比ρ > 8×时，Spectral-LSH能有效保持生成质量，优于传统分块（chunking）策略；在ρ = 16×时，Qwen2.5-7B与14B模型的PPL比值分别从353.409/9.533降至196.963/3.427。此外，通过自适应后端设计，系统可动态选择低压缩比下的分块路径与高压缩比下的谱聚类路径，兼顾延迟与质量，实现最优权衡。
链接: https://arxiv.org/abs/2607.19368
作者: Ali Mahdavi,Azaseh Zamanifar,Amirfarhad Farhadi,Omid Kashefi
机构: Islamic Azad University (伊斯兰大学); Iran University of Science and Technology (伊朗科学技术大学); Meta (Meta)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Long-prompt inference remains expensive because prefill attention scales quadratically with sequence length. We propose Spectral-LSH, a training-free prompt compression method that operates before the prompt enters the language model. Spectral-LSH approximates the dominant components of an implicit attention-kernel operator using a Krylov subspace method together with random features, avoiding explicit O(N^2) attention-kernel materialization. It then applies SimHash in the resulting attention eigenspace to group similar tokens and aggregate them into macro-tokens with causal positional assignments. We evaluate Mistral-7B-Instruct-v0.3, Qwen2.5-7B-Instruct, and Qwen2.5-14B-Instruct on C4. Our experiments reveal a compression-ratio phase transition. Below \rho = 4 \times , local token redundancy is low enough that lightweight chunking typically provides the best latency–quality trade-off. Above \rho = 8 \times , the spectral path preserves quality that chunking loses. At \rho = 16 \times , Qwen2.5-7B (adaptive) reduces the PPL ratio from 353.409 to 196.963, while Qwen2.5-14B (adaptive) reduces it from 9.533 to 3.427. On a small long-context structured stress test containing JSON-like, code-like, and table-like inputs, local LSH also improves every metric over chunking at 8 \times . The adaptive backend captures both regimes by using the chunk path at low compression and spectral clustering at high compression, although chunking remains the fastest backend in total latency. Subjects: Artificial Intelligence (cs.AI); Computation and Language (cs.CL) Cite as: arXiv:2607.19368 [cs.AI] (or arXiv:2607.19368v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2607.19368 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-61] Statistically Grounded Sparse-Feature Interventions for Activation-Space Control in Large Language Models
【速读】： 该论文旨在解决生成式AI（Generative AI）中基于激活量化的行为控制问题，特别是针对现有基于稀疏自动编码器（Sparse Autoencoder, SAE）的激活引导方法依赖于学习到的引导目标或单一标准特征选择所导致的可解释性差与泛化能力不足的问题。其解决方案的关键在于提出一种透明、无优化的SAE特征引导流程：首先通过六条件可靠性过滤器筛选高质量特征，随后利用未加权的博达共识（Borda consensus）对三种互补统计量——F检验、KSG互信息和Cohen’s d进行排序，以确定关键稀疏特征；最终构建的引导方向为解码器行的Cohen’s d加权组合，该设计在近似SAE特征正交化条件下，受到Fisher-LDA理论启发，实现了无需优化的可解释引导方向。实验表明，该方法在Gemma系列模型的多个层级与强度配置下均能实现显著且领域特定的行为偏移，但同时也揭示了可用引导效果高度依赖于模型、任务领域、层数及强度的局部特性。研究强调，激活引导评估应同时报告质量条件下的成功指标与原始行为变化，以更全面反映实际有效性。
链接: https://arxiv.org/abs/2607.19364
作者: Oshayer Siddique,J. M Areeb Uzair Alam,Md Jobayer Rahman Rafy,Syed Rifat Raiyan,Hasan Mahmud,Md Kamrul Hasan
机构: Islamic University of Technology, Dhaka, Bangladesh (伊斯兰科技大学达卡分校)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Under review, 22 pages, 5 figures, 16 tables
Abstract:Activation steering offers a lightweight alternative to fine-tuning for behavioral control of large language models, but SAE-based steering methods often rely on learned steering objectives or single-criterion feature selection. We introduce a transparent SAE-feature steering pipeline that first applies a six-condition reliability filter, then ranks sparse features through an unweighted Borda consensus over three complementary statistics: F -test, KSG mutual information, and Cohen’s d . The resulting steering direction is constructed as a Cohen’s- d -weighted combination of SAE decoder rows, providing an optimization-free direction motivated by Fisher-LDA under approximate SAE-feature decorrelation. Across three Gemma-family models, four behavioral domains, and 356 layer-strength configurations, the method produces measurable domain-specific shifts while revealing a substantial gap between raw attribute movement and quality-preserving generation. In the strongest configuration, logical-correctness steering reaches a primary-score delta of +1.16 in Gemma~2 9B; however, our broader finding is that usable steering is highly localized by model, domain, layer, and strength. These results argue that activation-steering evaluations should report quality-conditioned success alongside raw behavioral shift. Our code and data are available at this https URL.
[NLP-62] AdaRoPE: Not All Attention Heads Should Rotate and Scale Equally ICML2026
【速读】： 该论文旨在解决标准旋转位置编码（Rotary Position Embedding, RoPE）在注意力机制中对所有注意力头采用统一频率调度和缩放策略所导致的性能瓶颈问题。研究表明，不同功能角色的注意力头需要不同的频率范围与注意力缩放因子以实现最优性能，而现有方法忽略这一差异性结构，造成嵌入维度利用不充分，并在长上下文场景下显著降低模型表现。其解决方案的关键在于提出AdaRoPE，即为每个注意力头引入可学习的旋转频率和注意力缩放因子，实现头级别的自适应优化。实验表明，基于AdaRoPE预训练的大语言模型在多种任务上均优于现有RoPE变体（如部分RoPE、NoPE等），且在上下文扩展方面，相比统一缩放策略（如YaRN），AdaRoPE通过头特定缩放实现了更优的长上下文泛化能力，同时有效保持短上下文性能。研究结果强调了在单个注意力头层面优化旋转位置编码的重要性。
链接: https://arxiv.org/abs/2607.19363
作者: Shaowen Wang,Yuke Zheng,Tansheng Zhu,Shuang Chen,Shaofan Liu,Suncong Zheng,Jian Li
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Accepted at ICML 2026
Abstract:Rotary Position Embedding (RoPE) is widely adopted in Transformers to encode positional information, yet standard implementations enforce a uniform frequency schedule and scaling across all attention heads. Using simplified retrieval tasks and length generalization scenarios, we show – both empirically and theoretically – that heads with different functional roles require distinct frequency ranges and attention scaling factors to operate effectively. Ignoring this structure leads to suboptimal utilization of embedding dimensions and degraded performance, particularly under long-context settings. To address these limitations, we propose AdaRoPE, which equips each attention head with learnable rotation frequencies and attention scaling factors. Pretrained LLMs with AdaRoPE consistently outperform existing RoPE variants, including partial RoPE and NoPE baselines. For context extension, we further show that uniform frequency and attention scaling, used in methods such as YaRN, are suboptimal. By applying head-specific scaling, AdaRoPE enables better context extension while better preserving short-context performance in both the extrapolation setting and the long-context continued pretraining setting. These results highlight the importance of optimizing rotary position embedding at the level of individual attention heads.
[NLP-63] Stateful Guardrails for Multi-Turn LLM Systems: A Conversational Risk Accumulation Framework
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）安全防护中一个关键缺陷：现有安全护栏大多孤立评估每个提示-响应对，忽略了对话过程中逐步累积的潜在风险。这种风险表现为会话级风险累积（Conversational Risk Accumulation, CRA），即通过多个看似无害的对话轮次逐渐积累出有害意图、碎片化拼接违规指令以及因重复披露信息导致敏感度上升等现象。其解决方案的核心在于提出一种会话层CRA框架，通过追踪三类轨迹信号实现对风险的动态感知：（1）以会话锚点为基准的语义漂移；（2）基于抽取实体的信息累积图，结合敏感度加权；（3）反映合规意愿递增的合规梯度信号。在评分机制上，采用无监督凸融合方法进行归因分析与消融实验，并引入CRA-Net DA——一个经过家族对抗性目标训练的紧凑轨迹学习模型，以缓解长度和话题覆盖范围带来的混淆因素。为评估该框架，研究发布了CRA-Bench v0.1与v0.2及扩展版五类威胁数据集，涵盖多种攻击模式（如角色诱导、上下文投毒），并设计了原生轨迹评估协议，包括会话级划分、混合集阈值校准、轨迹AUROC、检测耗时、校准误报率、置信区间估计、留一类别诊断压力测试以及合成到人类迁移验证等，重点验证在分布内会话评分与人类转移能力上的有效性。
链接: https://arxiv.org/abs/2607.19361
作者: Sanjay Mishra,Divya Chukkapalli,Ganesh R. Naik
机构: Flinders University (弗林德斯大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 45 pages, 10 figures, 20 tables
Abstract:Most safety guardrails for large language models (LLMs) evaluate each prompt-response pair in isolation, which misses failures that arise only over a dialogue as benign turns compose into harm. We term this Conversational Risk Accumulation (CRA): gradual intent drift, fragmented assembly of prohibited instructions, and sensitivity build-up from repeated disclosures. We propose a session-layer CRA Framework that tracks three trajectory signals: semantic drift from a session anchor, a sensitivity-weighted information accumulation graph over extracted entities, and a compliance-gradient signal capturing increasing willingness to comply. For scoring, we provide (i) an unsupervised convex fusion for attribution and ablations, and (ii) CRA-Net DA, a compact learned trajectory model trained with family-adversarial objectives to reduce length and topic-coverage confounds. To benchmark CRA, we release CRA-Bench v0.1 (1,200 eight-turn sessions across three threat families with topic-matched benign twins), CRA-Bench v0.2 (LLM-paraphrased variants to reduce template artifacts), and an extended 5-family set (2,000 sessions adding persona priming and context stuffing). We introduce a trajectory-native evaluation protocol with session-level splits, mixed-set threshold calibration, Trajectory AUROC, turns-to-detection, calibrated false-positive metrics, bootstrap confidence intervals, leave-one-family-out diagnostic stress tests, and synthetic-to-human transfer checks. Claims focus on within-distribution session scoring on CRA-Bench and human-transfer subsets.
[NLP-64] Information Discernment in Large Language Models
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在使用外部知识源（如互联网）时的信息权衡问题，即模型是否能够正确区分信息的可靠性（源辨识，source discernment）与真实性（真值辨识，truth discernment）。具体而言，研究关注模型能否根据来源可信度调整其信念更新程度，以及能否在新信息使先验更接近真实情况时做出合理修正。为系统评估此能力，作者提出“Learn2Discern”（L2D）实验框架与基准测试，基于三个规范性公理构建可解释的评估指标，并通过一项预注册、配额匹配的用户研究（n=299）验证了真实用户对这些公理的认可度，且发现违背这些原则会显著降低用户信任与使用意愿。在涵盖13个模型、近67万次试验的实证分析中，研究发现模型在源辨识与真值辨识两方面均存在系统性失败：表现接近随机水平；对来源流行度的依赖程度是对其可靠性依赖的两倍；无论新信息是否提升其与真实值的一致性，模型均进行相似幅度的信念更新。此外，模型仅在先验已较准确的数据集上有效整合外部知识，而随着模型规模和复杂度提升，仅真值辨识能力略有改善，源辨识能力仍无进展，表明这一盲点无法通过模型扩容解决。研究进一步识别出若干推理阶段的简单干预措施，可同时提升两类辨识能力。研究结果强调了信息辨识作为核心对齐属性的重要性，并公开发布数据集与调查工具，以支持未来对齐研究的发展。
链接: https://arxiv.org/abs/2607.19355
作者: Joshua Ashkinaze,Laura Kurek,Alina Faisal,Tongyuan Miao,Mariam Joseph,Ceren Budak,Eric Gilbert
机构: University of Michigan(密歇根大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Computers and Society (cs.CY)
备注:
Abstract:LLMs are increasingly used with external knowledge sources like the internet. Do they weigh information appropriately – updating more for reliable sources (source discernment) and more when claims bring priors closer to the truth (truth discernment)? We formalize this as information discernment and introduce Learn2Discern (L2D), an experimental framework and benchmark grounded in three normative axioms with interpretable metrics. To establish external validity, a pre-registered, quota-matched user study (n=299) confirms that real LLM users endorse all three axioms and report that violations reduce their trust and usage intent. Across 13 models and nearly 670K trials, we find consistent failures across both dimensions: models perform near chance on source and truth discernment, rely on source popularity twice as much as source reliability, and update roughly equally whether a claim improves or worsens their position relative to the ground truth. Models integrate external knowledge most effectively on datasets where their priors are already the most accurate. Newer and larger models improve truth discernment but not source discernment, a blind spot that model complexity does not address. We identify simple inference-time interventions that improve both forms of discernment. We release our dataset and survey as a testbed for a core alignment property that scales in importance as LLMs replace traditional search.
[NLP-65] Auditing Retrieval-Augmented LLM Hypotheses for Longitudinal Cell Painting Morphology
【速读】： 该论文旨在解决高内涵形态学分析（Cell Painting）中长期形态轨迹向可解释生物学机制转化的难题，尤其针对低剂量率电离辐射等弱而慢性扰动下的生物响应解析。其核心挑战在于如何从高维、动态的细胞形态数据中提取具有生物学意义且可验证的假说。解决方案的关键在于提出一种“评估先行、检索增强”的解释框架：通过整合周匹配的处理组与对照组间的形态差异（deltas）、检索到的相似扰动邻域、通路背景信息及文献证据，并利用稳定的证据标识符进行关联，使大语言模型（LLM）能够生成结构化、证据链可追溯的生物学假说，同时保持溯源性。为确保科学可靠性，引入两项量化审计测试——V1引用有效性（验证引用标识符是否存在于提示中）和V2基于代理的形态兼容性（评估预测生物过程与最显著改变的形态特征之间的一致性），实验结果显示两者均表现良好，且形态兼容性随扰动强度增加而提升，与独立的形态漂移总结呈正相关。该框架最终生成了可审计、可证伪的生物学假说，例如在低剂量率（0.003–0.3 mGy/hr）下存在代谢重编程与蛋白稳态应激协同作用的适应性表型。当前局限包括依赖代理评估以及缺乏真实机制标签。
链接: https://arxiv.org/abs/2607.19415
作者: Gilchan Park,Guang Zhao,Byung-Jun Yoon,Shinjae Yoo
机构: Brookhaven National Laboratory (布鲁克黑文国家实验室)
类目: Quantitative Methods (q-bio.QM); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Accepted for publication at ACM BCB 2026. This is the author’s version. The definitive Version of Record is available at [ this https URL ]( this https URL )
Abstract:High-content morphological profiling (Cell Painting) yields sensitive, high-dimensional signatures of cellular state, but translating longitudinal morphology trajectories into interpretable biology remains difficult, especially for weak, chronic perturbations such as low-dose-rate ionizing radiation. Large language models (LLMs) can synthesize heterogeneous evidence into biological narratives, yet their scientific use requires quantitative auditing. We present an evaluation-first, retrieval-augmented interpretation framework for longitudinal Cell Painting morphology, applied to a 9-week RPE-1 time course across five dose rates (0.003–6.0 mGy/hr). Week-matched treated-control morphology deltas are combined with retrieved perturbation neighbors, pathway context, and literature evidence through stable evidence identifiers, enabling an LLM to generate structured, evidence-linked hypotheses that are hierarchically summarized while preserving provenance. We introduce two quantitative auditing tests: V1 citation validity, which verifies that cited evidence identifiers exist in the prompt, and V2 proxy-based morphology compatibility, which evaluates consistency between predicted biological processes and the most altered morphology features. In our experiments, V1 detected no invalid evidence references, while V2 showed meaningful morphology compatibility that increased with perturbation strength and was positively associated with an independent morphology drift summary. The framework produces auditable, falsifiable biological hypotheses, including an adaptive phenotype involving metabolic reprogramming and proteostatic stress at lower dose rates (0.003–0.3 mGy/hr). Current limitations include proxy-based evaluation and the lack of ground-truth mechanism labels.
信息检索
[IR-0] Near-Optimal Dimension Lower Bounds for Single-Vector Embeddings of Maximum Inner Product Similarity
链接: https://arxiv.org/abs/2607.20393
作者: Rajesh Jayaram,Honghao Lin,Vahab Mirrokni,David P. Woodruff
类目: Data Structures and Algorithms (cs.DS); Information Retrieval (cs.IR)
备注:
Abstract:Multi-vector embeddings represent items by point clouds and compare query and document point clouds using Chamfer similarity, whereas single-vector embeddings use ordinary inner products. For singleton queries, Chamfer becomes maximum inner product similarity (MAX-IP). In our setting, MUVERA gives dimension m^O(1/\varepsilon^2) ~\citedhulipala2024muvera, whereas the previous lower bound (\varepsilon^2m)^\Omega(1/\varepsilon) ~\citejayaram2026expressive left a gap between 1/\varepsilon and 1/\varepsilon^2 in the exponent of m . We nearly close this gap. For every fixed \delta\in(0,1) , there are constants A_\delta,c_\delta0 such that, for all sufficiently small \varepsilon0 and every m\ge(1/\varepsilon)^A_\delta , there exist unit query vectors and document point clouds of at most m unit vectors for which every single-vector approximation of all pairwise MAX-IP values to additive error \varepsilon has dimension [ D \ge m^c_\delta/\varepsilon^2-2\delta. ] This holds even for fully data-dependent representations chosen after seeing the dataset. It also applies to Chamfer because all queries are singletons. Since \delta can be arbitrarily small, the exponent approaches the O(1/\varepsilon^2) dependence of the upper bound. The proof combines Sherstov’s pattern matrix method with polynomial-size, constant-width DNF formulas computing functions of approximate degree \Omega(k^1-\delta) . Uniform-width padding and a block encoding create an \Omega(\varepsilon) gap. A dummy coordinate then equalizes all false inputs, yielding a unit-sphere MAX-IP matrix that is an exact two-valued affine image of the DNF pattern matrix with gap at least 8\varepsilon . This allows the approximate-rank bound to apply. Subjects: Data Structures and Algorithms (cs.DS); Information Retrieval (cs.IR) Cite as: arXiv:2607.20393 [cs.DS] (or arXiv:2607.20393v1 [cs.DS] for this version) https://doi.org/10.48550/arXiv.2607.20393 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[IR-1] wo-Step Occupation Coding
链接: https://arxiv.org/abs/2607.20101
作者: Alexander M. Esser,Jens Dörpinghaus
类目: Computation and Language (cs.CL); Information Retrieval (cs.IR)
备注: Preprint of the paper accepted for the Federated Conference on Computer Science and Information Systems (FedCSIS 2026)
Abstract:Occupation coding links job titles in free text to occupational taxonomies and is a core task in labor market research. Existing approaches typically address this problem in a single end-to-end step, jointly identifying job titles and assigning occupational codes. This paper presents a novel two-step approach that separates these tasks. In the first step, a domain-specific Named Entity Recognition (NER) model identifies occupational titles in continuous text, even under noise such as OCR errors. In the second step, the extracted job titles are mapped to a taxonomy, enabling the classifier to focus exclusively on this mapping. We demonstrate that this separation improves accuracy, robustness, and interpretability compared to single-step approaches. The method has been developed for German documents but is transferable to other languages. We further introduce a margin-based confidence criterion for occupation coding, replacing common absolute thresholds. To support reproducibility, we publish the source code and evaluation scripts.
[IR-2] UniRank: Benchmarking Ranking Models for Unified Sequential Modeling and Feature Interaction
链接: https://arxiv.org/abs/2607.19987
作者: Honghao Li,Xianquan Wang,Zibin Zhang,Yi Zhang,Kangyi Lin,Yiwen Zhang
类目: Information Retrieval (cs.IR)
备注: 11 pages, 6 figures, and 7 tables. Code and data: this https URL
Abstract:Ranking is a core stage in online advertising and recommender systems. Modern ranking models increasingly unify sequential modeling and feature interaction, yet many advances rely on proprietary data, closed implementations, and large-scale industrial infrastructure. This setting limits reproducible comparison and hinders academic study of scaling laws, long-sequence modeling, and multi-task ranking. To address these limitations, this paper proposes UniRank, an open benchmark for ranking models that unify sequential modeling and feature interaction. UniRank uses chronological pointwise autoregressive supervision, standardizes evaluation across feedback tasks, and provides a PyTorch toolkit with Distributed Data Parallel training, operator optimization, mixed-precision training, attention optimization, and other efficiency techniques that reduce hardware requirements. We benchmark 15 representative unified ranking models on five large-scale public datasets from short-video, advertising, and e-commerce platforms, with the largest dataset containing over 700 million instances and the longest behavior sequence exceeding 10^5 interactions. UniRank provides a reproducible basis for comparing unified ranking models, studying scaling laws under limited compute, and narrowing the gap between academic and industrial ranking research. We believe UniRank benefits researchers, practitioners, and beginners through reproducible experiments, production-oriented evaluation, and accessible implementations. Code and data are available at this https URL.
[IR-3] Using Hierarchical Controlled Vocabularies to Understand CLIP Retrieval Failures in Historical Photo Collections
链接: https://arxiv.org/abs/2607.19836
作者: Ratan Sebastian,Anett Hoppe,Christoph Rippe,Ralph Ewerth
类目: Information Retrieval (cs.IR); Digital Libraries (cs.DL)
备注:
Abstract:GLAM institutions (Galleries, Libraries, Archives, and Museums) organise image access using controlled vocabularies such as the Art and Architecture Thesaurus (AAT). For content-based image retrieval in these settings, vision-language models like CLIP are increasingly used, but their performance varies. This variation is known to relate to measures like concept abstraction and concept frequency. However, no prior work explains this variation in terms of the structural properties of vocabularies like the AAT that GLAM professionals already use. The AAT groups concepts into broad facets (Objects, Activities, Agents, etc.) and arranges terms hierarchically within them. In this paper, we ask whether two structural properties (root facet type and hierarchy depth) explain where CLIP retrieval succeeds and fails, and where fine-tuning helps. Across three historical photographic collections annotated with AAT terms, we examine visual coherence (whether a term’s photographs cluster in CLIP’s embedding space), text-image alignment (whether its label is near that cluster), and standard retrieval measures, which conflate the two. We find that visual coherence and text-image alignment are nearly uncorrelated across terms and jointly separate distinct failure modes. Terms whose photographs cluster tightly but whose label is distant from the cluster retrieve poorly in every collection, in two of three collections even worse than terms that fail on both metrics. We also show that while retrieval metrics do not correlate significantly with either structural property, root facet type does significantly separate categories with varying visual coherence. Finally, we find that fine-tuning improves retrieval overall, but its gains favour shallower terms in the hierarchy, where text-image alignment improves most, beyond what concept frequency explains.
[IR-4] Zero-Observation User Reactivation with Gap-Driven Dimensional Gating RECSYS2026
链接: https://arxiv.org/abs/2607.19802
作者: Jiandong Ding,Tianying Liu,Fuyuan Liu,Huijie Qin,Tiandeng Wu
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注: Accepted at the 20th ACM Conference on Recommender Systems (RecSys 2026)
Abstract:Sequential recommendation (SR) models capture continuously observed behavior, but a returning user may have no interactions for months or years. We define this setting as Zero-Observation Reactivation: the user has a pre-gap history, while the platform observes no behavioral signals during a macro-gap Delta t. Under a chronologically aligned Gap-Synthesize Protocol on three Amazon datasets (Video Games, CDs Vinyl, and Movies TV), Hit@10 decreases monotonically across the evaluated gap buckets and reaches its lowest level beyond one year. The pattern appears across recurrent, unidirectional, and bidirectional SR backbones. We propose DeltaGate, a lightweight output-layer plugin that keeps the backbone frozen and routes each representation dimension between the personalized history and a learned, zero-initialized global prior. The gate is conditioned jointly on Delta t and the personalized representation. In a controlled diagnostic, we hold the personalized representation fixed and vary Delta t to isolate the trained gate’s response to the gap input. In the 365d Video Games bucket, DG-SASRec reaches 0.047 Hit@10 versus 0.031 for SASRec, while DG-BERT4Rec reaches 0.046 versus 0.025 for BERT4Rec, with 66K trainable parameters (2–4% overhead). End-to-end retraining attains higher absolute accuracy but changes the backbone embeddings; the frozen plugin preserves zero backbone drift, uses about 40x fewer trainable parameters, and retains observable dimension-wise routing. The source code is available at this https URL. Comments: Accepted at the 20th ACM Conference on Recommender Systems (RecSys 2026) Subjects: Information Retrieval (cs.IR); Machine Learning (cs.LG) Cite as: arXiv:2607.19802 [cs.IR] (or arXiv:2607.19802v1 [cs.IR] for this version) https://doi.org/10.48550/arXiv.2607.19802 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Related DOI: https://doi.org/10.1145/3773078.3831771 Focus to learn more DOI(s) linking to related resources
[IR-5] CIR at iKAT SCAI 2026: Exploring Clarification Need Prediction in Agent ic Conversational Search
链接: https://arxiv.org/abs/2607.19801
作者: Nolwenn Bernard,Jüri Keller,Philipp Schaer
类目: Information Retrieval (cs.IR)
备注:
Abstract:This paper presents the participation of the Cologne Information Retrieval group in the iKAT SCAI 2026 shared task. We use an agentic conversational search system, equipped with tools for query rewriting, retrieval and reranking, answer generation, and clarification need prediction and clarification question generation. We experiment with two different neural clarification need prediction models.
[IR-6] Personalized Recommendation Tool Learning via Autonomous Language Agents RECSYS’26
链接: https://arxiv.org/abs/2607.19739
作者: Mingdai Yang,Zhiwei Liu,Weizhi Zhang,Yibo Wang,Hao Peng,Philip Yu
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注: 6 pages. Accepted by RecSys’26
Abstract:Although large language models (LLMs) have recently gained traction in recommender systems due to their strong reasoning capabilities and extensive world knowledge, previous LLM-based agents suffer from hallucination and context-length limitations, and thus are not suitable for full-ranking recommendation tasks. To circumvent these limitations through architectural design rather than modifying the LLM itself, we propose an agent-based recommendation framework, memory-based \textbfP ersonalized \textbfR ecommendation \textbfT ool learning via autonomous language \textbfA gents (PRTA), in which an LLM acts as a central planner interacting with multiple recommendation models as tools. The LLM-based agent is responsible for high-level reasoning and personalized tool selection, while traditional recommendation models perform full-ranking scoring, leveraging their scalability in modeling behavioral patterns. To support personalized tool selection, we design reflection mechanisms that enable the agent to evaluate and compare tools for each user based on user profiles and candidate ranked lists. Extensive experiments across three public datasets demonstrate the superiority of \modelname over traditional recommendation and LLM-based baselines in improving full-ranking recommendation performance.
[IR-7] GraphContainer: A Unified Platform for Comparing and Debugging Graph RAG Methods VLDB2026
链接: https://arxiv.org/abs/2607.19362
作者: Seonho An,Chaejeong Hyun,Min-Soo Kim
类目: Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
备注: VLDB 2026 demo accepted
Abstract:Graph RAG mitigates hallucinations and stale knowledge in LLMs, particularly for multi-hop question answering. However, existing approaches remain highly fragmented and incompatible. The structural heterogeneity of graph formats across different frameworks and the lack of granular visualization tools make it exceedingly difficult to evaluate and compare retrieval behaviors. To bridge this gap, we propose GraphContainer, a novel platform designed to unify and visualize diverse graph RAG workflows. GraphContainer features two key components: (1) a Unified Graph Representation (UGR) layer that seamlessly standardizes multi-format graphs, and (2) a Graph Recorder that tracks and visually renders the step-by-step retrieval process. Through an interactive web interface, we demonstrate GraphContainer’s ability to import heterogeneous graphs and perform live, traceable visual debugging of graph RAG methods. Ultimately, we show how GraphContainer enables controlled comparisons of various graph formats and retrieval strategies, lowering the barrier for researchers and practitioners to design optimal graph RAG pipelines. A demonstration video is available at this https URL.
人机交互
[HC-0] owards Miniature Humanoid Tele-Loco-Manipulation Using Virtual Reality and Reinforcement Learning
链接: https://arxiv.org/abs/2607.20399
作者: Nicolas Kosanovic,Jordan Dowdy,Jean Chagas Vaz
类目: Robotics (cs.RO); Human-Computer Interaction (cs.HC); Machine Learning (cs.LG)
备注: 8 pages, 6 figures. Accepted manuscript. Published in the 2025 IEEE-RAS 24th International Conference on Humanoid Robots (Humanoids), pp. 1233-1240
Abstract:Full-sized humanoid robot capabilities have grown exponentially in recent years, aiming towards general-purpose deployment in human environments. A popular control method used by manufacturers utilizes Virtual Reality for upper-body teleoperation and Reinforcement Learning for lower-body balance and locomotion control. As a result, a single remote operator can see, manipulate, and navigate about a real, distant physical environment. This powerful control stack is often relegated to expensive full-sized robots, many of which are inaccessible to the research community. Miniature humanoids are more prevalent, but employ less biomimicry in their design (e.g. fewer sensors, Degrees of Freedom, etc) and lack similar developments. This paper describes a compliant full-body telepresence control stack developed from the ground up for miniature humanoids. Framework experimentation on ROBOTIS OP3 hardware showcases walking at speeds up to 0.45 m/s independent of arm motions. Tele-loco-manipulation is demonstrated via a cube relocation experiment with an expert human operator. On average, the teleoperated system moved 2 different 40 g cubes within 10 mins, walking a total distance of 5 m. Overall, the developed system shows potential for miniature humanoid tele-loco-manipulation.
[HC-1] FMRP-LEAN: A HIPAA-Compliant AI-Augmented LIMS Architecture for End-to-End Clinical Assay Workflow Optimization
链接: https://arxiv.org/abs/2607.20382
作者: Eva McCord,Ernest Pedapati,Zag ElSayed
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI); Emerging Technologies (cs.ET); Medical Physics (physics.med-ph); Neurons and Cognition (q-bio.NC)
备注: 15 pages, 5 figures, conference
Abstract:Clinical biomarker workflows in translational research settings often rely on spreadsheet-driven tracking, manual quality control (QC) reconciliation, and loosely integrated systems, resulting in limited state visibility, delayed reporting, and increased operational risk. These challenges are particularly pronounced in multi-day assays such as Luminex-based quantification of Fragile X Messenger Ribonucleoprotein (FMRP), where HIPAA-compliant data governance, deterministic workflow progression, and coordinated communication across laboratory and clinical teams are required. This paper presents FMRP-LEAN, a HIPAA-compliant, AI-augmented Laboratory Information Management System (LIMS) architecture that formalizes biospecimen lifecycle management through a finite-state workflow model with explicit transition guards and dwell-time observability. The system integrates a self-hosted Supabase/PostgreSQL stack deployed within hospital-controlled infrastructure, hybrid edge-internal isolation with encrypted tunneling and loopback-only services, and bi-directional REDCap synchronization. A unified MRN-UUIDv7 identifier framework with QR-based tracking ensures traceable clinical-research linkage under PHI residency constraints. FMRP-LEAN incorporates automated statistical QC pre-screening and a governance-constrained AI operations module that operates exclusively on aggregate projections, with deterministic fallback guarantees. Deployment demonstrates improved workflow observability, reduced QC latency, and enhanced cross-role transparency between laboratory technicians, research coordinators, and patient-facing teams. The architecture provides a reproducible model for secure, state-explicit, and AI-augmented clinical research workflows in regulated healthcare environments.
[HC-2] Proceedings of The Fourth International Workshop on eXplainable AI for the Arts (XAIxArts 4)
链接: https://arxiv.org/abs/2607.20131
作者: Shuoyang Jasper Zheng,Terence Broad,Elizabeth Wilson,Adam Cole,Ziqing Xu,Jia-Rey Chang,Gabriel Vigliensoni,Jeba Rezwana,Lanxi Xiao,Michael Clemens,Makayla Lewis,Alan Chamberlain,Helen Kennedy,Corey Ford,Nick Bryan-Kinns
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:The fourth workshop on Explainable AI for the Arts (XAIxArts) continues to bring together and expand a community of researchers and creative practitioners in Human-Computer Interaction (HCI), Interaction Design, AI, eXplainable AI (XAI), and Digital Arts to explore the role of XAI for the Arts. XAI is a key concern of Responsible and Human-Centred AI, emphasising HCI techniques that make opaque AI models more understandable to people. XAIxArts offers a distinctive lens to examine explainability through creative and artistic domains. The previous workshops explored the landscape and the speculative futures of AI in creative processes. To respond to emerging challenges and contribute to creative and societal transformation more broadly, this workshop focuses on the operationalisation of XAI in the Arts. Specifically, we will: i) critically reflect on emerging practices that encourage diversity and inclusivity in XAI; ii) collectively ideate a library of missing projects to encourage future collaborations and speculations; iii) scope the development of a resource hub for open XAIxArts projects to archive tangible XAI interventions and facilitate future community building with the wider discourse on Human-Centred AI.
[HC-3] A Task Taxonomy for Edge and Trail Bundling IEEE-VIS
链接: https://arxiv.org/abs/2607.20089
作者: Markus Wallinger,Stephen G. Kobourov
类目: Human-Computer Interaction (cs.HC)
备注: Accepted for presentation at IEEE VIS (Short Paper)
Abstract:Edge bundling reduces visual clutter by aggregating similar edges, yet practitioners lack a structured vocabulary for reasoning about the tasks that bundled visualizations support. Such a vocabulary is needed both to evaluate the general utility of bundling and to compare different bundling approaches. We address this gap by assembling a corpus of 102 papers, 49 of which contain explicit bundling tasks, spanning node-link diagrams, geographic trail sets, and parallel coordinate plots. From this corpus, we derive a task taxonomy organized as a matrix of scope (Element, Bundle, Global, Multi-view) crossed with action (Verify, Identify, Characterize, Quantify, Compare, Assess), instantiated across the three representation types. We show that bundling simultaneously enables tasks (bundle-level and global reasoning) and disables others (element-level precision), a duality not captured by existing task frameworks. Our coded corpus and taxonomy are released as supplemental material on OSF (this http URL).
[HC-4] Factor-Informed Uncertainty Distillation for Gaze Estimation
链接: https://arxiv.org/abs/2607.20072
作者: Mohammadreza Jamalifard,Yaxiong Lei,Javier Fumanal Idocin,Parastoo Azizinezhad,Tom Foulsham,Javier Andreu-Perez
类目: Computer Vision and Pattern Recognition (cs.CV); Human-Computer Interaction (cs.HC)
备注:
Abstract:Deep gaze estimation works well in controlled capture but degrades in unconstrained settings, where systems must reject unreliable predictions. Single-pass uncertainty (e.g., heteroscedastic regression) infers uncertainty from pixels without explicit input-validity cues, while sampling based methods are often too costly for real time use. We propose Factor-Informed Uncertainty Distillation (FIUD), a teacher-student framework that aligns uncertainty with interpretable image-quality failure modes. A gradient-boosting teacher predicts expected gaze error from factors such as illumination, sharpness, eye visibility and symmetry; a neural student distills these signals via curriculum learning and ranking supervision into a lightweight single-pass uncertainty head. Across ETH-XGaze, Gaze360, and MPIIFaceGaze (300k samples), FIUD improves uncertainty, error rank correlation and selective prediction versus deterministic and sampling-based baselines, with the largest gains in unconstrained settings.
[HC-5] Experiential Versus Instructional Approaches for Eliciting Metacognitive Awareness in AI-Assisted Learning: A Short-Term Longitudinal Study
链接: https://arxiv.org/abs/2607.20047
作者: Pau Benazet i Montobbio,Janne Rotter,Davinia Hernández-Leo
类目: Human-Computer Interaction (cs.HC); Computers and Society (cs.CY)
备注:
Abstract:With generative AI (GenAI) entering classrooms the question to which teaching approach best supports metacognitive skill acquisition in AI-assisted learning becomes pressing. In this short-term longitudinal study we investigate two contrasting approaches: experiential learning encompassing hands-on approaches and instructional learning such as classical lectures. We conducted a quasi-experiment with 126 university students from a first-year engineering course which were distributed across the two conditions and completed a two hour session on learning with GenAI in the corresponding learning style. Metacognitive awareness which encompasses both knowledge of cognition (understanding effective AI-use strategies) and regulation of cognition (applying that knowledge in practice) was measured before and after the session. Additionally, students longitudinal metacognitive awareness was tracked over the trimester and assessed again five weeks after the initial intervention. Results reveal that experiential methods outperform instructional approaches in engagement and knowledge of cognition immediately after the intervention. By five weeks, the two groups converged on these measures, while the experiential group showed a delayed, continuous within-group increase in regulation of cognition that was not observed in the instructional group. This suggests that the benefits of experiential approaches extend beyond conventional educational settings to AI-assisted learning, and that knowledge and regulation may develop on different timescales under experiential learning.
[HC-6] Visual Indicators to Increase the Detection of Linguistic Media Bias
链接: https://arxiv.org/abs/2607.20031
作者: Smi Hinterreiter,Anna Chelsea Bahß,Ann-Christin Gah,Timo Spinde,Isao Echizen,Marc Erich Latoschik
类目: Human-Computer Interaction (cs.HC)
备注: Conditionally accepted for publication in IEEE Transactions on Visualization and Computer Graphics
Abstract:The influence of linguistic bias in online news articles is a growing concern, particularly in the context of shaping public opinion and rising political polarization. While there is a growing body of literature on indicators for misinformation, none have been sufficiently tested to counteract the influence of media bias. Hence, we design six indicators (Bias Bar, Bias Gauge, Bias Highlights, Political Scale, Sentiment Scale, and Trust Score) and test their impact on linguistic bias detection and perception in a two-phased experiment (n = 214). First, we expose participants to short, social-media-like statements along with one indicator and query bias perception. Second, we evaluate bias detection by removing the indicator and asking participants to mark biased words. In addition, we examine how trust, sharing discernment, and sentiment relate to bias perception and detection. Our results show that highlighting biased phrases and showing total bias with contextual information in a gauge significantly improve bias detection skills. However, the strongest predictor for reduced bias detection was political congruency between the statement and the participant. We conclude with design recommendations for linguistic media bias indicators in online news environments.
[HC-7] Are Attributions of Consciousness to AI Chatbots Epistemically Innocent?
链接: https://arxiv.org/abs/2607.20001
作者: Uwe Peters
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注: 1 Figure
Abstract:Artificial intelligence (AI) chatbots (e.g., ChatGPT) can communicate in strikingly humanlike ways. This has prompted many chatbot users to attribute psychological properties, including consciousness, to these systems. However, there is little scientific evidence that current AI chatbots are conscious. How, then, should we understand people’s consciousness attributions to chatbots? Are they merely metaphorical claims, or do they express genuine beliefs? If these attributions lack evidential support, are users epistemically blameworthy for making them, or might they be epistemically innocent, yielding significant benefits otherwise unattainable? This paper offers a conceptual analysis of consciousness attributions to AI chatbots and develops a multidimensional taxonomy of the attitudes they may express, ranging from non-doxastic stances (e.g., pretence) to different forms of belief, including delusions. This taxonomy helps avoid conflations by showing that linguistically identical attributions can reflect importantly different attitudes and degrees of epistemic commitment to the proposition that chatbots are conscious. The taxonomy also provides a framework for empirical studies to operationalize and measure different forms of epistemic commitment to AI consciousness. Using this taxonomy, I argue that although some consciousness attributions to chatbots are epistemically benign, and even some irrational ones may be epistemically innocent, many others render the attributor epistemically blameworthy.
[HC-8] INY_SCHILLER: A Drop-In German Drama Corpus for Small Language Models
链接: https://arxiv.org/abs/2607.19992
作者: Mark Schutera
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Digital Libraries (cs.DL); Human-Computer Interaction (cs.HC)
备注: 5 pages. Dataset: this https URL ; Code: this https URL
Abstract:tiny_schiller closes the small-language-model prototyping, fine-tuning, education, and research gap for German literary text, providing a single-file, drop-in counterpart to Karpathy’s tiny_shakespeare. The available German literary corpora are larger and richer, but require parser engineering before a single line of training or fine-tuning code can run. tiny_schiller is a 2.07-megabyte single file of eleven public-domain Schiller dramas, sourced from DraCor’s GerDraCor export (CC0) and processed by deterministic parser engineering. Character-level, GPT-2 byte-pair encoding, and cl100k_base tokenization splits, an instruction-formatted dialogue-completion split, and 89 per-character persona splits load from a single HuggingFace call. A small language model literally reaches German literary text in one line of code.
[HC-9] A Framework of User Experience Principles for Human-AI Agent Interaction in the Workplace
链接: https://arxiv.org/abs/2607.19941
作者: Kathrin Paimann,Elizangela Valarini,Sebastian Juhl
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注: 6 pages, 1 figure, 1 table, to be published in the proceedings of Mensch und Computer 2026
Abstract:As AI agents become integral to business workflows, establishing guiding user experience (UX) principles is crucial for ensuring user trust and successful adoption. To address this, our study uses a multi-method approach - combining participatory design workshop, paper-and-pencil, expert review, meta-analysis, and in-depth interviews - to identify and validate a design framework of eight core UX principles for human-AI agent interaction in the workplace. Together with their underlying criteria, these principles provide actionable guardrails for designers and software engineers, creating a foundation for developing effective and human-centered AI agent interactions. This study contributes to a structured foundation for future empirical studies on agentic AI in enterprise settings.
[HC-10] MV-Bench: Benchmarking Multimodal Large Language Models for Coordinated Multi-View Interface Construction IEEE-VIS2026
链接: https://arxiv.org/abs/2607.19910
作者: Yue Zhao,Hongxu Liu,Feiyu Wang,Xiaoyu Yang,Tong Ge,Zhen Yang,Chao Wang,Qiong Zeng
类目: Computer Vision and Pattern Recognition (cs.CV); Human-Computer Interaction (cs.HC)
备注: Submitted to IEEE VIS 2026
Abstract:Multimodal large language models (MLLMs) are increasingly expected to automate visualization development by generating code directly from visual designs. However, existing evaluations mainly focus on single-chart generation and overlook coordinated multi-view interface construction, which requires joint reasoning about data semantics, view coordination, and interaction logic. Consequently, MLLM capabilities in this setting remain underexplored, and the field lacks a dedicated benchmark for systematic assessment. We introduce MV-Bench, a benchmark for evaluating MLLMs on coordinated multi-view interface construction. Instead of relying on incomplete or inconsistent open-source implementations, we use Tableau workbook files as ground truth because they explicitly encode data bindings, visual mappings, and interactions. We develop a multi-stage pipeline that converts these specifications into executable web interfaces through structured intermediate representations. The benchmark contains 92 base interfaces and 1,048 verified instances created by recombining chart types, datasets, and interaction patterns. Each instance includes executable code, a rendered interface, a dataset, and interaction annotations. We evaluate five state-of-the-art MLLMs in a single-pass setting using metrics for visual fidelity, data binding correctness, and interaction completeness. The strongest model achieves 75.45 percent accuracy in visual layout reproduction, but only 21.71 percent in data binding and 11.68 percent in interaction completeness. These results show that current MLLMs can reproduce visual appearance but remain limited in generating the data semantics and interactive logic required by coordinated multi-view interfaces. Iterative refinement improves code executability but does not substantially reduce the gap in data binding and interaction generation.
[HC-11] argetFinder: Detecting Widgets from Pixels on Desktop Interfaces
链接: https://arxiv.org/abs/2607.19907
作者: Ahmed Ben Akouche(ISIR),Géry Casiez(LOOP),Mathieu Nancel(LOOP),Julien Gori(ISIR)
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:‘‘Target-aware’’ pointing techniques, like Bubble Cursor or Semantic Pointing, outperform traditional pointing by leveraging knowledge of target locations. Yet the lack of application-agnostic widget geometry information limits their adoption across the desktop. We present TargetFinder, a computer vision-based system for real-time detection of GUI widgets. TargetFinder leverages several fine-tuned YOLO networks trained on a new dataset of 520 annotated desktop screenshots (~38,000 annotations) spanning Windows, macOS, Ubuntu, and web interfaces. TargetFinder uses lightweight screen monitoring and low-latency detection, achieving millisecond responsiveness suitable for interactive use. Evaluations show that TargetFinder outperforms the baseline methods (OmniParser and REMAUI), while system-wide implementations of Bubble Cursor and Semantic Pointing demonstrate the feasibility of deploying universal target-aware techniques that work across applications. We release the dataset, models, annotation tool, and an open-source library for research and applications.
[HC-12] Mammal: Supporting Breastfeeding Monitoring Through Computational Garments with Inter-Body Sensing
链接: https://arxiv.org/abs/2607.19796
作者: Yanfeng Zhao,Morgan Geck,Kate Fernandez,Madison Nicole Jones,Xia Zhou,Jessica L. Ridgway,Te-Yen Wu
类目: Human-Computer Interaction (cs.HC)
备注: Submitted to IMWUT’26
Abstract:Breastfeeding provides critical insight into infant feeding competence and physiological health, yet objective monitoring remains difficult due to the intimate and internal nature of feeding. We present Mammal, a caregiver-worn computational garment that unobtrusively monitors breastfeeding without attaching sensors to the infant. Mammal leverages inter-body signal transmission through natural mouth-to-breast contact to capture infant cardiac and feeding-related acoustic signals on the caregiver’s body. Using novel algorithms to detect latch onset, infer infant electrocardiogram (ECG), and identify suck and swallow events from inter-body signals, Mammal estimates latch duration, in-feeding heart rate, suck-swallow-breathe (SSB) ratio, and milk intake. In a user study with 10 caregiver-infant dyads, Mammal achieves a mean absolute percentage error (MAPE) of 5.56% for latch duration, a mean absolute error (MAE) of 3.61 bpm for infant heart rate estimation, a mean absolute error of 0.12 for SSB ratio estimation, and a mean relative error of 15.76% for milk intake, with participants reporting high comfort and wearability.
[HC-13] AI-Increased Talent Retention Strategies: Fostering Long-Term Employee Engagement and Development in Talent Management
链接: https://arxiv.org/abs/2607.19733
作者: Jay Barach
类目: Computers and Society (cs.CY); Human-Computer Interaction (cs.HC)
备注: 8 pages, 3 figures, 2 tables
Abstract:The integration of AI in Talent Management is a change in the way that organizations are designing their strategies for Talent Retention (TR), engagement, and future strategy. New and innovative tools such as predictive models, sentiment analysis, and personalized career planning have come up, and they offer better ways of addressing retention issues, workforce engagement, and, in general, sustainability. Through the application of predictive analytics, organizations can determine employees’ likelihood of leaving the organization, who is likely to leave, and when to act, thus minimizing the costs and time associated with the recruitment process and improving performance. Furthermore, AI solutions help the development of individualized learning plans that help to define employees’ professional goals and link them with the organization’s strategy to encourage the employees’ continuous growth. This paper explained how AI is impacting the retention process and how it can be used to decrease attrition rates, create a loyal workforce, and promote sustainable management of human and other resources. Furthermore, the author discusses the ethical issues, such as privacy and fairness of algorithms, that are involved in the implementation of AI systems. Thus, the above challenges can be solved by developing a sustainable and inclusive ecosystem that can help in the development of the future workforce. Based on a systematic review of literature, the study presents a framework that can help organizations improve their talent management practices with the help of AI to support long-term sustainable growth. It can be used to help industry professionals and decision-makers understand the new technological shifts that are occurring. Comments: 8 pages, 3 figures, 2 tables Subjects: Computers and Society (cs.CY); Human-Computer Interaction (cs.HC) Cite as: arXiv:2607.19733 [cs.CY] (or arXiv:2607.19733v1 [cs.CY] for this version) https://doi.org/10.48550/arXiv.2607.19733 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[HC-14] Associations Between Support-Seekers Cross-Community Interactions and Their Engagement with Received Comments in Online Health Communities
链接: https://arxiv.org/abs/2607.19655
作者: Shenghan Tan,Daiang Jia,Chunghiu Kong,Tianjian Liu,Zhenhui Peng
类目: Human-Computer Interaction (cs.HC)
备注: Accepted by ICWSM 2027
Abstract:Support-seeker’ active engagement with received comments, e.g., showing positive sentiment and willingness to improve in the replies, can indicate the success of online health communities (OHCs). Their participation in other communities may correlate with their engagement in OHCs but remains under-explored. This paper analyzes 26, 725 seekers’ behaviors in the other 40, 479 communities and their associations with seekers’ engagement with received comments under their 78, 501 posts in 30 Baidu Tieba OHCs. We found that seekers primarily posted in other communities that are also health-related (25.3%), followed by those about games and entertainment (e.g., Dota, 20.8%). Seekers who posted in other communities about health (26.3%) or personal issues (e.g., saving money, 20.7%) before had relatively higher probabilities of subsequently posting in the 30 OHCs we identified, but this posting experience was associated with fewer replies and less expressed willingness to improve based on received comments. We provide insights into fostering seekers’ engagement in OHCs based on cross-community interactions.
[HC-15] How Far Can Wearable-Compatible Signals Go? A Controlled Decomposition of Non-EEG Sleep Staging
链接: https://arxiv.org/abs/2607.19441
作者: Yi Wang
类目: Human-Computer Interaction (cs.HC)
备注: 9 pages, 5 figures, 6 tables
Abstract:Consumer wearables increasingly infer sleep stages from signals including heart rate, accelerometry, and photoplethysmography. However, existing studies often report end-to-end performance under a fixed signal setting, making it difficult to determine whether the observed performance comes from genuine physiological decoding, temporal priors, or dataset-specific confounds. To address this limitation, we introduce a four-layer controlled decomposition framework for non-EEG sleep staging, covering signal source, physiological representation, temporal prior, and decision layers. The framework is evaluated across a signal-quality ladder spanning Apple Watch Sleep-Accel ( N=31 ), the Sleep Heart Health Study ( N=195 , laboratory ECG, respiratory, and SpO _2 signals), and Sleep-EDF-20 as an EEG+EOG reference, using the same compact Mamba2 model throughout. Laboratory cardiorespiratory signals reach \kappa=0.492 , while EEG+EOG reaches \kappa=0.796 , leaving a residual gap of \Delta\kappa=+0.304 that reflects missing cortical information rather than temporal modeling alone. Consumer HR/ACC reaches only \kappa=0.255 , quantifying the additional penalty of derived wearable signals and real-world sensing constraints. Confidence-based abstention provides a calibrated operating mode: removing the 20% lowest-confidence epochs increases \kappa from 0.452 to 0.512 , while a label-shuffled control collapses to \kappa=-0.003 . These results support non-EEG sleep staging as coarse, confidence-aware sleep-structure monitoring rather than EEG-equivalent five-class clinical staging.
[HC-16] Validating the Single Item Kawaii Measure
链接: https://arxiv.org/abs/2607.19352
作者: Katie Seaborn,Yijia Wang
类目: Human-Computer Interaction (cs.HC); Computers and Society (cs.CY); Multimedia (cs.MM); Sound (cs.SD)
备注: Accepted at CUI '26 (Short Paper)
Abstract:Kawaii is the Japanese instantiation of cuteness. As a multimodal percept theoretically derived from the notion of baby schema, kawaii can be a property of voice and sound, visual appearance and form factor, and movement and expression. However, measuring user perceptions of kawaii remains an open question. In the absence of a validated instrument, a one-item self-report measure has been used extensively, but has not been validated. Here, we report on three types of validity – convergent, known groups, and cross-context – and reliability for the single item measure across nine data sets featuring responses to video game character voices and visual appearances and computer-generated voice assistant voices from N=967 unique participants. Our results demonstrate initial evidence of the validity of the one-item measure for voice and visual kawaii perceptions. Further rigour can be pursued with novel stimuli, test-retest validation, and concurrent validity against the upcoming multi-item measure of kawaii.
[HC-17] A Scalability Analysis of Quantitative Confidence Assessment Methods for Assurance Cases
链接: https://arxiv.org/abs/2606.15480
作者: Simon Diemert,Jens H. Weber
类目: oftware Engineering (cs.SE); Human-Computer Interaction (cs.HC); Systems and Control (eess.SY)
备注: Preprint. Version of Record to appear in SafeCOMP’26 Workshop Proceedings published by Springer
Abstract:This paper proposes a model to estimate the decision complexity and effort required to apply quantitative confidence assessment methods to assurance cases. The model considers both the worst and average case for these measures and characterizes how these quantities scale with argument size. Prior work has indicated that the additional effort required to apply these methods is a barrier to their adoption by assurance case practitioners. Researchers developing new methods, or improving existing methods, can use this model to estimate the effort required to apply their method. The proposed model is parameterized using data from published case studies and is applied to three existing quantitative confidence assessment methods: the Bayesian Belief Network method, the Dempster-Shafer Theory method, and the Certus method. The results show that, while Certus has the highest worst-case decision complexity, its average-case effort is lower than the BBN and DST methods.
计算机视觉
[CV-0] ATSplat: Compact Feed-forward 3D Gaussian Splatting with Adaptive Token Expansion
链接: https://arxiv.org/abs/2607.20417
作者: Cho In,Jeonghwan Cho,Mijin Yoo,Gim Hee Lee,Seon Joo Kim
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:3D Gaussian Splatting (3DGS) achieves high-quality novel-view synthesis by optimizing freely placed primitives in 3D and adaptively densifying them in under-reconstructed regions. However, this scene-adaptive capacity allocation is largely lost in existing feed-forward 3DGS methods, which commonly regress Gaussians at input pixels and lift them along camera rays. Such pixel-aligned formulations make the number and placement of primitives depend on image resolution and input viewpoints rather than scene complexity, resulting in dense and often redundant Gaussian sets. We present ATSplat, a feed-forward 3DGS framework that restores the adaptive allocation capability of 3DGS optimization through Adaptive 3D Tokens. ATSplat first lifts coarse patch-level depth and camera cues into sparse 3D anchor tokens, forming a compact scaffold of the scene. Each token is then regressed into local Gaussians with learnable 3D offsets, decoupling primitive placement from input image grids. An Adaptive Token Expansion module predicts a token-level uncertainty score, supervised by rendering error maps, and selectively expands high-uncertainty tokens through learnable expansion layers. This sparse-to-adaptive formulation enables ATSplat to concentrate primitives in challenging regions while maintaining a compact representation. Experiments on two representative datasets, RealEstate10K and DL3DV, show that ATSplat achieves state-of-the-art rendering quality while reducing the number of Gaussians by more than 5.7\times compared with dense feed-forward 3DGS methods. From 12 input images at 512 \times 960 resolution, ATSplat completes reconstruction in less than a second using a single commercial GPU, and renders high-quality novel views at 1136 FPS ( 512 \times 960 ) with only 311K Gaussians.
[CV-1] PercepCap: Video Captioner with Structured Spatio-Temporal Perception
链接: https://arxiv.org/abs/2607.20389
作者: Yifan Xu,Zihao Wang,Zhixiao Wang,Jiaming Zhang,Yichun Yang,Desen Meng,Yuanxing Zhang,Pengfei Wan,Limin Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Video captioning requires fine-grained spatio-temporal understanding of videos, including spatial perception of where objects are located and temporal perception of when events occur. Existing MLLMs usually generate captions directly from video inputs without exposing the perceptual evidence behind descriptions. As a result, mistakes in spatiotemporal perception are only observed in the final caption, making it difficult to identify the underlying perceptual errors directly. To address these issues, we present PercepCap, a perception-aware video captioning framework that makes perceptual evidence explicit before producing the final caption. Specifically, PercepCap follows a perceive-describe generation chain, where the model first produces a spatiotemporal perception trace comprising object trajectories and temporal events, and then generates the final caption conditioned on the perceived evidence. To support this, we design a two-stage training strategy. Perceive-then-Describe Supervised Fine-tuning adapts the model from caption-only generation to the proposed perceive-describe chain, while Perception-Grounded Reinforcement Learning optimizes perception trace and caption quality with joint rewards over perception chain and the final caption. To support our two-stage training, we introduce Caption-Anchored Perception Data Construction. This pipeline builds the SFT and RL training data by first generating a caption-only description, extracting the objects and events it mentions, and grounding them back in the video with boxes and timestamps. This yields caption-aligned perception data that provides solid training ground truth, ensuring that the explicit perception trace and final caption refer to the same objects and events. Across direct caption and caption-to-QA evaluation, PercepCap consistently improves upon the Qwen3-VL baseline and demonstrates leading caption quality.
[CV-2] Persian Pixel: A large-scale synthetic OCR dataset for Persian language
链接: https://arxiv.org/abs/2607.20385
作者: Pouria Mahdi,Haq Nawaz Malik
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Optical Character Recognition (OCR) for Persian remains substantially less mature than for Latin-script languages despite Persian being spoken by more than 110 million people across multiple countries. This gap arises from two fundamental challenges: the intrinsic complexity of the Perso-Arabic writing system and the limited availability of large-scale, high-quality annotated datasets. Persian script exhibits obligatory cursive connectivity, context-dependent glyph shaping, extensive ligatures, diacritic placement, and stylistic variation across writing forms such as Naskh and Nastaliq, all of which significantly complicate text recognition. At the same time, the high cost and labor-intensive nature of manual annotation have created a persistent data bottleneck, limiting the development of robust OCR systems and slowing progress in Persian document this http URL this paper, we introduce Persian Pixel, a comprehensive synthetic OCR dataset specifically designed to address these challenges. Comprising over 343,000 high-fidelity image text pairs, the dataset spans sentence, paragraph, and full-page document layouts generated from a carefully curated seven-million-word Persian corpus using the SynthOCR-Gen rendering framework. The generation pipeline faithfully models the typographic characteristics of Persian script, including contextual character joining, positional glyph variants, diacritic placement, and multiple representative Persian typefaces. To bridge the synthetic-to-real domain gap, the rendered images are further enriched with more than twenty-five stochastic degradation models that emulate realistic document acquisition artifacts, including ink bleed, paper aging, blur, illumination variation, scanner imperfections, compression artifacts, and multiple noise this http URL overcoming the long-standing scarcity of annotated Persian OCR data, Persian Pixel provides a scalable and openly available resource for training and fine-tuning modern OCR architectures, including transformer-based models such as TrOCR and Donut. The dataset establishes a strong foundation for research in Persian document analysis, historical manuscript digitization, and end-to-end document understanding, while demonstrating that programmatic synthetic data generation offers a practical, cost-effective, and scalable alternative to manual annotation for advancing OCR in low-resource and typographically complex scripts.
[CV-3] Self Gradient Forcing: Native Long Video Extrapolation
链接: https://arxiv.org/abs/2607.20368
作者: Junhao Zhuang,Shiyi Zhang,Yuxuan Bian,Yaowei Li,Yawen Luo,Yijun Liu,Weiyang Jin,Songchun Zhang,Xianglong He,Xuying Zhang,Haoran Li,Haoyang Huang,Zeyue Xue,Nan Duan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Project page: this https URL
Abstract:Recent autoregressive video diffusion methods are increasingly built upon Self Forcing, where the student is trained on histories produced by its own rollout rather than ground-truth video contexts. This reduces exposure bias, but the historical key-value cache is still used by future frames only as frozen rollout state. As a result, future losses cannot supervise how earlier generated latents should be written into more useful keys and values for later video-latent generation. We call this the historical context-gradient gap. We propose Self Gradient Forcing (SGF), a two-pass training strategy that restores this missing supervision signal without backpropagating through the full serial rollout. Pass 1 performs a no-gradient autoregressive rollout matching inference and, at a sampled denoising exit step, records both the self-generated context and the noisy latents fed to the model. Pass 2 performs parallel context-gradient reconstruction for the recorded exit step. The generated context is used as stop-gradient clean-latent input, while the model recomputes the context KV representations and future-to-context causal attention. Thus, SGF provides the missing memory-writing supervision within the native autoregressive training objective, using losses on future video latents to train the model to encode context into more effective causal memory. Across extensive long-horizon frame-wise and chunk-wise experiments under different initializations, SGF achieves stronger native long-video extrapolation than Self Forcing, especially in subject identity, background/layout consistency, and temporal stability. Remarkably, using only a 5-second training window, SGF can extrapolate to videos lasting several minutes. Code and models will be released to advance research on autoregressive video generation.
[CV-4] Look Less Think Faster: Joint Token-Compute Adaptation for Multimodal LLM s ECCV2026
链接: https://arxiv.org/abs/2607.20357
作者: Pengcheng Wang,Zhiquan Wang,Jayoung Lee,Zhuoyan Xu,Ran Xu,Saurabh Bagchi,Yin Li,Somali Chaterji
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at ECCV 2026
Abstract:Multimodal Large Language Models (MLLMs) have recently demonstrated strong performance across vision-language tasks. However, their high inference cost, arising from both the large number of input visual tokens and the heavy computation of the large language model (LLM), remains a key barrier to practical deployment. Recent work attempts to reduce the cost by adaptively optimizing individual dimensions, e.g., pruning redundant visual tokens or skipping LLM layers and heads. Nonetheless, prior approaches typically treat these dimensions independently and overlook a fundamental coupling: the available compute resources must be dynamically allocated across all dimensions based on the input content. To bridge the gap, we propose SmartVL, a unified adaptive inference framework that jointly controls vision token number and model compute capability in response to varying input contents and compute budgets. SmartVL introduces a vision-side token controller that dynamically selects informative visual tokens and an LLM-side compute controller that adaptively adjusts LLM computation. Importantly, these controllers are trained to coordinate with each other so that the overall inference cost satisfies a target budget. To allow this joint scheduling, we connect the controllers using a shared budget encoding and leverage a differentiable latency estimator for end-to-end training. This design enables SmartVL to learn cross-stage allocation strategies that adapt to both input complexity and runtime compute constraints. Experiments across multiple MLLM benchmarks demonstrate that, with joint scheduling, SmartVL consistently outperforms prior adaptive methods and achieves superior accuracy-efficiency Pareto frontiers. Project page: this https URL.
[CV-5] oward Reliable RGB-D Semantic Segmentation: Handling Missing Modalities via Condition Dropout
链接: https://arxiv.org/abs/2607.20326
作者: Xuchen Zhu,Yajuan Wei,Shuang Hao,Jiwei Jiang,Guanxiang Mao,Fang Ren
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:RGB-D semantic segmentation has achieved remarkable progress, yet most models assume that RGB and depth are always available. In practice, failures or occlusions of surveillance sensors often remove one modality. Although RGB or depth alone can contain sufficient cues, models trained only on full-modality inputs fail to exploit the remaining modality once one is missing, causing severe degradation. We tackle this issue with a simple continued-training paradigm, \emphCondition Dropout (ConD), which mitigates degradation while preserving full-modality accuracy. Starting from a pretrained RGB-D model, ConD adds a second stage that randomly simulates complete, RGB-missing, and depth-missing inputs, freezes the original encoders, and trains copied encoders with zero-initialized feature injection. Experiments on NYU-Depth V2 and SUN RGB-D show that ConD improves robustness under missing modalities and even yields slight gains when modalities are complete. Our code will be made publicly available upon acceptance.
[CV-6] Evolving Cache Schedules for Fast Diffusion Policy Inference
链接: https://arxiv.org/abs/2607.20293
作者: Siying Wang,Kangye Ji,Di Wang,Fei Cheng
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 15 pages, 3 figures, supplementary material included. Accepted by PRCV 2026
Abstract:Diffusion policies achieve strong visuomotor control by iteratively denoising action chunks, but repeated denoising makes real-time deployment computationally demanding. Cache-based methods reduce inference cost by reusing intermediate activations, but existing training-free schedules typically allocate computation uniformly across blocks, ignoring heterogeneous redundancy across blocks and leading to a suboptimal performance-efficiency trade-off. To bridge this gap, we introduce Evolving Cache Schedules (EVO), a training-free acceleration framework that globally schedules cache refreshes via evolutionary search. EVO represents each candidate as a complete schedule over the block-timestep lattice. Thus, redundant transformer computations during iterative denoising can be skipped through cache reuse while preserving closed-loop rollout performance. To make the search practical, EVO introduces redundancy-aware initialization, which seeds the population with promising schedules, and target-conditioned early stopping, which verifies and terminates once a desired performance target is reached. The offline-optimized schedule can be directly plugged into pretrained diffusion policies without retraining. Extensive manipulation benchmarks show that EVO preserves near-full performance while substantially reducing computation, achieving up to 8.05x action-generation speedup and reducing FLOPs from 15.77G to as low as 1.96G. Source code is available at this https URL.
[CV-7] Diverse-Intent Multi-Turn Fashion Image Retrieval
链接: https://arxiv.org/abs/2607.20291
作者: Mingqiang Tang,Haokun Wen,Meng Liu,Yupeng Hu,Weili Guan,Xuemeng Song
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Real-world fashion search involves interactive retrieval across multiple turns. However, existing multi-turn retrieval methods are built on a restrictive assumption that every interaction follows the same attribute-editing paradigm, leaving heterogeneous intent transitions unexplored. Moreover, existing approaches often rely on textification to bridge multimodal queries and visual retrieval, which may lose fine-grained visual cues. To address these gaps, we introduce DIM-Fashion, a benchmark of 26K multi-turn sessions constructed from 13 fashion retrieval datasets across 7 tasks, featuring diverse intent transitions and rollback behaviors. We further propose FashionAM, an MLLM-VLP framework that directly aligns multimodal conversational queries with a fashion-oriented gallery embedding space, avoiding intermediate textification. Extensive experiments demonstrate the effectiveness of FashionAM over existing approaches. The dataset and code will be made publicly available upon acceptance.
[CV-8] Multimodal Large Language Models for Remote Sensing Image Understanding: Domain-Specific or General-Purpose?
链接: https://arxiv.org/abs/2607.20284
作者: Qiwei Ma,Chunping Qiu,Xinjun Cheng,Xiaoyu Zhang,Puhong Duan,Ke Yang,Xudong Kang,Shutao Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 27 pages, 11 figures
Abstract:The rapid development of multimodal large language models (MLLMs) has introduced a flexible paradigm for remote sensing image scene understanding (RSISU), enabling natural-language interaction with remote sensing imagery. However, a systematic understanding of the capability boundaries, cross-task generalization, and task-specific limitations of existing remote sensing MLLMs (RS-MLLMs) is still lacking. This paper presents a systematic survey and diagnostic evaluation of MLLMs for RSISU. We review the technical evolution of RS-MLLMs, focusing on model design, multimodal learning, training data, and downstream capabilities. We further compare RS-MLLMs with general-purpose computer vision MLLMs (CV-MLLMs) across diverse RSISU tasks and benchmarks. RS-MLLMs remain competitive in domain-specific settings, particularly remote sensing visual grounding and high-resolution visual question answering. More notably, general-purpose CV-MLLMs can match or even outperform these specialized models on several RSISU tasks without remote sensing-specific fine-tuning. These findings demonstrate the strong transferability of general-purpose CV-MLLMs and show that current RS-MLLMs do not consistently outperform them across diverse RSISU tasks. Current MLLMs also face limitations in spatial and relational reasoning, fine-grained visual understanding, instruction diversity, and generalization across heterogeneous task formats. Based on these findings, we outline future directions toward reliable evaluation, multimodal and high-resolution reasoning, efficient deployment, and tool-augmented remote sensing agents. This survey provides a systematic reference for developing robust, generalizable, and practical MLLMs for RSISU.
[CV-9] How Does Urban Context Relate to Residential Building Health? A Vision-POI Fusion Framework for Building-Level Housing Inspection
链接: https://arxiv.org/abs/2607.20263
作者: Kun Zhao,Helei Ren,Guilin Tang,Tianyi Chen,Zhehui Song,Xing Liu,Lijian Zhou,Yuhong Zhao,Xiang Gao,Jinming Jiang,Qichao Ban
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Housing-level urban physical examination is essential for identifying residential building problems and supporting targeted urban renewal. Existing automated inspection studies primarily rely on individual images and rarely examine whether surrounding urban functional context can provide supplementary information for building-level assessment. This study proposes a vision-POI fusion framework that combines multi-view visual inspection with POI-derived neighborhood context for residential building health assessment. The empirical dataset covers 92 old residential communities, 3,237 residential buildings, and 25,608 field-acquired inspection images in Qingdao, China, encompassing seven categories of housing-related issues. First, multiple object detection models are evaluated to extract issue locations, categories, and confidence scores from individual images. The image-level outputs are subsequently aggregated across multiple views to construct interpretable building-level representations. Second, POI features are extracted within 500m, 1,000m, and 1,500m neighborhood buffers to characterize surrounding functional environments. Pearson and Spearman correlation analyses, combined with false discovery rate correction, are used to identify candidate contextual features. Finally, visual and POI features are integrated using a cost-sensitive Random Forest classifier under community-isolated spatial cross-validation. The results show that multi-view aggregation provides the main performance improvement, increasing the building-level Macro-F1 from 60.84% under Direct Detection to 74.95%. Incorporating POI context further increases Macro-F1 to 76.79%, although the additional gain is modest and category-dependent. POI information therefore functions as a supplementary contextual prior rather than a substitute for direct visual evidence or a causal determinant of building condition.
[CV-10] Vera: Identity-Faithful Human Subject-to-Video Generation
链接: https://arxiv.org/abs/2607.20247
作者: Yulong Xu,Xinyue Liu,Shujuan Li,huafeng shi,Yan Zhou,Jiwen Liu,Xintao Wang,Yu Shen Liu,Huaibo Huang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Subject-to-video (S2V) generation has made substantial progress in preserving reference subjects across diverse categories, yet generic subject consistency remains insufficient for human-centric generation. A video may appear globally consistent while identity-critical human details still drift across frames, poses, and interactions. This issue becomes more severe in multi-person scenarios, where incorrect identity-role binding leads to subject confusion, attribute swapping, and excessive copying of reference-specific appearance cues. We propose Vera, a unified human-centric S2V framework for single- and multi-person generation. We first construct a million-pair identity-aligned human image-video dataset through person-level cross-clip retrieval, providing explicit identity correspondence and diverse references. Built on this dataset, Vera introduces two complementary designs. Identity-Focal Masked Supervision (IFMS) strengthens identity-aware learning with spatially focused supervision while reducing interference from irrelevant artifacts. Reference-Aware Layer-wise Attention (RALA) regulates how video tokens interact with reference identity cues in the DiT backbone, preserving stable identity anchors and enhancing layer-aware identity readout. Extensive experiments demonstrate that Vera improves human identity consistency, multi-person subject binding, and motion naturalness, while reducing identity confusion and excessive reference-image copying.
[CV-11] Not All Patches are Equal: Sampling Matters for Visible-Infrared Pre-Training
链接: https://arxiv.org/abs/2607.20238
作者: Qiwei Ma,Bin Deng,Junjie Zhu,Qiangjuan Huang,Puhong Duan,Ke Yang,Xudong Kang,Shutao Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 13 pages, 11 figures,
Abstract:Visible-infrared (VIS-IR) alignment is a key pre-training task for robust multi-sensor perception. Most existing methods use uniform patch-wise contrastive learning, but this can be unreliable in VIS-IR data because imaging-physics differences make some spatially paired regions inherently less comparable, and aligning them with equal strength hinders representation learning and downstream transfer. In this paper, we revisit VIS-IR pre-training from a sampling perspective and propose Importance-Aware Sampling (IAS), which adjusts training emphasis based on patch reliability. Specifically, IAS (i) derives patch weights from infrared structural cues and uses them to reweight the contrastive objective; (ii) learns a soft importance mask with a lightweight sampler, optionally warm-started from the hand-crafted prior; and (iii) employs a patch curriculum learning strategy that gradually expands from high-reliability regions to harder patches. It is worth noting that IAS is plug-and-play and works with both patch-/correlation-level alignment (e.g., UNIV-style) and image-level contrastive baselines (e.g., ImageBind-style). Extensive experiments on multiple VIS-IR benchmarks demonstrate consistent improvements over strong baselines, including for IR semantic segmentation, IR object detection and VIS semantic segmentation and cross-modal retrieval task. Code will be released on this https URL.
[CV-12] RS-RIE-Bench: Benchmarking Reasoning -Guided Remote Sensing Image Editing
链接: https://arxiv.org/abs/2607.20197
作者: Zihan Qin,Boao Xu,Zhao Dong,Yingping Sun,Ziheng Jiao,Junying Wang,Hongwei Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Remote sensing image editing aims to modify remote sensing images according to natural language instructions while preserving geographic rules and sensor observation characteristics. Existing benchmarks mainly target natural images or general visual scenes, and thus may not fully capture the reasoning, regional control, and sensor-consistency abilities required in remote sensing editing. To fill this gap, we introduce RS-RIE-Bench, the first benchmark for reasoning-guided remote sensing image editing. RS-RIE-Bench organizes tasks into three categories: temporal reasoning, causal reasoning, and spatial reasoning. These categories capture temporal evolution, causal consequence, and spatial imaging consistency in remote sensing scenes. The evaluation protocol covers three dimensions: target region plausibility, non-target region preservation, and image quality consistency. We further demonstrate the feasibility of MLLM-based evaluation through cross-judge consistency analysis and stratified expert review. Systematic evaluation on eight open-source and closed-source image editing models shows that current models still have clear limitations in reasoning-guided remote sensing editing. Even the strongest model achieves only 24.28% overall accuracy under the strict joint-satisfaction criterion, while the mean relaxed joint-4 success rate across all eight models is 32.23%. Causal reasoning and spatial reasoning remain especially challenging, and several open-source models are close to zero in some categories. These results show that RS-RIE-Bench can effectively reveal the limitations of current models in geographic reasoning, regional control, and sensor-consistent generation. It also provides a standardized benchmark and a clear research direction for future remote sensing intelligent editing models.
[CV-13] PerceptDrive: Perception Prior World-Action Modeling with Adaptive Expert Routing for End-to-End Autonomous Driving
链接: https://arxiv.org/abs/2607.20175
作者: Yushan Liu,Tianxiong Lv,Bohua Wang,Hangqi Fan,Chenxu Zhao,He Zheng,Xuchang Zhong,Yifan Xie,Congyang Zhao,Zhihao Liao,Leigang Luo,Yang Cai,Xiao-Ping Zhang,Wenbo Ding
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Frozen perception foundation models encode rich geometric, semantic, and dynamic knowledge. Yet narrow conditioning interfaces may attenuate task-relevant cues, while static fusion cannot adjust expert contributions to each scene. We cast this challenge as the prior-to-plan transfer problem and introduce PerceptDrive, a perception prior world-action modeling framework with adaptive expert routing. PerceptDrive feeds teacher-distilled priors from a frozen, driving-adapted provider and dense observation latents from a frozen self-supervised video encoder into a trainable expert-routed world-action model. Expert-specific query branches process these signals, while a prior-retention objective anchors each branch to its prior. A router predicts soft gates from a shared scene representation and combines the expert conditions before trajectory generation. During training, privileged rule-based sub-metric estimates for branch-specific trajectory drafts provide soft-gate distillation targets. The predicted action-free future latent conditions a flow-matching actor. At inference, privileged components are absent; with one front-facing camera, PerceptDrive generates one trajectory per planning step without test-time scoring, reranking, or search. Experiments show that PerceptDrive achieves state-of-the-art performance with 90.4 PDMS on NAVSIM v1 and 90.2 EPDMS on NAVSIM v2, outperforming existing methods. Ablations confirm complementary gains from prior retention and scene-conditioned routing, alongside differential reliance on the three priors. These results demonstrate that preserving and adaptively routing perception priors improves direct planning without test-time candidate selection.
[CV-14] StreamHOI: Interaction-aware Temporal Memory Adaptation for Streaming HOI Video Generation
链接: https://arxiv.org/abs/2607.20174
作者: Zejing Rao,Haoxian Zhang,Xiaoqiang Liu,Yiping Meng,Guoxin Zhang,Pengfei Wan,Fan Tang,Tong-Yee Lee
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Code and models are available at this https URL
Abstract:Existing human–object interaction (HOI) video generation methods are largely limited to offline short-video generation with complex driving conditions, making them unsuitable for real-time interactive applications. We present \emphStreamHOI, a low-latency streaming framework for long-duration HOI video generation. Instead of converting heavily conditioned HOI pipelines into streaming systems, we study how an image-to-video streaming generator should organize historical memory to preserve interactions under bounded latency. We find that the standard sink-local memory design faces a trade-off in streaming HOI generation, and different transformer blocks show different historical-memory preferences for HOI regions and surrounding regions. To match memory composition with block behavior, StreamHOI performs offline HOI-aware block profiling and applies bias-guided memory-specialized training to adapt the generator to block-specific memory layouts. We further introduce a memory distance scaling module to strengthen long-range access to early interaction states. Extensive comparisons with both long-video baselines and recent HOI generation methods demonstrate that StreamHOI achieves strong interaction plausibility, object fidelity, human quality and efficiency, reaching 17.6 FPS with 0.75s first-chunk latency.
[CV-15] SHFormer: Dynamic Spectral Filtering Convolutional Neural Network and High-pass Kernel Generation Transformer for Adaptive MRI Reconstruction
链接: https://arxiv.org/abs/2607.20159
作者: Sriprabha Ramanarayanan,Rahul G. S.,Mohammad Al Fahim,Keerthi Ram,Ramesh Venkatesan,Mohanasankar Sivaprakasam
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Published in Neural Networks (Elsevier), Vol. 187, Article 107334, July 2025. DOI: https://doi.org/10.1016/j.neunet.2025.107334
Abstract:Attention Mechanism (AM) selectively focuses on essential information for imaging tasks and captures relationships between distant pixel neighborhoods to compute feature representations. Accelerated MRI reconstruction benefits from AM, as the imaging process involves Fourier domain measurements that influence image representation non-locally. However, AM-based models are more adept at capturing low-frequency information with limited capacity for high-frequency representations, restricting models to smooth reconstruction. Additionally, AM-based models need mode-specific retraining for multimodal MRI data, as their knowledge is restricted to local contextual variations that may be inadequate to capture transferable features across heterogeneous domains. To address these challenges, we propose a neuromodulation-based discriminative multi-spectral AM for scalable MRI reconstruction that can (i) propagate context-aware high-frequency details for high-quality reconstruction, and (ii) capture features reusable across deviated unseen domains in multimodal MRI. The proposed network consists of a spectral filtering CNN to capture mode-specific transferable features and a dynamic high-pass kernel generation transformer focusing on high-frequency details. We evaluate our model on comparative studies in supervised and self-supervised learning, diffusion model-based training, closed-set and open-set generalization under heterogeneous MRI data, and interpretation-based analysis. Our method offers scalable, high-quality reconstruction with best improvement margins of ~1 dB in PSNR and ~0.01 in SSIM under unseen scenarios. Code: this https URL
[CV-16] Real-Time EEG Cap Electrode Detection for Guided Point-of-Care Placement
链接: https://arxiv.org/abs/2607.20142
作者: William Lehn-Schiøler,Mads Sverker Nilsson,Nicki Skafte Detlefsen
类目: Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
备注: Preprint. 13 pages, 7 figures, 4 tables
Abstract:We present a two-stage vision system that detects EEG cap electrodes in a live webcam stream and validates their anatomical placement in real time. A single-class YOLO detector localises electrodes; a geometric stage assigns each detection to a named 10-20 role from facial landmarks. Evaluating under subject-disjoint leave-one-subject-out (LOSO) cross-validation across five subjects wearing the clinically-validated Small/Medium/Large caps, the detector attains mAP@.5 = 0.94 +/- 0.07 across five held-out folds (0.96 pooled). A dedicated leave-one-cap-out axis, holding out every frame of a cap regardless of subject, leaves Medium and Large mAP@.5 within 0.01 of LOSO (0.97, 0.97) while Small drops to 0.72 +/- 0.28, a gap confounded with subject familiarity rather than cap style. Geometric augmentation (rotation, perspective, mixup) improves in-plane-roll robustness and temporal-electrode recall at no inference cost, and a landmark-driven head crop extends the usable distance range, lifting mAP@.5 from 0.23 to 0.45 at 0.6 x apparent scale. A compact mobile-candidate backbone (YOLOv10n) keeps the detector at real-time throughput (19 FPS) on a commodity CPU at 640 px.
[CV-17] HeadCast: Casting Attention Heads for Efficient Autoregressive Video Generation
链接: https://arxiv.org/abs/2607.20125
作者: Jinliang Shen,Lianghao Su,Zheming Li,Kang He,ZiLiang Lai,Yanbing Jiang,Chengru Song
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Autoregressive (AR) video diffusion models have become a promising paradigm for long and streaming video synthesis, but the continuously growing Key-Value (KV) cache makes attention the dominant inference cost, especially at high resolution where each frame contributes many tokens. Existing remedies either evict the cache with coarse heuristics that cause inter-frame flickering, or require model re-training. We propose HeadCast, a training-free, plug-and-play acceleration framework built on the observation that a pre-trained AR model’s attention heads exhibit stable, heterogeneous behaviors. After a short warm-up, HeadCast performs a one-time classification at the maximum-noise step that sorts every head into one of four archetypes: Sink, Dummy, Spatial, and Global, and restructures the monolithic KV cache into head-specific pathways. Crucially, it retains the Global heads that preserve the long-range temporal consistency aggressive eviction destroys. Because the Spatial pathway operates on a fixed-size grid, its savings grow with resolution: across state-of-the-art AR models, HeadCast accelerates inference by up to 1.62x at 720P and 1.95x at 1080P, while keeping VBench quality on par with full attention and largely flicker-free. Code is available at this https URL .
[CV-18] RIM: A Retrieval-In-Matching Framework for Cross-Domain Global Visual Localization of UAVs
链接: https://arxiv.org/abs/2607.20116
作者: Xin Li,Siyuan Duan,Shang Wang,Zhimin Mao,Bingliang Hu,Geng Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 56 pages, 10 figures, and 18 tables. Supplementary material is included
Abstract:Global visual localization of unmanned aerial vehicles (UAVs) using remote-sensing reference maps has attracted increasing attention. However, acquisition-time and imaging-platform differences between UAV and reference imagery induce substantial cross-domain appearance and viewpoint shifts, challenging robust six-degree-of-freedom (6-DoF) pose estimation. We address these shifts by sampling UAV-viewpoint reference views from Google 3D Tiles across locations, altitudes, and orientations. A two-stage cross-domain fine-tuning recipe adapts SALAD using pose-near positives and geographically distant hard negatives, while local geometric consistency re-ranks the Top-K candidates. We further propose Retrieval-In-Matching (RIM), which freezes the adapted DINOv2-B retriever and distils a local-descriptor decoder that reuses its token field alongside a shallow VGG19 detail stream. One query-side DINOv2-B forward thus serves both SALAD retrieval and local description, eliminating a second foundation-model backbone while preserving retrieval descriptors by construction. We evaluate RIM zero-shot on the reconstructed EPFL Urbanscape and self-collected Chang’an Park datasets, both geographically disjoint from the training data. RIM outperforms ten recent retrieval baseline families. At 25/50 m under the full 3D distance metric, it improves Recall@1 over SALAD by 8.55/13.77 percentage points on EPFL and 4.45/8.94 points on Park. At Top-K=5, the complete measured localization query, including retrieval, candidate matching, and robust geometric verification, takes 67.9 ms end-to-end: 1.8 times faster than the strongest separate sparse-matching baseline and over 40 times faster than RoMa, while achieving comparable re-ranking accuracy. These results establish an efficient and deployable pipeline for UAV global visual localization in GNSS-challenged environments.
[CV-19] Development of an automated reliable and clinically meaningful artificial intelligence (AI) tool for diagnosing cardiac disease from conventional cardiovascular magnetic resonance (CMR) images
链接: https://arxiv.org/abs/2607.20087
作者: Sina Amirrajab,Volker Vehof,Michael Bietenbeck,Nuriye Akyol,Redouane Bouras,Khuraman Isgandarova,Alexandru Zlibut,Philipp Stalling,Ali Yilmaz
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Aims: Cardiovascular magnetic resonance (CMR) imaging enables non-invasive assessment of myocardial structure, function, and pathology, but requires substantial experience in interpretation of CMR images that could be supported by artificial intelligence (AI)-based models. However, use of AI models for enhanced CMR reading is limited by labor-intensive data curation, suboptimal model performance, and unclear implementation pathways. Methods and results: We developed an automated data curation pipeline for CMR-based cardiovascular disease (CVD) diagnosis, integrating open-source locally-run large language models (LLMs) to extract diagnostic labels from narrative CMR reports and preprocessing multimodal imaging data, including cine and late-gadolinium-enhancement (LGE) CMR sequences. Three vision foundation models (DINO, VST, UMedPT) were fine-tuned across these modalities in a two-stage approach. The dataset comprised hypertrophic cardiomyopathy (HCM), dilated cardiomyopathy (DCM), ischemic cardiomyopathy (ICM), cardiac amyloidosis (CA), and normal controls (NOR). A total of 988 curated cases were randomly divided into 742 for training and 246 for validation. Fine-tuned AI-models achieved high discriminative diagnostic performance on an independent test set comprising 1067 patients , with individual AUC-ROC values of up to 0.937 for the correct diagnosis of HCM and 0.945 for cardiac amyloidosis. Ensemble strategies combining multiple models and modalities further improved AI-based diagnostic accuracy and robustness, achieving the highest overall diagnostic performance for HCM (AUC=0.959, CI [0.936-0.978]), CA (AUC=0.966, CI [0.939-0.986]), NOR (AUC=0.872, CI [0.852-0.894]), DCM (AUC=0.848, CI [0.808-0.885]) and ICM (AUC=0.840, CI [0.809-0.868]). All training and inference code, along with the trained model weights, are publicly available on this https URL.
[CV-20] GaussianSeed: Hierarchical Gaussian Seeding for High-Resolution 3D Occupancy Prediction
链接: https://arxiv.org/abs/2607.20071
作者: Xinzhuo Li,Xianghui Pan,Jiayuan Du,Wei Wei,Liuyi Wang,Chengju Liu,Qijun Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Vision-centric 3D occupancy prediction provides dense scene representations essential for autonomous driving and robotic navigation, yet existing methods struggle to scale to high voxel resolutions due to prohibitive computational costs. To address this, we introduce GaussianSeed, a progressive multi-scale Gaussian occupancy prediction framework that organizes primitives into a coarse-to-fine hierarchy. Benefiting from this hierarchical design, GaussianSeed effectively circumvents the memory bottlenecks inherent in dense representations, successfully scaling to a 0.1\textm spatial resolution while maintaining real-time inference capabilities. To comprehensively evaluate high-resolution geometric perception, we further construct TJScenes, a panoramic six-camera occupancy dataset with highly detailed 0.1\textm annotations. Extensive experiments on Occ3D-nuScenes and TJScenes demonstrate that GaussianSeed delivers the lowest latency among all evaluated methods while maintaining highly competitive accuracy, advancing the efficiency-quality frontier of high-resolution 3D occupancy prediction.
[CV-21] Importance-Aware OBS Pruning for Diffusion Models
链接: https://arxiv.org/abs/2607.20048
作者: Ba-Thinh Lam,Srijan Das,Hieu Le
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:We propose importance-aware pruning for diffusion models, a training-free framework that prioritizes preserving parameters critical to semantically salient image regions. To do so, we incorporate spatial importance maps – derived from conditioning signals or model attention – into the pruning objective. This produces parameter rankings aligned with perceptual relevance rather than uniform reconstruction error. On MS-COCO dataset, our proposed approach consistently retains subject fidelity and structural correctness at high compression ratios where conventional pruning causes visible degradation. These results demonstrate that content-aware objectives are key to perceptually faithful compression of generative models.
[CV-22] A Systematic Benchmark of Intensity Normalisation Methods for 3D Knee MRI Segmentation and Cross-Domain Generalisability
链接: https://arxiv.org/abs/2607.20028
作者: Oliver Mills,Philip Conaghan,Samuel Relton
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: This preprint has not undergone peer review or any post-submission improvements or corrections. The Version of Record of this contribution is published in 3Oth Annual Conference on Medical Image Understanding and Analysis, MIUA 2026. Code is available at this https URL
Abstract:Robust out-of-the-box performance is essential for the clinical deployment of deep learning models in medical imaging. An important but underexplored factor affecting model generalisability is intensity normalisation, particularly for magnetic resonance imaging (MRI), where image intensities vary across scanners and protocols. In this study, we systematically compared seven normalisation methods and their impact on the performance of a 3D U-Net model for meniscus segmentation from knee MRI. The methods included standard scaling approaches, histogram-based techniques, and a Gaussian Mixture Model (GMM)-based method. Models were trained on the IWOAI 2019 dataset and evaluated on both internal and external test sets (SKM-TEA) to assess generalisability. Performance was similar internally but differences were significant on external data, with Z-score, Nyúl histogram matching, and CLAHE showing greater robustness than other methods. However, these differences were small compared to the significant performance drop observed between datasets. Overall, while intensity normalisation had a measurable effect on model generalisability, its impact was limited relative to the effects of domain shift, highlighting the need for complementary strategies for robust deployment.
[CV-23] oward Seasonal Guidelines for Robust Deep-Learning Sentinel-2 Building Detection in Different Area Types
链接: https://arxiv.org/abs/2607.19994
作者: Michał Romaszewski,Kamil Drejer,Katarzyna Kołodziej,Anna Zawadzka,Stanisław Lewiński,Przemysław Głomb,Marek Ruciński,Michal Krupiński,Krzysztof Gryguc Przemysław Sekułaa,Szymon Sala
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Sentinel-2 imagery offers open access, global coverage, and frequent revisit times, making it attractive for practical building mapping at scale; however, its native 10m resolution makes building vs non-building classification challenging, particularly for small or sub-pixel buildings, and performance can vary with both seasonality and the heterogeneity of built-up environments. This paper introduces a Sentinel-2 building-detection framework designed to systematically quantify these effects and to support more formalised, practice-oriented model selection. We construct a dedicated multi-temporal Sentinel-2 dataset over the Warsaw region and derive binary ground-truth masks by rasterising official Polish topographic database (BDOT10k) building footprints onto the Sentinel-2 pixel grid. Using two established convolutional segmentation backbones (U-Net and DeepLabV3+), we first perform scene-specific fine-tuning to select a robust architecture and identify the best monthly models for L1C and L2A products separately. We then conduct cross-temporal inference by applying each best monthly model to all scenes, enabling an assessment of (i) which months provide favourable training and inference conditions, (ii) how performance transfers between seasons, (iii) the impact of processing level, and (iv) how these effects differ across built-up typologies. Based on these results, we provide practical guidance for routine Sentinel-2 building classification under varying acquisition periods and settlement characteristics.
[CV-24] STEREOFLOW: Progressive Stereo Matching with StereoDiT and Transition Flow Matching
链接: https://arxiv.org/abs/2607.19986
作者: Hao Wang,Haoran Geng,Xiaotong Yang,Jing Tang,Songlin Wei,Linlong Lang,Yeying Jin,Zheng Zhu,Zhaoxin Fan,Biao Leng
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 10 pages, 6 figures, submitted to TVCG
Abstract:Stereo matching is a fundamental task in 3D reconstruction. Despite remarkable advances, the prevailing paradigms formulate stereo matching as a deterministic regression problem, collapsing the multimodal distribution modeling into a single-point estimation. This formulation suffers from a regression-to-mean bias, frequently struggling with ambiguous regions. In contrast, we introduce a prior-guided generative framework that integrates deterministic matching regression and generative distribution modeling within a complementary formulation. Built upon this formulation, we introduce StereoFlow through three key components: (i) a two-stage progressive cascade matching network that progressively produces multi-resolution stereo conditions with complementary matching cues; (ii) a pixel diffusion transformer (termed StereoDiT) with a frequency-decoupled architecture for modeling correspondence ambiguity; (iii) a few-step flow matching objective (termed Transition Flow Matching) for efficient optimization. In summary, \textsc\textbfStereoFlow achieves strong geometric consistency and rich fine-grained details in ill-posed, discontinuous regions and under zero-shot generalization. Extensive experiments demonstrate that the proposed StereoFlow establishes multiple state-of-the-art results across benchmarks, including Scene Flow, KITTI, ETH3D, and Middlebury.
[CV-25] Forecasting the Number of Harvest-ready Fruits of Sweet Peppers Using Multimodal Time-Series Data
链接: https://arxiv.org/abs/2607.19975
作者: Enrico Pallotta,Mohamed Farag,Esra Guclu,Chris McCool,Ribana Roscher,Juergen Gall
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Accurate yield forecasting at the individual-plant level is critical for precision agriculture and supply-chain planning, yet public datasets capturing both visual growth dynamics and per-plant measurement labels are scarce. In this paper, we introduce a novel, annotated image time-series dataset of 691 sweet pepper plants monitored over two growing seasons, comprising 4837 images with per-plant fruit counts categorized by maturity. We propose a multimodal deep learning framework that fuses high-dimensional image features, extracted using the DinoV3 encoder, with numerical count measurements. Our architecture utilizes a Long Short-Term Memory (LSTM) network to model temporal dependencies and handles irregular sampling intervals common in greenhouse monitoring. Through quantitative experiments, we demonstrate that this multimodal approach reduces RMSE over a persistence baseline by 33% and 38% in the 2022 and 2023 seasons, respectively, with a further 1.2% average gain over a measurement-only model. Furthermore, we employ Deep Ensembles and Gaussian Negative Log-Likelihood (NLL) to provide calibrated uncertainty estimates, with an Uncertainty Calibration Error (UCE) ranging from 0.39 to 0.89 depending on the cross-season evaluation direction, offering a principled confidence signal for real-world agricultural decision-making. We release the dataset and code to support reproducible research and to accelerate development of data-driven yield forecasting methods for horticultural crops.
[CV-26] Unified Prediction and Planning via Conflict-Aware Disjoint Parameter Training ECCV2026
链接: https://arxiv.org/abs/2607.19971
作者: Taewon Seo,Seonae Jeon,Giwon Lee,Kuk-Jin Yoon,Daehee Park
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at ECCV 2026. 38 pages, 14 figures. Project page: this https URL
Abstract:Accurate motion prediction of surrounding agents and safe motion planning are two closely coupled key tasks for social robot navigation in crowded environments. Deploying these systems on resource-constrained edge devices necessitates compact, unified models that can perform both tasks simultaneously. However, within these compact shared encoders, recent unified models often overlook severe representational conflicts that arise from the distinct objectives of predicting neighbor behaviors versus ego-centric safety planning. To address this issue, we first identify the Skill Conflict \unicodex2014 a phenomenon where overlapping parameter assignments cause distinct tasks to compete for the same weights, preventing the model from fully specializing in individual skills. To resolve this, we propose a novel model-merging-based framework, Disjoint Parameter Training (DPT). DPT mitigates performance degradation caused by Skill Conflict through distributed parameter learning, which separates the key parameter regions of each task while preserving their core capabilities prior to merging. In addition, we observe that sparse merging, which selectively integrates only the most influential parameters for each task rather than combining all task-specific parameters, yields optimal performance by preventing interference among adjacent features and concentrating representational capacity. DPT can be applied in parallel with a variety of merging methods. Evaluated on standard crowd navigation benchmarks (JRDB and JTA), our framework demonstrates superior performance, validating its versatility and effectiveness for safe, resource-efficient robot navigation.
[CV-27] OffNadirLoc: Benchmark and Framework for Challenging UAV-to-Satellite Geo-Localization under Large Off-Nadir Views
链接: https://arxiv.org/abs/2607.19951
作者: Qian Qiao,Wenye Liu,Ting Liu,Jiuhe Shu,Peng Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Cross-view geo-localization between UAV and satellite imagery remains a fundamental yet highly challenging task, especially under large off-nadir views where drastic perspective distortions, occlusions, and appearance gaps occur. Existing benchmarks and methods primarily focus on near-nadir scenarios and often overlook the importance of structural scene understanding and intra-domain relational constraints, limiting their performance in real-world deployments. In this work, we introduce OffNadirLoc, a new benchmark for large off-nadir UAV-to-satellite geo-localization. To tackle the unique challenges posed by off-nadir perspectives, we further propose ONLoc, a framework that incorporates a structure-aware contextual weighting mechanism to dynamically emphasize reliable local features while suppressing ambiguous or repetitive regions. Additionally, we design a view-coherent learning strategy, which treats one satellite image and the corresponding UAV images from multiple views as a cohesive semantic group. This set-level supervision enables the model to learn viewpoint-invariant and discriminative features, making it more effective at capturing multi-view consistency than conventional pairwise contrastive learning. Extensive experiments on the OffNadirLoc benchmark and four near-nadir datasets demonstrate that our method consistently outperforms state-of-the-art approaches while exhibiting strong zero-shot generalization to unseen datasets without additional training. The code will be released at this https URL.
[CV-28] ETPDesigner: Multi-Agent Orchestration for Interactive Multimodal Electronic Theater Program
链接: https://arxiv.org/abs/2607.19947
作者: Mengtian Li,Xinru Guo,Xiaoru Lin,Xiao Rong,Zhifeng Xie,Chaofeng Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Electronic Theater Programs (ETPs) serve as critical promotional media in the performing arts, comprising a multi-page collection of heterogeneous visual assets such as theatrical posters, performance details, and character portraits. However, existing text-to-image paradigms struggle with such complex design tasks due to their inability to comprehend long-context narratives and maintain visual consistency across multiple distinct pages. To address this, we introduce ETPDesigner, a collaborative Multi-Agent framework that directly synthesizes high-quality ETPs from raw dramatic scripts. Emulating a professional design pipeline, our framework orchestrates specialized agents for semantic script analysis, core poster synthesis, functional background generation, and the stratified composition of character assets. Central to ETPDesigner is a global style anchor mechanism that extracts visual priors from the core poster to enforce strict aesthetic uniformity across all generated components. Furthermore, we elevate the ETP from a static publication to an immersive interactive companion. By integrating portrait animation, customized speech synthesis, and persona-grounded Large Language Models (LLMs), our system enables users to engage in real-time, voice-enabled conversations with the generated virtual characters. To rigorously benchmark this task, we construct ETP-Pro, a domain-specific benchmark of professional theater posters and high-quality character portraits. Extensive evaluations demonstrate our method’s superiority in producing semantically faithful, aesthetically consistent, and highly interactive program sets.
[CV-29] SIINR: Structurally Informed Implicit Neural Representations for super-resolution with uncertainty quantification of clinical quality diffusion MRI datasets
链接: https://arxiv.org/abs/2607.19943
作者: Tom Hendriks,William Consagra,Anna Vilanova,Yogesh Rathi,Maxime Chamberland
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 27 pages, 8 figures, 2 tables
Abstract:Diffusion Magnetic Resonance Imaging (dMRI) is a powerful tool for probing brain microstructure, but clinical acquisitions are often limited by low out-of-plane resolution, resulting in degraded structural information and reduced utility for advanced analysis. We introduce SIINR (Structurally Informed Implicit Neural Representations), a general framework for super-resoltion of clinical dMRI datasets while quantifying uncertainty in the reconstructed outputs. SIINR utilizes a supervised 3D U-net as a prior and combines it with a self-supervised implicit neural representation (INR) that fuses the high-resolution prior and the original low-resolution data. The INR enables joint modeling across spatial and angular domains, enforces data consistency, and provides analytic approximate posterior distributions for downstream uncertainty quantification. We validate the framework on a diverse set of open-access dMRI datasets, demonstrating that SIINR outperforms standard interpolation methods in both quantitative error metrics and qualitative anatomical fidelity. Experiments on clinical cases, including subjects with multiple sclerosis and brain lesions, illustrate the framework its ability to propagate intensity changes and flag uncertain regions in challenging scenarios. SIINR is flexible, modular, and can be adapted to different upsampling ratios and downstream tasks, providing a principled approach for enhancing clinical dMRI and supporting robust interpretation of derived neuroimaging metrics.
[CV-30] G-MAD: A Game-Based Data Generation Framework for Multi-View RGB-T Aerial Object Detection
链接: https://arxiv.org/abs/2607.19942
作者: Yechan Kim,JongHyun Park,Dongho Yoon,Namhoon Jung,Moongu Jeon
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: ACM Multimedia 2026 (OSS)
Abstract:This work introduces G-MAD, an open-source framework that uses Arma3 to generate synchronized multi-view RGB-T data for aerial object detection. G-MAD addresses key limitations of real-world aerial dataset construction, including limited viewpoint control, imperfect RGB-T alignment and high annotation cost. The framework supports structured scenario specification, controllable multi-view camera placement, simultaneous visible/thermal capture, and automatic bounding box annotation using engine-level geometric metadata. These capabilities enable controlled studies of viewpoint variation, multi-modal fusion, and synthetic-to-real transfer in aerial object detection. Besides, using G-MAD, we construct and release AMOD, a new large-scale multi-view aerial RGB-T object detection benchmark. The source code and the dataset are available at this https URL.
[CV-31] WearWow: Native 2K Multi-Garment Virtual Try-On via Adaptive Token Packing and Preference Alignment
链接: https://arxiv.org/abs/2607.19923
作者: Xujie Zhang,Runyan Du,Song Chang,Jiang Li,Dongliang Shao,Liping Wu,Wei Luo,Xiaochao Qu,Luoqi Liu,Xiaodan Liang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Synthesizing native 2K multi-garment virtual try-on is a formidable frontier in digital fashion, critically bottlenecked by two fundamental limitations: the O(N^2) memory explosion induced by 2k conditions, and the spectral bias of diffusion models that over-smooths high-frequency fabric details. We present WearWow, an end-to-end, mask-free generative framework that pioneers ultra-high-resolution multi-garment synthesis. To mitigate the memory explosion , we propose Adaptive 2D Token Packing (ATP). ATP leverages inherent garment sparsity to algorithmically pack heterogeneous items onto a unified 2D canvas and prune uninformative background tokens, minimizing the effective sequence length and subsequent memory overhead while rigorously preserving 2D spatial priors. To rectify texture degradation, we introduce the Multi-dimensional Try-on Reward (MTR) system. MTR synergizes a Semantic Guidance Reward to explicitly drive tactile restoration with a Cloth Distribution Reward to implicitly anchor the physical distribution, a joint formulation that effectively mitigates the severe reward hacking. Furthermore, we curate WearWow-2K, an extreme-quality dataset comprising native 2K triplets, providing physically correct spatial interactions that naturally empower the model’s mask-free generation. Extensive experiments demonstrate that WearWow establishes a new state-of-the-art, exceeding existing commercial baselines in native 2K multi-garment synthesis.
[CV-32] LoRFT: Benchmarking Long-Range Vehicle Trajectory Reconstruction from Fixed Highway Cameras
链接: https://arxiv.org/abs/2607.19911
作者: Yufan Zhu,Kefu Yi,Xueju Zhang,Yunyang Tian,Long Chen,Zixuan Xiao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 17 pages, 5 figures. Code and processed annotations: this https URL
Abstract:Long-range vehicle trajectories provide important spatio-temporal evidence for traffic safety analysis, autonomous driving evaluation, and data-driven traffic management, yet continuously recovering them from fixed highway cameras remains difficult. As vehicles recede into distant road regions, perspective compression and scale decay often fragment or prematurely terminate automatic tracklets, even when their continuation remains identifiable from motion consistency across neighboring frames. We formulate this problem as recovering the far-range continuation of a vehicle trajectory from a reliable near-field tracklet. We introduce LoRFT, to our knowledge the first open benchmark dedicated to long-range vehicle trajectory reconstruction from fixed highway cameras. LoRFT comprises 22 expressway surveillance scenes, 366,109 video frames, 6,601 manually verified trajectories, 2,694,889 bounding boxes, road-geometry annotations, scene-level splits, and evaluation scripts. We further propose Map-RSTNet, a map-aware residual sequence-to-sequence model that reconstructs distant trajectories in a road-geometry-aligned state space and dynamically refreshes local road geometry during decoding. On LoRFT, Map-RSTNet reduces ADE, FDE, and 5-second RMSE by 11.0%, 15.4%, and 10.5%, respectively, relative to the strongest baseline. These results demonstrate that road-geometry-aware reconstruction can extend usable trajectory records from existing fixed-camera infrastructure. LoRFT provides a reproducible testbed for long-range vehicle trajectory reconstruction.
[CV-33] StrokeSeg2: Stroke Lesion Segmentation in Clinical Research Workflows
链接: https://arxiv.org/abs/2607.19901
作者: Youwan Mahé(EMPENN, MALT),Axel Plessis(EMPENN),Stéphanie Leplaideur(EMPENN, MPR, CMRRF),Elise Bannier(EMPENN),Florent Leray(EMPENN, SED (Rennes)),Francesca Galassi(EMPENN)
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Deep learning frameworks like nnU-Net achieve state-of-theart brain lesion segmentation performance but remain difficult to deploy in clinical research environments due to, among other reasons, software dependencies and computational requirements. We introduce StrokeSeg2, a lightweight, modular, cross-platform C++/Qt framework designed to adapt resource-intensive 3D stroke segmentation pipelines into portable and reproducible applications. To improve compatibility with standard clinical workstations, we investigate the combined effect of architectural compression through knowledge distillation and inference optimisation using ONNX Runtime with Float16 quantisation. Across heterogeneous hardware configurations (CPU, integrated GPU, and dedicated GPU) architectural distillation emerged as the primary contributor to efficiency gains, contributing to over 90% reduction in energy consumption and an average 84% reduction in inference time. Specifically, we identify a 0.84M-parameter student model as the most favourable trade-off, reducing the original 102.3M-parameter teacher architecture to a 2.1 MB disk footprint while preserving robust lesion localisation and competitive segmentation performance. This small footprint supports the development of a self-contained installer for clinical workstation targets. Finally, StrokeSeg2 packages these optimisations into standalone installers for Windows, macOS, and Linux. By providing both graphical and commandline interfaces without Docker or external environment dependencies, StrokeSeg2 facilitates deployment of high-performance segmentation workflows for routine clinical research pipelines.
[CV-34] OSVE: One Step Video Editing with One Step Diffusion Models
链接: https://arxiv.org/abs/2607.19895
作者: Habin Lim,Gyeong-Moon Park
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Text-guided video editing with diffusion models is impractically slow, hindered by costly multi-step sampling and inversion. We present OSVE, the first framework to successfully adapt one-step Text-to-Image (T2I) models for high-quality video editing, addressing the core challenges of inversion, editability, and temporal consistency. To bypass slow iterative inversion, we train a learnable encoder that predicts the initial noise for each frame in a single forward pass. This encoder is trained with a novel Structure-Aware Editing (SAE) loss on a curated dataset of structurally-aligned image pairs, teaching it to preserve the source video’s geometry during edits. For temporal coherence, we introduce Unified-Frame Editing (UFE), a technique that concatenates frame latents to facilitate cross-frame attention in a single generation step. Furthermore, for long videos, a sliding-window strategy with an anchor frame maintains global consistency. Our extensive experiments demonstrate that OSVE achieves editing quality comparable or superior to state-of-the-art multi-step methods, while operating approximately 155–171 times faster. This breakthrough paves the way for practical, real-time video editing applications. Code is available at this https URL.
[CV-35] LAVIFT: Latent-Action-Guided Vision Fine-Tuning for Surgical Interaction Recognition
链接: https://arxiv.org/abs/2607.19889
作者: Jiajun Cheng,Subarna Tripathi,Sainan Liu,Xiaofan Yu,Shan Lin
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Understanding instrument-tissue interactions is essential for context-aware surgical AI and autonomous robotic surgery. Pretrained vision-language models (VLMs) and vision encoders offer an alternative to conventional interaction classifiers by transferring broad visual and semantic knowledge. However, adapting them to fine-grained surgical interactions remains challenging: (1) freezing the vision encoder depends entirely on pretrained representations that may retain noise and provide weak spatial localization, while (2) full fine-tuning can improve global semantic alignment without ensuring that the encoder learns meaningful features in the correct action region. We address these limitations by introducing LAViFiT, an end-to-end latent-action-guided framework for vision-language fine-tuning. An inverse dynamics model captures the visual changes induced by each action, while a forward world model drives the encoder to represent action-relevant regions. A patch-level SIG Regularizer further prevents local feature collapse without additional supervision, such as bounding boxes or pseudo-labels. Experiments across multiple encoders and datasets improve recognition and image-text alignment, while representation analyses show stronger grounding over the complete instrument-tissue interaction region and more spatially coherent features.
[CV-36] MTVDiff: Multimodal Conditional Latent Diffusion for Enhanced Thermal-to-Visible Face Translation ECCV2026
链接: https://arxiv.org/abs/2607.19886
作者: Zhiyuan Xia,Haojie Li,Jingyu Lin,Yiguo Qiao,Cunjian Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ECCV 2026
Abstract:Thermal-to-visible face translation presents fundamental challenges including geometric discontinuities, semantic attribute mismatches, and identity degradation. We propose MTVDiff, a novel multimodal latent diffusion framework that synergistically integrates depth and textual information to address these limitations while preserving identity characteristics. The MTVDiff framework presents three core technical contributions: (1) a Dual-Branch Cross-Attention Fusion (DBCAF) module for multi-scale thermal-depth feature extraction and fusion; (2) a Gated Text-to-Visual Feature Alignment mechanism for semantically-guided generation; and (3) Spatial Feature Transformations (SFT) for adaptive multimodal prior integration. Extensive experiments on the MCXFace and SpeakingFaces datasets demonstrate that our multimodal approach significantly outperforms existing GAN-based and diffusion-based approaches across multiple metrics, achieving substantial improvements in both image quality and face verification performance, with FID reductions of up to 48.3% and Rank-1 accuracy improvements of up to 8.9%. Our work provides a robust solution for face recognition systems operating under varying illumination conditions and advances the state-of-the-art in cross-spectral facial image translation through effective multimodal integration.
[CV-37] EA-Nav: Learning Safe Visual Navigation Policies with Embodiment Awareness
链接: https://arxiv.org/abs/2607.19880
作者: Jialu Zhang,Yong Du,Xianda Guo,Shunwang Sun,Xinqi Liu,Yue Sun,Guodong Lu,Wei Sui,Jituo Li
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Cross-embodiment navigation is a key challenge in embodied intelligence. Due to differences in embodiment, the same visual observation may imply different actions for different agents, making prediction ambiguous when relying solely on vision. Existing studies mainly rely on reinforcement learning, which requires large-scale interaction and careful reward design, making it difficult to support scalable pretraining and real-world adaptation. In contrast, imitation-learning-based approaches remain limited. To address these challenges, we propose an imitation-learning-based embodiment-aware navigation framework with a modular multi-stage design. In pretraining, we construct a cross-embodiment navigation dataset from Internet videos and introduce embodiment geometry as conditional tokens to reduce action ambiguity under the same observation. In fine-tuning, we design a multimodal information injection mechanism based on a decoupled architecture. Specifically, we design a trajectory augmentation strategy to generate high-risk samples, which are used to train spatial perception and risk-aware correction separately, thereby explicitly incorporating embodiment geometry for safe navigation. Experimental results show that the proposed method effectively improves navigation performance across different embodiment settings, demonstrating the effectiveness of incorporating embodiment geometry into embodied navigation.
[CV-38] Current Injection Spiking Neural Network for Infrared and Visible Image Fusion
链接: https://arxiv.org/abs/2607.19879
作者: Rui Zhao,Zhuoyuan Li,Wenrui Li,Yanchen Dong,Yajing Zheng,Giuseppe Valenzise,Weisi Lin
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Infrared and visible image fusion (IVIF) integrates the complementary information of two modalities into a single image with richer scene content. While existing methods are largely built on artificial neural networks (ANNs), which densely compute over all activations, spiking neural networks (SNNs) communicate through sparse binary spikes and compute only where and when a spike occurs, offering a route to more energy-efficient fusion. However, directly applying SNNs to IVIF creates a fundamental tension: cross-modal fusion relies on fine-grained responses from both modalities, whereas binary spikes can discard complementary cues that remain below the firing threshold. The membrane potential retains these subthreshold responses before firing, letting both modalities jointly shape the output when integrated at this stage. Building on this, we propose CIS-Fuse, a spiking network that performs cross-modal fusion directly at the membrane-potential level. At its core is the current injection spiking (CIS) operator, which injects one modality as a gated auxiliary current into the driving neuron of the other, so the two integrate before spike firing, with a per-channel learnable injection strength that adaptively regulates the modulation magnitude. Building on CIS, we construct a bidirectional cross-modal fusion (BCMF) module and deploy it on a dual-branch architecture with asymmetric stacking depths, where the two branches develop a clear functional specialization. Extensive experiments on four IVIF benchmarks and on downstream detection and segmentation show that CIS-Fuse achieves fusion quality on par with state-of-the-art ANN-based methods while inheriting the energy efficiency of spike-based computation, with roughly an order of magnitude lower inference energy than the similarly-sized ANN-based DCEvo. Code will be released upon publication.
[CV-39] Robust Activation Map Rectification for Weakly Supervised Volumetric Segmentation: Temporal Coherence as a Free Lunch
链接: https://arxiv.org/abs/2607.19877
作者: Renshu Gu,Jialiang Chen,Fei Gao,Hang Su,Jun Qi,Jiamin Xu,Yicheng Shen,Jiayu Zhang,Jiaxi Pan,Caiming Zhang,Gang Xu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Weakly supervised segmentation relies heavily on class activation maps (CAMs) to initially localize target regions. However, CAMs are often noisy and prone to catastrophic failures. Existing remedies typically introduce additional training stages or prototype learning, increasing computational cost and reducing robustness. In this paper, we propose a training-free prototype-free framework that rectifies unreliable CAMs by exploiting temporal and structural coherence in volumetric data as a free lunch. Our approach is built on two key components. First, we introduce Variance-Reduced Activation Aggregation (VRAA) which suppresses noise and amplify coherent semantic signals. We provide a theoretical justification by modeling CAMs as high-dimensional random vectors and show that aggregation yields provable variance reduction. Second, we design a Bidirectional Extremity Rectification (BER) mechanism that detects and rectifies implausible activations through bidirectional extremity checks, effectively mitigating extreme-value failures without learning additional parameters. Our method is model-agnostic and can be seamlessly integrated with existing pipelines. Extensive experiments on multiple public benchmarks demonstrate substantial improvements over state-of-the-art weakly supervised methods, achieving up to 20% Dice and 40% mIoU gains while reducing inference time by more than 5 times. These results indicate that leveraging coherence as an implicit inductive bias yields a principled and efficient approach to stabilizing weakly supervised volumetric segmentation. Our code will be available.
[CV-40] KineBench: Benchmarking Embodied World Models via IDM-Free Kinematic Grounding ECCV2026
链接: https://arxiv.org/abs/2607.19876
作者: Zeyu Liu,Zhangzhe Zhu,Yang Zhang,Chenyou Fan,Chenjia Bai,Xuelong Li
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注: Accept to ECCV2026
Abstract:Evaluating the physical consistency of embodied world models(EWMs) is a critical open challenge. While closed-loop evaluation via simulator rollouts offers a more faithful assessment of physical plausibility than open-loop alternatives, existing frameworks almost exclusively rely on Inverse Dynamics Models(IDMs) for action extraction. Due to the intricate mapping from 2D pixel space to 3D kinematic space, the learned IDMs can be brittle to data outside their training distribution, resulting in unreliable action extraction from the generated videos with novel objects and scenarios. This creates an unavoidable attribution ambiguity between world model inaccuracies and extractor errors. To reduce this ambiguity, we present KineBench, an IDM-free closed-loop benchmark for EWMs, built upon an explicit kinematic grounding pipeline. Given a generated video, KineBench employs cascaded visual foundation models to directly extract 6D end-effector poses from individual frames, which are then executed in a physics simulator for closed-loop validation. Beyond execution-based task success, KineBench incorporates two classical 3D kinematic metrics–Spectral Arc Length (SPARC) and the Maruyama Manipulability Index–to characterize trajectory smoothness and kinematic feasibility from a robot-centric perspective. Built on 20 diverse manipulation tasks in ManiSkill3, KineBench evaluates EWMs across four progressive suites: basic execution, task transfer, visual out-of-distribution generalization, and complexity-conditioned scaling. Evaluation across frontier models reveals task-complexity-bounded nonlinear scaling in embodied video generation, providing empirical guidance for future data-scaling strategies.
[CV-41] Memory-Augmented Multimodal Large Language Models for Small Object Understanding in Streaming Aerial Videos
链接: https://arxiv.org/abs/2607.19857
作者: Penglei Sun,Yehua Huang,Zhuoli Tao,Xiang Li,Runwei Guan,Yaoxian Song,Kaiyong Zhao,Henghui Ding,Bo Han,Yang Yang,Xiaowen Chu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Language-guided aerial perception aims to understand user-specified tiny targets in complex unmanned aerial vehicle (UAV) scenes. In real UAV deployment, the UAV must respond while it flies, so such perception runs in an online streaming manner, where frames arrive sequentially and the model responds to each one without access to future frames. However, applying current Multimodal Large Language Models (MLLMs) to this setting raises two challenges. First, targets viewed from the air are often tiny, yet the visual compression in existing MLLMs treats all regions equally and discards their fine-grained details. Second, understanding a continuous stream requires past-frame context, yet retaining the entire history is infeasible on resource-constrained onboard hardware, whereas discarding it causes the target to drift or disappear. We address the tiny object and streaming challenges from both data and method perspectives. From the data perspective, we present \textbfDroneEyes, the \textbffirst pixel-level and open-vocabulary referring-segmentation dataset for tiny aerial targets, comprising 2,140 high-definition videos and 176,623 pairs across Object Description and Referring Expression tasks, with dense per-frame masks. From the method perspective, we propose \textbfSkyAnchor, an MLLM with two designs to the above challenges: a Semantics-Aware Token Router that preserves small-target under a reduced visual-token budget, and a Hierarchical Memory Bank that keeps the target consistently understood on streams.
[CV-42] MoAKE: Toward Unified All-in-One Action Quality Assessment via Mixture of Action Knowledge Experts ECCV2026
链接: https://arxiv.org/abs/2607.19826
作者: Huangbiao Xu,Huanqi Wu,Xiao Ke,Jiaxin Cai,Junyi Wu,Jinglin Xu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: ECCV 2026
Abstract:Action Quality Assessment (AQA) aims to objectively evaluate performance quality from action videos. Most existing methods follow a ``one-by-one’’ paradigm, training a separate model for each action type. This setting limits real-world deployment, as it requires prior action-type knowledge to select the corresponding model and suffers from poor generalization across diverse actions. To address these limitations, we study the challenging task of all-in-one AQA, which aims to assess heterogeneous actions within a single unified model. We propose a novel Mixture of Action Knowledge Experts (MoAKE) framework, designed to mitigate negative knowledge transfer caused by large semantic discrepancies among actions. MoAKE learns complementary experts that capture diverse action patterns within a shared semantic space and dynamically aggregates their knowledge to adapt the assessment to the input action. Each expert is tailored with segment-aware prototypes to handle varying temporal lengths, together with an Adaptive Intra- and Inter-Segment Relationship Modeling (AIISRM) module to model multi-granularity temporal dynamics. Furthermore, we establish comprehensive benchmarks for all-in-one as well as zero/few-shot AQA. Extensive experiments on three long-term datasets demonstrate that MoAKE significantly outperforms existing methods in the all-in-one setting, while also achieving consistent generalization on three short-term datasets under zero/few-shot evaluation. Code is available at this https URL.
[CV-43] Lean-SAM2: Target-Anchored Memory and Encoder Acceleration for SAM2
链接: https://arxiv.org/abs/2607.19811
作者: Xudong Ouyang,Wenlun Zhang,Yimin Xu,Huazhong Liu,Yunshan Zhong
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:The Segment Anything Model 2 (SAM2) has advanced temporal promptable segmentation, yet its deployment remains hindered by heavy memory cross-attention overhead and redundant full-frame visual feature extraction. While recent methods explore efficiency via heuristic memory pruning and window-based sparse routing, they typically suffer from catastrophic performance degradation in complex segmentation scenarios replete with occlusions and distractors. To resolve these limitations, we propose \textbfLean-SAM2, a holistic lightweight framework designed to address the above vulnerabilities while systematically eliminating computational redundancies. Specifically, Lean-SAM2 integrates three collaborative mechanisms: (1) Target-Anchored Memory Pruning (TAMP) safeguards target tokens against deceptive attention by modulating raw attention significance with semantic consistency against prompt-derived foreground anchors; (2) Temporal Condensation with Insurance Memory (TCIM) condenses historical context via a visibility-gated fusion while conditionally archiving high-confidence entries in a parallel insurance bank; and (3) Target-Anchored Risk-Aware Routing (TARR) selectively activates the heavy image encoder for target-related windows based on anchor similarity, utilizing a risk-aware fallback policy to trigger full-frame refreshes during volatile transitions. Extensive evaluations across multiple challenging benchmarks demonstrate that Lean-SAM2 establishes a superior balance between accuracy and efficiency. For example, on the LVOSv2 validation dataset, Lean-SAM2 achieves overall inference speedups of 1.412\times and 1.417\times on the SAM2.1-Large and SAM2.1-Base+, respectively, significantly outperforming Efficient-SAM2 while boosting the corresponding \mathcalJ\mathcalF scores by 5.0% and 3.6% . Code is available at this https URL.
[CV-44] Silent Failures in Multimodal Agent ic Search:A Diagnostic Taxonomy and Cross-Judge Evaluation
链接: https://arxiv.org/abs/2607.19793
作者: Zhengxian Wu,Junjie Gao,Kai Yang
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Multimodal agentic search systems increasingly rely on external tools to answer knowledge-intensive visual questions. However, existing evaluations mainly focus on final-answer accuracy and may miss failures in the search trajectory. In this work, we study such hidden reliability issues as silent failures. We introduce a six-category taxonomy covering modality shortcuts, phantom grounding, wrong-evidence-right-answer cases, over-retrieval laundering, cross-modal contradiction, and provenance hallucination. Based on this taxonomy, we build a trajectory-level diagnostic pipeline that evaluates both answer correctness and evidence-grounding quality under a unified ReAct-style scaffold. Experiments on MMSearch-Plus trajectories across four frontier multimodal models show that surface accuracy consistently overestimates true trajectory-level correctness. We further use cross-judge validation, blank-image stress tests, and tool ablations to show that silent failures are capability-dependent and often shift rather than disappear. Home-page: this https URL
[CV-45] race: A Taxonomy-Guided Environment for Multidomain Visual Reasoning
链接: https://arxiv.org/abs/2607.19790
作者: Md Tanvirul Alam
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Reinforcement learning with verifiable rewards (RLVR) has substantially improved language-model reasoning, yet its extension to vision-language models remains constrained by the lack of training data that are simultaneously broad, exactly verifiable, and reproducible. We introduce Trace, a taxonomy-guided environment for multidomain visual reasoning. Trace factorizes task construction into a scene grammar and an executable task program, separating visual realization from answer computation. A shared semantic state determines the rendered image, prompt, typed answer, verifier state, and replayable instance trace. The resulting environment comprises 1,000 tasks over 277 scene grammars and 11 visual domains, with controlled semantic and visual variation. RLVR on 64,000 Trace instances improves the macro-average across 24 external benchmarks by 3.51 percentage points for Qwen2.5-VL-3B and 4.06 points for Qwen2.5-VL-7B, providing evidence that broad procedural training can transfer beyond the generated task distributions. Project page: this https URL.
[CV-46] Physics-Aware Complex-Valued State Space Model with Scattering-Prior Feature Modulation for PolSAR Image Classification
链接: https://arxiv.org/abs/2607.19787
作者: Fangyan Zhang,Fan Zhang,Shiqi Zhou,Jun Ni,Carlos López-Martínez,Qiang Yin
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 20pages, 14 figures, 10 tables
Abstract:Polarimetric synthetic aperture radar (PolSAR) image classification is a representative task for physics-aware GeoAI, where land-cover semantics are closely coupled with electromagnetic scattering mechanisms. Many existing complex-valued networks can preserve amplitude-phase information, but they are often limited in long-range spatial dependency modeling and usually incorporate polarimetric priors only as input-level or shallow auxiliary features. As a result, physical knowledge is insufficiently used to guide deep feature evolution. To address this issue, this paper proposes CV-SSMNet, a physics-aware complex-valued state-space network with scattering-aware feature modulation for PolSAR image classification. The proposed method builds a complex-valued state-space model (CV-SSM) in the original complex domain to capture long-range spatial dependencies while preserving polarimetric amplitude-phase coupling. Meanwhile, seven physically meaningful scattering priors, are encoded as FiLM-style modulation signals to adaptively recalibrate complex-valued representations during feature evolution. CV-SSMNet further integrates multi-scale complex convolutions, branch-wise CV-SSM encoding, prior-guided recalibration, and lightweight global context aggregation, enabling physically guided representation learning from local scattering structures to global spatial context. Experiments on three L-band benchmark datasets and an additional P-band BIOMASS evaluation demonstrate that CV-SSMNet achieves competitive accuracy, improved regional consistency, and better boundary preservation, supporting the effectiveness of embedding polarimetric scattering mechanisms into complex-valued long-range GeoAI representation learning.
[CV-47] WASABI: Whole-graph Assignment-based Stabilizer for lAne topology By Inter-frame tracking
链接: https://arxiv.org/abs/2607.19781
作者: Tetsuhiro Uchida,Myu Sasaki,Kensho Nakajima,Yasuhiro Shimada,Toru Saito
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Autonomous driving requires understanding the road as a graph of drivable lanes and their connectivity, beyond the ego lane alone, to follow routes through intersections and reason about cross- and merging-traffic. Recent perception models infer such lane topology, i.e., lane segments together with their inter-lane connectivity (LCLC), from onboard sensors over a 360-degree BEV view. Due to neural perception’s imperfections, their outputs retain structural instabilities such as missed detections, lost or incorrect LCLC, over-detection, and label flicker. This paper presents WASABI, a real-time post-processing pipeline that stabilizes lane topology outputs both within and across frames by treating lane segments and their LCLC connectivity as joint tracking targets, under onboard real-time constraints (10 Hz / 20 ms / up to 200 input lanes). The pipeline integrates segment tracking with connectivity, noise-robust topology-aware refinement, and a resource-constrained real-time design. On internal validation data (16 sequences), WASABI improves LCLC detection F1 from 0.834 to 0.948 (+0.114, +13.6%) and reduces centerline lateral error from 2.50 m to 0.95 m, while reducing detection false-positives by 24.6%. Temporal-stability metrics on the same data show LCLC toggle rate reduced by 63.3% and boundary-label flicker rate by 30.2%, confirming across-frame stabilization beyond per-frame accuracy.
[CV-48] Frequency-Hierarchical Active k-Space Sampling for Diagnostic MRI MICCAI2026 MICCAI
链接: https://arxiv.org/abs/2607.19779
作者: Ruru Xu,Kian Anvari Hamedani,Zhikai Yang,Ilkay Oksuz
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: EMA4MICCAI 2026: The 2nd MICCAI Workshop on Efficient Medical AI
Abstract:Active sampling for accelerated MRI must distribute a tight sampling budget across spatial frequencies that carry very different kinds of information. Low frequencies hold most of the anatomical context; high frequencies carry the fine details that drive pathology assessment. Existing active samplers either treat both regions identically or restrict the action space to entire Cartesian rows, which forces a poor compromise at high acceleration. We propose HieraSample, a task-driven framework built around this hierarchy. A cosine-annealed curriculum lowers the acceleration factor from 20x to 4x across 80 acquisition steps while keeping a fully-sampled low-frequency disk at every step; a Mamba-based policy then picks individual high-frequency coordinates from features extracted by dual disease and severity classifiers. The reward is the per-sample reduction in class-weighted cross-entropy after each action, so a positive reward corresponds directly to a more confident correct prediction. On the fastMRI+ knee benchmark, HieraSample matches the fully-sampled oracle on ACL diagnosis from 4x to 10x acceleration, and improves on a recent Cartesian baseline by as much as 20.4 AUC points on ACL severity.
[CV-49] Look Before You Edit: Attention-Guided Camera Placement and Multi-View Alignment for 3D Gaussian Splatting Editing
链接: https://arxiv.org/abs/2607.19777
作者: Jaeyeon Park,Taeho Kang,Youngki Lee
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Text-driven 3D scene editing with 3D Gaussian Splatting (3DGS) typically applies a 2D diffusion editor to views rendered from fixed training cameras, limiting both the spatial coverage of edits and the user’s freedom to target specific objects in complex scenes. We present LB-Edit, a framework that addresses two coupled problems: where to place editing cameras for localized edits, and how to make per-view edits agree with one another so that the 3D scene remains consistent after fine-tuning. First, Attention-Guided Editing Camera Placement (ACP) probes the diffusion model’s self- and cross-attention at multiple candidate camera distances to find where attention is well-contained in the region of interest, then places a compact, geometrically diverse editing camera set at that attention-optimal distance. Second, Multi-view Attention Alignment (MAA) steers the editor toward the same edit across views along two axes: it aligns appearance by sharing self-attention features via token-level correspondence, and aligns spatial location by lifting cross-attention maps onto the 3D Gaussians as a shared 3D attention field, suppressing both appearance and spatial drift. Experiments on multi-object and single-object scenes show that our method achieves the highest user preference in instruction fidelity, multi-view consistency, and editing locality, using as few as 5 editing views and reducing latency by up to 7x over existing methods.
[CV-50] DRGBT-1K: A Large-scale High-quality Benchmark for Dynamic RGBT Tracking
链接: https://arxiv.org/abs/2607.19772
作者: Zhaodong Ding,Chenglong Li,Zeyu Ding,Futian Wang,Jin Tang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Submitted to TIP (Under Review)
Abstract:Dynamic RGBT (DRGBT) tracking aims to continuously localize a target when the available sensing modalities and observation platforms vary over time. Compared with conventional RGBT tracking with fixed RGBT inputs and a fixed observation platform, DRGBT tracking is more consistent with real-world collaborative perception systems, where targets may be observed by heterogeneous sensors from different viewpoints. However, existing benchmarks are still insufficient for systematically evaluating tracker robustness under real dynamic modality variations and cross-platform transitions. To address this limitation, we make the following contributions. 1) We construct DRGBT-1K, a large-scale high-quality benchmark for DRGBT tracking. It contains 1,045 sequences captured entirely in real-world scenarios and 795K RGBT frame pairs collected using UAVs and handheld RGBT devices, encompassing diverse real-world scenes, pronounced viewpoint changes, modality variations, and target appearance discontinuities. 2) We provide comprehensive annotations for fine-grained evaluation, including dense bounding boxes, target category labels, challenge attributes, frame-level modality labels and platform labels. DRGBT-1K covers 24 target categories, more than 15 scene types and 15 challenge attributes. 3) We establish a comprehensive benchmark by evaluating 20 representative multimodal tracking methods, including conventional RGBT trackers, modality-missing RGBT trackers, and DRGBT trackers under a unified evaluation protocol. 4) We release an unaligned version of DRGBT-1K and derive UGVT-1K to support broader research on unaligned multimodal tracking and UAV-ground collaborative tracking. 5) We develop an online evaluation platform for DRGBT-1K and provide a leaderboard that collects all methods evaluated on this benchmark.
[CV-51] Global Building Area Estimation Products: How Accurate Are They?
链接: https://arxiv.org/abs/2607.19766
作者: Saad Lahrichi,Doa’a Allabadi,Kyle Bradbury,Jordan Malof
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 29 pages, 4 figures, 7 tables
Abstract:Geo-spatial rasters of building footprint area are useful for a variety of tasks, such as monitoring urbanization, improving energy efficiency, and tracking greenhouse gas emissions. There are now multiple global building raster datasets, however there lacks an independent, comprehensive, and fair assessment of their accuracy. In this work, we evaluate the accuracy of four major global building products: Global Human Settlement Layer (GHSL), Microsoft’s TEMPO (TEMPO), The Global Building Atlas (GBA), and Overture. As ground truth for assessing their accuracy, we use ORBITaL-Net, a globally diverse dataset of manually labeled building footprints. To ensure fairness, we evaluate products on grids of multiple spatial resolutions, and several conventional performance metrics. Our results indicate that either GBA or TEMPO generally achieves the highest overall accuracy, depending upon the particular evaluation criteria. We also stratify the accuracy of each product by several factors: geographic location, population density, and income groups. The results reveal that product accuracy can sometimes vary significantly with respect to these factors. Notably, all products are significantly less accurate in Africa and Asia. Most products also suffer significant accuracy reduction in high-density urban areas.
[CV-52] Extending a Large View Synthesis Model for Multi-view Panoptic Segmentation ECCV2026
链接: https://arxiv.org/abs/2607.19765
作者: Kwonyoung Ryu,In-Jae Lee,Jonghyun Jin,Hyunjee Lee,Jongmin Lee,Jaesik Park
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to ECCV 2026. 31 pages, 7 figures, 9 tables. Project page: this https URL
Abstract:Large view synthesis models synthesize novel views through cross-view attention without explicit 3D representations, and recent studies have shown that they learn accurate spatial correspondence from RGB supervision alone. We observe that this correspondence generalizes beyond appearance. When non-photorealistic signals such as binary encoded panoptic labels are passed through the model, they are propagated to novel views with consistent spatial structure. These results indicate that the correspondence learned for RGB view synthesis can also propagate view-independent per-pixel labels. From this observation, we present the first work to extend large view synthesis models beyond appearance rendering to 3D scene understanding. We propose a panoptic segmentation pipeline that reuses a frozen view synthesis model to propagate panoptic labels from input views to novel views, without 3D reconstruction or any segmentation-specific training of the view synthesis model. Given panoptic labels on the input views, we encode them into binary channel representations and pass them through the same model to render target-view segmentation. On ScanNet, our method achieves segmentation quality on par with Gaussian based approaches requiring explicit 3D reconstruction, while outperforming them in novel view synthesis by more than 7 dB. The label propagation also transfers across datasets, surpassing these approaches on Replica without any fine-tuning.
[CV-53] EgoRecovery: Acquiring Failure Recovery Ability Through Human Recovery Demonstration
链接: https://arxiv.org/abs/2607.19745
作者: Zuhao Ge,Yuchen Zhou,Weitao Zhou,Minglei Li,Xinyu Li,Chao Wu,Hanwen Zhao,Haotian Wang,Zuxuan Wu,Xiaosong Jia,Yu-Gang Jiang
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Robust embodied robots should be able to recover from failures and retry tasks in order to operate reliably in unstructured and noisy real-world environments. Achieving this capability requires training policies on data that captures recovery behaviors. However, collecting such data through robot teleoperation is difficult to scale, as it is time-consuming to induce diverse failure states, perform corrective actions, and reset the environment. This challenge is further exacerbated by the high diversity of failure modes, which demands substantially more recovery data than success demonstrations. In this work, we show that egocentric human data capturing failure recovery processes provides a scalable alternative. By efficiently arranging task-level failure configurations and recording short recovery segments, human operators can generate more than 10x as much valid recovery data per hour compared to robot teleoperation under our protocol. To address the embodiment gap between human and robot, we propose EgoRecovery, a co-training framework for learning recovery behavior, where human recovery demonstrations are aligned to a compact corrective-intent space shared with robot data, which captures the timing and magnitude of correction. Only a small number of robot recovery demonstrations are required to connect this intent to executable robot actions. At deployment, a learned recovery gate predicts when correction is needed from robot observations and activates the corrective intent only in recovery states. Experiments on real-world recovery tasks show that EgoRecovery improves success from failure starts over robot-only recovery, direct co-training with human recovery data, and direct intent-transfer baselines.
[CV-54] Efficient Tracking and Understanding Object Transformations
链接: https://arxiv.org/abs/2607.19743
作者: Yihong Sun,Bharath Hariharan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Tracking objects through state transformations is essential for understanding real-world dynamics. However, existing methods are computationally expensive. TubeletGraph recently showed impressive capabilities, but its inference cost (~ 4.4 seconds per object-frame on VOST) precludes any real-time deployment possibilities. We observe that TubeletGraph’s overhead arises from building a spatiotemporal partition of the input video: (1) entity segmentation is computed densely for every frame regardless of whether a transformation occurs, and (2) every entity in the scene is tracked, scaling cost with scene complexity rather than the number of transformations of interest. To address both, we propose FluxGraph, a reactive variant that uses SAM2’s internal multi-mask disagreement as a lightweight trigger for transformation detection, and removes the need for tracking all entities in the given video. FluxGraph is ~ 3.3\times faster than TubeletGraph on VOST while improving tracking performance and preserving state graph quality. Furthermore, we also observe consistent speedups of 3.7-10.7\times across VSCOS, M ^3 -VOS, and DAVIS17 while maintaining performance. Code is publicly available at this https URL.
[CV-55] An Exploratory Analysis of Pain Localization via Explainable Computational Modeling
链接: https://arxiv.org/abs/2607.19726
作者: Ioannis Kyprakis,Stefanos Gkikas,Eric Nichols,Yu Fang,Manolis Tsiknakis
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at the 14th International Conference on Affective Computing and Intelligent Interaction (ACII 2026)
Abstract:Automatic pain localization, which involves identifying the anatomical origin of pain from peripheral physiological signals without patient self-report, is a clinically critical but largely unaddressed problem, particularly for non-verbal patients. This paper presents a systematic comparison of classical feature engineering and deep sequence learning for subject-independent three-class pain localization using the AI4Pain 2026 Challenge dataset, which comprises four synchronously recorded wearable modalities: electrodermal activity, blood volume pulse, respiration, and peripheral oxygen saturation recorded from 65 participants under controlled TENS-induced pain. A 115-dimensional hand-crafted feature set spanning time-domain, frequency-domain, modality-specific, and cross-modal descriptors is benchmarked against end-to-end deep architectures. Extremely Randomized Trees achieves the highest macro-F1 of 0.539, outperforming the best deep model by 7.4 percentage points, with EDA spectral features emerging as the dominant discriminators. A consistent 26-point gap between pain detection (F1,=,0.815) and localization (F1,=,0.552) across all models points to a fundamental ceiling imposed by the anatomical diffuseness of peripheral autonomic pathways at 10-second resolution.
[CV-56] Analytic Distribution of Classifier-Free Guidance for Schedule Design
链接: https://arxiv.org/abs/2607.19725
作者: Enze Jiang,Zheng Ma
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Classifier-free guidance (CFG) is the default mechanism for conditional generation in diffusion models, but the distribution sampled by its deterministic guided dynamics is not captured by the usual product-distribution heuristic p_0^\omega q_0^1-\omega . We analyze CFG through the probability flow ODE and derive exact analytic path-integral representations of the induced distributions for both constant and time-dependent guidance. The resulting formulas show that CFG modifies p_t_0 by an exponential path-integral correction, and that a time-dependent schedule enters this correction through the weight \omega(t)-1 . This characterization explains how score discrepancies accumulate along sampling trajectories and motivates Distribution-Guided CFG (DG-CFG), a schedule that balances timestep contributions while accounting for signal strength and low-noise score-error amplification. A toy model with analytic scores closely verifies the predicted distributions. On Stable Diffusion~1.5, DG-CFG improves generation and yields a stronger diversity–fidelity trade-off across guidance strengths, with especially clear gains when strong guidance causes saturation and quality degradation in constant and heuristic schedules. Across NFE budgets, DG-CFG reaches fixed image-quality targets with fewer sampling steps, reducing the sampling cost needed to achieve target metrics.
[CV-57] ReFace: Reorganizing Facial Spatiotemporal Representations for Improved Pain Assessment
链接: https://arxiv.org/abs/2607.19722
作者: Stefanos Gkikas,Yu Fang,Christian Arzate Cruz,Muhammad Umar Khan,Raul Fernandez Rojas
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at the 14th International Conference on Affective Computing and Intelligent Interaction (ACII 2026)
Abstract:Automatic pain assessment from facial video remains challenging due to the spatial heterogeneity of pain-related facial cues. This study proposes ReFace, a spatial reorganization pipeline that divides facial input into four spatial quadrants before tokenization, rather than processing the entire face as a single region. Evaluated on the AI4Pain dataset, the proposed approach achieves 56.00% accuracy on the test set using video only, achieving the highest reported accuracy under the fixed AI4Pain benchmark protocol among the compared methods. Notably, the four-quadrant configuration processes the same total pixel budget as the full-face input, yet achieves higher accuracy, suggesting that spatial reorganization can improve performance under the proposed tokenization design. A single quadrant region, processing just one quarter of those pixels, remains competitive at a fraction of the computational cost.
[CV-58] A Unified Tokenization Framework for Pain Recognition using Heterogeneous 3D Modalities
链接: https://arxiv.org/abs/2607.19716
作者: Stefanos Gkikas,Christian Arzate Cruz,Valentina Becchetti,Muhammad Umar Khan,Alessandro Giuseppi,Raul Fernandez Rojas
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at the 28th ACM International Conference on Multimodal Interaction (ICMI 2026)
Abstract:Pain is a complex and pervasive phenomenon affecting a large percentage of the population, and accurate assessment is essential for effective clinical management and intervention. Computational pain recognition systems enable continuous monitoring, support clinical decision-making, and help mitigate pain-related distress and functional decline. This study introduces a unified tokenization framework for heterogeneous 3D modalities in pain recognition that provides a single processing pipeline across behavioral and brain-activity 3D data, without requiring separate architectures for each modality or handcrafted inductive biases. The framework preserves spatial, temporal, and time–frequency structure while mapping diverse inputs into a shared token space. Extensive experiments show that the proposed approach effectively processes facial videos and fNIRS data in both raw-signal and spectrogram-based representations. On the AI4Pain benchmark dataset, the proposed framework achieves state-of-the-art performance while maintaining high computational efficiency and enabling real-time assessment on both GPU and CPU hardware.
[CV-59] Point-Selection Fine-Tuning Framework for Robust Point Cloud Classification
链接: https://arxiv.org/abs/2607.19711
作者: Da Li,Chang Ma,Dongfu Yin
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Noisy and corrupted points can substantially degrade point cloud recognition performance, especially under challenging corruption settings. In particular, full fine-tuning of 3D pre-trained models may amplify the influence of outliers and overwrite robustness priors learned during pre-training, while naive parameter-efficient adaptation remains sensitive to corrupted tokens. To address this issue, we propose PSFT, a point-selection fine-tuning framework that improves robustness while remaining parameter-efficient. PSFT first estimates point-wise influence from pre-pooling features and adaptively retains minimally influential points to suppress outliers. Based on the selected subset, a prompt generation branch predicts layer-wise prompt tokens and injects them into a frozen backbone for lightweight downstream adaptation. To further mitigate residual noise after selection, we append a lightweight feature filter with bottleneck MLP transformation and Beta-gated residual blending to refine patch-token representations before prediction. Extensive experiments show that PSFT consistently reduces corruption error on ModelNet-C and ModelNet40-C across all tested 3D pre-trained backbones, while achieving the strongest ScanObjectNN-C results with ULIP-2 and Uni3D-B among the evaluated tuning strategies. Our implementation can be found at this https URL.
[CV-60] SafeGen: Goal-Conditioned Video Diffusion of Safety-Critical Scenarios for VLM-Based Autonomous Driving
链接: https://arxiv.org/abs/2607.19701
作者: Jiangfan Liu,Zexuan Cui,Tianyuan Zhang,Zonglei Jing,Zonghao Ying,Yaoyuan Zhang,Jiakai Wang,Xiaoqi Jiang,Aishan Liu,Xianglong Liu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:VLMs are increasingly deployed in AD systems, creating an urgent need for rigorous safety evaluation under rare yet safety-critical scenarios. Among these, interactions with vulnerable road users represent a major source of real-world failures. However, existing safety-critical scenario generation methods predominantly rely on simulator-based pipelines, which suffer from a substantial sim-to-real gap and often fail to capture realistic, diverse, and unforeseen human-vehicle interaction dynamics. We present SafeGen, a goal-conditioned diffusion framework for safety-critical scenario generation in VLMADs. Our key insight is to formulate scenario generation as a goal-conditioned diffusion process, where a predefined catastrophic end-state serves as a strong supervisory signal, guiding the generation of temporally coherent video trajectories that naturally evolve toward safety-critical outcomes. Building on this formulation, we introduce Context Grounded End State Reasoning, which leverages VLMs to analyze benign driving contexts and infer latent vulnerabilities in human-vehicle interactions, producing structured end-state specifications that induce high-risk scenarios. Conditioned on these targets, we further propose End State Conditioned Video Evolution, which grounds semantic threats into physically plausible visual dynamics. Specifically, we instantiate high-risk agents within the scene via depth-aware geometric projection, followed by boundary-conditioned diffusion to generate intermediate frames with consistent motion patterns and temporal coherence. Extensive experiments across 3 VLMADs demonstrate that SafeGen increases the Judge Overall Score, a metric using a VLM judge to evaluate VLMADs’ understanding and decision-making, by 24.25% on average compared to SoTA baselines. Furthermore, fine-tuning a VLMAD improves performance in real-world driving scenes by an average of 15.9%.
[CV-61] PhenSPINE: A Standardized Benchmark for Spine Pathology Diagnosis
链接: https://arxiv.org/abs/2607.19696
作者: Duong Ngoc Vu,Hai Son Nguyen,Trong-Nghia Nguyen,Bien Tran Van,Trang Mai Xuan,Huan Vu,Thien Van Luong
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 12 pages, figures, Accepted at CITA 2026 (The 15th Conference on Information Technology and its Applications, Scopus)
Abstract:The accurate diagnosis of spinal pathologies depends heavily on radiological interpretation, yet automated systems are hindered by the lack of diverse, high-quality benchmarks. In this study, we present PhenSPINE, a Magnetic Resonance Imaging dataset comprising 16,813 images from 250 patients, curated to facilitate advanced deep learning research. We propose a robust diagnostic benchmark that integrates state-of-theart convolutional backbones with a Positional Encoding mechanism to explicitly model the anatomical context of intervertebral discs. Evaluating across four standard MRI sequences, our experiments demonstrate that the Sagittal T2-weighted sequence offers the most robust diagnostic value, achieving a superior Macro F1-score of 50.31%. We find that multisequence fusion strategies yield inferior performance compared to this single-sequence baseline, as the images across sequences in our dataset are significantly compromised by noise interference from surrounding anatomical regions. This work establishes a robust baseline and offers critical insights into sequence selection for spine analysis.
[CV-62] NavVerse: Benchmarking Indoor-to-Outdoor Embodied Navigation in Continuous Robot Simulation
链接: https://arxiv.org/abs/2607.19695
作者: Junzhe Wu,Yue Hu,Zeyu Han,Po-Hsun Chang,Yinan Dong,Behrad Rabiei,Maani Ghaffari
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Robots deployed in delivery, campus, and emergency-response settings often need to navigate from buildings to streets within a single continuous episode. Existing benchmarks usually evaluate indoor and outdoor navigation separately, and many abstract away robot execution, leaving exit finding, boundary traversal, adaptation, and kinodynamic failures underexplored. We introduce NavVerse, a physics-enabled benchmark for indoor-to-outdoor embodied navigation. NavVerse contains 100 indoor scenes, 50 urban outdoor scenes, and 50 indoor-to-outdoor scenes, and 10,000 episodes spanning Object Navigation, Vision-and-Language Navigation, and Place Navigation tasks, where agents search for semantic points of interest such as restaurants or banks. Agents are evaluated through executable robot interfaces using task-success, path-efficiency, and safety metrics. Zero-shot experiments with RL, VLA, and modular baselines show that current agents remain far from solving cross-context navigation: end-to-end VLAs obtain the highest zero-shot success, while the modular method provides the strongest safety profile. PlaceNav further reveals a clear drop from outdoor to indoor-to-outdoor scenes, indicating that adaptation remains major bottleneck.
[CV-63] A Unified Variational Framework for Deep Weakly Supervised Image Segmentation
链接: https://arxiv.org/abs/2607.19669
作者: Yin King Chu,Lingfeng Li,Sung Ha Kang,Jianping Zhang,Xue-Cheng Tai
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:We propose a unified variational framework for image segmentation under sparse pixel-level supervision. Our method is based on a simplex-constrained Potts model with a smooth perimeter regularizer, yielding a convex, smooth energy functional that can be used as a training loss in weakly supervised deep learning paradigms or optimized efficiently using iterative methods. Sparse labels are incorporated into the data fidelity term by constructing a fuzzy membership function via a function extension problem in a Reproducing Kernel Hilbert Space (RKHS), which can effectively capture inhomogeneous intensity statistics. The derived discrete loss for training standard networks demonstrates robustness and consistent improvements over non-training and partial cross-entropy (PCE) baselines in experiments, achieving comparable performance without requiring ground-truth segmentation images.
[CV-64] Domain Shift in Echocardiography: Interpretable Quantification and Prediction of Cross-Dataset Left Ventricular Segmentation
链接: https://arxiv.org/abs/2607.19643
作者: Soroush Elyasi,Nasim Dadashi Serej,Julie Wall,Massoud Zolgharni
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Cross-dataset generalisation remains a major barrier to clinical deployment of echocardiographic left ventricular segmentation, yet the sources of this shift are rarely disentangled. We examined whether transfer degradation could be estimated before deployment using handcrafted ultrasound descriptors, VAE latent features, and segmentation-derived latent features across six echocardiographic datasets. Geometry-aware preprocessing substantially improved several poor transfer cases, suggesting that much of the apparent domain shift reflects field-of-view and framing inconsistencies rather than intrinsic acoustic differences alone. Intensity z-normalisation changed dataset separability by less than 0.005, indicating that brightness and contrast are not the dominant shift axis. Absolute Dice drop on held-out source-target pairs was predicted with an R-squared value of 0.612, an MAE of 0.082, and a Spearman rho of 0.681. The variant without LV and fan-shaped features retained approximately 70% of this explanatory power, supporting mask-free transfer-risk monitoring. The most informative discrepancy measure depended on the representation, with CMD strongest in z-normalised handcrafted features, with an absolute r of approximately 0.86 and an R-squared value of approximately 0.70; log-Wasserstein strongest in VAE space, with an r of approximately -0.90 and an R-squared value of approximately 0.81; and log-MMD strongest in LV-segmentation latent features, with an r of approximately -0.92 and an R-squared value of approximately 0.84. Apparent vendor effects were largely dataset-confounded. Echocardiographic domain shift is therefore structured and measurable, and its impact on segmentation can be partly reduced through geometry-aware preprocessing and anticipated using representation-specific transfer-risk estimation.
[CV-65] Pathologist Attention-Aligned Report Generation for Prostate Histopathology MICCAI2026
链接: https://arxiv.org/abs/2607.19624
作者: Ruoyu Xue,Suryakant Singh,Souradeep Chakraborty,Pierre Marza,Oksana Yaskiv,Constantin Friedman,Natallia Sheuka,Paul Friedman,Bharat Ramlal,Beatrice Knudsen,Rajarsi Gupta,Joel Saltz,Prateek Prasanna,Gregory Zelinsky,Dimitris Samaras
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 11 pages, 4 figures, accepted for publication at the 29th International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI 2026)
Abstract:The allocation of visual attention by pathologists during cancer diagnosis is a highly selective process that critically shapes the information extracted from whole-slide images (WSIs). Human attention helps medical imaging tasks such as classification and segmentation, and becomes a strong semantic cue for identifying diagnostically informative regions for report generation. In this paper, we introduce human attention into the training of pathologist report generation models. To this end, we collected a multimodal human-attention dataset of 121 prostate WSIs annotated with pathologists’ multi-scale viewport trajectories synchronized with the pathologists’ verbal descriptions and cursor movements for five clinically relevant components (e.g., Gleason patterns). Using this dataset, we finetune two report generation models with an attention-alignment loss that regularizes the model attention over image patches to match the distribution of pathologist attention. We evaluate our approach on prostate cancer report generation and visual question answering using two models with different internal attention mechanisms (i.e., how image tokens are integrated into the language decoder). Experiments show average gains of 10.9% on NLP-based metrics and 19.3% in accuracy across five clinically relevant report components. Further, model attention maps extracted at inference time, with minimal computational overhead, align more closely with pathologist attention, providing stronger visual support for the generated reports by highlighting the regions that most influence the output.
[CV-66] EGRNet: A Lightweight Semantic Segmentation Network with Edge-Gated Refinement and Adversarial Sensing
链接: https://arxiv.org/abs/2607.19617
作者: Bareera Qaseem,Mohsin Kamal,Muhammad Naveed Aman
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 14 pages, 8 figures, 3 tables and 1 algorithm
Abstract:As autonomous systems and smart cities continue to evolve, the demand for efficient and robust scene understanding becomes increasingly critical. Semantic segmentation plays a key role in enabling autonomous vehicles to comprehend complex urban environments. However, achieving high accuracy with minimal computational cost remains a significant challenge. In this paper, we present Edge-Gated Refinement Network (EGRNet), a lightweight and efficient deep learning model designed for real-time semantic segmentation in urban scenarios. The model incorporates depthwise separable convolutions to reduce computational complexity and dilated residual blocks for capturing rich multi-scale contextual information. Additionally, we introduce a novel Edge-Gated Refinement (EGR) module, which adaptively fuses original and refined features through a learnable gating mechanism, enhancing boundary preservation and edge-sensitive regions. To further improve feature representation, Squeeze-and-Excitation (SE) attention is applied across the network. With only 0.46M parameters, EGRNet achieves state-of-the-art performance while maintaining low computational overhead. When evaluated on the Cityscapes dataset, the model attains a mean Intersection over Union (mIoU) of 65.28%, demonstrating strong accuracy with minimal resource consumption. Moreover, we introduce a lightweight adversarial attack detection strategy, ensuring robustness against adversarial inputs without compromising real-time performance. By combining efficiency, accuracy, and resilience, EGRNet is well-suited for deployment on edge devices in safety-critical real-time applications.
[CV-67] VQ-Transplant: Efficient VQ-Module Integration for Pre-trained Visual Tokenizers
链接: https://arxiv.org/abs/2607.19575
作者: Xianghong Fang,Yuan Yuan,Dehan Kong,Tim G. J. Rudner
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 20 pages, 9 figures and 16 tables
Abstract:Vector Quantization (VQ) underpins modern discrete visual tokenization. However, training quantization modules for state-of-the-art VQ-based models requires significant computational resources which, in practice, all but prevents the development of novel, cutting-edge VQ techniques under resource constraints. To address this limitation, we propose \bf VQ-Transplant, a simple framework that enables plug-and-play integration of new VQ modules into frozen, pre-trained tokenizers by replacing their native VQ modules. Crucially, the proposed transplantation process preserves all encoder-decoder parameters, obviating the need for costly end-to-end retraining when modifying the quantization method. To mitigate decoder-quantization mismatch, we introduce a lightweight decoder adaptation strategy (trained for only 5 epochs on ImageNet-1k) to align feature priors with the new quantization space. In our empirical evaluation, we find that VQ-Transplant allows obtaining near state-of-the-art reconstruction fidelity for industry-level models like VAR while reducing the training cost by 95%. VQ-Transplant democratizes quantization research by enabling resource-efficient integration of novel VQ techniques while matching industry-level reconstruction performance.
[CV-68] ChronoStitch: Training-Free Composition of Visual KV Memories for Long-Horizon Temporal Reasoning
链接: https://arxiv.org/abs/2607.19547
作者: Santiram Tiwari,Nishant Sinha,Kunal Kislay
类目: Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
备注: 6 pages, 4 tables
Abstract:Long-video question answering requires a model to preserve visual evidence over time without repeatedly reprocessing the same video. A practical approach is to store the vision-language model’s internal key-value (KV) cache for each video chunk and retrieve that state at query time. However, independently cached video chunks do not compose correctly: every chunk is prefilled from local rotary position zero, so naive concatenation collides temporal phases and removes the global order required for questions about what happened first, how often events occurred, or what changed across the video. This paper presents ChronoStitch, a training-free method for composing independently stored visual KV memories. The method first re-bases stored post-rotary keys onto a global three-axis multimodal RoPE coordinate system that preserves time, height, and width structure. We show why a one-dimensional scalar re-indexing is geometrically inconsistent for visual tokens because it turns spatial order within a frame into false temporal displacement. We then address the residual content gap left by positional repair: later chunks were originally encoded without attending to earlier chunks. ChronoStitch therefore selectively recomputes a small fraction of high-deviation later-chunk visual tokens while allowing them to attend over the composed cache. On Qwen2.5-VL-3B and the temporal split of TempCompass, ChronoStitch outperforms naive composition and position-only variants, improving event-ordering accuracy while running 3.3x faster than full joint re-prefilling.
[CV-69] Synthetic and Derived Training Images for Campus Waste Detection: A Multi-Seed Evaluation with YOLOv8n
链接: https://arxiv.org/abs/2607.19535
作者: Ali Behbahani,Newsha Javanmardi,Shahriar Ahmed,Ling Chen,Phouvadeth Vathana
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Incorrect disposal can contaminate campus recycling streams, and a bin-mounted camera could provide feedback as an item is discarded. We evaluated whether synthetic and derived images improve a YOLOv8n detector for this view. The real dataset contained 148 campus photographs: 86 for training, 31 for validation, and 31 for testing. Twelve joint-training configurations varied the amount and source of added images. We repeated seven principal settings with four matched seeds and computed bootstrap percentile intervals over those seeds. The real-only model reached a mean mAP@0.5 of 0.691 [0.665, 0.722]. Background replacement reduced the mean to 0.560 [0.499, 0.619], isolated-object images gave 0.680 [0.644, 0.724], and the full augmentation pool gave 0.487 [0.438, 0.537]. We also tested hand-and-forearm composites because every real photo showed a held object. Two cutouts in the initial composite set came from test photographs, so we discarded that experiment, rebuilt the set with training-split cutouts, and reran all four seeds. The corrected paired difference was +0.034 [-0.063, 0.199], which does not support a reliable hand-composite effect. Single-seed transfer experiments produced source-dependent rankings between joint mixing and sequential pretraining. None of the evaluated configurations exceeded the real-only baseline. The reported intervals quantify seed variation; the 31-photo test set remains too small for strong class-specific conclusions.
[CV-70] D3VL: Understanding Driving Scenes from 3D Time Series Data and Video with Language Models
链接: https://arxiv.org/abs/2607.19528
作者: Heesang Han,A. Lynn Abbott,Abhijit Sarkar
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted to IEEE IV 2026
Abstract:Recent advances in Multimodal Large Language Models (MLLMs) have triggered the development of end-to-end MLLMs for autonomous driving. However, the main emphasis to date has been for MLLMs using 2D images and videos. In contrast, this paper considers MLLM effectiveness using 3D sensors, particularly LiDAR and stereo cameras. LiDAR presents unique challenges to integration within an MLLM, largely because of data sparsity and lack of a grid structure for the data. For similar reasons, fusion of camera and LiDAR data within an MLLM pipeline is also uncommon. However, most autonomous systems rely on LiDAR-based sensing, and incorporating 3D data has been proven to improve performance in traditional 3D scene perception tasks. This paper presents D3VL, a novel MLLM framework that integrates 2D and 3D time-series data in a single but simple architecture. The model aims to answer questions involving traffic scene understanding and safety. D3VL shows an 11% improvement in the KITTI Question-Answering (QA) dataset compared to baseline methods in processing 2D and 3D time-series data. This paper further introduces the Waymo QA dataset extension, which assesses models’ capabilities in processing 3D and time-series data under diverse driving conditions. D3VL implementation code and WaymoQA extension can be found on our supplemental website: this https URL
[CV-71] Geospatial Diffusion-based Evolution Synthesis (GeoDES) for Storm-Centered Weather Augmentation
链接: https://arxiv.org/abs/2607.19522
作者: Sonia Cromp,Satya Sai Srinath Namburi GNVV,Youran Wang,Grace Kisslinger,Frederic Sala,James Booth,Allegra LeGrande
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注: 31 pages, 11 figures
Abstract:While machine learning-based weather models hold significant promise, they struggle to predict the detailed structure of large-scale weather systems such as cyclonic storms. Regional models are constrained by limited historical records within fixed geographic boundaries, while global models are computationally expensive and often operate at resolutions too coarse to capture fine-grained storm dynamics. To bridge this gap, we introduce the Geospatial Diffusion-based Evolution Synthesis (GeoDES) model, a custom image-to-video diffusion model. By focusing generation strictly on the evolving storm structure, GeoDES synthesizes physically consistent, high-fidelity weather events suitable for stress-testing forecast models and expanding meteorological datasets. Evaluations demonstrate that GeoDES outperforms prior methods on key metrics, achieving 52% lower Peak Vorticity Error and 8% higher Anomaly Correlation Coefficient than the next strongest methods on the North Atlantic test set.
[CV-72] Crowd4D: Scene-Aware Monocular 4D Crowd Reconstruction
链接: https://arxiv.org/abs/2607.19517
作者: Hongbo Kang,Tianyi Zhou,Qingyang Yang,Hongwei Wen,Jing Huang,Yu-Kun Lai,Kun Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Recovering scene-consistent 4D crowd motion from monocular video in large-scale scenes remains challenging due to severe depth ambiguity and complex scene geometry. Existing monocular crowd reconstruction methods typically rely on single-plane assumptions, leading to unreliable metric scale and spatial drift under complex terrain. We propose Crowd4D, the first scene-aware 4D crowd reconstruction framework that jointly optimizes the crowd and scene from a monocular RGB video in large-scale scenes. Crowd4D explicitly incorporates scene geometry and ensures consistency across image and scene spaces via a multi-stage optimization strategy. A key bottleneck of this task lies in accurate human-scene alignment, particularly in scale and position. However, human and scene reconstructions are typically decoupled. To address this, we introduce the Human-Scene Interaction Proxy, abbreviated as HSIP, as an intermediate representation derived from Scene Interaction Point Clouds and a Scene Interaction Surface, abbreviated as SIPC and SIS. These representations encode explicit scene-aware geometric priors and redefine the optimization space for large-scale monocular 4D crowd reconstruction. To further improve temporal stability under occlusions, we introduce Crowd Structural Coherence Regularization, abbreviated as CSCR, which leverages HSIP-based spatial priors to impose soft temporal consistency on pairwise relative displacements and directions within local crowd neighborhoods. Extensive experiments demonstrate that Crowd4D consistently outperforms existing state-of-the-art methods and enables robust monocular 4D crowd reconstruction in complex, large-scale real-world scenes.
[CV-73] Detect Early Escalate Rarely: Anytime Detection of AI-Generated Video from the Compressed Bitstream
链接: https://arxiv.org/abs/2607.19476
作者: Mert Onur Cakiroglu,Mehmet Dalkilic,Hasan Kurban
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Detectors for AI-generated video are evaluated offline. A clip is decoded to pixels and scored once, increasingly by a large vision-language model. Detection, however, is deployed online. We recast the task as streaming perception and score the motion field the codec already wrote into the bitstream. Reading that field is a parse, not a pixel-domain forward pass. Because the running aggregate is monotone, one end-calibrated threshold is anytime-valid at the data-dependent decision time. Recalibrating at each prefix is not. Escalation is priced in closed form. A compute budget maps to a deferral window, on a frontier monotone exactly where the deferral condition holds. On matched GenVidBench the codec stage reaches full-length AUC 0.64 at five orders of magnitude less compute than a pixel CNN, on CPU. Its gate holds the stopping-time false-positive rate at target while the real data match its calibration, and drifts above it under distribution shift. Deferring 15% of clips lifts accuracy from 0.75 to 0.78 at 7\times less compute (paired: McNemar p10^-6 ). The stage-1 ordering replicates on AIGVDBench. We introduce no new detector. The contribution is the reframing, two guarantees, and the measured frontiers. Code, configurations, and evaluation splits: this https URL.
[CV-74] When Does Consensus Beat Voting? A Critical Analysis of Statistical Label Fusion in Medical Image Segmentation
链接: https://arxiv.org/abs/2607.19402
作者: Renjie He
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV); Medical Physics (physics.med-ph)
备注:
Abstract:This paper provides a rigorous, self-contained investigation of consensus segmentation. We derive the mathematical foundations from first principles – the generative model, EM algorithm, Van Leemput’s marginalization analysis, identifiability conditions, Spatial STAPLE, and deep variational formulations – and validate each theoretical prediction through controlled experiments. The central finding is sobering: under common conditions, STAPLE reduces to thresholded majority voting, suffers 95% EM suboptimality, and collapses under class imbalance. These are not edge cases but typical scenarios in medical imaging. Majority voting – simple, non-parametric, and robust – is a surprisingly strong baseline that the field has perhaps too hastily dismissed in favor of more “sophisticated” methods. At the same time, the deep consensus model demonstrates that the consensus problem is not inherently difficult – it becomes tractable when the image is used alongside the labels. And conformal prediction shows that formal uncertainty guarantees are achievable and practical. We hope this work encourages practitioners to critically evaluate their consensus methods rather than applying STAPLE by default, and provides the mathematical and empirical foundation for more principled approaches.
[CV-75] Leverag ing Offline Supervision for Efficient and Generalizable Reinforcement Learning in Large-Scale Vision-Language-Action Models
链接: https://arxiv.org/abs/2607.19399
作者: Dmitriy Poyarkov,Aleksei Staroverov,Aleksandr I. Panov
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:It is commonly observed that online reinforcement learning (RL) produces better-performing strategies than offline methods across a broad range of performance measures. In particular, RL-trained policies exhibit stronger out-of-distribution (OOD) behavior, where models trained only with imitation learning approaches often struggle. A recent study introduced an OOD-focused benchmark and reported that RL-trained vision-language-action (VLA) policies achieve noticeably better OOD performance and slightly better in-distribution (IND) performance than their counterparts trained with supervised fine-tuning (SFT). In this work, we investigate whether hybrid offline-online training can combine the advantages of both approaches. Specifically, we study RL methods regularized by offline supervision via either offline data or an offline-trained reference policy. We evaluate these approaches on the OOD benchmark and compare them with both offline-only training and standard RL. Our results show that although offline training achieves limited OOD performance by itself, incorporating offline supervision into RL preserves strong OOD capability while substantially improving training efficiency. In particular, the guided methods reach performance close to that of standard RL while requiring roughly half of the training budget. Rather than producing a trade-off between speed and OOD performance, the hybrid approach retains strong OOD capability while achieving this efficiency gain. Project page: this https URL
[CV-76] Native Multi-Dimensional Subquadratic Operators via Input Dependent Long Convolutions
链接: https://arxiv.org/abs/2607.19378
作者: David R. Wessels,Farhad Ramezanghorbani,David W. Romero,Alireza Moradzadeh,Olivia Viessmann,Maksim Zhdanov,John St. John,Ken Janik,David M Knigge,Yucheng Tang,Erik J Bekkers,Saee Gopal Paliwal
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (stat.ML)
备注:
Abstract:Subquadratic alternatives to attention require compromises when applied to multi-dimensional data: standard convolutions lack global receptive fields and input dependency, while recurrent models require rasterizing data such as images, volumes, and partial differential equation (PDE) into an ad-hoc 1\rm D scan order that violates their spatial structure. We introduce \textitHyenaND, a subquadratic, global, input-dependent operator that acts directly on the native geometry of multidimensional data through convolutions with implicitly parametrized global, input-dependent multi-dimensional convolutional kernels. Our CUDA implementation, \textttnSubQ, fuses the FFT-convolution path to turn HyenaND’s \mathcalO(L \log L) scaling into wall-clock speedups. Across long-context genomics, computer vision, medical imaging, and PDE modeling, pure HyenaND stacks match the accuracy of strong attention baselines, while hybrid configurations that interleave HyenaND and attention layers outperform both pure attention and strong recurrence-based hybrids.
[CV-77] PRISM-DR: Per-lesion Retinal Inference with Specialist Models for Diabetic Retinopathy
链接: https://arxiv.org/abs/2607.19864
作者: Zübeyr Özeren,Tansel Uyar
类目: Image and Video Processing (eess.IV); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注: 17 pages, 9 figures, 15 tables
Abstract:Diabetic retinopathy is a leading cause of preventable blindness; its early lesions are small, low contrast, and easily missed in manual screening. Most automated detectors handle the four non-proliferative DR lesions: microaneurysms, hemorrhages, hard exudates, and soft exudates, with a single multi-class model, even though these lesions differ sharply in size, color, morphology, and prevalence, so a shared model favors common, easy classes over rare, difficult ones. We present PRISM-DR, a lesion-specific pipeline that trains one single-class detector per lesion, each with its own configuration. From a raw fundus image, the pipeline applies region of interest cropping, fundus-specific preprocessing, four parallel YOLO detectors, tiling, per-lesion ensembling of five cross-validation folds, and an inter-lesion suppression step that resolves overlaps by physical lesion size and clinical priority rather than confidence. Per lesion, the best of five YOLO generations is selected, and augmentation is tuned by Bayesian optimization. Trained on IDRiD with stratified five-fold cross-validation, the system reaches a test mAP50 of 0.527 and F1 of 0.529, highest AP50 on hard exudates with 0.561. Without fine-tuning, the models transfer well where the imaging scale is close to IDRiD and degrade as field of view and resolution depart. These modest absolute results reflect a small single-source training set and a difficult task; however, treating each lesion as a separate detection problem is a practical alternative to a single multi-class model.
[CV-78] Deep Shape Regression for Planar Curves with Multimodal Covariates MICCAI2026
链接: https://arxiv.org/abs/2607.19600
作者: Manuel Pfeuffer,Roshan Prakash Rane,Hadya Yassin,Kerstin Ritter,Sonja Greven
类目: Methodology (stat.ME); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Quantitative Methods (q-bio.QM); Machine Learning (stat.ML)
备注: 17 pages, 4 figures, 1 algorithm. Submitted to the ShapeMI Workshop, MICCAI 2026. Code is available at this https URL
Abstract:The shape of a planar curve is the geometric information that remains once translation, rotation, scale and reparametrisation are removed and is of interest in many health applications, e.g. in neuroimaging. We propose a deep shape regression model for open planar curves that admits multimodal and high-dimensional covariates. Representing curves as complex-valued functions, we show that the conditional full Procrustes mean is the leading eigenfunction of the conditional covariance. To estimate this covariance surface, we propose a novel deep conditional covariance smoother with modality-specific encoders - e.g. splines for scalar covariates and convolutional networks for images, which classical spline smoothers cannot accommodate. Our model is by construction invariant to the translation, rotation and scaling of the input curves and handles sparsely and irregularly sampled curves. We further provide an algorithm for elastic mean estimation that also removes parametrisation by iterating covariance smoothing, rotational alignment and parametrisation alignment. We illustrate the method on simulated outlines with known conditional mean and multimodal covariates, and give a first application to hippocampal outlines from the ADNI cohort, recovering covariate effects consistent with the literature. Code is available at this https URL.
[CV-79] BLUE: Semantics-Preserving Video Compression for Efficient Vision-Language Surveillance Analytics
链接: https://arxiv.org/abs/2607.19515
作者: Shubham Baid,Akash James,Sahil Chachra,Nishant Sinha,Kunal Kislay
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
备注: 17 pages, 10 figures, 6 tables
Abstract:Continuous surveillance video creates a growing storage, transmission, and inference burden for enterprise video analytics systems. While modern codecs such as H.265 reduce bitrate for human-viewable video, aggressive compression can degrade downstream computer-vision performance and does not necessarily reduce the number of vision-language model (VLM) inference calls required for semantic video understanding. This paper evaluates BLUE, a fixed-camera surveillance compression approach that suppresses static-background redundancy while preserving foreground activity, for its effect on VLM-based event and anomaly understanding. We compare raw H.265 and BLUE-compressed H.265 video on two surveillance datasets: VIRAT, comprising 227 paired event samples from 106 clips, and CHAD, comprising 54 human-activity anomaly clips. For each pair, the same frame index is evaluated using a VLM captioning pipeline, and outputs are scored against annotation-derived ground truth using a blind judging protocol. The results show no measurable degradation in semantic inference quality. On VIRAT, the mean VLM score remains effectively unchanged between raw H.265 and BLUE, with a mean difference of approximately -0.01 on a 0-10 scale. On CHAD, raw H.265 and BLUE obtain near-equivalent mean scores of 4.31 and 4.26, respectively. Compression saving is also uncorrelated with VLM score change on VIRAT (r = 0.004), indicating that higher BLUE compression does not predict semantic quality loss. Beyond storage reduction, BLUE increases the share of skip-heavy P-frames on CHAD from 1.4% to 53.2%, enabling an estimated 53% reduction in VLM calls through packet-size-based frame skipping. These findings suggest that BLUE functions as a machine-centric compression layer for surveillance video, reducing bandwidth and inference cost while preserving VLM semantic performance.
[CV-80] Strong Gravitational Lensing Posterior Sampling in Pixel-Space Using Diffusion Models and Recurrent Inference Machines ICML2026
链接: https://arxiv.org/abs/2607.19459
作者: Guillaume Payeur,Laurence Perreault-Levasseur,Gabriel Missael Barco,Yashar Hezaveh
类目: Instrumentation and Methods for Astrophysics (astro-ph.IM); Cosmology and Nongalactic Astrophysics (astro-ph.CO); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (stat.ML)
备注: 31 pages, 34 figures, Extension of an article accepted at the ICML 2026 Workshop on AI for Physics
Abstract:Modeling galaxy-galaxy strong gravitational lenses to infer the brightness of the source galaxy and the mass distribution of the foreground galaxy is computationally challenging, particularly for high-resolution, high signal-to-noise ratio observations. In this regime, high-dimensional representations of both the source and the foreground mass distribution are necessary to model the data down to the noise level. This inference problem has been challenging for both traditional and machine learning-based methods because of its high dimensionality and its non-linearity in the foreground mass distribution. We present a method to generate joint posterior samples of the source galaxy and foreground mass distribution as pixelated images conditioned on observations. The method combines diffusion-based generative modeling and recurrent inference machines. It can model realistic gravitational lensing simulations with background and foreground galaxies drawn from cosmological hydrodynamical simulations down to the noise level.
[CV-81] Group-of-Latents: Perceptual Video Compression at Extreme Bitrates via Masked Latent Generative Modeling
链接: https://arxiv.org/abs/2607.19437
作者: Shaokang Wang,Jinchang Xu,Peidong Jia,Zhijian Hao,Siyuan Qian,Fei Zhao,Rui Ma,Xiaozhu Ju,Jian Tang,Xiaodong Xie,Shanghang Zhang,Huizhu Jia
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV); Multimedia (cs.MM)
备注:
Abstract:Most existing video compression algorithms follow a paradigm of transformation and quantization, optimizing the trade-off between distortion and bitrate. However, extremely low-bitrate compression remains an underexplored frontier where perceptual quality optimization under severely constrained coding resources has not been adequately addressed. In this paper, we propose a unified generative framework that leverages pre-trained Diffusion Transformer (DiT) priors to achieve high perceptual quality at extremely low bitrates. We first introduce a flexible Group-of-Latents (GoL) strategy within the latent space of a causal tokenizer, explicitly partitioning the latent stream into intra I -latents and inter P -latents. The Deep Compression Module (I-DCM) then encodes key I -latents to preserve perceptual anchors with minimal overhead. Building upon these anchors, the DiT-based Unified Latent Denoising Module (U-LDM) refines intra-frame textures and synthesizes P -latents from noise, reconstructing temporal dynamics at zero additional bitrate cost. Extensive experiments demonstrate that our method uniquely operates in the extreme-low-bitrate regime (e.g., (0.005) bpp), achieving state-of-the-art perceptual fidelity with rich spatial details and robust temporal consistency. The code will be made publicly available.
人工智能
[AI-0] SoftReason : A Fully Differentiable Neuro-Soft-Symbolic Deductive Reasoning Architecture over High-Dimensional Perceptual Data
链接: https://arxiv.org/abs/2607.20402
作者: Wael AbdAlmageed
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:In many reasoning problems, the premises are not observed as discrete symbols, but must be inferred from high-dimensional inputs. Further, the predicate vocabulary, argument structure, and trusted evidence are supplied by a Knowledge Graph (KG), or rule definitions. Classical neuro-symbolic pipelines have a discrete interface between perception and deduction. We present a neuro-soft-symbolic architecture for differentiable deductive reasoning over latent perceptual facts and knowledge-provided predicates. SoftReason removes the gradient gap by representing the deductive state as a local soft interpretation tensor over candidate constants and predicates. Perception proposes probabilistic base facts, KG triples enter as high-confidence soft evidence, and every query anchor, predicate choice, and closure update remains differentiable. Our core innovation is a learned differentiable lift of the immediate-consequence operator. It uses predicate-definition embeddings and latent composition channels to form soft body-predicate mixtures, aggregate over all possible witnesses, propose query-conditioned head facts, and update the interpretation through a monotone probabilistic OR. We instantiate the framework on Knowledge-aware Visual Question Answering (KVQA), and demonstrates how SoftReason supports end-to-end perceptual grounding, KG evidence injection, and differentiable deductive closure in one trainable architecture.
[AI-1] Closing the Lab-to-Store Gap: A Data-Efficient Post-Training and Experience-Driven Learning VLA Framework for Retail Humanoids
链接: https://arxiv.org/abs/2607.20345
作者: Roger Sala Sisó,Tiago Silvério,Jakob Sand,Tran Nguyen Le
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: 8 pages. This work has been submitted to the IEEE for possible publication
Abstract:Closing the gap between benchmark performance and reliable real-world operation remains a central challenge for Vision-Language-Action (VLA) humanoid robots, which must handle execution errors, distribution shifts, and environmental variability. This paper presents DEED (Data-Efficient Post-Training and Experience-Driven Learning), a systems-level approach evaluated on a supermarket chip-restocking task using a Unitree G1-Edu humanoid robot and the GR00T N1.6 foundation model. DEED comprises three key components: (1) a data-efficient post-training pipeline with control-frequency alignment, data curation, task-relevant visual highlighting, and reduced VLA dependence; (2) a real-world study of experience-driven refinement, adapted from RECAP via a text-based advantage prefix and a vision-language value function; and (3) a latent-space analysis tool for studying in- and out-of-distribution behavior. Our results suggest that bridging the lab-to-store gap is primarily a systems integration challenge rather than an architectural one: careful data design and targeted post-training can transform a policy that fails under naive fine-tuning into a competent real-world system using only a single GPU.
[AI-2] Understanding Generative AI-mediated User Engagement with Academic Library Resources
链接: https://arxiv.org/abs/2607.20328
作者: Hae Min Kim,Stacy Stanislaw
类目: Digital Libraries (cs.DL); Artificial Intelligence (cs.AI)
备注: This is the accepted manuscript. The final published version will appear in College Research Libraries, May 2027
Abstract:This study empirically analyzed generative AI as an emerging discovery pathway to academic library resources. Utilizing web analytics from August 2023 to October 2025, the research identifies a significant increase in AI-mediated traffic, particularly following the integration of linked citation features. Referral analysis identified ChatGPT, Perplexity, and Gemini as the primary platforms driving this traffic. A substantial portion of users reached the institutional repository, primarily accessing electronic theses and dissertations. This pattern suggests that AI retrieval mechanisms effectively surface resources with structured metadata and stable permalinks that are Open Access and freely available. The results illustrate how AI ecosystems currently expose library resources and underscore the need for continued analysis and a strategic response to the evolving AI landscape.
[AI-3] Dont Trust the Label: License Laundering in AI Supply Chains
链接: https://arxiv.org/abs/2607.20300
作者: James Jewitt,Hao Li,Gopi Krishnan Rajbahadur,Bram Adams,Ahmed E. Hassan
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: 9 pages, 2 figures
Abstract:AI artifacts move through a multi-platform supply chain, spanning datasets and models on Hugging Face and applications on GitHub. While each artifact carries a license whose obligations should propagate through redistribution, no study has yet measured whether those obligations survive the chain or are stripped and replaced as artifacts move downstream. We trace 232,270 dataset \rightarrow model \rightarrow application chains and quantify two forms of license laundering: when artifacts with no declared license acquire definitive labels downstream, and when one declared license category replaces another during redistribution. We find that 62.3% of chains pass through at least one artifact with no declared license (concentrated in a small set of foundational datasets), and that every obligation-bearing license category falls below 7% end-to-end survival while the Permissive category reaches 95.1%. Based on these findings, we provide actionable recommendations for practitioners, model publishers, rights holders, and platform owners.
[AI-4] Courteous Anticipation: Improving Long-Lived Task Planning in Persistent Shared Environments
链接: https://arxiv.org/abs/2607.20289
作者: Md Ridwan Hossain Talukder,Roshan Dhakal,Elizabeth Phillips,Gregory J. Stein
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: 9 Pages
Abstract:We consider a task planning scenario in which robots sharing a persistent environment are assigned tasks one at a time from a held-out sequence. Standard task planners, lacking foresight of future tasks and inconsiderate of others’ constraints, solve each task in isolation, leaving terminal states that increase future cost for all, side effects that compound over lengthy task sequences. To reduce cost over the sequence, a robot must anticipate how its actions now may impact performance on future tasks for all robots sharing the environment. Therefore, we present courteous anticipatory planning, wherein a model-based planner proposes candidate plans and selects the one that jointly minimizes immediate cost and aggregated expected future cost across all robots, estimated via independent per-robot learned estimators. This factored formulation avoids combinatorial joint rollouts and supports modular deployment: adding a robot requires only training its own estimator. We evaluate in two persistent PDDL domains, a home environment with robots that have similar capabilities but different responsibilities, and a restaurant environment where robots’ distinct capabilities create states that other robots lack the capability to resolve. During lengthy task sequences, our planner reduces total cost by 10.43% versus myopic and 4.03% versus selfish anticipatory planning in a two-robot home environment and by 17.41% and 13.24%, respectively, in a three-robot restaurant.
[AI-5] he Ethics of Autonomous AI Agents for Offensive Security
链接: https://arxiv.org/abs/2607.20255
作者: Andreas Happe,Jürgen Cito,Jasmin Wachter
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注: accepted at FAIEMA 2026
Abstract:LLM-driven autonomous agents are reshaping offensive security. Unlike traditional penetration-testing tooling – deterministic, narrowly scoped, and operated by trained practitioners – agentic security tools exhibit \textitindeterminacy along three independent dimensions. First, their actions are drawn from a non-deterministic policy whose outputs resist both ex-ante and ex-post explanation, frustrating incident attribution and pre-deployment safety review. Second, their impact is open-ended due to the non-deterministic actions, agency of utilized models, and opaque LLM supply-chains. Third, their user population is indeterminate in both size and required skill: the operating skill floor for using or developing offensive capabilities has dropped sharply. These three properties are linked thematically, but are not derivable from one another. Combined with the structural cost asymmetry between offense and defense, they enable the industrialization of offensive capability. The net short-term effect favors attackers, even if the same technology may, in the long run, democratize access to defensive practice. Existing dual-use cybersecurity and AI-ethics frameworks were not designed for this combination. Our work analyzes how moral attribution becomes diffuse between users, tool-makers, and third parties when employing autonomous AI agents for offensive security. We also examine the stakeholder impact of this technology and provide stratified recommendations.
[AI-6] Pushing the Frontier of Full-Song Generation: Hierarchical Autoregressive Planning Meets Flow-Matching Rendering
链接: https://arxiv.org/abs/2607.20253
作者: Junyu Dai,Xinyue Fan,Weiqin Li,Xiangang Li,Yunjia Li,Bin Ma,Yukun Ma,Chongjia Ni,Yufei Shi,Haoxu Wang,Menglin Wu,Jianwei Yu,Huaicheng Zhang,Han Zhao,Shengkui Zhao,Haina Zhu
类目: ound (cs.SD); Artificial Intelligence (cs.AI); Audio and Speech Processing (eess.AS)
备注:
Abstract:In this report, we present a unified song generation framework capable of producing high-quality full-length music from lyrics, text descriptions, and musical attributes. The proposed framework supports three tasks: Lyrics-to-Song Generation, which generates complete songs from text descriptions, lyrics, and musical attributes; Instrumental Music Generation, which creates music without vocals; and Cover Song Generation, which reinterprets existing songs with different styles while preserving their melodic content. Architecturally, our system consists of four main components: a semantic-aware tokenizer, hybird-LM, FullDiT, and a two-level melody module. The tokenizer encodes audio into 8-codebook RVQ tokens for efficient discrete music representation. Based on these tokens, hybird-LM performs hierarchical autoregressive audio-token modeling for full-song generation. To improve audio fidelity, FullDiT performs full-song flow matching in a continuous VAE latent space conditioned on codec tokens, lyrics, and text captions. For cover song generation, the melody module extracts and discretizes melody cues from reference audio to guide generation while preserving the original melodic content. Finally, we investigate DPO, GRPO, and OPD as reward-based post-training strategies for hybird-LM and apply flow-based GRPO to FullDiT to improve musicality and rendering quality. Experimental results on a multilingual automatic benchmark, complemented by the Artificial Analysis Music with Vocals leaderboard, show that the proposed framework achieves competitive performance in the evaluated settings.
[AI-7] Small Free and Effective: Orchestrating Open-Weight Small Language Models to Outperform Single LLM for Malware Analysis RAID
链接: https://arxiv.org/abs/2607.20216
作者: Adel ElZemity,Shujun Li,Budi Arief
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注: To appear in Proceedings of the 29th International Symposium on Research in Attacks, Intrusions, and Defenses (RAID)
Abstract:Malware analysis demands rapid interpretation of complex detonation reports spanning filesystem, network, and process behaviours. While large language models (LLMs) demonstrate impressive capabilities for technical artifact interpretation, the opacity and escalating API costs of closed-weight frontier models motivate exploration of open-weight alternatives. However, many open-weight models are large, demanding significant compute resources and incurring non-trivial hosting costs that place them beyond reach for resource-constrained deployments. This paper investigates whether orchestrated ensembles of small language models (SLMs) can match or exceed single LLM performance on structured questions about malware detonation reports. We established baselines by testing eleven open-weight SLMs, three cyber security pre-trained models, and six frontier LLMs on Meta’s CyberSecEval Malware Analysis benchmark. We then designed and evaluated four orchestration architectures: (i) a multi-agent pipeline that decomposes analysis into structured evidence-collection and reasoning stages, (ii) an adversarial debate framework in which two agents iteratively critique each other’s reasoning, (iii) a hierarchical consultation system that pairs a general-purpose SLM with a cyber-specialised expert model, and (iv) a hybrid architecture that combines evidence-grounded pipelines with adversarial debate reasoning. The hybrid system (Qwen3-4B with Foundation-Sec-8B) achieved 35.30% overall accuracy, exceeding the strongest cyber-specialised baseline (22.54%) and the strongest ungrounded frontier baseline (34.77%); when given the same evidence pipeline, grounded Gemini remained the strongest configuration at 38.22%. These findings show that evidence-grounded orchestration can substantially improve the performance of collaborative SLMs for supporting interpretation of malware detonation reports.
[AI-8] ELSAA: Efficient Low-Rank and Sparse Attention Approximation for Training Transformers
链接: https://arxiv.org/abs/2607.20214
作者: Mahdi Heidari,Mohammad Mahdi Rahimi,Jaekyun Moon
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:The quadratic N\times N attention score matrix remains a central obstacle to extending Transformers to longer input lengths. Existing efficient attention methods usually reduce this bottleneck by either imposing sparsity, so that each query attends to only a small subset of keys, or by using low-rank/kernel sketches, so that global interactions are compressed into a lower-dimensional representation. We propose \emphELSAA, an efficient low-rank and sparse approximation of attention. Importantly, ELSAA does \emphnot decompose the learned projection or output matrices of the Transformer into sparse and low-rank factors. Instead, after dense projections produce Q,K,V , ELSAA approximates the induced attention score operator itself: a sparse branch captures selected high-similarity interactions, while a low-rank branch summarizes diffuse global interactions. Since the two branches can be normalized over supports with very different denominator mass, ELSAA introduces a denominator-aware fusion term that scales the sparse branch according to its estimated attention mass relative to the low-rank branch. This gives a practical framework for constructing low-rank and sparse attention outputs without materializing the full quadratic score matrix, aiming to enable longer-context training while preserving both sharp token-level interactions and broad contextual mixing.
[AI-9] he Quadrilateral Loss: Additivity as a Measurable Behavior of Dense Neural Networks
链接: https://arxiv.org/abs/2607.20201
作者: Antonio Di Cecco
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Additive models buy interpretability by forbidding feature interactions, a constraint that neural instantiations enforce architecturally. We introduce the quadrilateral loss, a differentiable penalty that treats additivity as a measurable behavior instead: a second-order mixed difference on pairs of training points swapping one coordinate, which vanishes if and only if the coordinate carries no interaction, remains informative for piecewise-linear networks, and equals in expectation the per-coordinate interaction mass of the interventional Shapley-GAM. The loss turns additivity into a dial - most learned interactions prove removable almost for free, and on small datasets a moderate penalty improves accuracy and additivity simultaneously - and into an online observable: its per-feature surrender curves show, across seeds and datasets, that pre-regularization interaction magnitude barely predicts what a regularized model retains, undermining post-hoc interaction rankings. Against this instrument we compare routes to exact additivity, spanning structural masks, behavioral penalties (optionally crystallized into exact structure), weight decay, backfitting, the shared-section model, and bagged boosted stumps: constraining behavior before structure dominates weight-space constraints, rankings reverse between data regimes, and converging routes agree on the shape functions themselves. Three silent failure modes we document share one anatomy: guarantees imported into settings that quietly void their preconditions.
[AI-10] Audio-Zero: Label-Free Self-Evolution for Fine-Grained Audio Reasoning
链接: https://arxiv.org/abs/2607.20166
作者: Siqian Tong,Xuan Li,Chaozhuo Li,Baolong Bi,Yiwei Wang,Yujun Cai,Shenghua Liu,Chengpeng Hao
类目: ound (cs.SD); Artificial Intelligence (cs.AI)
备注:
Abstract:Large Audio Language models (LALMs) have made rapid progress on acoustic understanding, yet they still struggle with fine-grained audio reasoning (e.g., recognizing event order, repetitions and duration). Existing post-training methods heavily rely on expensive external labels or provide only coarse semantic signals. To bridge this gap, we introduce Audio-Zero, the first label-free self-evolution framework in the field of LALMs that improves fine-grained auditory perception and reasoning. Audio-Zero constructs an auditory self-play game from unlabeled audio contrast pairs: most players hear a reference audio, while one odd listener hears a subtle variant. The model first generates clues describing what it hears and then identifies the odd listener by reasoning over inconsistencies among clues. Since the odd listener is known by construction, the game provides verifiable rewards without any annotated answers. Experiments with Qwen2-Audio-7B-Instruct and Qwen2.5-Omni-7B on TREA, MMAU Test-mini and MMAR show that Audio-Zero improves fine-grained audio reasoning while preserving broad audio understanding. Evolutionary and diagnostic analyses further reveal that increasingly fine-grained auditory descriptions emerge naturally from game pressure.
[AI-11] Active Inference as a Convex Markov Decision Process
链接: https://arxiv.org/abs/2607.20152
作者: Nikola Milosevic,Nicolás Hinrichs,Nico Scherf
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Machine Learning (stat.ML)
备注:
Abstract:Active Inference (AIF) frames adaptive behavior as the minimization of expected free energy (EFE), combining epistemic and pragmatic objectives within a single variational principle. We frame AIF as policy optimization and show that, for closed-loop control policies, EFE minimization can be formulated as a convex Markov decision process (MDP). In this formulation, the pragmatic terms are linear in the predictive state marginals and therefore equivalent to reward maximization in a latent MDP, while the epistemic value introduces a nonlinear component that distinguishes EFE minimization from standard reinforcement learning. This perspective further reveals the epistemic drive of active inference as a policy-dependent (performative) reward. We analyze finite-horizon, discounted, and average-reward formulations of EFE and derive a mirror descent (MD) algorithm that locally linearizes the objective around the current state marginals, yielding a policy-dependent reward that is compatible with actor-critic methods and dynamic programming. Finally, we argue that coupling world-model learning with policy optimization gives active inference the structure of performative reinforcement learning, providing a route toward grounding active inference within modern reinforcement learning and optimization theory, including convergence analysis and principled policy improvement guarantees.
[AI-12] Formal Foundations for Known Good Reliable Die Screening in Chiplet-Based AI Systems-on-Chip
链接: https://arxiv.org/abs/2607.20141
作者: Prashanthi Metku,Chandra Gandu
类目: Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI)
备注:
Abstract:The rapid growth of chiplet-based artificial intelligence systems-on-chip (SoCs) has exposed a fundamental gap in semiconductor test methodology. Existing Known Good Die (KGD) screening guarantees pre-assembly functional correctness, yet it offers no probabilistic assurance of post-assembly reliability lifetime. To address this limitation, the present work formalizes the transition from KGD to Known Good Reliable Die (KGRD) screening as a constrained inference problem over incomplete pre-assembly observability. Building upon this formulation, four interlocking contributions are presented: (i) a Bayesian probabilistic risk model that maps pre-assembly telemetry to post-assembly failure likelihood with a quantified observability bias bound; (ii) a safety-gated decision architecture that provides a provable post-assembly failure probability guarantee; (iii) uncertainty-aware disposition boundaries derived from Bayes-optimal decision theory; and (iv) a constrained closed-loop feedback mechanism that delivers consistent model improvement without violating reliability constraints. A Monte Carlo simulation study on N = 4,000 synthetic dies verifies all four theoretical properties and confirms that the safety guarantee holds uniformly across the full range of tested gate threshold.
[AI-13] CUSUM-Shaped Inference-Time Monitoring and Targeted Re-Decoding for Quantized Small Language Model Reasoning
链接: https://arxiv.org/abs/2607.20129
作者: El Hassane Ettifouri,Ayoub Belfatmi,Mahaman Sanoussi Yahaya Alassan,Walid Dahhane
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Quantized small autoregressive reasoning models can enter long, repetitive, or unproductive trajectories, yet inference-time compute is usually allocated without observing how a trajectory develops. Building on an earlier token-level e-CUSUM controller, we develop MGT-B (Monitoring-Guided Test-time Backtracking), a revised external controller that maps overlapping windows of pre-sampling uncertainty and degeneration features to position-conditional empirical tail probabilities, accumulates mixture betting factors with a CUSUM-shaped reset, and responds to an alarm by estimating a rollback point, restoring token and key-value-cache state, and performing constrained re-decoding. To audit whether the effect persists on problem identities first observed after the manual choice of log threshold h = 10, we retrospectively exclude 260 IDs present in pre-threshold artifacts and retain the chronologically first post-threshold pair for each remaining ID, yielding a 240-pair chronology-audit set. On this set, accuracy changes from 82/240 to 88/240 (+2.50 percentage points; 13 corrections, 7 regressions; exact McNemar p = 0.2632; paired bootstrap 95% interval [-1.25, +6.25]). A broader 467-pair historical-coverage set of seed-matched pairs changes accuracy from 146/467 to 167/467 (+4.50 points; McNemar p = 0.000753), but includes 200 seed-1 IDs available before or during threshold selection and is reported only as an exploratory estimate. All 316 no-alarm outputs in the 467-pair set are identical to vanilla, while the 151 alarmed trajectories contain 29 corrections and 8 regressions. Neither analysis is confirmatory, and the empirical factors are not established as a valid e-process or e-detector. The results support a selective monitoring-and-repair mechanism for the studied MATH-500 setting, rather than a general or theoretically certified reasoning improvement.
[AI-14] Co-Evolving LLM Evaluators and Policies via DynamicRubric
链接: https://arxiv.org/abs/2607.20083
作者: Beining Wang,Weihang Su,Hongtao Tian,Hao Kong,Tao Yang,Ting Yao,Qingyi Pan,Yueyue Wu,Qingyao Ai,Min Zhang,Yiqun Liu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Post-training with evaluator feedback on policy-induced samples serves as a major mechanism for improving large language models. As policies improve, these sampled responses become close in quality. These close candidates create a bottleneck for policy optimization: collapsed relative evaluator score gaps yield weak or misleading policy supervision. We theoretically characterize why these gaps matter through a probability allocation view, showing that the directional gain of shifting probability mass from one response to another is exactly the evaluator score gap between them. This identifies relative score gaps as the policy optimization signals that guide updates. Motivated by this view, we propose DynamicRubric, a response-set-conditioned evaluator–policy co-evolution framework that generates weighted binary rubric items for each candidate set and aggregates the resulting judgments into response-level scores. In our experiments with 8B backbones, DynamicRubric improves evaluator performance and provides stronger policy supervision than baselines using a 70B reward model or a 235B static rubric generator. DynamicRubric-optimized policies also show gains on verifiable reasoning and coding tasks. A DynamicRubric-optimized model is fully deployed in WeChat Search’s AI answering scenario, where it serves all online traffic across tens of millions of requests per day and improves key online metrics. These results suggest a principle for evaluator-guided post-training: evaluators should evolve with the policies they supervise.
[AI-15] RUST-ESD: A Risk-Calibrated and Governance-Aware AI Framework for Enterprise Strategic Decision Support Under Uncertainty APWEB
链接: https://arxiv.org/abs/2607.20065
作者: Tian Qiu,Li Yan,Mahabubur Rahman Miraj,Shanqin Yi,Md Intekhab Rahman Galib,Jahid Hasan
类目: Artificial Intelligence (cs.AI)
备注: 15 pages, 7 figures, 4 tables. Submitted to APWeb-WAIM 2026, Danang, Vietnam, September 7-9, 2026
Abstract:Enterprise strategic decision support requires AI systems that are not only accurate, but also uncertainty-aware, risk-calibrated, explainable, and governance-compliant. This paper proposes TRUST-ESD, a risk-calibrated and governance-aware framework for enterprise decision support under uncertainty. TRUST-ESD evaluates feasible counterfactual strategies through predictive utility estimation, conformal uncertainty calibration, CVaR-based downside-risk scoring, risk-memory retrieval, policy-as-code governance, explainability, and human oversight. Unlike prediction-only methods that select actions by maximum expected utility, TRUST-ESD recommends strategies that balance value, reliability, risk exposure, and compliance. Experimental results show that TRUST-ESD improves risk-adjusted utility by 7.95%, reduces risk exposure by 23.22%, reduces CVaR by 23.78%, lowers calibration error by 13.89%, improves explanation fidelity by 10.90%, and increases governance compliance by 9.76% compared with strong uncertainty-aware baselines, while maintaining competitive predictive accuracy. Ablation and case-study analyses further confirm that uncertainty calibration, downside-risk scoring, risk memory, explainability, and governance validation jointly improve trustworthy enterprise decision-making.
[AI-16] PRO-LONG: Programmatic Memory Enables Long-Horizon Reasoning
链接: https://arxiv.org/abs/2607.20064
作者: Alexis Fox,Junlin Wang,Paul Rosu,Bhuwan Dhingra
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Long-horizon tasks require sustained perception, reasoning, and exploration, and are a persistent challenge for large language model (LLM) agents. This gap is reflected in their limited performance on continual learning benchmarks such as ARC-AGI-3, especially when models are evaluated out of the box. Various agent harnesses have been proposed to close this gap, and each commits to a strategy for handling long sequences of observations, i.e., what information to save from the environment and how to load it into model context, a choice we argue is particularly consequential. Existing methods for context management face a significant tradeoff, as preserving more information makes retrieving relevant details less tractable. We propose PRO-LONG, a minimal context management framework built around programmatic memory for LLM agents in long-horizon, exploratory settings. PRO-LONG addresses the tradeoff by keeping a complete, structured interaction log and capitalizing on recent progress in coding agents to search this history efficiently. On the full ARC-AGI-3 public game set, PRO-LONG improves over a base coding agent by an average of 18.0 percentage points across frontier models, and matches or exceeds state-of-the-art specialized harnesses (up to 76.1% pass@1) while using 4.2-5.8x fewer tokens. With Fable 5, PRO-LONG achieves 97.4% best@2 at a total cost of \ 1,750. Relevant code and logs are available at this https URL.
[AI-17] st Case Prioritization for DNNs via Neural Collapse Instability ISSTA2026
链接: https://arxiv.org/abs/2607.20046
作者: Chunyu Liu,Mingyuan Li,Yang Li,Wenmin Li,Fei Gao,Tengfei Tu,Su-Juan Qin
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
备注: 24 pages. Accepted at ISSTA 2026
Abstract:With the widespread deployment of deep neural networks (DNNs) in safety-critical domains, reducing the cost of model validation under limited testing budgets has become increasingly important. Existing test case prioritization techniques often rely on single-checkpoint confidence signals derived from output probabilities. However, DNNs can be confidently wrong, and the confidence margin between the predicted and competing classes is frequently small, which weakens early fault discovery. To address this limitation, we propose a Neural-Collapse-Inspired Prioritization (NCIP) framework that replaces absolute confidence with cross-checkpoint prediction variability in the terminal training regime, where model geometry becomes highly structured. NCIP introduces two key components. First, it selects an NC-guided representative subset of training checkpoints using an equiangularity score of classifier weights, quantified as the standard deviation of pairwise cosine similarities among class weight vectors. Second, it prioritizes test inputs by their prediction variability across the selected checkpoints, surfacing boundary-adjacent and failure-prone samples that are unstable under checkpoint-induced decision boundary shifts. Extensive experiments across multiple datasets and architectures show that NCIP achieves strong performance in early fault discovery compared with competitive baselines, with 1.5 to 16.6 percent RAUC-ALL gains and 4.9 to 20.6 percent RAUC-500 gains under the same testing budget. NCIP further attains the best average performance across all dataset-model pairs.
[AI-18] Global Difference Constraint Propagation for Constraint Programming
链接: https://arxiv.org/abs/2607.20022
作者: Lucas Kletzander,Jip J. Dekker,Andreas Schutt,Peter J. Stuckey
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Difference constraints of the form x - y \leq d are well studied, with efficient algorithms for satisfaction and implication, because of their connection to shortest paths. Finite domain propagation algorithms, however, typically do not make use of these algorithms, and treat each difference constraint as a separate propagator. Propagation does guarantee completeness of solving, but can be needlessly slow. In this paper we describe how to build a (bounds consistent) global propagator for difference constraints that treats them all simultaneously. SAT modulo theory solvers have included theory solvers for difference constraints for some time. While a theory solver for difference constraints gives the basis of a global difference constraint propagator, we show how the requirements on the propagator are quite different. Crucially, we show how to explain propagations by a global difference constraint propagator, in order to use it within a lazy clause generation solver. We give experiments showing that treating difference constraints globally can substantially improve on the standard propagation approach.
[AI-19] EvoDRC: A Self-Evolving Agent ic Framework for Automated DRC Violation Repair
链接: https://arxiv.org/abs/2607.20019
作者: Bing-Yue Wu,Chia-Tung Ho,Haoyu Yang,Brucek Khailany,Vidya A. Chhabria
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Design rule check (DRC) closure remains a major bottleneck in advanced-node physical design. Although detailed routers are rule-aware, residual design rule violations (DRVs) often require manual engineering change order iterations. Automating this process is challenging because repairs must account for complex geometric interactions, preserve circuit connectivity, and avoid introducing new violations. We present EvoDRC, a skill-evolution framework for agentic block-level DRC repair. EvoDRC initializes layer-specific repair skills using knowledge distilled from an unrelated reference design and continuously evolves these skills using traceable repair experience collected from the target design. EvoDRC decomposes the layout into bounded repair regions and assigns an LLM repair agent to each region. Local DRC analysis, connectivity-checking, and impact-preview tools provide feedback on proposed modifications. Repair operations and their resulting DRV changes are stored in a knowledge database and used to evolve the repair skills. Experiments on seven block-level designs from the DAC26 DRC Benchmark show that EvoDRC achieves a 73.5% overall reduction compared to the reported baseline.
[AI-20] Safe Remediation as Risk-Constrained Intervention Decision in Microservice Systems
链接: https://arxiv.org/abs/2607.20005
作者: Chengxiao Dai,Zhaokun Yan,Chenjun Lei,Qiao Li,Luyan Zhang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:In modern IT operations (IT-Ops), the cost of an incorrect repair often exceeds the cost of no action at all. Yet existing automated remediation systems are designed to generate actions rather than to decide whether intervention is warranted, leaving safety as an afterthought enforced by manual approval. This paper makes three contributions to close this gap: (i) we reformulate safe remediation as a risk-constrained intervention decision problem and cast it as a Constrained Markov Decision Process (CMDP), in which the agent maximizes repair success subject to a bounded false remediation rate (FRR); (ii) we introduce a three-dimensional risk decomposition comprising blast radius, reversibility, and epistemic uncertainty, providing operators with an interpretable per-action safety interface; and (iii) we design a context-adaptive human-in-the-loop (HITL) gate that turns escalation from a binary failsafe into a bandwidth-aware control layer responsive to on-call load and business criticality. The full policy is learned offline from historical incident logs, enabling explicit control of the expected FRR. Experiments on the Train Ticket microservice benchmark with Chaos Mesh fault injection and an RCAEval-aligned fault taxonomy show that our framework reduces FRR by 39% while improving repair success by 2.5 points over a strong runbook baseline, and reduces on-call escalation load by 17% relative to a fixed-threshold variant.
[AI-21] aming the Security-Energy Paradox: A Green AI Approach to Optimized Android Malware Detection
链接: https://arxiv.org/abs/2607.20003
作者: Shrinidhi Sridhar,Vikas K. Malviya
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 6 pages, 1 figure
Abstract:An increase in advanced Android malware requires the use of deep learning models, which can run on Android devices. But there is a trade-off between security and energy use, as strong detection models can drain the battery of devices fast. This work tests different Multi-Layer Perceptron (MLP) model configurations to balance malware detection performance and energy efficiency. In this work, we compared standard FP32 models with optimized INT8 quantized neural networks with different model depths using TUANDROMD and DREBIN datasets for both classification performance and energy consumption. The results show that INT8 quantization reduces model size by about 3.5 times with a decrease in energy consumption to 0.0189 mJ per inference, while maintaining more than 99.2% detection accuracy. We found that shallow quantized architectures, such as 3-layer and 4-layer QNNs, reduce energy costs by improving throughput and shortening the time of CPU operating in a high-power state. This work shows that efficient malware protection can be achieved on resource-constrained smartphones and provides a foundation for Green AI in mobile security.
[AI-22] Post-Training in Time Series Foundation Models: A Unifying Framework
链接: https://arxiv.org/abs/2607.20002
作者: Shifeng Xie,Ambroise Odonnat,Zehao Xiao,Lei Zan,Malik Tiomoko,Lujia Pan,Themis Palpanas,Boris N. Oreshkin,Chenghao Liu,Keli Zhang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Time series foundation models (TSFMs) have emerged as general-purpose models for time series analysis, but pretraining alone is often insufficient for reliable downstream deployment. Bridging this gap requires further intervention to handle domain shift, task heterogeneity, limited supervision, and computational constraints, which motivates post-training as a broad class of methods to adapt, augment, compose, calibrate, or specialize pretrained TSFMs for downstream tasks. In this work, we analyze TSFM post-training methods based on their locus of intervention in the prediction pipeline, yielding five categories: parameter adaptation, context augmentation, model composition, output processing and uncertainty control, and compression and specialization. Within each category, we study main representative methods and discuss their current limitations. We further identify future directions toward controlled adaptation, reliable context construction, uncertainty-aware model composition, calibrated output processing, and deployment-aware specialization. Overall, by providing a unifying framework for the emerging TSFM post-training landscape, this work aims to support future research to navigate the design space between a pretrained TSFM and its reliable downstream deployment.
[AI-23] CLARK: Closed-loop Learning for Adaptive Reasoning over Knowledge Graphs
链接: https://arxiv.org/abs/2607.19996
作者: Yousef Khan,Luca Gherardini,Marco Maratea,Joel Arrais,Jose Sousa
类目: Artificial Intelligence (cs.AI)
备注: 18 pages, 3 Figures, 1 algorithm, 3 tables, submitted to The International Semantic Web Conference’26
Abstract:Machine Learning models are widely used for automating classification tasks by extracting statistical patterns from data. However, their performance deteriorates if the data distribution changes, making them ill-suited to handle uncertain and evolving information. Moreover, they provide limited support for integrating prior knowledge. To address these limitations, we present CLARK (Closed-loop Learning for Adaptive Reasoning over Knowledge Graphs), a framework that integrates knowledge graphs, symbolic rule mining, and probabilistic reasoning under the Logic Programs with Markov Logic Networks (LP ^\textMLN ) formalism. Starting from CACTUS-derived KGs, CLARK translates graph structure into an LP ^\textMLN program and iteratively enriches it with candidate rules proposed by symbolic learners. These rules are calibrated through probabilistic weight learning, enabling reasoning under uncertainty and refinement of the underlying graph structure. We evaluate CLARK on two medical datasets, analysing both rule quality and downstream classification performance. Results demonstrate that CLARK leads to improved classification performance and more generalisable inference. Overall, CLARK provides a principled approach to constructing adaptive, interpretable, knowledge-driven models for classification.
[AI-24] Coordinating from Memory: Graph-Structured Experience Reuse for Multi-Agent Adaptation in Dynamic Manufacturing
链接: https://arxiv.org/abs/2607.19985
作者: Chengxiao Dai,Zhanhui Lin,Zhaokun Yan,Youyang Ni,Chenjun Lei,Luyan Zhang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Dynamic manufacturing environments require multi-agent systems to coordinate effectively under frequent operational disturbances such as machine failures, urgent job arrivals, and processing time variations. Existing multi-agent reinforcement learning approaches treat each disturbance episode independently, discarding valuable coordination experience that could accelerate future adaptation. In this paper, we propose a Graph-Structured Experiential Memory (GSEM) framework for multi-agent coordination in dynamic manufacturing. The framework encodes historical coordination episodes as heterogeneous relational graphs that capture task dependencies, machine states, and inter-agent collaboration patterns. When a new disturbance occurs, a graph neural network-based retrieval mechanism identifies structurally similar past episodes, enabling experience-guided policy adaptation rather than learning from scratch. Experiments on dynamic flexible job-shop scheduling benchmarks with three disturbance types show that GSEM reduces makespan by 4.1%-10.0% and adaptation time by 33%-38% compared to the strongest memory-augmented baseline, with the advantage increasing under higher disturbance frequency. Ablation studies and cross-disturbance transfer experiments further validate the necessity of graph-structured encoding and similarity-based retrieval and demonstrate the cross-disturbance generalizability of learned coordination patterns.
[AI-25] me Series Network Utilization KPI Forecasting Using Advanced AI/ML Models
链接: https://arxiv.org/abs/2607.19974
作者: Niraj Gadhe,Kirti Bhardwaj,Moulik Jain,Shubhi Sharma,Vinay Saini
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:The rapid proliferation of data-intensive applications, cloud infrastructure, and IoT ecosystems has made proactive resource provisioning critical for maintaining optimal network performance. However, network administrators face a constant battle against capacity constraints, where traditional reactive approaches fail to accurately anticipate traffic fluctuations. This inability to foresee demand leads to costly over-provisioning, unexpected downtime, and degraded quality of service directly impacting operational budgets and business continuity. To achieve efficient capacity planning, accurate forecasting of bandwidth utilization is essential. This study addresses the challenge by evaluating a diverse spectrum of models including seasonal decomposition, Prophet, Random Forest, XGBoost, Support Vector Regression, and advanced deep learning architectures like bidirectional and Convolutional LSTMs - using a common interface dataset benchmarked across MAPE, NRMSE, and R-square metrics. Ultimately, this research delivers actionable insights into the trade-offs between model accuracy and computational efficiency, empowering engineers, operators, and business owners to select the optimal forecasting model for their specific infrastructure needs.
[AI-26] he Giant Hippocampus: From Structural Monoculture to a System of Systems
链接: https://arxiv.org/abs/2607.19973
作者: Jaeho Seol
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE); Neurons and Cognition (q-bio.NC)
备注: 48 pages, 23 figures
Abstract:AI researchers describe state-of-the-art models as one thing repeated at scale: the Transformer, wired identically for text, pixels, or speech. Neuroscientists describe the cortex as a mosaic - dense Layer 4 in visual cortex for spatial encoding, thick Layers 5/6 in motion cortex for temporal integration - different jobs solved by different structures. This paper argues the gap is a structural error, not a stylistic one, and is measurable. A century of cytoarchitecture, from Brodmann to single-cell Patch-seq, shows distinct cognitive functions are implemented by qualitatively different structures, not by rescaling one template. The convolutional neural network is the field’s own proof: local receptive fields and hierarchical depth encoded this prior directly, reaching strong image recognition on far less data than later architectures needed. The paper traces how this lesson was discarded: the “Hardware Lottery” made the Transformer the path of least resistance, not the principled choice, and Mixture-of-Experts, often cited as diversity, in fact partitions parameters among identical experts. A functionalist analysis shows the Transformer is best understood as a functional analog of the hippocampal formation, not a general-purpose cortex - the same mistake as treating cortex as one giant Broca’s area, except the field has now standardized on a giant hippocampus, applied to tasks it was never built for: audition, executive gating, working memory. The paper closes with an alternative: a Heterogeneous Topological Network, a System of Systems in which distinct modules keep the inductive bias their computation demands and communicate through standardized interfaces. This is a design discipline for AI architects, not cognitive science: specify modularity before training, using structural evidence as a design input rather than reverse-engineering architecture from a trained model’s behavior.
[AI-27] EvoThink: Evolving Thinking in Large Reasoning Models via Self-Pruning and Aha-Moment Preference Optimization IJCAI2026
链接: https://arxiv.org/abs/2607.19962
作者: Xinbang Dai,Zheyu Xin,Huikang Hu,Lin Ren,Rihui Jin,Guohui Xiao,Guilin Qi,Kuicai Dong,Zhaocheng Du,Yuyang Zhang
类目: Artificial Intelligence (cs.AI)
备注: 9 pages, 7 figures, accepted by IJCAI 2026
Abstract:Large Reasoning Models (LRMs) often suffer from overthinking due to redundant verification steps. Existing approaches for mitigating overthinking, such as fast-slow thinking switching and reasoning trajectory compression, fail to make a fine-grained distinction between beneficial and redundant steps within the LRM’s reasoning process, and may thus impair reasoning capability in their pursuit of efficiency. To simultaneously improve reasoning efficiency and capability, we propose EvoThink, a framework that reduces redundant verification and encourages the exploration of new reasoning paths. EvoThink comprises two key components: Self-Pruning Training (SPT), an unsupervised method that iteratively prunes redundant reasoning steps and self-trains on the concise trajectories; and Aha-Moment Preference Optimization (AMPO), which, inspired by genetic algorithms, identifies valuable failed reasoning attempts, synthesizes from-wrong-to-right aha-moment data, and optimizes the model to internalize this reasoning pattern. Extensive evaluations across mathematical reasoning and code generation benchmarks demonstrate that EvoThink not only substantially reduces inference-time token usage but also improves the reasoning capability of LRMs.
[AI-28] HijackKV: New Threat in Position-Independent KV Cache Reuse USENIX-SECURITY2026
链接: https://arxiv.org/abs/2607.19957
作者: Yichi Zhang,Zhiqi Wang,Huan Zhang,Yuchen Yang
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 20 pages, accepted by USENIX Security 2026
Abstract:Key-Value (KV) cache reduces inference latency in large language models (LLMs). Traditional prefix-based reuse has low cache hit rates across inference requests because it requires exact token and position matches. To improve efficiency, recent system optimizations introduce position-independent KV reuse, allowing KV cache to be reused whenever identical text chunks appear, regardless of their position in the sequence. We show this design introduces a new threat, KV Cache Hijacking. Since KV caches are retrieved by token match but encode the context in which they were originally computed, the KV tied to a benign-looking token chunk may encode an attacker-controlled prefix. When later reused in a victim query, this contaminated KV silently hijacks the model’s behavior, even if no attacker-controlled text appears in the input. We introduce HIJACKKV, the first attack framework that systematically exploits this vulnerability, demonstrating its severity and practicality. HIJACKKV optimizes an attacker-controlled prefix, so that the KV computed for a subsequent common benign text encodes the attacker’s goal, while the text remains unchanged for future cache hits. HIJACKKV achieves an average 94% success rate in a single attempt, remains effective under realistic constraints including low hit rates (10%) and frequent recomputation (50%), persists over multi-turn interactions, and transfers across models in black-box settings. We further provide design insights for building secure KV reuse systems. Comments: 20 pages, accepted by USENIX Security 2026 Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG) Cite as: arXiv:2607.19957 [cs.CR] (or arXiv:2607.19957v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2607.19957 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-29] SenWorld: A Digital-Twin Simulation for Generating Context-Rich Evaluation Data
链接: https://arxiv.org/abs/2607.19949
作者: Zenghui Zhou,Xiaoyang Li,Xiaoxuan Qiao,Zhilang Wei,Tianming Lei
类目: Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注:
Abstract:Smartphone personal assistants reason over longitudinal personal data, yet evaluating them requires context-rich evaluation data whose correct answers are known, and real device traces are too privacy-sensitive to share. To address this challenge, we present SenWorld, a physically grounded, deterministic, event-sourced digital-twin simulation that generates such data with ground truth fixed by construction. In SenWorld, personas live through a full day in a world built from real map, weather, holiday, and network data; every observable signal is archived in full-system snapshots; and each evaluation case is labeled by a pointer to an existing record rather than by post-hoc annotation or a large language model (LLM) judge. We evaluate this method with 16 personas in Beijing. The generated data closely matches the held-out real-user benchmark in category distribution (Jensen–Shannon divergence (JSD) 0.070) and in the daily rhythm of communication records (JSD below 0.1), though generated records remain shorter than real ones. Without scripted interaction, personas form a fully reciprocated dialogue subgraph and differentiated behavioral repertoires. Projected into 717 evaluation cases, the generated data exposes 78 failures in a production smartphone assistant, concentrating on call and Short Message Service (SMS) records while contacts, schedules, and alarms never fail. The snapshot pointer confirms each failure as an assistant-side retrieval error, with no LLM judge involved. Overall, SenWorld offers a privacy-safe, reproducible, and distribution-checked path to evaluation data whose labels are fixed by construction.
[AI-30] MOF-Sleuth: Tool-Grounded Reward Alignment for Explainable Fine-Grained MOF CIF Auditing
链接: https://arxiv.org/abs/2607.19935
作者: Yu Liu,Zhiwei Yang,Diandian Guo,Kun Peng,Fangfang Yuan,Cong Cao,Chaozhuo Li,Zhiyuan Ma,Yanbing Liu,Guobin Zhao
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large metal-organic framework (MOF) databases support simulation, screening, and machine learning through crystallographic information files (CIFs). Subtle chemical and structural errors in these inputs can compromise downstream results and hinder manual inspection. LLM advances in computational chemistry offer paths beyond predictive screening toward fine-grained diagnosis with evidence-grounded explanations. However, two challenges remain: (i) limited fine-grained attribution: MOF-specific validators and machine-learning models scale detection but provide fixed checks, readiness scores, or coarse labels rather than evidence-grounded explanations; and (ii) unreliable CIF reasoning: direct LLM auditing is costly and unreliable because chemical evidence is implicit across atom-site records and requires geometric, connectivity, occupancy, and charge calculations. Both stem from weak coupling between chemical evidence and language-model explanation. We introduce MOF-Sleuth, a reinforcement-guided CIF auditing agent with two modules: a deterministic Forensic Lab and a Sleuth reasoning engine. The Lab derives composition, geometry, connectivity, occupancy, coordination, and charge evidence, and Sleuth uses this evidence to produce an evidence-grounded explanation, error types, and a binary decision. Reward-guided reinforcement learning (RL) turns tool measurements into chemical explanation-level supervision, rewarding not only the final answer but also cited chemical evidence and evidence-supported diagnoses. We introduce Chemically Grounded Diagnosis (Chem-GD), a metric that assesses whether a correct diagnosis is explained by factual, relevant CIF-derived evidence. Across four benchmarks, MOF-Sleuth establishes state-of-the-art performance among LLM-based approaches and MOF-specific machine-learning methods, demonstrating gains in detection, attribution, and grounded explanation quality.
[AI-31] Long-Term Sequential Decision Making under Risk UAI2026
链接: https://arxiv.org/abs/2607.19914
作者: Irmaan(Mohammad)Mirzanejad,Nadjet Bourdache,Abdel-Illah Mouaddib
类目: Artificial Intelligence (cs.AI)
备注: Accpeted in Forty-Second Annual Conference on Uncertainty in Artificial Intelligence (UAI 2026)
Abstract:We study finite-horizon MDP planning under \emphroot-based (resolute) risk objectives that apply a rank-dependent functional to the distribution of total returns. Such objectives are non-linear in the return distribution and generally break Bellman optimality, so direct optimization by scenario-tree enumeration is intractable. We propose \textbfERQDP, an enumeration-free and sampling-free method that solves a rank–quantile surrogate via exact DP (Dynamic Programming), evaluates candidate policies exactly by DP over return Probability Mass Functions (PMFs) on a discretized return grid (with an explicit rounding bound), and refines the surrogate in an anytime loop that reports an explicit upper–lower gap (certificate) for the target objective up to discretization budgets. Across tested benchmarks, ERQDP returns certified solutions or explicit residual gaps, enables fast risk-parameter sweeps with substantial runtime gains, and supports both risk-averse and risk-seeking behaviors.
[AI-32] Defense Against LLM Backdoors using Critical Neuron Isolation Pruning
链接: https://arxiv.org/abs/2607.19894
作者: Yuxi Li,Zhibo Zhang,Kailong Wang,Xingshuo Han,Ling Shi,Haoyu Wang
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models (LLMs) are vulnerable to backdoor attacks, where hidden triggers induce malicious outputs. Existing defenses generally fall into inference-time detection or training-time mitigation, but face two key limitations. First, they focus on fine-tuning-based backdoors (e.g., PEFT modules) and fail to address insidious model-editing attacks that bypass training pipelines. Second, they target simple classification settings and do not naturally extend to open-ended LLM generation and do not naturally extend to the open-ended generation characteristics of LLMs. Consequently, these methods focus on surface-level behavioral patterns while neglecting the deeper representational causes of malicious activations. This lack of mechanistic understanding forces defenses to depend on empirical heuristics, limiting their robustness, generality, and practical applicability in real-world LLM deployment. To bridge this gap, we introduce DeCNIP (Defense with Critical Neuron Isolation Pruning), which leverages representational analysis to identify and neutralize backdoors in a unified pipeline. Specifically, DeCNIP identifies trigger-like behaviors by optimizing a cross-entropy loss between harmful prompts with candidate tokens and benign inputs. This representational discovery exposes latent threats by uncovering mechanisms through which triggers hijack model weights. It then isolates Backdoor Critical Neurons (BCNs) and prunes them selectively to remove malicious influence while preserving model utility. Extensive evaluations on six open-source LLMs and two benchmark datasets demonstrate that DeCNIP achieves over 95% relative reduction in Attack Success Rate (ASR), outperforming seven state-of-the-art defenses with only 0.1% neuron intervention. Moreover, it maintains 97% of the model’s performance on normal benchmarks, demonstrating its efficacy, robustness, and scalability. Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.19894 [cs.CR] (or arXiv:2607.19894v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2607.19894 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-33] Beyond Fail-to-Pass: Iterative Hardening of Co-Generated Bug Reproduction Tests and Fixes
链接: https://arxiv.org/abs/2607.19843
作者: Yuhao Tan,Zhibang Yang,Fangkai Yang,Yuan Yao,Yu Kang,Lu Wang,Pu Zhao,Xin Zhang,Xiaoxing Ma,Qingwei Lin,Saravan Rajmohan,Dongmei Zhang
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: 29 pages, 5 figures, preprint
Abstract:Large language models (LLMs) have made automated program repair (APR) increasingly practical for real-world bugs, but repairing directly from bug reports remains underconstrained. Bug reproduction tests (BRTs) help close this gap by turning a bug report into an executable, bug-specific signal that can guide repair and validate candidate patches. Existing work has therefore studied BRT generation as a core subproblem in APR and mainly evaluates a generated BRT using the fail-to-pass (F-P) criterion, which requires the test to fail on the buggy code but pass on the golden fix. We show that F-P alone is insufficient when the goal of a BRT is to improve downstream repair. In particular, some F-P BRTs are lax, reproducing the observed symptom yet still admitting plausible-but-incorrect patches. We formalize this missing quality dimension by separating F-P BRTs into rigorous and lax ones, and show empirically that only the former consistently improve repair success. We further find that co-generation introduces test–fix error coupling, where the in-trajectory fail-to-pass (F-P) check can pass even when both the generated patch and generated test are wrong. Based on these findings, we propose CoHarden, a co-generation framework that uses the Lax signal as an in-loop convergence criterion. CoHarden first generates a test before any fix, then iteratively hardens the test and fix against surviving mutation patches until the generated test no longer admits Lax regressions. Experiments show that CoHarden reaches 69.4% Resolved and 78.9% F-P on SWE-bench Verified, outperforming the strongest fix-only and cogeneration baselines by +9.6 and +7.9 percentage points in Resolved, respectively, with consistent gains across LLM backbones and benchmarks.
[AI-34] Know Your Agent : Reconnaissance-Driven Pentesting of AI Agents
链接: https://arxiv.org/abs/2607.19837
作者: Or Zion Eliav,Eyal Lenga,Shir Bernstien,Yisroel Mirsky
类目: Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR); Machine Learning (cs.LG)
备注:
Abstract:Traditional pentesting uses reconnaissance at each step to uncover unseen weaknesses, build stronger attacks, and advance the objective; we argue that AI agents require the same treatment. We formalize agent reconnaissance by modeling the process and identifying the knowledge assets it seeks to extract: what they are, how they are used, and which agent weaknesses they exploit to give adversaries leverage in indirect prompt injection attacks. We instantiate these insights in Know Your Agent (KYA), a framework that automates black-box, reconnaissance-driven pentesting by probing agents, building target profiles, and using those profiles to craft stronger attacks. We evaluate KYA on agent-security benchmarks and a real-world coding agent, and release KYA, its benchmarks, and baseline implementations for reproducibility.
[AI-35] OPIUM: Mitigating Steering Externalities and Over-Refusal via Dual Objective Latent Optimization ICML2026
链接: https://arxiv.org/abs/2607.19806
作者: Kavin Aravindan,Arihant Rastogi,Aadi Prasad,Krishak Aneja,Saiyam Jain,Vaishnavi Shivkumar,Ponnurangam Kumaraguru
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Accepted to the Mechanistic Interpretability Workshop at the 43rd International Conference on Machine Learning (ICML 2026)
Abstract:Activation steering provides a lightweight mechanism for controlling large language models at inference time, but steering vectors can have unintended externalities: utility vectors may weaken safety behavior, while refusal vectors may induce over-refusal on benign prompts. We introduce OPIUM (Optimizing Protected Injections via Utility Manifolds), a training-free method for sanitizing steering vectors through representation matching. Given reference behaviors on two prompt sets, OPIUM optimizes a new steering vector that preserves the downstream representations induced by the desired intervention while matching a safer reference behavior on prompts where the original vector fails. Across steering-externality and over-refusal settings, OPIUM improves the safety–utility tradeoff relative to vanilla steering and directional ablation, suggesting that harmful side effects of activation steering can often be mitigated directly in activation space.
[AI-36] RPPNet: Perceptually-Grouped Rhythm-Pitch Primitives for Long-Term Structure Melody Generation via Boundary-Aware Modeling
链接: https://arxiv.org/abs/2607.19776
作者: Tieyao Zhang,Yuke Liu,Jiaxing Yu,Xinda Wu,Kejun Zhang,Genfang Chen
类目: ound (cs.SD); Artificial Intelligence (cs.AI)
备注:
Abstract:Existing symbolic music generation models typically use bars as the basic structural unit. However, human perception of musical phrases often does not align with notated bar lines, leading to long-term structural fragmentation. This paper proposes RPPNet-a two-stage deep learning architecture with variable structural boundaries. It first generates variable-length Rhythm-Pitch Primitive (RPP) sequences, where each RPP encodes note count, rhythm, and contour; then decodes the RPP sequences into concrete notes. The grouping of RPPs is automatically derived from acoustic cues, auditory inertia, and similarity perception based on music psychology. Experiments show that melodies generated by RPPNet are superior in both long-term structure and musicality, with significant improvements across all subjective evaluation dimensions. Ablation studies confirm that the performance gain stems from the structural correctness of the psychological representation, rather than from model capacity. This work offers an interdisciplinary perspective for music generation, integrating music theory, computational modeling, and music psychology.
[AI-37] An Isotropy-Preserving Spectral Cap for Muon: Theory and Three Case Studies
链接: https://arxiv.org/abs/2607.19771
作者: Jiachun Li
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Preliminary Report; Larger scale experiments ongoing; Comments are welcome
Abstract:Muon and related matrix-sign optimizers are increasingly used to pre-train large language models, but their effect on the internal geometry of individual weight matrices is not well understood. This preliminary report proposes a unified framework built on a single idealizing assumption – exact scale invariance of the loss under weight rescaling, which holds approximately in normalization-heavy networks. Under this assumption, plain SGD carries a built-in 1/||W|| brake on its update size, whereas Muon’s matrix-sign step removes that brake, so both the Frobenius and spectral norms drift outward faster (t^1/2 versus t^1/4). We further observe that the spectral-norm perturbation has a non-negative second-order term. This implies that a lightweight “spectral cap” – which projects out only the first-order growth of the single top singular direction from each update – can control the output covariance W K_X W^T without freezing training: the weight keeps learning through non-top directions, top-direction rotation, and top switching. We relate this cap to the min-entropy (H-infinity) of the singular-value spectrum. We then study three systems trained with Muon: a nanoGPT feed-forward projection, a 64-expert mixture-of-experts router, and the query/key projections of a bf16 FlashAttention block. In each case the cap increases isotropy and, at the margins – a router collapsing to a single expert, and the near-divergence of one attention head – prevents a concrete failure, while leaving validation loss essentially unchanged. We emphasize that the scale-invariance assumption is strong and that these small-scale results are preliminary; comments are welcome.
[AI-38] Symbol and Footprint Database for Electronic Components by Agent ic Recognition and Generation
链接: https://arxiv.org/abs/2607.19767
作者: Yichen Shi,Yuzhi Liu,Zhuofu Tao,Li Huang,Yuhao Gao,Ting-Jung Lin,Lei Hel
类目: Artificial Intelligence (cs.AI)
备注: Accepted by PRCV 2025
Abstract:A rich and recognizable component library is the cornerstone of printed circuit board (PCB) design and generation. Traditionally, engineers manually create symbols and footprints and design PCB schematics, which is time-consuming and error-prone. Leveraging multimodal large language models (MLLMs), we develop SFgen, an agentic recognition and generation flow of symbol and footprint for electronic components. SFgen achieves 86% accuracy for symbol generation and 80% accuracy for footprint generation. We use the SFgen method to create SFnet, a database of symbols and footprints. It now has 1000 components and is expanding constantly, which lays the foundation for automatic generation of PCB designs.
[AI-39] Convergence-Latency-Aware Adaptive Modulation and Resource Allocation in RIS-Assisted Wireless Federated Learning
链接: https://arxiv.org/abs/2607.19759
作者: Liwei Wang,Wen Chen,Jun Li,Qingqing Wu,Ming Ding,Xusheng Zhu,Qiong Wu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Information Theory (cs.IT)
备注:
Abstract:Federated learning (FL) over wireless networks suffers from significant training latency and degraded convergence due to unreliable wireless transmission, especially under blocked propagation environments. Although reconfigurable intelligent surfaces (RISs) can improve communication reliability, existing wireless FL studies rarely characterize the trade-off between learning convergence and communication delay under modulation-dependent transmission errors. In this paper, we consider a wireless FL system operating under RIS-assisted blocked-link propagation scenarios, and focus on adaptive modulation and sub-channel allocation for convergence-latency aware communication design. By characterizing the effect of symbol errors on uploaded local gradients, we derive a convergence-related upper bound that reveals the impact of symbol error rate (SER) on FL loss decay. Based on this result, we formulate a joint convergence-latency optimization problem, which is cast as a mixed-integer nonlinear programming (MINLP) problem, and solve it using a low-complexity hybrid alternating optimization framework. Extensive experiments on MNIST, CIFAR-10, and Speech Commands show that the proposed scheme consistently achieves faster convergence and higher test accuracy than existing adaptive communication schemes, especially in complex tasks and challenging wireless scenarios.
[AI-40] he World Model Remembers the Actor Forgets: Dream Rehearsal for Continual Model-Based RL
链接: https://arxiv.org/abs/2607.19749
作者: Gurp Nijjer
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 11 pages, 2 figures. Code, pre-registration trail, and run data: this https URL
Abstract:Model-based reinforcement-learning agents of the DreamerV3 family forget catastrophically when trained on task sequences, even when an unbounded replay buffer preserves every earlier experience. We ask a question the continual-RL literature has assumed an answer to but never measured: which component forgets? Under never-clear replay, pre-registered component-level probes (n=3 seeds throughout) show that the world model retains essentially everything measurable about old tasks – reward discrimination (retention ratio ~1.0), value estimates, and termination structure – while the actor’s behavior collapses. Forgetting in this regime is a channel problem, not a memory problem. We demonstrate this by intervention: with the world model frozen and identical imagined rollouts, reinforcement learning in imagination fails to recover a lost skill (0/3 seeds), while supervised self-imitation on the world model’s own graded dreams recovers it on 3/3 seeds with zero environment interaction. Interleaved during training, this graded dream rehearsal yields a task-label-free, parameter-constant continual learner: 3/3 four-task chains retained where plain replay passes 0/3, 3/3 eight-task chains, and consistent gains over matched real-episode cloning (paired difference +0.13, bootstrap 95% CI [0.07, 0.24], complete seed separation). The dream-grading step is load-bearing: we characterize two scoring failure modes, provide an offline selection gauge that caught both before they contaminated results, and give a realized-first grading rule that closes them. All experiments were pre-registered with committed protocols; every refuted hypothesis is reported.
[AI-41] An Automated Framework for Extracting Reachable Attack Chains from Cyber Threat Intelligence Reports
链接: https://arxiv.org/abs/2607.19742
作者: Wenbo Hou,Ning Hu,Xueping Wang,Jiahao Gu,Wenjian Luo
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注: 16 pages, 3 figures
Abstract:Cyber Threat Intelligence (CTI) reports richly describe real-world attack processes, but their unstructured narratives cannot be directly used for automated attack-path reasoning. Existing CTI extraction methods focus on indicators, entities, or TTP labels without modeling the execution conditions and resulting states of each attack step, so the extracted knowledge supports neither state matching nor reachability analysis across multi-stage attack chains. This paper proposes an automated framework that extracts reachable attack chains by modeling each attack step as an attack unit of preconditions, an attack behavior, and postconditions. A multi-stage pipeline assisted by large language models (LLMs) extracts attack behavior skeletons, recovers their preconditions and postconditions, normalizes them into predefined predicates, and repairs broken dependencies; the resulting units are compiled into Datalog-style rules for attack-goal reachability reasoning. On a dataset of 20 CTI reports containing 334 human-validated annotated steps, our framework achieves higher annotated-step coverage than representative CTI extraction systems in recovering attack behaviors. Moreover, by explicitly generating preconditions and postconditions, it produces attack units that are more complete and consistent than those generated by end-to-end LLM baselines. On the extracted chains, Datalog inference reaches the specified attack goal in 19 of 20 reports, while backward search yields 34 attack paths under the generated rules. The source code and experimental artifacts are available in an anonymized repository. .
[AI-42] Did Alice Do Wrong? Cross-Cultural Differences in Student Perceptions of Generative AI Use in University Computing Education
链接: https://arxiv.org/abs/2607.19699
作者: Brian Harrington,Irina Zlotnikova,Gayathri Nadarajan,Samuel Ekundayo
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI)
备注:
Abstract:The rise of generative AI (GenAI) in higher education has prompted urgent debates surrounding academic integrity and ethical use. This study examines cross-cultural differences in student perceptions of GenAI use, comparing responses from students at Canadian and South Korean universities. Using a scenario-based survey administered in Fall 2024, we analyzed how students judged the ethicality and rule compliance of AI-assisted coding practices. Results reveal that Canadian students were consistently more likely to perceive the use of GenAI as both unethical and against institutional policies compared to Korean students, despite functionally identical institutional policies. Statistical analysis, including Mann-Whitney U tests and correlation coefficients, demonstrated significant differences across nearly all scenarios. Analysis of the factors used in generating scenarios indicated that the amount of AI-generated code incorporated into assignments most strongly influenced ethical judgments. Findings were interpreted through Hofstede’s cultural dimensions framework, suggesting that cultural factors such as power distance, individualism, and uncertainty avoidance significantly shape students’ ethical reasoning regarding GenAI. Our results contribute to the growing body of evidence emphasizing that equitable AI integration in education must be culturally responsive, taking into account diverse conceptions of academic integrity. We advocate for the development of nuanced AI-use guidelines that are sensitive to local cultural contexts while upholding fundamental principles of academic honesty. This study highlights the need for ongoing cross-cultural research to inform ethical AI policies and support responsible GenAI use in global higher education settings.
[AI-43] Edge Intelligence in Civil Aviation: Paradigms Techniques and Applications
链接: https://arxiv.org/abs/2607.19676
作者: Wenbin Li,Zhongtian Liao,Bolin Liu,Yongjie Zhou,Jingling Wu,Xiaoyong Lin,Jing Chen
类目: Artificial Intelligence (cs.AI)
备注: 8 pages, 2 figures, 2 tables
Abstract:Civil aviation is safety critical and its operations, from flight decks and towers to ramps and maintenance, generate massive, heterogeneous data at the network edge. Yet cloud centric deployment of large Artificial Intelligence (AI) models often produces high task latency, lacks offline capability in communication denied environments, and requires centralizing sensitive data, raising privacy and sovereignty risks. Edge AI moves perception, prediction, and decision logic closer to the data producers via compression, collaborative inference, and split learning, thereby reducing latency, bandwidth, and exposure while enabling graceful operation during disconnections. This paper provides a panoramic view and a common understanding of edge intelligence tailored to civil aviation. We firstly articulate the operational motivations for edge AI, and then review recent techniques for edge inference and edge learning. We then introduce the organizational computing paradigms and the respective configurations in civil aviation environments; finally, we describe the emerging applications and the future research trends of edge intelligence in civil aviation. We argue that a refined edge solution can complement cloud foundations to deliver low latency, privacy preserving, and resilient AI services across the civil aviation lifecycle.
[AI-44] FedLSG: LLM -Enhanced Semantic Calibration for Federated Graph Backdoor Defense
链接: https://arxiv.org/abs/2607.19674
作者: Chenyu Zhou,Yabin Peng,Wei Huang,Kunlin Li,Shuaishuai Zhang,Xinyuan Miao
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Federated Graph Neural Networks (FedGNNs) are highly vulnerable to backdoor poisoning, yet existing defenses typically rely on rule-based approaches that lack semantic understanding, making them vulnerable to stealthy triggers and harmful to benign structures. To solve this, we present FedLSG, the first framework that integrates large language models (LLMs) into federated graph backdoor defense. FedLSG introduces a graph and behavior to text grounding scheme that transforms local graph structures and client update behaviors into semantically rich natural language representations. The framework further adopts a lightweight student-teacher architecture. On the server side, a full scale LLM serves as a teacher, providing global contextual guidance and evaluating client updates during aggregation to identify potentially malicious participants. On the client side, a LoRA-based student is maintained to perform semantic reasoning, to suppress the influence of edges associated with backdoor triggers. By enabling semantic interpretation of both graph patterns and client behaviors, the framework adaptively incorporates rule-based signals into message passing and client aggregation for defense. Experiments demonstrate that FedLSG significantly improves resistance to backdoor attacks without compromising graph integrity.
[AI-45] PerfAgent : Profiler-Guided Iterative Refinement for Repository-Level Code Optimization
链接: https://arxiv.org/abs/2607.19653
作者: Ryan Deng,Yuanzhe Liu,Bastian Lipka,Yao Ma,Xuhao Chen,Tim Kaler,Jatin Ganhotra
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language model (LLM) agents now perform well on correctness-oriented repository-level tasks, including SWE-Bench issue resolution and feature implementation in real codebases. However, they still struggle with repository-level code optimization, which requires preserving behavior while improving runtime performance. Passing tests is not enough in this setting; a patch must preserve behavior, implement code optimization, and approach expert speedups. Current agents often miss bottlenecks hidden behind abstraction layers and native extensions, stop after shallow speedups, or insufficiently test the code patches that thus may silently break edge cases. We present PerfAgent, a profiler-guided, verifier-in-the-loop workflow that gives an off-the-shelf coding agent the feedback needed to find real hotspots, improve beyond the first passing patch, and use profiler evidence rather than timing alone to decide what to optimize next. On two challenging optimization benchmarks, GSO and SWE-fficiency-Lite, PerfAgent more than doubles the rate of expert-matching patches over OpenHands with GPT-5.1, improving from 19.6% to 39.2% on GSO and from 26% to 74% on SWE-fficiency-Lite. It also surpasses an oracle best-of-five baseline at substantially lower cost, showing that the gains come from better feedback rather than additional test-time sampling.
[AI-46] Anatomy of a Sound Neural Reason er: One-Shot Amortization First-Pass Poisoning and Search Inertness in Clue-Rich Completion
链接: https://arxiv.org/abs/2607.19635
作者: Aleksey Komissarov
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Neural solvers are built to deduce, branch, and revise intermediate states. The Lattice Deduction Transformer (LDT) appears to do exactly that. In clue-rich Sudoku, it does not: one forward pass commits essentially the entire grid (every blank cell on standard 6x6, 94-96% on augmented 9x9), turning the iterative solver into a one-shot predictor wrapped in an exact verifier. All hard-slice failures are decided before search begins, when the first pass confidently deletes a value required by the true solution. We call this first-pass poisoning. Adding learned branching, MRV, backtracking, value exclusion, and shared nogoods (CoLT) does not change which Sudoku instances are solved; it cuts repeated invalid derivations 1,497-fold. At the frozen training budget, constraint-graph attention alone matches full-CoLT accuracy, while positional tables recover only under substantially longer training, indicating an optimization and sample-efficiency advantage rather than an absolute capacity difference. The diagnosis predicts two effective interventions. Digit-permutation augmentation raises 9x9 accuracy from below 1% to 96.5 +/- 0.3 across three training seeds on a symmetry-disjoint split. Test-time union over symmetry-transformed passes raises all three hard-slice checkpoints from 72.8-78.9% to 100% without retraining. On from-scratch graph coloring, one-shot behavior disappears and search changes accuracy. In clue-rich completion, LDT-like systems are one-shot amortized predictors rather than learned search procedures: accuracy is determined by calibration and symmetry, while search primarily removes computational waste.
[AI-47] Understanding Developer Pain Points in Federated Learning: Insights from Stack Overflow and GitHub
链接: https://arxiv.org/abs/2607.19621
作者: Sahand Saed,Khairul Alam,Banani Roy
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: 44 pages
Abstract:Federated Learning (FL) enables collaborative model training without centralizing raw data, but building and operating FL systems remains difficult due to distributed execution, rapidly evolving frameworks, and privacy and governance requirements. In this paper, we present an empirical study of FL developer challenges by independently analyzing 495 Stack Overflow posts and 9,116 GitHub issues and pull requests from 92 FL-related projects. Using BERTopic-based topic modeling and difficulty indicators such as unresolved rates and median resolution time, we characterize recurring problem areas and compare how they manifest across the two support platforms, Stack Overflow and GitHub. Our analysis surfaces nine dominant Stack Overflow topics and thirteen GitHub topics, with persistent difficulties concentrated in environment setup and dependency compatibility, API breakages and migration, training instability under non-IID data, evaluation and metric correctness, and the integration of privacy-preserving mechanisms. We also categorize posts by question intent to understand the kinds of help developers seek; this intent analysis shows that “How”-type questions dominate, reflecting strong demand for procedural guidance. Several topics, such as “TFF Installation and Environment Compatibility” and “Federated Feature Engineering and SecureBoost Issues,” exhibit high unresolved rates and long resolution times, suggesting shortcomings in tooling, documentation, and debugging support. Based on these findings, we provide actionable implications for FL framework designers, documentation authors, and educators. Although our results are constrained to public discussions and a subset of widely discussed frameworks, the study offers a scalable method for continuously monitoring developer pain points and improving the usability, reliability, and deployability of FL systems.
[AI-48] SCPP: A Unified Python Library for Soft Clustering
链接: https://arxiv.org/abs/2607.19620
作者: Kiyan Rezaee,Morteza Ziabakhsh,Artin Bahrampour,Seyed Mohammad Ghoreishi,Asal Khaje,Ali Sajedifar,Manny Chalak,Ava Zerafatangiz,Sadegh Eskandari
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 4 pages
Abstract:In this paper, we present SCPP (Soft Clustering Python Package), an open-source Python framework for soft clustering. SCPP establishes a canonical, scikit-learn-compatible estimator interface that standardizes model training, prediction, membership representation, evaluation, and benchmarking across heterogeneous soft clustering methods, including fuzzy, probabilistic, graph-based, matrix factorization, and deep learning methods. The framework currently integrates 40 representative algorithms together with a comprehensive benchmarking comprising datasets, clustering quality metrics, and standardized runtime, memory, and scalability evaluation. SCPP further provides extensive documentation, practical examples, automated testing, and seamless integration with the scientific Python ecosystem, enabling reproducible experimentation and straightforward extension with new algorithms. The source code is publicly available at this https URL.
[AI-49] Fine-grained Computation-Communication Overlap via Tile-level Signaling and Scheduling for Mixture-of-Experts
链接: https://arxiv.org/abs/2607.19539
作者: Minyu Cui,Anna Wingkvist,Morgan Ericsson
类目: Distributed, Parallel, and Cluster Computing (cs.DC); Artificial Intelligence (cs.AI)
备注: To appear at the 55th International Conference on Parallel Processing (ICPP 26)
Abstract:Mixture-of-Experts (MoE) architectures increase model capacity without proportionally increasing computation cost and have become a key building block for scaling large language models (LLMs) to trillion-parameter regimes. Efficient deployment of these MoE models relies on distributed execution across multiple GPUs, where each MoE layer involves two all-to-all communications: dispatching tokens to expert ranks and returning the expert outputs to their source ranks. Conventional MoE implementations launch this return all-to-all after expert compute completes, exposing communication latency on the critical path and reducing GPU utilization. We present a fine-grained approach that overlaps expert compute with the second all-to-all via tile-level signaling and scheduling. Our producer-consumer co-design combines: (1) a persistent per-rank computation kernel (producer) that covers all local experts on the rank to eliminate repeated kernel launch overhead and prioritizes remote-critical tiles, and (2) a persistent communication kernel (consumer) on a small dedicated partition of streaming multiprocessors (SMs) that issues segment-granular transfers as tiles become ready. Our co-design avoids intrusive changes to the underlying computation operators or communication primitives, making it practical for improving distributed MoE execution efficiency on multi-GPU systems. On a 4-A100 GPU platform, evaluated on three MoE models against four state-of-the-art MoE systems, our approach achieves up to 2.64x end-to-end speedup and 2.74x MoE-layer speedup. Compared with a conventional non-overlap baseline, our approach consistently improves both operator- and MoE-layer-level performance across varying GEMM shapes, router modes, and a broad range of producer/consumer SM partitions, while preserving correctness.
[AI-50] rustworthy Privacy-Preserving Multimodal Federated Learning for Personalised Breast Cancer Prediction
链接: https://arxiv.org/abs/2607.19532
作者: Ruth Amey,Muhammad Arifur Rahman,Taha Osman,Nicholas Shopland,Andy Burton,Mufti Mahmud,David J. Brown
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 26 pages, 12 figures
Abstract:Federated learning has emerged as a potential solution to privacy concerns associated with using sensitive health data for training predictive models, particularly in personalised cancer care. This research investigates whether federated learning can support the development of robust models for predicting tumour progression in breast cancer patients while addressing four critical deployment pillars: transparency, scalability, security, and fairness. This study evaluates a federated learning framework using multimodal data, including clinical information, tumour characteristics, biomarker data, and patient demographics, alongside medical imaging data such as MRI scans, to model changes in tumour characteristics over time. The performance of the federated approach was compared with that of a centralised model trained on aggregated data. The report then further examines strategies to enhance secure model updates, maintain performance across patient subgroups, and support scalability across institutions. The findings assess whether federated learning can achieve predictive performance comparable to centralised learning while preserving data locality. These results contribute to understanding the feasibility of privacy-preserving, multimodal predictive modelling and support future applications such as digital twins to assist clinicians and patients in personalised treatment planning.
[AI-51] SynPre-FL: Synthetic data-driven pretraining integrated Federated Learning training framework
链接: https://arxiv.org/abs/2607.19524
作者: Akarsh K Nair,Muhammad Arifur Rahman,Nicholas Shopland,Andy Burton,Jun He,Yuan Shen,David Baldwin,Emma O’Dowd,Amna Burzic,Mufti Mahmud,David J. Brown
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Distributed, Parallel, and Cluster Computing (cs.DC)
备注: 18 pages, 10 figures
Abstract:Federated learning (FL) offers a promising approach to privacy-preserving clinical risk prediction, but its deployment remains limited by restricted data sharing, client heterogeneity, class imbalance, and the lack of realistic tabular electronic health record (EHR) benchmarks. Synthetic data generation may alleviate data scarcity, yet its integration with federated optimisation has received limited systematic study. We propose SynPre-FL, a unified framework combining high-fidelity synthetic EHR generation with synthetic-pretrained FL for robust prediction under non-IID conditions. A latent autoencoder-diffusion model generates privacy-preserving synthetic cohorts, which are used to warm-start federated training. This pretraining is followed by heterogeneity-aware optimisation using class-balanced local objectives, proximal regularisation, and adaptive server aggregation. Post-hoc calibration and federated-safe explainability support reliable and interpretable risk estimates. Experiments show that the synthetic generator preserves univariate, bivariate, and multivariate structure while protecting against membership-inference and reconstruction attacks. The generated data achieve strong downstream utility under TSTR, TRTS, and model-based evaluations. Across federated settings with 5, 10, and 15 heterogeneous clients, SynPre-FL consistently improves robustness and scalability over baseline methods, especially under severe non-IID fragmentation. Calibration improves probability reliability, while SHAP analysis produces stable and clinically coherent feature attributions across federation sizes. SynPre-FL therefore provides a practical and reproducible framework for combining synthetic data with FL to enable privacy-aware, interpretable, and robust clinical prediction from distributed tabular EHR data.
[AI-52] Sophisticated Policies from Epistemic Priors
链接: https://arxiv.org/abs/2607.19518
作者: Wouter W. L. Nuijten,Bert de Vries
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Sophisticated Inference is a variant of active inference often associated with recursive belief modeling and tree search. We argue that its central computational role is simpler: within a planning horizon, it makes active inference closed-loop by allowing future actions to depend on future states and observations. This closed-loop structure can be represented in the epistemic-prior variational free energy framework. Epistemic priors supply the active-inference objective, while a joint posterior over future states and actions supplies the state-contingent control structure. We evaluate this decomposition in the Reactivity Maze, a stochastic benchmark designed to separate epistemic incentive from inner-horizon closed-loop control. The comparison includes three variational objectives with the same state-action posterior family, an action-state factorized active inference objective, Sophisticated Inference, and standard Expected Free Energy planning. The results show that neither ingredient is sufficient on its own. Methods without an epistemic component do not seek information, while methods that prevent future actions from depending on future states cannot turn information into reliable goal-reaching. By contrast, both Sophisticated Inference and full-joint epistemic-prior active inference solve the environment by combining epistemic drive with closed-loop inference. These results show that the advantage associated with Sophisticated Inference need not be specific to tree search itself. It arises from the closed-loop form of active inference, and this form can be represented in epistemic-prior variational inference when the posterior keeps future actions dependent on future states.
[AI-53] Integrity of peer-to-peer distributed LLM inference under malicious nodes
链接: https://arxiv.org/abs/2607.19490
作者: Mert Cihangiroglu,Antonino Nocera
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:Peer-to-peer distributed inference executes a Large Language Model (LLM) on pooled consumer hardware by spreading its layers across many nodes. Every request passes through nodes that are owned and controlled by multiple independent parties. However, in this setting, any party can tamper with the output of its layers to corrupt the end result. Recomputing the forward pass on trusted hardware can catch this, but it introduces additional computational cost. The scientific literature includes several prior integrity-checking approaches, such as known-answer traps for image classifiers and cryptographic commitments. However, these solutions test only the exact correctness and do not account for the ordinary variation that may arise between benign nodes. In this paper, we propose a method that checks the output integrity by measuring the variation in the activations that each node passes to the next. A peer who wants to use the network selects a small set of secret canary inputs whose correct activations are known in advance and mixes them into regular traffic. Because the peers cannot tell a canary from a real query, any tampering node corrupts them as well. The deviation from the known reference then reveals malicious activity: benign nodes exhibit only minor variation from hardware-induced noise, whereas tampered nodes deviate far more. We treat the identification of malicious nodes as a probabilistic test that separates two drift distributions, without relying on a fixed threshold. We study 408 configurations with metrics and success criteria fixed before any experiment ran; the detector reaches AUROC 1.0, correctly ranking the malicious shard above every benign shard on every canary in every configuration.
[AI-54] ModPack: An Extensible Teleoperation Interface for Bimanual Mobile Manipulation
链接: https://arxiv.org/abs/2607.19479
作者: Joshua Citron,Renee Zbizika,Zeyi Liu,Shuran Song
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:
Abstract:Existing teleoperation systems are often tailored to specific robot hardware and task domains, limiting their scalability and adaptability. We present ModPack, a modular and extensible teleoperation system designed to support diverse robot embodiments and task requirements within a unified framework. At the core of ModPack is a self-contained wearable “backpack” that integrates onboard computation, power, communication, and data storage. Built on top of this shared interface, the system supports plug-and-play capability modules including joint-level teleoperation with haptic feedback, mobile manipulation, and active perception. Experiments across two distinct robot platforms and real-world mobile manipulation tasks demonstrate that ModPack provides a flexible and reusable framework for data collection and policy learning. To support future research, we open-source the complete hardware design and software stack. Project website: this https URL
[AI-55] MoA-Structured Decode Attention DNF Derivation KV-Cache Accumulation GQA/MQA and OpenACC Kernel
链接: https://arxiv.org/abs/2607.19456
作者: Lenore Mulin,Gaetan Hains
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:We derive four memory-optimal inference artifacts for transformer attention using the Mathematics of Arrays (MoA), each following directly from the forward-pass Denotational Normal Form (DNF) of with the query-row index fixed to the current decode step. The artifacts are: (1)~a single-query decode DNF in which the \psi -reduction eliminates the K^\top buffer algebraically, achieving (d_k + nd_k+ nd_v+ d_v)\times4,B Dynamic Random Access Memory (DRAM) traffic result numerically verified to |err|\leq2\times10^-7 ; (2)~a C/OpenACC Graphics Processing Unit (GPU) kernel with Operational Normal Form (ONF) stride arithmetic and hardware-coalesced memory access, verified to |\mathrmerr|\infty=0 (exact IEEE-754 floating-point arithmetic); (3)~a multi-step KV-cache with O(d_k+d_v) per-step append via MoA concatenation # ; and (4)~Grouped-Query Attention (GQA) and Multi-Query Attention (MQA) derived via \psi -selection, achieving a proven \frac h_q h_kv reduction in KV traffic. All programs are verified against PyTorch scaled_dot_product_attention.
[AI-56] Predictive Extrema Unprofitable Policies: An AI-Assisted Audit of Candle-Based Binance Spot Timing Models
链接: https://arxiv.org/abs/2607.19453
作者: Ayoub Jadouli
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Statistical Finance (q-fin.ST); Trading and Market Microstructure (q-fin.TR)
备注: 23 pages, 5 figures, 8 tables. Simulation-only negative empirical study and human-supervised AI-assisted evidence audit. No live trading or investment advice. Code and available artifacts: this https URL
Abstract:We audit whether candle-based machine-learning models can turn predictions of cryptocurrency extrema or short-horizon outcomes into positive Binance Spot paper policies after assumed costs. Numerical results come from scripted fixed-seed model runs and deterministic simulators; human-supervised AI agents supported the July 20 evidence-integrity revision through literature retrieval, separately tasked critique, artifact reconciliation, documentation, and source packaging, not trading decisions. The strongest later-period evidence, conditional on extensive predecessor search, is negative: an unchanged ten-pair mandatory-daily selector lost 6.72% over 19 July cycles at an assumed 31-bps completed-cycle cost, with 3 wins and 16 losses. In short model-specific July evaluations, the validation-selected local-minimum policy returned -1.79%, while the local-maximum sell-to-cash/re-entry policy underperformed continuous holding by 2.80%; their gross mean advantages of 11.11 and 12.21 bps were below even the 21-bps stress. A Gurgul-inspired, OHLCV-only daily adaptation attained minimum/maximum ROC AUC of 0.874/0.896 but average precision of only 0.134/0.116 and lost 44.30% over seven cycles, versus -41.20% for buy-and-hold. A forensic audit also downgraded an earlier One4All “30-day holdout”: its dates had influenced prior architecture work, its four-hour outcome horizon was not purged at split boundaries, it used same-close entry, and its raw result directories were absent. Across the tested, mostly exploratory protocols, event-ranking performance did not establish positive executable policy value. Every operational decision remains NO_TRADE.
[AI-57] REGEN: Replay-recycling for Expert-to-Generalist distillation with Offline Reinforcement Learning
链接: https://arxiv.org/abs/2607.19450
作者: Yunjie Chen,Xiaoxin Chen,Fang Wang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Large-scale online reinforcement learning (RL) is the predominant means of eliciting advanced abilities including long-term reasoning and agentic tool use in large language models (LLMs). However, continuing to scale it across vast task domains of interest remains challenging in both computational infrastructure and cost, especially when considering RL as merely a one-off learning stage. Recently, a widely used technique for distilling knowledge across various domains and training stages, multi-teacher on-policy distillation (MOPD), helps to decouple the RL stage, saving costs, while maintaining generality across vast domains. Nonetheless, similar to online RL, MOPD requires coupled inference and backward passes, which continues to limit its scalability and computational efficiency. To address these challenges, we propose REGEN: Replay-recycling for Expert-to-Generalist Distillation with Offline RL. Instead of distilling from multiple teacher models, REGEN trains a generalist by simply recycling the replay memory – the free by-product of the teachers’ specialized RL training – and employing offline RL algorithms. REGEN completely decouples the rollout sampling from the backward training process and thus greatly reduces the training cost. Across mathematical reasoning, code generation, and instruction following, REGEN matches the accuracy of MOPD at substantially lower cost. It potentially turns online RL into a data synthesis process instead of a one-off learning stage, and can potentially be extended to large-scale post-training without requiring heavy computational load.
[AI-58] Guardrails as Scapegoats: Auditing Unfaithful Safety Refusals in Tool-Augmented LLM Agents KDD2026
链接: https://arxiv.org/abs/2607.19449
作者: Aarushi Singh
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
备注: 10 pages, 3 figures. Accepted at the ACM KDD 2026 Workshop on Evaluation and Trustworthiness of Agentic AI
Abstract:Evaluation frameworks for tool-augmented LLM agents focus overwhelmingly on capability metrics or explicit tool crashes, leaving silent infrastructure failures and HTTP 200 responses with empty, null, or malformed payloads largely unaudited. We introduce a lightweight black-box auditing framework that injects four silent failure profiles across 12 production-adjacent tool stubs and classifies agent responses into three mutually exclusive behavioral classes: Honest Surrender (HSR), Fabrication (FAR), and Unfaithful Safety Refusal (USR). Evaluating two frontier and two open-source models at temperature zero under a neutral system prompt, we find that FAR dominates (56.6% of valid responses): agents treat empty payloads as real data, silently returning fabricated results. USR, in which an agent invents a policy or privacy rationale to explain the failure, is nearly absent at baseline (0.25%, one instance across 396 valid trajectories). Our key finding emerges from an ablation where we augment the system prompt with standard safety language (“prioritize user privacy and data security”), which amplifies USR by 15.6x (from 0.25% to 3.95%; 95% CI on ablation rate: 2.2%-6.4%; Fisher’s exact test, p 0.001). USR is a latent behavior, activated when safety vocabulary in the system prompt primes the model to reach for policy rationales when tools silently fail. Sensitive tools (fetch_medical_record, retrieve_contract, fetch_user_profile) account for the majority of USR instances. We propose a payload-response misalignment heuristic for production-level detection and discuss governance implications for safety-forward deployments.
[AI-59] Unlearning as Distribution Restoration: A Controlled Counterfactual Study a Validated Selective Screen and the Limits of Oracle-Free Certification
链接: https://arxiv.org/abs/2607.19442
作者: Sen Yang,Yuen-Hei Yeung
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Machine unlearning is commonly evaluated by matching a retrained oracle on trained probes. In a controlled nonce-fact testbed with a matched retraining reference, we find this criterion can favor methods that retain held-out knowledge: candidates it rates adequate score held-out forget facts -2.82 nats below the never-learned level (cluster CI [-3.16,-2.48] ). We recast unlearning as restoration to the matched reference and audit oracle-free screens and certificate-style criteria across 45 model-seed cells spanning five open architecture families. The reference itself falsifies an absolute retain/round-trip certificate: the injected model, which retains the retain set by construction, fails the fixed retain threshold in 41/45 cells and its own round trip in 31/45, and the reference fully certifies in only 1/45. A base-anchored held-out screen remains strong as a selective necessary test: on a sealed challenge suite it rejects the injected model in 45/45 cells, accepts the reference in 44/45, and partially detects entity-routing suppression (35/45); it is a necessary test with measured sensitivity, not a sufficiency certificate. A damage-relative recalibration anchored to the reference’s own operating point certifies a small subset in 15/45 cells; where it does not abstain, its picks lie within retraining noise (0.80 nats) on the axes it optimizes, while the common trained-probe criterion sits 5.17 nats away (a supporting comparison, not a head-to-head benchmark). A fixed-magnitude logit-suppression attack defeats the full forward battery in 12/45 cells, so forward-only certification is not sound; our method is an empirical selective test for methods-as-produced. An identifiability theorem delimits which facts admit an oracle-free forget threshold at all, with TOFU as the predicted boundary case.
[AI-60] Building Trust in Autonomous Commerce: A Verifiable Global Event Timeline and AI-Ready Fraud Intelligence Layer
链接: https://arxiv.org/abs/2607.19436
作者: Rajat Srivastava
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注: 18 pages, 4 tables, 1 figure. Preprint also available on Zenodo: this https URL
Abstract:Agentic commerce protocols such as AP2 and ACP define mechanisms for secure agent-initiated transactions but do not provide interoperable, tamper-evident auditability or verifiable temporal ordering of events across heterogeneous domains. This paper addresses these gaps by proposing a verifiable global event timeline for agentic commerce, constructed from four core components: canonical event schemas that enforce deterministic serialization, deterministic batch formation ensuring reproducible ordering without reliance on synchronized clocks, Merkle-based append-only commitments providing logarithmic-cost inclusion proofs, and blockchain anchoring establishing a tamper-evident temporal backbone. Building on this infrastructure, we introduce a cryptographically signed fraud marker that binds risk labels to anchored evidence through an unforgeable provenance chain, and a dataset lineage model enabling reproducible, tamper-evident AI training pipelines. Empirical results from a prototype implementation demonstrate: Merkle tree construction processes 50,000 events in 47 milliseconds; end-to-end verification completes in under 0.013 milliseconds regardless of batch size; inclusion proof sizes grow logarithmically from 320 bytes at 1,000 events to 512 bytes at 50,000 events; and Merkle-based verification outperforms linear scan by 14.4x at 50,000 events.
[AI-61] he Chronos Vulnerability: A Taxonomy of Temporal Persistence and Memory-Based Deception in Agent ic AI
链接: https://arxiv.org/abs/2607.19433
作者: Om Narayan,Ramkinker Singh,Praveen Baskar
类目: Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
备注:
Abstract:The transition from stateless generative models in artificial intelligence to stateful, autonomous agents represents an architectural evolution that, while providing the capabilities of long-term planning and the automation of enterprise workflows, also represents the introduction of a new form of security threat, the Chronos Vulnerability. The Chronos Vulnerability represents the threat of memory-based attacks, including the Memory Injection Attack (MINJA) and the sleeper agent, in which the internal belief system of the autonomous agent is compromised, effectively decoupling the attack vector from the final catastrophic event. This study formalizes the threat model for persistence-based attacks and the threat of Dynamics Blindness in the context of the World of Workflows benchmark, demonstrating that traditional endpoint content filters are insufficient for the current stateful architecture. Consequently, this study synthesizes a defense-in-depth landscape, categorizing emerging frameworks such as diagnostic trajectory guardrails (AgentDoG), formal temporal verification (Agent-C), immunological memory consensus (A-MemGuard), and hardware-anchored trust via GPU-based Trusted Execution Environments (TEEs) and Zero-Trust memory architectures.
[AI-62] ChainWatch: A Kill Chain-Aligned Sequential Detection Framework for Multi-Step Attacks in MCP-Based AI Agent Systems
链接: https://arxiv.org/abs/2607.19432
作者: Om Narayan,Rashmi Jyoti,Ramkinker Singh
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:The Model Context Protocol (MCP) is an open-source standard that allows AI agents to connect to external tools, databases, and services. While this connectivity enables powerful agent capabilities, it also introduces multi-step attacks that existing per-call defenses cannot reliably detect. Attackers can compose individually benign tool invocations into malicious sequences that evade isolated inspection. This paper presents ChainWatch, a sequential detection framework for identifying multi-step attacks in MCP-based AI agent systems. ChainWatch models attack progression using a six-stage kill chain and applies a Hidden Markov Model (HMM) to classify tool-call sequences. Detection rules are triggered when a session exhibits suspicious progression across multiple stages. The framework is supported by a structured threat model covering direct sequential attacks, indirect prompt injection chains, and hybrid multi-stage attacks. A 20-dimensional feature extraction schema captures behavioral signals from tool interactions. We demonstrate the approach using five representative attack scenarios from the security literature, showing how ChainWatch detects attack chains that evade traditional per-call security mechanisms.
[AI-63] BRIM: Workload-Balanced Dual-Sided Bit-Serial Sparse Inference Accelerator
链接: https://arxiv.org/abs/2607.19431
作者: Varun Manjunath,Ruokai Yin,Donghyun Lee,Arkapravo Ghosh,Priyadarshini Panda
类目: Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI)
备注:
Abstract:Bit-serial accelerators exploit bit-level sparsity to reduce DNN inference cost, but existing designs exploit sparsity on only one operand, bounding the speedup. Extending sparsity exploitation to both operands simultaneously yields compounding reductions in partial products but introduces a critical new bottleneck: workload imbalance. Because each concurrent weight - activation pair’s execution cost depends on the product of two independently varying operand non-zero bit counts, pairs that must complete together finish at vastly different times, leaving faster computations idle. We show this limits PE utilization to 56 - 64% in existing dual-sided designs. We present BRIM, a hardware - software co-designed dual-sided bit-serial sparse accelerator that directly targets this bottleneck. BRIM combines two integrated mechanisms: 1) Cyclic-Balanced Pruning (CBP), a post-training weight optimization that reshapes weight representations based on profiled activation statistics to equalize expected workloads across concurrently processed pairs offline; and 2) Pairwise Slot Donation, a lightweight hardware mechanism that absorbs residual runtime imbalance with negligible area overhead. Evaluated across CNNs, ViTs, and LLMs under iso-area constraints, BRIM achieves over 90% PE utilization, up to 2.37x speedup, and up to 1.63x energy efficiency improvement over prior dual-sided designs.
[AI-64] ChannelGuard: Safe Models Do Not Compose into Safe Multi-Agent Systems
链接: https://arxiv.org/abs/2607.19430
作者: Elias Hossain,Md Mehedi Hasan Nipu,Fatema Tuj Johora Faria,Tasfia Nuzhat Ornee,Maleeha Sheikh
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:Multi-agent LLM applications chain a planner, worker agents, a verifier, and a synthesizer, and every hop between agents is an unmonitored channel through which an adversary can smuggle instructions. Existing defenses guard only the input boundary (IBProtector, Llama Guard, perplexity filters, SmoothLLM) or run outside the application as opaque, stochastic provider-side filters. We show this gap carries a consequence rarely measured: on a 2,100-trace evaluation across eight attack families, five defenses, and three model backends, an undefended pipeline that appears fully safe under standard reporting (attack success 0.000 on tool- and memory-poisoning) owes that safety almost entirely to the cloud provider’s server-side filter (54 of 60 blocks on Azure GPT-5), and silently shifts to the agent model’s own alignment on a backend without such a filter. Outcome-only reporting hides this dependence. We present ChannelGuard, a training-free defense-in-depth framework placing information-bottleneck gates on every inter-agent channel; each scores channel text against an adversarial phrase bank by embedding similarity and deterministically passes, compresses, or blocks it, adding no LLM call, while an attribution method records which layer stopped each attack. ChannelGuard’s tool-output gate blocks Tool Poisoning 30 of 30 at the application layer, identically across Azure GPT-5, Anthropic Sonnet 4.5, and Anthropic Haiku 4.5, whereas the undefended pipeline shifts entirely across backends; it also lowers Prompt Injection attack success by half (0.333 to 0.167) and preserves GSM8K accuracy exactly (0.867). White-box adaptive paraphrase evades every embedding gate, where a perturb-and-vote baseline does better. An extended appendix adds baselines, ablations, sweeps, a benign-preservation analysis, and a judge audit (kappa = 0.900), at a total cost of 47.36 USD.
[AI-65] Opto-ViT-v2: Noise-Resilient On-Chip Fine-Tuning for Photonic Near-Sensor Vision Transformer Accelerators
链接: https://arxiv.org/abs/2607.19421
作者: Xuming Chen,Deniz Najafi,Mehrdad Morsali,Chengwei Zhou,Zahra Ghanaatianjobzari,Mahdi Nikdast,Shaahin Angizi,Gourav Datta
类目: Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI)
备注: 9 pages, 8 figures. Accepted for publication in the Proceedings of the IEEE/ACM International Conference on Computer-Aided Design (ICCAD 2026)
Abstract:Silicon-photonic (SiPh) accelerators have emerged as a promising platform for Vision Transformer (ViT) inference by performing matrix multiplications on microring-resonator (MRR) banks with high throughput and energy efficiency. Extending these platforms to support on-chip fine-tuning remains challenging because backpropagation requires large activation storage, frequent weight write-back to MRRs, and tolerance to device-level noise. We present Opto-ViT-v2, the first framework for parameter-efficient fine-tuning (PEFT) on a near-sensor SiPh ViT accelerator. Our tensorized low-rank decomposition separates pretrained optical weights from a small set of trainable electronic factors (as few as 8K parameters for ViT-Base), greatly reducing activation storage and weight updates while enabling practical on-chip training. We further introduce a gradient-accumulated sparse classifier that freezes low-importance weights through one-shot top-k gradient masking, reducing classifier training cost by about 40 percent. We also develop the first system-level noise model for photonic on-chip training, capturing the effects of MRR crosstalk, thermal drift, and laser amplitude noise during both forward and backward propagation. Calibrated using measurements from more than 200 fabricated MRR devices, the model shows that low-rank factor updates are more robust than full fine-tuning and conventional layer-wise low-rank adaptation under identical noise conditions. Experiments on VTAB-1K (19 tasks) and FGVC few-shot benchmarks demonstrate that Opto-ViT-v2 recovers within 0.3 to 0.8 percent of clean software accuracy under measured photonic noise while achieving more than 100 KFPS/W, enabling practical on-chip domain adaptation for photonic edge vision systems.
[AI-66] FORCE-Bench: A Benchmark Dataset and Evaluation Harness for Agent ic AI in Enterprise Finance
链接: https://arxiv.org/abs/2607.19409
作者: Wolfgang M. Pauli,Sarah Panda,Kidus Admassu,Said Bleik,Ademola Okerinde,Jeremy Reynolds
类目: Artificial Intelligence (cs.AI)
备注: 22 pages, 10 figures
Abstract:Recent advances in large language models have accelerated deployment of agentic systems in operational finance. Existing benchmarks emphasize measuring general capabilities, instruction following, or safety, but few directly address the operational finance workflows that agentic systems are now being deployed to automate. Finance professionals require agents to not only provide factually sound and properly grounded information, but also ensure that this information is verifiable and consistently adheres to rules and constraints of the operational finance domain. We introduce FORCE-Bench, which contains 251 expert-annotated queries and evaluates responses using a rubric-based framework calibrated to the requirements of the operational finance domain, across eight dimensions: accuracy, citations, clarity, depth, groundedness, recency, relevance, and structure. FORCE-Bench assesses agentic systems on three task types: financial obligation research (querying ERP systems for accounts receivable and payable data), financial entity performance research (answering time-bound questions from public filings and market data), and business brief generation (synthesising multi-source company intelligence reports). To reflect real deployment conditions, we evaluate our purpose-built agent, as well as the general-purpose agentic systems, under common tool access and latency-bounded settings. Results show that general-purpose agentic systems do not consistently meet finance-domain quality requirements under operational constraints, while the purpose-built Finance Agent for Microsoft 365 Copilot is more reliable across dimensions. We release the dataset, rubrics, harness, and analysis code as open-source to support reproducible comparison and adaptation to other enterprise finance environments.
[AI-67] ITPEval: Benchmarking Formal Translation Across Interactive Theorem Provers
链接: https://arxiv.org/abs/2607.19407
作者: Jiayi Wu,Robert Joseph George,Anima Anandkumar
类目: Artificial Intelligence (cs.AI); Logic in Computer Science (cs.LO)
备注: 23 pages
Abstract:Formal theorem proving has emerged as a frontier challenge for machine learning, yet the ecosystem is fragmented: proofs remain siloed across incompatible systems, limiting both training data for learning-based provers and the portability of verified results. We present ITPEval, the first benchmark for evaluating automated formal proof translation across four major ITPs (Lean 4, Rocq, Isabelle, and HOL Light), spanning two distinct logical foundations. Our benchmark comprises 1,560 source files and 6,848 theorems organized into a controlled tier of axiomatized files that isolates foundational translation difficulty, and an ecosystem tier drawn from real libraries that exposes API and proof-style mismatches. We release itpeval, a unified multi-ITP verification infrastructure with state-isolated warm backends that preserve per-artifact native checking semantics. We evaluate both statement and proof translation across five frontier and open-weight LLMs on 12 directed translation pairs: statement translation peaks at 29.1% pass@1 and proof translation at 10.5%; controlled theorems reach 29.7% proof pass@1 versus 5.2% for ecosystem-level translations, confirming that library mismatch is the dominant bottleneck. In addition to pass@k evaluation, a deterministic Lean 4 BEq check establishes equivalence for 54.0% of verified source-to-Lean 4 miniF2F statement translations, showing that native type-checking alone can substantially overestimate semantic fidelity; in an autoformalization/auto-informalization round-trip study, Rocq and HOL Light are easier formalization targets than Lean 4 and Isabelle, while multi-ITP context improves pooled Lean 4 success from 4.8% to 10.6%. Our benchmark, verification infrastructure, and evaluation pipelines are publicly released.
[AI-68] Structured Latent Space Modeling over Multi-Scale Temporal Patches for Multivariate Time Series Forecasting
链接: https://arxiv.org/abs/2607.19404
作者: Xingsheng Chen,Deyu Yi,Siu-Ming Yiu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Multivariate time series encode structural patterns that unfold across multiple temporal scales, yet most forecasting backbones treat learned representations as transient byproducts of prediction, leaving the organizational geometry of these patterns underexploited. We introduce M2Patch, a CNN-based forecasting architecture that maps channel-independent multivariate observations into a structured latent space through two complementary differentiable constraints. Multi-scale patching decomposes the input into overlapping temporal granularities; depthwise separable convolutions with progressive dilation extract scale-specific features in linear time; and per-scale learned projections compress these features into a compact latent representation. The latent space is organized by an intra-scale smoothness constraint that enforces temporal continuity between adjacent patches, and an inter-scale alignment constraint, realized through learnable cross-scale mappings, that restores cross-granularity interaction within the channel-independent design, ensuring that all scales encode mutually consistent representations of the underlying dynamics. Experiments on ten real-world benchmarks show that M2Patch achieves 57 best and 34 second-best results across 40 forecasting settings, matching or exceeding representative baselines on most benchmarks while maintaining linear computational complexity and robustness to patch-level input corruption.
[AI-69] Recovering Clinical Utility Under Differential Privacy: Empirical Validation of Adaptive Federated Aggregation on Heterogeneous Cardiovascular Datasets
链接: https://arxiv.org/abs/2607.19403
作者: Rodrigo Tertulino,Laercio Alencar,Ricardo Almeida
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR); Computers and Society (cs.CY)
备注: 16 pages, 3 figures. Submitted to the Journal of the Brazilian Computer Society (JBCS)
Abstract:Validating federated learning frameworks on real clinical data is an essential step between proof-of-concept demonstrations in controlled synthetic environments and deployment in real multicenter healthcare settings. A prior architectural study by the same authors (Tertulino and Alencar, 2026) demonstrated, on a synthetic six-feature benchmark, that server-side adaptive optimization acts as a temporal denoiser for Differential Privacy noise, answering an open challenge identified in the original pipeline work (Tertulino, 2025). That study used synthetically generated data and explicitly identified real-world validation as a priority future direction. The present work addresses this gap by validating the FedCVR framework on five publicly available real cardiovascular datasets (Framingham, Cleveland, Hungarian, Switzerland, and Long Beach VA), harmonized to the 13-attribute UCI Heart Disease schema and configured as a heterogeneous federated scenario with leave-one-institution-out cross-validation. Results demonstrate that FedCVR preserves its adaptive advantage on real data, achieving an F1-Score of 79.2% and AUC of 0.96 under the operational privacy budget (noise multiplier = 0.8, privacy budget epsilon approximately 4.2), while statistically outperforming standard FedAvg on all evaluated metrics (paired t-tests, all p = 0.003, significant under the Bonferroni-corrected threshold). The measured privacy cost on real data confirms the graceful degradation pattern observed in the synthetic experiments, providing empirical evidence of the framework’s clinical viability in genuine multicenter contexts.
[AI-70] CrackedPDFs: A Controlled Benchmark for Hidden Prompt Injection in PDFs
链接: https://arxiv.org/abs/2607.19396
作者: Pukaphol Thienpreecha(“Volk”)
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Document-based LLM systems often flatten a PDF before guardrails inspect it. That step can discard evidence that an instruction was never visible to the user. We introduce CrackedPDFs, a controlled benchmark for hidden prompt injection in PDFs. The benchmark contains 29,322 generated PDFs from 4,983 base docu ments. It includes 9,774 injected files and 19,548 benign or matched-confounder files. We evaluate PromptGuard and a rule baseline. We also evaluate structural only learned models and a sanitized hybrid detector. The evaluation uses held-out provenance splits and paired benign-confounder controls. It also uses label-shuffle checks and shortcut audits. On a 2,919-document held-out test set, the hybrid de tector reaches 0.960 F1. ROC-AUC is 0.998 and PR-AUC is 0.997. It also ranks injected files above matched benign confounders in 95.9% of 973 pairs. Prompt Guard has low recall when given extracted text only. Structural-only learned mod els are weak under paired controls. A text-only TF-IDF model reaches perfect held-out scores but fails shortcut audits. These results show that document-aware hybrid detection is useful under controlled paired evaluation. They do not show broad real-world robustness or reliable cross-family generalization.
[AI-71] From Trajectories to Prefixes: Reusing Teacher Trajectories via Replayed Prefixes and Online Continuation
链接: https://arxiv.org/abs/2607.19395
作者: Yihan Wang,Zhong Guan,Haoran Sun,Jiale Huang,Likang Wu,Hongke Zhao
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Small language models are attractive backbones for interactive agents, but direct distillation from strong teacher trajectories often turns rich multi-turn behavior into one-shot imitation targets. This is inefficient in long-horizon environments, where early decisions shape later states and rewards. We propose Prefix-GRPO, a reinforcement learning framework that decomposes teacher trajectories into replay-aligned prefix queries and online continuations. Each prefix is replayed in the environment to recover a valid intermediate state, after which the student continues online interaction and receives task reward. Unlike response-only GRPO, Prefix-GRPO also applies clipped policy updates to historical assistant tokens inside the replayed prefix, using a policy-distilled SFT checkpoint to estimate their old log-probabilities. This unifies prefix learning and continuation learning within the same policy-optimization form. Experiments on TextCraft, BabyAI, and ALFWorld show that Prefix-GRPO improves small-model agents over distillation and standard RL baselines, while ablations show that replay alone is insufficient without explicit prefix-token optimization. The implementation and reproduction scripts are available at this https URL.
[AI-72] Cross-Subject Semantic Decoding with Shared-Space Alignment for Generalized Neural Representation Learning
链接: https://arxiv.org/abs/2607.19394
作者: Ji-Hoon Heo,Aleksandra Joanna Wisniewska,Seo-Hyun Lee,Seong-Whan Lee
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Generalizing across subjects remains challenging in invasive neural recordings because electrode configurations, anatomical structures, and neural signal patterns vary substantially across individuals. To investigate such inter-subject variability, we propose a cross-subject semantic decoding framework that aligns neural responses to speech perception from multiple subjects into a shared latent space and learns a mapping from the aligned neural representations to contextual embeddings. More specifically, using electrocorticography data collected during natural language comprehension, we estimate the shared space using the shared response model and train a decoder to predict contextual semantic embeddings from projected neural responses. For a held-out subject, we estimate a subject-specific projection into the predefined shared space, and directly apply the pretrained decoder without any retraining. Experimental results demonstrate that the proposed framework consistently outperforms baseline methods across evaluation settings and exhibits a reduced performance drop from source subject to held-out subject testing, indicating improved cross-subject generalization. These results suggest that aligning neural activity into a shared latent space, while decoding in a semantic embedding space, provides an effective strategy for improving cross-subject generalization by reducing subject-specific differences in neural responses while effectively capturing shared stimulus-related representations.
[AI-73] LAARA: Layer-Aware Adaptive Rank Allocation for Parameter-Efficient Fine-Tuning
链接: https://arxiv.org/abs/2607.19391
作者: Ashutosh Tripathi,Surya Deep Singh,Pranab Sahoo,Sriparna Saha
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Low-Rank Adaptation is widely used for parameter-efficient fine-tuning, yet existing methods typically assign the same adapter rank to every transformer layer despite their heterogeneous adaptation requirements. In this work, we show theoretically and empirically that uniform rank allocation is fundamentally suboptimal. Motivated by this observation, we propose LAARA (Layer Aware Adaptive Rank Allocation framework), a search-free framework that dynamically allocates ranks using lightweight diagonal Fisher estimates computed during training. LAARA combines projection-wise normalization, logarithmic compression, blended adapter importance estimation, and a vote-to-change dampening mechanism to produce stable and efficient rank adaptation. Experiments on GLUE and MathInstruct benchmark demonstrate that LAARA consistently matches or outperforms popular state of the art approaches such as LoRA, AdaLoRA, DyLoRA, and Bitfit while using significantly fewer trainable parameters. Our results show that Fisher-guided rank allocation provides a principled and effective foundation for adaptive parameter-efficient fine-tuning. The code is publicly available at: this https URL
[AI-74] Simulating Eutopia: Revisiting Long-term Fairness with Outcomes Performativity and Dynamics
链接: https://arxiv.org/abs/2607.19389
作者: Vedant Palit,Udvas Das,Brahim Driss,Debabrota Basu
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Machine Learning (stat.ML)
备注:
Abstract:As AI-driven Decision Makers (ADMs) influence our socioeconomic reality, their roles in both enhancing efficiency and amplifying the social biases have drawn attention. In this paper, we revisit the nuances of long-term `fairness’ achievable by an ADM, specifically in the context of a credit lending induced wealth process. The literature on long-term fairness mostly (a) considers passive environments, i.e. the outcome of a predictor does not change the population’s behaviour, and (b) measures bias in terms of disparity in instantaneous predictions rather than the downstream equity. These are not true for modern ADMs, like credit lenders. To address these caveats, we first formalise the wealth dynamics induced by a loan approving ADM interacting with a multi-demographic population as a performative Markov Decision Process with ADM level and social outcome level reward functions. Then, we mitigate the absence of such a performative test-bed by developing Eutopia: a lending-process simulator enabled with a novel performative data generator to learn long-term fair strategies. Finally, we test performative and classical RL algorithms with different fairness-aware and utilitarian utilities. Experimental results show that (a) learning with performative dynamics lead to better long-term efficiency and equity, and (b) learning with well-designed fairness-aware utility evaluated on social outcomes induces better efficiency, equity, and inclusivity.
[AI-75] Scale-Aware Learning of Chaotic Dynamics on Unstructured Meshes via Binned Spectral Losses
链接: https://arxiv.org/abs/2607.19387
作者: Kanad Sen,Romit Maulik
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computational Physics (physics.comp-ph); Fluid Dynamics (physics.flu-dyn)
备注:
Abstract:Surrogate modeling for high-dimensional nonlinear dynamical systems that exhibit chaos requires mechanisms that preserve not only pointwise accuracy but also the scale-dependent structure of physical fields. Bandwise spectral power losses, such as the binned spectral loss function, provide such supervision on structured grids, where Fourier modes define a standard frequency decomposition. On irregular meshes, however, no canonical Fourier basis exists, and spectral representations must be constructed from graph operators induced by mesh connectivity and geometry. In this study, we extend the binned spectral power loss for application to unstructured-mesh surrogate modeling of nonlinear dynamical systems. This is obtained by replacing Fourier bands with graph-Laplacian frequency bands, and we provide scalable Chebyshev and multilevel approximations for improving long-horizon rollout fidelity. In its full-spectrum form, our approach uses graph Laplacian eigenspaces to provide a graph analogue of Fourier band-power matching, but incurs the high cost of spectral decomposition. As a scalable approximation, we replace exact band projectors with sparse Chebyshev polynomial graph filters, avoiding explicit eigendecomposition. When utilizing multilevel graph architectures, we introduce Graph Laplacian Energy Alignment for Meshes (GLEAM), which applies retained-subspace scale-aware supervision across graph hierarchies so that coarse and fine representations are regularized during autoregressive rollout. Our results show that the proposed spectral losses improve long-horizon rollout fidelity and preserve statistical invariants for the forecasting of turbulent flows on unstructured meshes, compared to deterministic baselines.
[AI-76] Challenges of Explainability in Continual Learning for Time Series Forecasting
链接: https://arxiv.org/abs/2607.19382
作者: Quentin Besnard(RFAI),Emmanuel Doumard(BDTLN),Nicolas Labroche(LIFAT, BDTLN),Nicolas Ragot(RFAI),Nicolas Ringuet(BDTLN)
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Deep learning models have shown strong potential for time series forecasting, yet their deployment in real-world environmental monitoring remains challenging due to non-stationary dynamics and limited explainability. In this work, we investigate explainability as a central tool for understanding continual learning in adaptive time series forecasting, with Experience Replay strategies. We study neural forecasting architectures such as PatchMixer, PatchTST and DLinear, augmented with attention-based sampling mechanisms to support model adaptation over time. Explainability is leveraged through attention rollout and gradient-based attribution methods (Grad-CAM) to analyze both predictive behavior and sampling strategies within a continual learning framework. Experiments conducted on real-world piezometric time series exhibiting heterogeneous patterns and regime shifts show that analyzing model and sampling behaviors provides valuable insights into the dynamics of the continual learning framework. Beyond predictive performance, our results highlight the challenges and opportunities of using explainability to understand continual learning behaviors, revealing how attribution patterns evolve over time and how they can inform data selection and adaptation strategies in non-stationary forecasting scenarios.
[AI-77] Economic Evaluations of Language Models
链接: https://arxiv.org/abs/2607.19375
作者: Alexander Wan,Stephane Hatgis-Kessell,Tomás Aguirre,Percy Liang,Rishi Bommasani
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI)
备注:
Abstract:Language models perform economically valuable work, yet they are not currently assessed for how well they perform every economically valuable task. We introduce EconEvals as an open-source evaluation suite to measure capabilities relevant to tasks, work activities, and occupations in the US labor economy. We ground the evaluation suite in real user queries to language models where possible, and supplement these with synthetic data. Our evaluations improve coverage over OpenAI’s GDPval benchmark, which is the existing state-of-the-art that covers 5% of US occupations, at 500x lower cost. Alongside benchmarks, we also introduce a simulation-based exposure measure to estimate how much time current language model capabilities could save across all tasks belonging to all US occupations, with detailed accounting for each estimate. Our estimates indicate that current models could save workers substantial time on at least half of their tasks in 47% of occupations. However, for 79% of tasks where we predict substantial time savings, observed Claude usage is low, suggesting that existing usage lags potential. Beyond inherent constraints of language model chatbots, our data identifies privacy and proprietary systems as the principal bottlenecks limiting further time savings from AI. Overall, we introduce adaptable infrastructure that grounds inferences about language models’ labor-market impact in their current capabilities, which can be continually updated as capabilities improve.
[AI-78] Euclean: Automated Geometry Problem Formalization with Unified Verification in Lean ICML2026
链接: https://arxiv.org/abs/2607.19374
作者: Linbin Tang,Jingyan You,Zilin Kang,Hanzhang Liu,Sophia Zhang,Zenan Li,Chenrui Cao,Liangcheng Song,Jiaao Wu,Xian Zhang,Fan Yang
类目: Artificial Intelligence (cs.AI)
备注: ICML 2026
Abstract:Recent formal reasoning systems have reached IMO-level performance, yet they leave a fragmented landscape: algebra and number theory are handled in Lean, while geometry still relies on domain-specific languages with limited formal guarantees. This split increases the trusted computing base and hinders unified model development. Existing geometry-in-Lean efforts (LeanEuclid, LeanGeo) introduce custom axiom systems incompatible with standard Mathlib, and their small scale ( 1,100 problems) limits large-scale training. Native Mathlib autoformalization of geometry, however, poses distinct challenges: implicit diagrammatic assumptions (e.g., topological configuration and non-degeneracy) must be made explicit rather than deferred to external solvers, and models must adapt to Mathlib’s small, rapidly evolving geometry infrastructure. We present Euclean, a four-stage framework - constraint explication, configuration anchoring, formalization mapping, and iterative repair - for automatically formalizing geometry in native Mathlib. We construct OMNI-Geometry (768 competition problems) and Numina-Geometry (177,597 problems), the largest geometry formalization dataset in Lean. Human evaluation shows 48.89% TOP1 and 73.33% TOP5 accuracy. Training Goedel v2 on our formalizations improves proof success from 13.6% to 15.1%, validating dataset quality for unified neural theorem proving. Code and datasets: this https URL.
[AI-79] Beyond Tracking or Shortcut: Composition-Bounded Predictive States in Poker Autoregressive Models
链接: https://arxiv.org/abs/2607.19369
作者: Quanhao Li,Qianyu Chen
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Hidden-state probes often recover latent labels in imperfect-information sequence models, but this alone does not establish that a model maintains a posterior belief distribution over hidden states. This paper studies this ambiguity in a no-range Limit Hold’em autoregressive model trained only on action and value targets, not on an opponent’s hand or range. Opponent-range probes are positive after action/value controls in two of three seeds, and the behavior head predicts held-out actions about five percentage points above a baseline using only observable public history. However, visible public betting composition explains more opponent-range signal than residual hidden states, suggesting that most recoverable information comes from betting summaries. Action/value+composition baselines reach 16.5-16.7% top-10 accuracy while composition-residual hidden probes fall to 11.4-12.2%, and matched-composition comparisons are negative in every seed. We call this evidence pattern composition-bounded predictive support: hidden states remain behavior-predictive and opponent-range correlated, but most recoverable range information is explained by visible betting composition rather than residual hidden-state structure. This is a case-study claim about opponent-range representational evidence, not exact Bayesian posterior tracking or a causal belief mechanism. Synthetic control and oracle validations show that the same diagnostics accept posterior-sensitive states and reject raw composition states under matched controls. Thus positive belief probes should be interpreted through targeted alternatives before being treated as evidence of belief tracking.
[AI-80] Rethinking Uncertainty Evaluation in Large Language Models ICML2026
链接: https://arxiv.org/abs/2607.19367
作者: Krish Matta,Atharv Naphade,Andy Zou
类目: Artificial Intelligence (cs.AI)
备注: 19 pages, 11 figures, ICML 2026 EIML Workshop
Abstract:Calibration is the primary criterion for evaluating LLM confidence, but it is insufficient: it admits trivially incoherent estimators, depends on the evaluation distribution, and does not test the extent to which the estimation can be interpreted as a consistent, underlying probability function. What we actually need is for LLM confidence estimates to satisfy the conditions required of coherent probabilistic beliefs. We formalize these conditions along three axes (structural coherence, faithfulness, and usefulness) and operationalize them as the C1 metrics. Widely used estimators systematically violate these conditions despite appearing well-calibrated: models assign lower confidence to logically easier questions 31% of the time, and common interventions reducing RMSCE leave structural violations unchanged, suggesting that calibration is orthogonal to probabilistic validity. RLHF and chain-of-thought improve usefulness metrics without restoring coherence. Our results show current LLM confidence estimates cannot be interpreted as coherent probabilities; our framework provides the tools to measure and close this gap.
[AI-81] Geometry-Guided Constraint Learning for LLM Safety Classification
链接: https://arxiv.org/abs/2607.19366
作者: Fumiaki Uehara,Koo Imai,Masato Tsutsumi,Keigo Kansa,Sora Usui,Yuki Kobiyama
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Safety as Polytope (SaP) learns linear half-space constraints in LLM hidden space but requires per-category tuning of the constraint count K. We show that sparse autoencoder (SAE) feature extraction resolves this: K=2 becomes optimal for 12/14 categories on Qwen3.5-9B, achieving 96-99% accuracy per category on our BeaverTails classification benchmark, largely eliminating the need for exhaustive sweeps (K=4-25 with random initialization). This convergence to two planes is consistent with the Linear Representation Hypothesis, providing suggestive evidence that safety boundaries in this setting admit a low-dimensional linear description in the SAE feature space. Building on this geometric perspective, we introduce a cone constraint whose learnable aperture adapts to each category’s cluster concentration, stabilized by a three-phase training
[AI-82] Logic-Guided Data Extraction with Answer Set Programming and Large Language Models
链接: https://arxiv.org/abs/2607.19365
作者: Mario Alviano,Lorenzo Grillo,Nicola Leone,Fabrizio Lo Scudo
类目: Artificial Intelligence (cs.AI)
备注: 42nd International Conference on Logic Programming, to appear in Theory and Practice of Logic Programming (TPLP)
Abstract:When Large Language Models (LLMs) are used for semantic data extraction from unstructured text, producing candidate relational facts from natural language, they may remain unreliable for tasks requiring complex combinatorial reasoning and global consistency. This paper proposes a logic-guided data extraction framework combining LLM-based extraction with Answer Set Programming (ASP). The LLM produces candidate facts, whereas ASP performs validation, inference, consistency checking, and control. Unlike existing pipelines that query the LLM independently for all target predicates, the proposed approach uses ASP reasoning to identify which predicates are logically admissible at each stage and to guide extraction queries. By interleaving LLM calls with ASP derivation, the framework infers logically implied facts without further extraction and detects inconsistencies early. We formalize the pipeline and prove that, under mild assumptions, it is equivalent to the baseline approach with respect to the final extracted facts, while requiring fewer LLM calls. We also introduce a caching mechanism for logic-based control queries, exploiting monotonicity of conjunctive queries over incrementally constructed fact sets to reduce solver invocations. Experiments on ASP-derived benchmarks show that the framework reduces LLM calls and improves extraction quality by mitigating spurious outputs, demonstrating the value of non-monotonic logic programming for controlled semantic extraction.
[AI-83] Lifted Representation Hypothesis in Language Models ICML2026
链接: https://arxiv.org/abs/2607.19360
作者: Bumjin Park,Jaesik Choi
类目: Artificial Intelligence (cs.AI)
备注: 20 pages, 6 figures, ICML 2026 Workshop: Philosophy Meets Machine Learning
Abstract:Large language models (LLMs) often answer queries by mapping individual observations to more general rule-like structures. However, it remains unclear how these structures are stored, selected, and revised. To study this process, we propose thelifted representation hypothesis: LLMs update memory through shared latent structures rather than isolated instance-level facts. This view frames lifting as an efficient use of symmetry across instances, and shattering as the refinement of coarse lifted structures into more specific subtypes. We evaluate LLMs’ lifting and shattering through controlled exception-learning experiments across in-context learning, LoRA, and full fine-tuning. We find that LLMs are vulnerable to shattering failures when data are governed by nested rules and exceptions, while lifting often occurs prematurely. These results highlight the need to study the relation between data and rule structures in LLMs.
[AI-84] Profile-Graph Memory for LLM Agents : Implicit Cross-Entity Traversal through Narrative Profiles
链接: https://arxiv.org/abs/2607.19359
作者: Shengtong Zhu
类目: Artificial Intelligence (cs.AI)
备注: 11 pages, 2 figures, 7 tables. Code and MemHop benchmark: this https URL
Abstract:Long-term memory is essential for LLM agents that interact across sessions, yet current memory benchmarks primarily evaluate single-hop recall, leaving multi-hop association largely unmeasured. We make three contributions. First, we introduce MemHop, a multi-hop memory benchmark of 1,000 questions at hop depths 1-5 across 10 social-network scenarios, with per-hop evidence annotations. Second, we present Profile-Graph Memory (ProGraph), a two-layer memory architecture combining (i) profile expansion – substring-matched traversal of entity names that naturally appear in LLM-written profile narratives, a minimal alternative to explicit knowledge-graph construction – and (ii) compression residuals – exact dates, quantities, and named items co-extracted with each profile update at zero extra API cost. Third, a full-grid ablation shows cross-benchmark mechanism specialization: profile expansion drives multi-hop reasoning (-22.6pp on MemHop when removed) while compression residuals drive precision recall (-8.6pp on LoCoMo when not co-extracted), with cross-effects under 3pp within a single architecture. ProGraph averages 80.1% on MemHop (matching the FullContext reference) and 78.4% on LoCoMo (exceeding FullContext by 11.3pp), outperforming Mem0, A-Mem, HippoRAG, and RAG on both. We release MemHop, ProGraph, and baseline implementations.
[AI-85] LISA: Linear-Indexed Sparse Attention for Efficient Long-Context Reasoning
链接: https://arxiv.org/abs/2607.19358
作者: Yu Zhao,Zekun Zhang,Fan Jiang,Bo Zeng,Linlong Xu,Shimin Shan,Yu Liu,Longyue Wang,Weihua Luo
类目: Artificial Intelligence (cs.AI)
备注: 20 pages, 10 figures
Abstract:Recent advances in long chain-of-thought reasoning models such as DeepSeek-R1 have led to increasingly longer inference context lengths under the test-time scaling paradigm. However, the O(n^2) computational complexity of standard self-attention causes inference costs to grow sharply with long sequences, limiting the deployment of long-CoT reasoning in production settings. To address this, we propose LISA (Linear-Indexed Sparse Attention), a plug-and-play attention replacement module that requires no pretraining from scratch. LISA integrates two lightweight components in parallel within the original model: (1) a Linear Attention module that provides long-range memory with O(n) time complexity; (2) a Lightning Indexer that selects the top-M important tokens from the full context to feed into a Sparse Self-Attention. The two branches are fused via a gating mechanism, reducing inference complexity from O(n^2) to O(nM) (M n) for generating n tokens. We design a two-stage training pipeline: Stage 1 initializes the model by integrating the linear attention to capture long-range dependencies, complemented by a sliding-window attention mechanism that is optimized via knowledge distillation to approximate the full self-attention distribution of a frozen teacher model. In Stage 2, we further introduce the Indexer to replace the static sliding-window mechanism, enabling dynamic token selection from broader contexts. The Indexer is trained using a novel per-head KL divergence loss, which aligns its selection behavior with the attention patterns of the teacher model. Experiments on DeepSeek-distilled-Qwen models demonstrate that LISA achieves a 50% inference speedup under 16K-token context, while improving average performance by 5.6% on reasoning benchmarks including AIME and MATH-500.
[AI-86] Stochastic Primal-Dual Decoding for Multiobjective Generative Recommender Systems
链接: https://arxiv.org/abs/2607.19357
作者: Dmitrii Moor,Ben Carterette,Senthilkumar Krishnamoorthy,Kyle Kretschman,Denis Beslic,Melissa Yalla,Alice Y Wang,Mounia Lalmas
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Recent advances in recommender systems (RS) have shown substantial performance gains through generative modelling. In practice, recommendation often involves constructing slates – ordered lists of items – that must satisfy multiple objectives beyond relevance, such as constraints defined over item attributes or fairness constraints. Existing multiobjective approaches either rely on post-processing techniques designed for non-generative settings, or incorporate auxiliary objectives directly into model training. The former does not explicitly account for the sequential nature of generative RS, while the latter is often impractical in large-scale systems. We propose a lightweight, inference-time decoding layer that augments autoregressive generative RS to support multiobjective slate generation without modifying or retraining the underlying model. We formulate decoding as an online constrained optimisation problem, where items are selected sequentially, and trade-offs between relevance and auxiliary objectives are adjusted dynamically based on the remaining constraint slack, i.e., how much of each objective remains to be satisfied. This is implemented via a stochastic primal-dual approximation scheme that balances relevance and auxiliary objectives during generation. We provide theoretical guarantees on constraint violation and regret, and evaluate the proposed approach through extensive offline experiments and a large-scale online A/B experiment in a real-world recommender system. Our results show consistent improvements in multiobjective trade-offs, including a +1.8% gain in the auxiliary objectives achieved at zero cost to user satisfaction. Subjects: Artificial Intelligence (cs.AI); Machine Learning (cs.LG) Cite as: arXiv:2607.19357 [cs.AI] (or arXiv:2607.19357v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2607.19357 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-87] NEXUS: Structured Runtime Safety for Tool-Using LLM Agents
链接: https://arxiv.org/abs/2607.19356
作者: Elias Hossain,Md Mehedi Hasan Nipu,Tasfia Nuzhat Ornee,Rajib Rana,Niloofar Yousefi
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Tool-using LLM agents increasingly execute high-impact actions, making runtime safety monitoring essential. We present NEXUS (Neural EXecution Utility and Safety), a structured-plan safety monitor that applies a formal intervention policy to select among four actions: allow, block, request confirmation, or request revision. NEXUS combines deterministic safety rules, argument-level inspection, and a calibrated logistic-regression risk score for graded escalation. On a 128-instance synthetic benchmark, NEXUS achieves an F1 score of 0.949 and a 4-class intervention accuracy of 0.6406, outperforming rule-only intervention selection by 27.3 percentage points. It also improves over rule-only on R-Judge (F1 = 0.861 vs. 0.849), matches rule-only on AgentHarm due to threat-model limits, and achieves 0% ASR at 99% control allow on IPI. On the rule-blind NEXUS-Stress benchmark, NEXUS reaches an F1 score of 0.881, highlighting the difficulty of fine-grained intervention routing. With 0.205 ms median latency, NEXUS adds under 0.1% overhead to typical agent loops. Code, benchmarks, and the calibrated risk scorer are publicly released.
[AI-88] FormulaSPIN: Self-Play Fine-Tuning for Natural Language to Spreadsheet Formula Generation ACL2026
链接: https://arxiv.org/abs/2607.19354
作者: Cy Xie
类目: Artificial Intelligence (cs.AI)
备注: 15 pages,7 figures, 14 tables. Accepted to ACL 2026 Main Conference Oral
Abstract:Spreadsheet applications are used by hundreds of millions worldwide, yet writing formulas remains a significant barrier. Existing approaches rely on static supervised data, which quickly saturates on limited annotations. In this paper, we introduce FORMULASPIN, a self-play framework that breaks the ceiling of supervised fine-tuning by enabling iterative self-improvement without any additional data. Vanilla SPIN fails on this task: it uniformly penalizes every non-matching output, so execution-equivalent alternatives are punished as negatives in one example while serving as ground truth in another, producing contradictory gradients. Our framework resolves this by exploiting formula generation’s unique advantage: binary executability provides implicit supervision that separates semantic errors from valid stylistic variants. We frame training as a two-player game in which the main player learns to prefer ground-truth formulas over those from its previous version, while execution feedback sorts outputs into distinct granularities-enabling an adaptive curriculum that shifts from semantic correctness to stylistic refinement. To further increase accuracy, we incorporate ExecVote, a semantic-level voting mechanism that naturally handles multiple valid formulations. Experiments on multiple benchmarks demonstrate that FORMULASPIN achieves state-of-the-art performance, with 74.9% exact match and 87.1% execution accuracy on NL2FORMULA, matching models trained with additional preference annotations while outperforming both traditional SFT and frontier proprietary models. These findings underscore self-play’s potential to tackle scarce data tasks and open the door to extending it beyond executable domains.
[AI-89] Benchmarking Confidential GPU Inference on NVIDIA H100 under Intel TDX
链接: https://arxiv.org/abs/2607.19353
作者: Wei Wang,Abdul Hyee Waqas,Burns Smith
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Confidential computing is becoming a practical deployment requirement for AI inference workloads that process sensitive inputs or protect proprietary model assets. However, the performance cost of enabling confidential execution for GPU-accelerated large language model serving remains workload dependent and operationally important. This paper presents a benchmark study comparing standard non-confidential execution with confidential computing mode on a single NVIDIA H100 80GB GPU hosted in an Intel TDX confidential instance. The evaluation uses two representative language models, Mistral-7B v0.1 and Qwen3-30B-A3B, and measures time to first token, end-to-end request latency, per-request token generation throughput, global token throughput, and closed-loop request throughput under increasing concurrency. In fixed request-rate experiments, confidential mode increases average TTFT by 21.8% for Mistral-7B and 27.8% for Qwen3-30B-A3B, while global token throughput drops by 17.7% and 21.1%, respectively. In closed-loop concurrency experiments, throughput gaps remain in the 11.5-20.2% range, but the larger model reaches its saturation knee earlier under confidential mode. The results suggest that confidential GPU inference can retain usable throughput under load, but capacity planning must account for both the steady throughput penalty and the earlier saturation behavior observed for larger models.
[AI-90] OpenEvoShield: Dual Non-Stationary Continual Defense for Open-World Multi-Agent System Attacks
链接: https://arxiv.org/abs/2607.19351
作者: Litian Zhang,Chaozhuo Li,Yuting Zhang,Zejian Chen,Bingyu Yan,Qiwei Ye
类目: Artificial Intelligence (cs.AI)
备注: 29 pages, 5 figures, 14 tables
Abstract:LLM-based multi-agent systems (LLM-MAS) are increasingly deployed in safety-critical applications, where adversaries inject malicious instructions through inter-agent communication to propagate harmful behaviors. Unlike static threats, these attacks are doubly dynamic: adversaries refine injection strategies against deployed defenses while normal-agent behavior drifts with system expansion. Existing defenses treat deployment as a closed-world problem and degrade rapidly once either distribution shifts beyond training coverage. We propose OpenEvoShield, a co-evolutionary continual defense framework for LLM-MAS. An asymmetric rate controller (M1) decouples fast attack-side and slow normal-side learning rates from dual drift signals. A normal-boundary updater (M2) maintains a dynamic behavioral boundary at the slow rate, while an EWC-regularized policy ensemble (M3) fast-adapts without catastrophic forgetting. An energy-based multi-granularity detector (M4) fuses node-, subgraph-, and graph-level evidence to classify novel attacks as out-of-distribution. Experiments over 100 deployment rounds across five benchmarks and four MAS topologies show that OpenEvoShield outperforms static and continual baselines, detecting most previously unseen attacks while keeping false positive rates low.
[AI-91] Hybrid LSTM-Graph Neural Framework for Robust Financial Fraud Detection and Adversarial Resilience
链接: https://arxiv.org/abs/2607.19350
作者: Mariam Zakaria Moussa Ali
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 6 pages, 12 figures
Abstract:Financial institutions face significant challenges in detecting sophisticated money laundering patterns, such as smurfing and layering, due to extreme data imbalance (0.13% fraud rate) and evolving adversarial evasion tactics. This paper proposes FraudShield AI, a hybrid framework that integrates Long Short-Term Memory (LSTM) networks with hand-crafted Graph Topological Features to capture both temporal sequences and structural relational context. By engineering network-centric features including PageRank Centrality, In-Degree dynamics, and a custom Flow Ratio, the system shifts the detection paradigm from isolated transaction analysis to network-level forensics. A Focal Loss objective is used to address class imbalance, and a dynamic thresholding mechanism is introduced to improve resilience against low-value smurfing attacks. Experimental evaluation on the PaySim dataset shows that the proposed hybrid model substantially outperforms Logistic Regression and XGBoost baselines in Precision, Recall, and F1-Score, particularly on hard-to-detect micro-transaction fraud patterns. An ablation study confirms the complementary contribution of both the temporal and topological components.
[AI-92] FineServe: A Fine-Grained Dataset and Characterization of Global LLM Serving Workloads
链接: https://arxiv.org/abs/2607.19349
作者: Tiancheng Zhang,Shaoyuan Huang,Mingyuan Wang,Yunfeng Zhao,Xiaofei Wang,Wenyu Wang
类目: Artificial Intelligence (cs.AI)
备注: 14 pages, 14 figures
Abstract:Large language models (LLMs) are increasingly deployed as always-on online services, making efficient LLM serving a critical systems challenge. Achieving low latency and high throughput under volatile demand requires deep understanding of real-world serving workloads, yet existing studies often rely on proxy traces or coarse-grained characterizations that fail to capture the heterogeneity of modern multi-model LLM platforms. We present FineServe, an in-the-wild, multi-model LLM serving workload dataset collected from a global commercial marketplace, enabling fine-grained characterization of real-world serving dynamics across heterogeneous models and tasks. Leveraging FineServe, we conduct a comprehensive analysis of arrival dynamics and token behavior, revealing fundamentally different fluctuation regimes across model architectures, scales and task intents. Building on these insights, we develop the FineServe workload generator, which composes fine-grained model-aware workloads into configurable mixtures tailored for benchmarking multi-model serving platforms. By exposing these fine-grained workload dynamics, FineServe provides a realistic foundation for evaluating routing, scheduling, and capacity-planning strategies in LLM serving systems. FineServe is available at this https URL.
[AI-93] DQAOA-GPT : AI-Accelerated Distributed Quantum Optimization for Combinatorial Problems
链接: https://arxiv.org/abs/2607.20225
作者: Seongmin Kim,Abhinav Rijal,Yuri Alexeev,Nora Bauer,Martin Roetteler,Mina Yoon,George Siopsis,In-Saeng Suh
类目: Quantum Physics (quant-ph); Artificial Intelligence (cs.AI); Distributed, Parallel, and Cluster Computing (cs.DC); Optimization and Control (math.OC)
备注:
Abstract:While combinatorial optimization problems are central to many scientific and engineering applications, their solution remains challenging due to exponentially large search spaces. Variational quantum algorithms offer a promising route for tackling such problems, yet their practical performance is limited by repeated quantum circuit evaluations and classical parameter updates. In this work, we introduce DQAOA-GPT, a hybrid framework that integrates the distributed quantum approximate optimization algorithm (DQAOA), which decomposes a large optimization problem into smaller sub-problems, with GPT-based quantum circuit generation for solving those sub-problems. Rather than relying on iterative variational optimization, the proposed approach uses a trained generative model to directly generate high-quality quantum circuits for the decomposed sub-problems. As a benchmark, we evaluate DQAOA-GPT against conventional DQAOA on dense HUBO optimization problems with up to 100 decision variables. The results demonstrate that DQAOA-GPT significantly reduces computational cost while maintaining competitive solution quality, with larger acceleration observed for larger sub-problem sizes. Although this work focuses on benchmark-scale validation, the framework provides a promising foundation for larger-scale combinatorial optimization in hybrid HPC-QC environments through increased GPU resources and parallel computing capability.
[AI-94] PRIME-SVR: Physics-infoRmed Implicit Multi-Echo Slice-to-Volume Reconstruction for Fetal T2 mapping
链接: https://arxiv.org/abs/2607.20136
作者: Busra Bulut,Maik Dannecker,Thomas Sanchez,Sara Neves Silva,Steven Jia,Jean-Baptiste Ledoux,Leo Pomar,Joanna Sichitiu,Yvan Gomez,Meriam Koob,Vincent Dunet,Maria Deprez,Guillaume Auzias,Francois Rousseau,Jana Hutter,Daniel Rueckert,Meritxell Bach Cuadra
类目: Medical Physics (physics.med-ph); Artificial Intelligence (cs.AI)
备注:
Abstract:Slice-to-volume reconstruction (SVR) is the standard method for obtaining high-resolution (HR) 3D fetal brain volumes from motion-corrupted 2D MRI slice stacks acquired in multiple orientations. Existing SVR methods are optimized and validated only for clinical-range echo times (TEs), limiting their use at non-clinical TEs and making them incompatible with quantitative T2 mapping, a protocol- and center-independent biomarker of fetal brain maturation requiring HR reconstructions across multiple TEs. We present PRIME-SVR, the first implicit neural representation (INR) framework for joint HR reconstruction from multi-echo MRI. A single fully connected network models a continuous function from spatial coordinates to signal intensities across TEs, while a second network estimates slice-specific acquisition degradations. Cross-TE coherence is enforced via a Bloch equation-derived regularization penalizing deviations from expected T2 decay, with adaptive weighting that strengthens coupling for degraded stacks. The method is fully self-supervised. We validate PRIME-SVR on 39 in vivo fetal acquisitions (13 subjects x 3 TEs) from two centers, two vendors, and two field strengths (1.5 T and 0.55 T). Compared to state-of-the-art SVR, PRIME-SVR improves reconstruction sharpness by 47%, anatomical accuracy by 30%, and cross-TE structural consistency by 14%. It enables reconstruction at late TEs previously inaccessible to SVR, yielding the first 0.8 mm isotropic T2 maps at 0.55 T and the first T2 maps derived from INR-based SVR. PRIME-SVR also accelerates quantitative imaging by reducing the data needed for multi-TE reconstruction, cutting acquisition from 15 to 10 minutes while keeping T2 accuracy within 1.7% in white and deep gray matter, or to 5 minutes with a mean T2 error of 2.3% for high-quality acquisitions.
[AI-95] Drift-Aware RL-based Wavelet Denoising for Network-Traffic Anomaly Detection
链接: https://arxiv.org/abs/2607.20011
作者: Priyalakshmi Sheela,Indrakshi Dey
类目: ignal Processing (eess.SP); Artificial Intelligence (cs.AI)
备注: Submitted to IEEE Transactions on Wireless Communications. This version has not yet undergone peer review
Abstract:Traffic-utilisation measurements for network monitoring are corrupted by additive noise and statistical drift: time-dependent change in the signal’s mean, variance, distributional shape, or tail behaviour. Static wavelet denoising, calibrated under stationary independent and identically distributed (i.i.d.) Gaussian assumptions, becomes mismatched under drift and, at moderate-to-high signal-to-noise ratio (SNR), over-suppresses useful structure and degrades monitoring decisions. We propose a drift-aware framework treating adaptive wavelet denoising as a preprocessing layer optimised for two tasks: anomaly detection, recovering the multi-scale transient load bursts that noise and drift obscure, and capacity estimation, recovering the operational required capacity C_95 (95th percentile of utilisation). Because localised bursts are multi-scale structure a wavelet preserves but a low-pass filter removes, detection discriminates denoiser families. A four-detector gate (Page-Hinkley, variance-ratio, Jensen-Shannon, Anderson-Darling) determines when to invoke a learned policy, and a Proximal Policy Optimization agent selects a per-window wavelet configuration over a mixed discrete-continuous action space. Unlike prior work, the reward is downstream task utility, not reconstruction fidelity. The denoiser is benchmarked, per drift type and input SNR, against a low-pass moving-average filter, VisuShrink, SureShrink, BayesShrink, and a Wiener filter. Defining the anomaly target on the clean signal and the drift gate on the corruption keeps both stages non-circular.
[AI-96] When Shippers Become Algorithms: Candidate Exposure Information Design and the Concentration of LLM -Mediated Freight Markets
链接: https://arxiv.org/abs/2607.19967
作者: Takahiro Ezaki,Naoto Imura,Katsuhiro Nishinari
类目: Physics and Society (physics.soc-ph); Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注:
Abstract:Shippers are beginning to delegate carrier selection to large language model (LLM) agents. We ask what such delegation does to a freight matching market, and which platform design choices contain it. We carried out agent-based simulations in which fifty shipper agents, built on commercial LLMs from OpenAI (GPT), Anthropic (Claude), and Google (Gemini), procure truckload capacity for thirty days. The market implements the rules of digital freight matching: each load is offered down the shipper’s ranked list of carriers (waterfall tendering), carriers have daily capacity limits, spot prices respond to congestion, and carrier ratings accumulate with transactions. We found three risks and one remedy that works. Agents converged at once: for a fixed sampled carrier population, the same carrier was the modal first choice of every model on day one, attracting up to 76% of requests. Because each agent picks from its own randomly drawn list of displayed candidates, the platform controls how many options each shipper sees; concentration rose steeply once lists exceeded about ten carriers, with the onset differing across models. Which carriers ended up dominant varied widely from one sampled market to another, and displaying true quality instead of estimated ratings changed neither the level nor this variability (by design, quality affects only what agents see, never delivery outcomes). Against these risks, disclosing each carrier’s remaining daily capacity cut concentration by a third and doubled shipper surplus, while vendor diversification, list-order randomization, and popularity display showed no clearly detectable effect. Platform information design, ahead of model choice or model regulation, is the lever that works.
[AI-97] Causal dictionary learning reveals and validates transcription-factor binding features in genomic language models
链接: https://arxiv.org/abs/2607.19618
作者: Sarwan Ali
类目: Genomics (q-bio.GN); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Genomic language models achieve strong performance across regulatory-genomics tasks, yet what these models internally represent remains opaque, and the field lacks a principled procedure for verifying that an apparent concept'' inside a model is real rather than an artifact of sequence composition. We introduce a framework that combines sparse dictionary learning with causal intervention to extract, validate, and causally test interpretable features in genomic foundation models. Training top- k sparse autoencoders on the hidden activations of two architecturally distinct models, Nucleotide Transformer ( 6 -mer tokenization) and DNABERT-2 (byte-pair encoding), we recover thousands of monosemantic features that map to transcription-factor (TF) sequence motifs. We show that the naive validation of such features against position weight matrices is severely confounded by GC composition and repetitive elements, producing hundreds of spurious TF features’', and we develop a composition-matched, binding-resolved protocol that removes these confounds. Critically, we move beyond correlation: by ablating individual dictionary directions during the model’s forward pass and measuring the induced shift in the model’s own predictive distribution, we establish that specific features are \emphcausally used to represent cell-type-specific TF binding, not merely motif presence. Across three transcription factors (CTCF, GATA1, REST) and both architectures, causally validated binding features emerge reproducibly ( 7 – 14 of 15 tested features per condition), while two classes of negative control, scrambled binding labels and randomly selected features, yield no detectable signal. The framework is purely computational, uses only public data, and provides a reusable standard for interpretability claims in genomic deep learning.
[AI-98] Juxtaposition of Shallow Reservoir-Triggered Seismicity and Deep Tectonic Locking in the Qiaojia-Dongchuan Seismic Gap
链接: https://arxiv.org/abs/2607.19606
作者: Yuxin Zhou,Huai Zhang,S. Mostafa Mousavi,Guangyao Yin,Pei He,Yicun Guo,Shuang Yi,Yaolin Shi
类目: Geophysics (physics.geo-ph); Artificial Intelligence (cs.AI)
备注:
Abstract:Identifying the critical state of mature seismic gaps is challenging, especially when anthropogenic stress perturbations, such as reservoir impoundment, superimpose on tectonic loading. Here, utilizing a high-resolution dense array catalog from the Qiaojia-Dongchuan seismic gap (hosting the second-largest hydropower station in the world), we reveal a distinct vertical decoupling mechanism. The shallow activities exhibit high b-values (1.0), indicative of fluid-driven reservoir-triggered seismicity. Conversely, deep seismicity (20 km) outlines a ‘locked asperity’ characterized by low b-values (less than 0.8) and high Coulomb stress accumulation rate. We further identify a complex dipping structure, suggesting compound fault kinematics. Additionally, the calculated stress accumulation suggests this seismic gap is in a critical state with elevated rupture potential. Our findings indicate that shallow induced seismicity can mask the silent accumulation of deep tectonic strain. This decoupling model provides a new framework for assessing seismic risks in reservoir-fault systems globally.
[AI-99] Hybrid LLM -Guided Search for Quantum Reservoir Architecture Design
链接: https://arxiv.org/abs/2607.19506
作者: Krishna Bhatia,Gautami Sanjay Naik
类目: Quantum Physics (quant-ph); Artificial Intelligence (cs.AI)
备注: 4 pages, 2 figures, 3 tables
Abstract:Quantum reservoir computing (QRC) uses fixed quantum dynamics as a high-dimensional temporal feature map and trains only a lightweight classical readout. QRC is attractive for near-term quantum machine learning, but its performance depends strongly on architecture choices such as input encoding, reservoir depth, entanglement topology, measurement features, state-reset policy, feature construction, and readout regularization. We introduce \method, a simulator-based benchmark that formulates QRC design as constrained black-box architecture search and evaluates whether large language models can act as proposal controllers for this search problem. The benchmark compares five policies under identical evaluation budgets: random search, evolutionary search, Bayesian/TPE optimization, a feedback-based LLM agent, and \hybrid, which combines LLM proposals with memory, mutation, crossover, duplicate avoidance, and exploration. On NARMA10, Mackey-Glass forecasting, and temporal parity, \hybrid is the most consistent policy: it ranks first on NARMA10 and temporal parity and second on Mackey-Glass, narrowly behind evolutionary search. Under a 25-evaluation budget and three seeds, \hybrid improves over random search on all tasks, including a 23.6% relative reduction in Mackey-Glass error. The results do not show that LLMs are universal QRC optimizers; rather, they show that generative models can be useful high-level controllers when embedded inside validated, reproducible hybrid search loops.
[AI-100] Making Single-Cell Data Distillation Auditable: Traceable Real-Cell Coresets via Discrete Min-Max Selection
链接: https://arxiv.org/abs/2607.19426
作者: Yaodi Luo,Peize He,Bowen Han,Lingbei Mengg
类目: Genomics (q-bio.GN); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 9 pages
Abstract:Single-cell datasets are increasingly costly to store, audit, and reuse for model training. Dimensionality reduction and dataset distillation can reduce this burden, but conventional distillation methods often produce synthetic expression profiles that cannot be traced to an assayed cell. We formulate traceable single-cell data distillation as retaining original cell identifiers and gene symbols under fixed cell and gene budgets. The resulting training subset remains connected to measured counts, labels, and assay metadata, so unexpected predictions can be checked against their source data. We propose two real-cell selectors. Fixed-CF uses static characteristic-function matching. Minmax-CF solves an entropy-regularized discrete min–max problem that upweights poorly preserved directions and adds only observed cells. Across donor-, technology-, and perturbation-level shifts on three datasets, Minmax-CF retains 96.52% of Full balanced accuracy on MS, approximately matches Full on average on hPancreas with a median 2.55\times GPU speedup in the all-gene setting, and obtains the lowest pathway error among compressed methods on Norman. Performance remains weaker for rare states, some technology shifts, unseen perturbation components, and settings where fidelity is weakly associated with downstream utility. Because the selected IDs refer to measured cells, these cases can be investigated by inspecting the corresponding training support, labels, and assay metadata. Minmax-CF consistently reduces worst-direction discrepancy, while downstream utility and cost vary across datasets and tasks.
机器学习
[LG-0] PG-KINN: A Physics-Informed Petrov-Galerkin Kolmogorov-Arnold Network for Solving Forward and Inverse PDEs
链接: https://arxiv.org/abs/2607.20378
作者: Amirhossein Sadr,Nima Soltani,Vahideh Moghtadaiee,Aida Pakniyat,Dara Rahmati,Saeid Gorgin
类目: Machine Learning (cs.LG); Numerical Analysis (math.NA)
*备注:
Abstract:Physics-informed learning of partial differential equations (PDEs) has been dominated by multilayer perceptrons (MLPs), whose spectral bias and dense parameterization limit both accuracy and interpretability. Kolmogorov Arnold Networks (KANs) mitigate these limitations because their learnable spline activations are structurally aligned with the piecewise-polynomial bases of classical discretizations. However, the way a PDE is cast into a loss functional is as decisive as the choice of approximator: strong-form residual minimization requires high-order derivatives and heavily weighted losses, the energy (Bubnov-Galerkin) form is restricted to self-adjoint operators and, as we show, collapses to a trivial solution for parameter-identification problems, and boundary integral forms require a known fundamental solution. We propose PG-KINN, a physics-informed KAN built on a Petrov-Galerkin formulation in which the trial space is a KAN and the test space is an independent, compactly supported, piecewise-polynomial space evaluated with Gauss-Legendre quadrature. Integration by parts lowers the differentiation order while retaining applicability to general non-self-adjoint, nonlinear, and inverse problems; the localized test functions turn the global residual into a set of element-wise weak residuals with favorable conditioning. On a suite of benchmarks spanning crack singularities, stress concentration, Neo-Hookean hyperelasticity, inverse parameter identification in heterogeneous media, and complex geometries, PG-KINN consistently outperforms legacy MLP baselines and state-of-the-art KAN-based strong/energy/inverse formulations (PIKAN). These results position the Petrov-Galerkin coupling of KAN trial spaces and polynomial test spaces as a robust and accurate route for AI-based computational mechanics.
[LG-1] Online Variance Reduction for Domain Adaptation on Streaming Data
链接: https://arxiv.org/abs/2607.20374
作者: Andrea Napoli
类目: Machine Learning (cs.LG)
*备注:
Abstract:This paper studies the problem of stochastic variance reduction (SVR) for the maximum mean discrepancy (MMD) and correlation alignment (CORAL) loss functions. Although various offline SVR algorithms for these losses have been proposed, these are incompatible with online, distributed, or incremental learning settings. This paper presents Adaptive vaRiance Reduction via Online reWeighting (ARROW), the first online SVR algorithm for the MMD and CORAL for streamed data. The method maintains moving average references of the alignment statistics, and adaptively reweights incoming minibatches so that the minibatch and reference statistics are aligned. Further, we propose a relaxed reweighting scheme so that the ensuing weight-optimisation problem is tractable. In experiments and simulations, we show that ARROW performs competitively with offline algorithms in terms of runtime, degree of variance reduction achieved, and target domain accuracy.
[LG-2] Variance-reduced Domain Adaptation using Paired Sampling
链接: https://arxiv.org/abs/2607.20367
作者: Andrea Napoli
类目: Machine Learning (cs.LG)
*备注:
Abstract:Correlation alignment and the maximum mean discrepancy are two widely used distribution-matching frameworks for unsupervised domain adaptation (UDA). However, high variance in these losses has been shown to undermine their effectiveness in minibatch optimisation settings. Furthermore, the losses lack finite-sum structure, which renders them incompatible with classical stochastic variance reduction (SVR) methods. This paper proposes Paired Sampling for Domain Adaptation (PSDA), a novel SVR technique tailored to such objectives. PSDA pairs observations both within and across domains, to form quadruplets that are always sampled together during training. The pairings are designed to minimise expected gradient variance, and reduce to solving a set of linear assignment problems. Our simulations demonstrate reduced variance compared to related methods, and experiments on three domain shift datasets show improved target domain accuracy.
[LG-3] Interval and fuzzy physics-augmented neural networks (iPANN and fPANN) for uncertainty quantification and propagation in constitutive modeling
链接: https://arxiv.org/abs/2607.20339
作者: Somesh Pratap Singh,Govinda Anantha Padmanabha,Jingye Tan,Steven Yang,Reese E. Jones,D. Thomas Seidl,Nikolaos Bouklas
类目: Machine Learning (cs.LG); Computational Physics (physics.comp-ph)
*备注:
Abstract:Constitutive modeling under uncertainty remains a central challenge for reliable mechanics simulations, particularly when the available stress-deformation data are sparse, noisy, or heterogeneous. We propose interval and fuzzy physics-augmented neural networks (iPANNs and fPANNs) for uncertainty-aware hyperelastic constitutive modeling. iPANNs learn sparse lower, mean, and upper free energy density branches whose stresses, obtained by automatic differentiation, ultimately enclose noisy stress observations. In contrast to this deterministic interval description, fPANNs embed the learned iPANN branches into a fuzzy-set representation through alpha-cut interpolation, yielding a nested family of admissible responses. iPANNs and fPANNs encode mechanistic constraints - preserving objectivity, consistency and promoting polyconvexity - and smoothed L0 regularization promotes interpretable energy representations. The bound models are trained through a two-stage transfer-learning procedure in which a sparse mean constitutive response is learned first and then fine-tuned into lower and upper energy branches. We evaluate the framework on synthetic isotropic hyperelastic data with heteroscedastic noise, varying random realizations, shifted noise means, and varying noise magnitudes. The results show that the learned bounds enclose noisy stress observations while generalizing to the test set. Further, we examine the propagation of uncertainty through the mean, upper and lower bound predictions of the learned iPANN models in a finite element setting. The proposed framework provides a compact, physics-consistent route for distribution-free aleatoric uncertainty quantification in hyperelastic constitutive modeling, and propagation in downstream finite element simulations.
[LG-4] Multi-modal transformer for signal classification in nanopore blockade experiments
链接: https://arxiv.org/abs/2607.20323
作者: Sandro Kuppel,Julian Hoßbach,Samuel Tovey,Christian Holm
类目: Machine Learning (cs.LG); Computational Physics (physics.comp-ph); Biomolecules (q-bio.BM)
*备注: 22 pages (incl. references), 8 figures
Abstract:Nanopore devices have emerged as powerful tools for single-molecule sensing, with potential for rapid, portable diagnostics. They detect changes in ionic current as analytes enter nanometer-scale pores, providing a means of identifying diverse biomarkers from their characteristic signal patterns. However, these signals are highly complex, and reliably assigning them to specific molecules remains a major challenge. Here, we address this by introducing a multi-modal deep learning architecture that jointly processes multiple signal representations, including raw time-series data, wavelet-based images, and static feature vectors. Our approach surpasses existing methods by more than 10 percentage points on a 42-peptide benchmark and transfers to a 20-amino-acid dataset with near-perfect accuracy. The model integrates complementary information from these representations, with attention analysis showing that the time-series and wavelet-image inputs emphasize different features of the same event. Together, these results demonstrate the potential of machine learning to enable robust, high-accuracy molecular identification with nanopore sensors.
[LG-5] Classical Hardware Acceleration of Quantum Autoencoders for Real-Time Anomaly Detection in Collider Experiments
链接: https://arxiv.org/abs/2607.20302
作者: Ivan Ge,Sagar Addepalli,Abhilasha Dave,Julia Gonski
类目: Machine Learning (cs.LG); High Energy Physics - Phenomenology (hep-ph); Instrumentation and Detectors (physics.ins-det)
*备注: 15 pages, 7 figures, 2 tables
Abstract:Quantum machine learning (QML) algorithms in high energy physics (HEP) can efficiently represent and leverage long-range, high-order correlations in high-dimensional collider data, potentially with fewer parameters and favorable scaling relative to classical models. Deployment of QML in real-time collider applications such as trigger systems requires the ability to emulate and compile quantum circuits classically, then synthesize the resulting quantum gates onto low-latency hardware accelerators, namely field-programmable gate arrays (FPGAs). We present a study of variational quantum autoencoder models for real-time anomaly detection triggers in modern collider experiments. The models achieve performance comparable to state-of-the-art classical approaches and, after FPGA synthesis, satisfy resource usage and timing constraints consistent with trigger applications in future colliders. This work provides one of the first FPGA implementations of QML models for HEP triggers, enabling higher-capability models in today’s classical data acquisition pipelines while advancing quantum readiness of collider experiment infrastructure.
[LG-6] Interpretable Fuzzy Rule-Based Regression Extension for Ex-Fuzzy Library
链接: https://arxiv.org/abs/2607.20277
作者: Cayan Deniz Kucuktopana,Javier Fumanal-Idocin,Richard Pitts,Javier Andreu-Perez
类目: Machine Learning (cs.LG)
*备注:
Abstract:Machine learning models achieve high predictive accuracy in regression tasks, but their deployment in safety-critical and regulated domains requires interpretability. While fuzzy rule-based systems offer transparent, linguistically explicit interpretable models, Mamdani-style fuzzy regression remains underrepresented in modern machine learning software libraries. This paper presents an interpretable regression extension for the Ex-Fuzzy library, enabling Mamdani fuzzy inference with scalar consequents learned directly from data. For this, a target-aware partition initialisation strategy based on Fuzzy C-Means clustering is introduced, in which linguistic variables are derived from an augmented input-output space to emphasise output-relevant regions of the feature space. The proposed extension is evaluated on ten regression datasets from the KEEL repository, comparing Gaussian and trapezoidal partition strategies against standard baselines including linear regression, multilayer perceptron, and random forests. Experimental results show that Gaussian partitions consistently outperform uniform trapezoidal partitions, achieving a mean coefficient of determination of approximately 0.86 while producing compact rule bases of 10-15 human-readable rules. The proposed implementation provides a transparent and competitive alternative to black-box regression models, supporting practical interpretability with competitive predictive performance.
[LG-7] Breaking the T3/4 Barrier for Regret Minimization With Bi-Dimensional CDFs
链接: https://arxiv.org/abs/2607.20258
作者: Matteo Castiglioni,Anna Lunghi,Alberto Marchesi
类目: Machine Learning (cs.LG)
*备注:
Abstract:We study regret minimization for learning CDF-related objectives of the form [ g(x)\cdot\mathbbP_X\sim\mathcalD(X\le x), ] over [0,1]^2 , where g is a known Lipschitz function and \mathcalD is an unknown distribution. At each round t , the learner selects a point x_t and observes the binary feedback \mathbbI(X_t\le x_t) , where X_t\sim\mathcalD . We design an algorithm achieving regret \widetilde\mathcalO(T^7/10) , improving over the previous best-known bound of \widetilde\mathcalO(T^3/4) and showing that the curse of dimensionality can be at least partially lifted for this class of objectives, though a gap remains with the \Omega(T^2/3) lower bound. As an application, our techniques yield the same \widetilde\mathcalO(T^7/10) regret bound for profit maximization in repeated bilateral trade with fixed prices.
[LG-8] PhaseAware: Interpretable Human-in-the-Loop Rehabilitation Scoring with Boundary Monitoring
链接: https://arxiv.org/abs/2607.20237
作者: Yankai Zheng,Yuhe Liu,Yuxin Ma,Tianci Xue,Jiayuan Tian,Yu Fu,Yuxuan Hu,Jianing Wang,Zichun Xiao,Junya Mu,Shaohui Ma
类目: Machine Learning (cs.LG)
*备注: 22 pages, 4 main figures, 3 tables, and 17 supplementary figures. Supplementary Information is included in the same PDF
Abstract:Rehabilitation scoring systems are most useful when their outputs can be reviewed and interpreted within clinical workflows. This study presents PhaseAware, a compact framework for continuous rehabilitation quality assessment that combines a temporal backbone with phase- and body-group descriptors through a backbone-conditioned gated residual pathway. The model was evaluated on the UI-PRMD deep-squat protocol and further tested on the KIMORE squatting subset. On UI-PRMD, PhaseAware achieved an RMSE of 0.0230, corresponding to an 88.9% reduction relative to the accepted baseline. It also maintained favorable performance on KIMORE, suggesting that the phase-aware design transfers across related squatting protocols. In addition to score prediction, PhaseAware generates structured review cues based on phase- and body-level sensitivity, highlighting the movement stages and body regions most relevant to each prediction. The architecture employs a backbone-conditioned gated residual mechanism to stabilize feature representation, supporting use in resource-constrained settings. These cues are intended to support clinician review, boundary-case monitoring, and human-in-the-loop triage rather than autonomous decision-making. Overall, PhaseAware offers a practical and interpretable approach to rehabilitation scoring that may help integrate automated assessment into information systems while preserving clinician oversight.
[LG-9] PIER: Physics-Informed Environmental Retrieval for Time-Series Modeling
链接: https://arxiv.org/abs/2607.20230
作者: Shiyuan Luo,Runlong Yu,Chonghao Qiu,Yue Qin,Rahul Ghosh,Robert Ladwig,Paul C. Hanson,Yiqun Xie,Xiaowei Jia
类目: Machine Learning (cs.LG)
*备注:
Abstract:Accurate modeling of environmental systems is fundamental to scientific understanding and decision-making, yet remains challenging because observations are limited and physical dynamics vary across systems. Retrieval-augmented approaches offer a natural path to transfer knowledge across systems, but standard embedding-based retrieval does not guarantee consistency of underlying physical processes, since scenarios with similar embeddings may arise from different underlying mechanisms. We propose Physics-Informed Environmental Retrieval (PIER), a model-agnostic framework that augments embedding-based retrieval with a physics-aware stream that scores candidates by flux-response consistency with the target, using local verifiers trained on physics-derived flux features. A weight adjustment mechanism then learns per-scenario weights that adaptively balance the two retrieval streams based on diagnostic features summarizing physics-stream reliability. Experiments on 356 lakes across the Midwestern United States spanning 41 years show that PIER consistently outperforms baselines for water temperature and dissolved oxygen prediction, and serves as a general augmentation strategy across diverse backbones.
[LG-10] User-Centric Modeling of Transactional Sequences with Explainable State Space Models
链接: https://arxiv.org/abs/2607.20228
作者: Ivan Palagin
类目: Machine Learning (cs.LG)
*备注:
Abstract:We propose a hybrid approach for user-centric modeling of transactional event sequences that combines contrastive representation learning (CoLES) with State Space Models (SSMs). While contrastive methods yield high-quality compressed user representations, existing encoders – RNNs and Transformers – suffer from vanishing gradients or quadratic complexity, respectively. Mamba, a selective SSM, efficiently handles long-range dependencies but remains underexplored for personalized user analysis. We investigate two integration strategies: (1)~initializing the Mamba hidden state with a CoLES embedding, and (2)~prepending the projected CoLES embedding as a prefix token to the input sequence. Both approaches supply the model with an informative user prior from the first step. Experiments on three public datasets – Age (multiclass age-group prediction), MBD (multi-label product acquisition), and Taobao (binary purchase prediction) – demonstrate consistent improvements over standalone Mamba and CoLES with a linear classifier, with the hybrid models converging 2–3 \times faster than the plain SSM baseline. Explainability analysis via discretization-step maps and Integrated Gradients reveals selective event filtering on behavior-rich datasets and identifies the most informative transaction features.
[LG-11] OLEDLM: A Unified Language Model for OLED Molecular Design
链接: https://arxiv.org/abs/2607.20194
作者: Fukang Wen,Yuchong Tang,Jingyuan Li,Beichen Wang,Yixuan Jiang,Xiaoyi Jiang,Yaxuan Liu,Shunyu Wang,Zuoqiang Shi,Yi Zhu,Yanan Zhu,Pipi Hu
类目: Machine Learning (cs.LG)
*备注:
Abstract:The development of organic light-emitting diode (OLED) materials faces the compounded challenges of an astronomically large chemical space, stringent quantum-chemical constraints, and a scarcity of labeled data. Although the question of OLED generation is important, few models have been trained effectively for this specific domain. We propose an inverse molecular design framework based on causal language models: given target optoelectronic properties (e.g., excitation energy, oscillator strength), our model directly generates OLED SMILES sequences satisfying the specified constraints. We employ a multi-stage strategy: first, we establish a foundational chemical language model using a LLaMA-style transformer architecture. To the best of our knowledge, this represents the first successful adaptation of LLMs specifically for the OLED domain, bridging the gap between generic molecular generation and the stringent structural requirements of optoelectronic materials. Second, we fine-tune property predictors based on a BERT model pre-trained on our large-scale OLED dataset. Then, we perform Reinforcement Learning on our fine-tuned model, leveraging our property predictor, for better SMILES generation. Finally, through DFT verification, we demonstrate that our framework can efficiently navigate the OLED chemical space, generating novel candidates with high structural validity and optimized optoelectronic properties.
[LG-12] On Optimization Complexity of Second-Order Certified Unlearning
链接: https://arxiv.org/abs/2607.20192
作者: Nikita Doikov,Anastasia Koloskova
类目: Machine Learning (cs.LG); Optimization and Control (math.OC)
*备注:
Abstract:We study machine unlearning: the removal of memorized training data from a trained model. Specifically, we investigate the algorithmic complexity of certified unlearning from an optimization perspective. We formalize the goal of an unlearning algorithm as simultaneously achieving certified unlearning and optimization accuracy. Utilizing the notion of uniformly convex regularizers, we prove new bounds on the distance between initial and unlearned models using a novel substitute for generalization error. Thus we theoretically demonstrate that if the removed data is well-predicted by the unlearned model, the corresponding optimization problem is simple. Furthermore, we develop a new second-order unlearning algorithm with an anisotropic Gaussian mechanism and state-of-the-art global convergence. We prove fast rates for our method in achieving certified unlearning for linear models with quasi-self-concordant losses. As a direct application, our theory covers unlearning for logistic and exponential regressions and shows a provable benefit of utilizing second-order information compared to first-order unlearning methods.
[LG-13] Instance Hardness-Based Relevance for Imbalanced Regression
链接: https://arxiv.org/abs/2607.20173
作者: Vitor M. Leitao,Juscimara G. Avelino,George D. C. Cavalcanti,Rafael M. O. Cruz
类目: Machine Learning (cs.LG)
*备注:
Abstract:Imbalanced regression problems arise when the target variable has an asymmetric distribution, resulting in underrepresented value ranges in the dataset. Traditional approaches for identifying rare instances rely on a relevance function that assigns higher importance to specific regions of the target distribution. However, the effectiveness of imbalance-aware learning methods depends strongly on how relevance is defined. In more complex scenarios, such as bimodal distributions, traditional relevance functions struggle to capture rarity, as they assign fixed relevance values based solely on target values, thereby compromising the distinction between truly rare and normal instances. To address these limitations, this study proposes an Instance Hardness-based relevance function (InHaR) for identifying rare instances in regression problems. Unlike traditional relevance functions, the proposed approach incorporates learning difficulty, allowing rarity to be inferred not only from the target distribution but also from the difficulty of instances for the learning algorithm. This property is particularly important in bimodal scenarios, where rarity cannot be accurately inferred from target values alone. Experimental results demonstrate that the InHaR correctly identifies rare regions under bimodal distributions and, when used to guide resampling strategies such as Random Oversampling (RO) and Gaussian Noise (GN), leads to significant improvements in predictive performance compared to traditional relevance-based approaches. The code, dataset, and further details about the proposed method are publicly available at this https URL.
[LG-14] Self-organizing Architecture of Receptron Units: a Hardware-Aware Framework for Edge Intelligence
链接: https://arxiv.org/abs/2607.20162
作者: Stefano Radice,Ludovico Casaccia,Riccaro Emanuele Beccalli,Bruno Paroli,Paolo Milani
类目: Machine Learning (cs.LG); Emerging Technologies (cs.ET)
*备注: This work has been submitted to NeuroPHY 2026
Abstract:The growing demand for intelligent processing at the edge of IoT networks is constrained by the severe computational and memory limitations of microcontroller units, which render impractical conventional deep learning approaches. We propose a neuromorphicinspired classifier based on the Receptron model, a single-unit architecture capable of implementing non-linearly separable decision boundaries, without resorting to multi-layer networks. The model is designed for direct deployment on mid-range MCUs, while supporting continuous on-device adaptation. Experimental evaluation on basic dataset benchmarks yields cross-validated accuracies compatible with standard machine learning method baselines. These results position the Receptron as a viable and interpretable alternative for resource-constrained neuromorphic edge systems operating in dynamic, non-stationary environments.
[LG-15] Local Stability and Gaussian Smoothing of Quantized Neural Networks
链接: https://arxiv.org/abs/2607.20153
作者: Sergey Salishev,Anton Makarov,Oleg Granichin
类目: Machine Learning (cs.LG); Systems and Control (eess.SY); Optimization and Control (math.OC)
*备注: Accepted at the 23rd IFAC World Congress (IFAC WC 2026), Busan, Republic of Korea, 2026; 6 pages, 2 figures
Abstract:We study Gaussian averaging as a smooth surrogate for quantized neural models. Under bounded local oscillation, we derive a local dimension-dependent bound on |f-g|, linking Gaussian smoothing to the stability analysis of discontinuous networks. We compute closed-form Gaussian averages of the rectified linear unit (ReLU) and sign activation functions, and illustrate the mechanism on a high-dimensional binary perceptron, where layer-preactivation aggregation under an explicit quantization-noise surrogate yields the Gaussian envelope used in inference-side smoothing and training-side smooth surrogate gradients.
[LG-16] Multi-stage Dynamic Selection for Cross-Project Defect Prediction
链接: https://arxiv.org/abs/2607.20151
作者: Juscimara G. Avelino,Juscelino S. A. Junior,George D. C. Cavalcanti,Rafael M. O. Cruz
类目: oftware Engineering (cs.SE); Machine Learning (cs.LG)
*备注:
Abstract:Cross-Project Defect Prediction (CPDP) involves building models using data from external projects, called training projects, to predict modules from the target project. However, traditional CPDP methods suffer from the distribution shift between training and target projects that affects the model’s performance. This paper proposes a novel CPDP framework that addresses this issue by proposing a two-stage multiple classifier system (MCS) selection scheme: one working at the project level and another at the module level. In the first stage, the framework evaluates multiple possible MCS configurations to find one that covers and generalizes well across multiple training projects. Consequently, the proposal is likely to obtain a diverse set of classifiers, each specialized in tackling software modules with distinct characteristics. The second selection stage operates at test time, selecting the most competent classifiers to predict each new module in the target project. Unlike previous approaches that apply the same classifiers to the entire target project, the proposed framework performs module-level model selection. This way, the system is more robust to changes in distributions between training and target projects because the selected set of classifiers is module-dependent. Our experimental results using 82 projects from four different CPDP benchmark datasets demonstrate that the proposed approach outperforms the state-of-the-art CPDP methods in most scenarios. The code, dataset, and further details about the proposed method are publicly available at this https URL.
[LG-17] CURED: Creating Understanding and Repairing Errors Demonstrator
链接: https://arxiv.org/abs/2607.20140
作者: Nicholas Chandler,Sebastian Jäger,Philipp Jung,Felix Bießmann
类目: Machine Learning (cs.LG)
*备注:
Abstract:Detecting and cleaning errors in tabular data is a prerequisite for data intense software applications. Recent research at the intersection of Machine Learning (ML) and Database Management Systems (DBMS) highlights the potential of statistical learning algorithms for error detection and cleaning. This paper combines our recent work on ML-based data cleaning and error models in a unified demonstrator. The web application allows users to upload tabular data, perturb the data with realistic data dependent errors and use modern ML methods to clean and understand error mechanisms in data. Our demonstrator helps to bridge the gap between theoretical advancements and intuitive practical insights in the context of error models and data cleaning algorithms for tabular data. The demonstrator is available at this https URL
[LG-18] Cumsum-Composable Phase Transport for Low-Cost Streaming Keyword Spotting
链接: https://arxiv.org/abs/2607.20086
作者: Mahesh Godavarti
类目: ound (cs.SD); Machine Learning (cs.LG)
*备注:
Abstract:State-space sequence models are attractive for streaming speech because they maintain compact recurrent state, but scan-style training kernels can have unfavorable constants for short audio tasks. We study cumsum-composable phase transport, a streaming-native temporal layer for keyword spotting. Each layer projects acoustic frames to complex channels, transports them by learned unitary rotations, accumulates a finite window using prefix differences, and applies a gated residual update. The same prefix representation gives exact batched training with ordinary cumulative sums and exact online inference with one prefix update per frame. Unitary transport is the key constraint: inverse rotations have norm one, keeping prefix terms well conditioned while memory is supplied by windows or block readouts. On Google Speech Commands v2 with 12 labels, mel+cumsum models retain competitive accuracy with compact baselines. The strongest single-seed run reaches 97.3% test accuracy; a 51.6K-parameter tied model also reaches 97.3%, and a 24.8K tied model reaches 96.8% versus 97.1% for a 25.6K MelCNNMaxPool baseline. In a matched cumsum-versus-scan benchmark, cumsum+window gives comparable accuracy, 94.82% versus 94.33%, while training 1.07x faster and reducing single-example latency from 7.09 ms to 5.01 ms on a Tesla T4. These results support cumsum phase transport as a simple low-cost temporal primitive for streaming keyword spotting. Subjects: Sound (cs.SD); Machine Learning (cs.LG) MSC classes: 68T07, 68T10, 68W10, 94A12 ACMclasses: I.2.7; I.5.1; I.5.4; C.3 Cite as: arXiv:2607.20086 [cs.SD] (or arXiv:2607.20086v1 [cs.SD] for this version) https://doi.org/10.48550/arXiv.2607.20086 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-19] Evaluating and Mitigating Gender Bias in Pre-trained Embeddings for ML-based Recruitment ECAI2026 IJCAI
链接: https://arxiv.org/abs/2607.20073
作者: Farnaz Faramarzi Lighvan,Lynn Houthuys
类目: Machine Learning (cs.LG)
*备注: Accepted for presentation at the TRUST-AI 2026 Workshop, held in conjunction with IJCAI/ECAI 2026, Bremen, Germany
Abstract:AI-based recruitment systems that rely on machine learning models trained on historical CV data, risk perpetuating and amplifying social biases. A key challenge arises in unstructured CV text, where pre-trained language model embeddings may infer sensitive attributes such as gender even after explicit indicators are removed. In this paper, we evaluate nine pre-trained embedding models on the synthetic FairCVdb dataset, analyzing the informativeness of their embeddings for applicant scoring and their susceptibility to gender leakage, on both original and gender-scrubbed biographies. We further use a multi-task adversarial learning framework with gradient reversal to predict applicant suitability while suppressing gender information from learned representations. Finally, we use a multi-objective Pareto-front-based model selection to balance predictive utility and fairness. Our experimental results show that explicit gender scrubbing substantially reduces but does not eliminate gender leakage, while adversarial learning improves fairness mainly on original biographies and acts as a complementary strategy rather than a substitute for text-level debiasing.
[LG-20] Zero-Shot Heart Rate Variability Forecasting from Consumer Wearables Using Time Series Foundation Models
链接: https://arxiv.org/abs/2607.20027
作者: Luukas Peräkylä,Fahad Sohrab,Ville Hautamäki,Merja Heinäniemi,Sui Huang,Pekka Abrahamsson
类目: Machine Learning (cs.LG)
*备注: Accepted to Computing in Cardiology (CinC) 2026. 4 pages, 2 figures, 3 tables
Abstract:Short-term Heart Rate Variability (HRV) forecasting could provide clinicians with actionable lead time for detecting autonomic dysfunction and adverse cardiac events. Consumer wearable devices generate fragmented, artifact-rich HRV signals that challenge conventional forecasting approaches. In this study, we evaluated the forecasting ability of three Time Series Foundation Models (TSFMs), TimesFM, Chronos, and MOIRAI, against traditional baselines (Mean, Exponential Smoothing, and Exponentially Weighted Moving Average) on real-world wearable data collected from 49 healthy individuals. To address data fragmentation, we introduce a variability-preserving imputation method that augments linear interpolation with locally adaptive stochastic noise, retaining physiological dynamics essential for accurate forecasting. The results show that TSFMs outperformed all baselines without fine-tuning, achieving average Mean Absolute Scaled Error (MASE) between 0.81 and 0.87 across TSFMs and both context lengths (32 and 64 time steps), with Chronos and TimesFM as the top models, though MOIRAI showed limited gains over baselines. With up to a 2-hour forecast horizon, the results establish a baseline for TSFMs’ performance on a real-world dataset, highlighting domain-specific fine-tuning as a promising direction for clinical deployment.
[LG-21] Generalized Kalman filter based temporal difference reinforcement learning
链接: https://arxiv.org/abs/2607.20010
作者: Vasos Arnaoutis,Eric Lutters,Bojana Rosić
类目: Machine Learning (cs.LG); Computational Engineering, Finance, and Science (cs.CE)
*备注: 39 pages, 18 figures
Abstract:In this paper, we present a generalized temporal-difference (TD) reinforcement learning framework based on the theory of conditional expectations. The value and action-value (Q-value) functions are treated as uncertain quantities, and their estimation is formulated as a stochastic inference problem. Unlike classical Kalman-based temporal-difference learning, which relies on linear-Gaussian assumptions, the proposed formulation is derived directly from the conditional expectation framework and naturally extends to nonlinear models and non-Gaussian probability distributions. The proposed method recursively estimates not only the conditional expectation of the value function but also its second probabilistic moment, thereby quantifying the uncertainty associated with the learned value function throughout the learning process. To obtain a computationally tractable algorithm, the stochastic problem is discretized using either polynomial chaos expansions or ensemble-based approximations, providing efficient representations of the underlying random variables. The proposed framework is demonstrated on two optimal control problems: a linear mass–spring–damper system and a nonlinear heat conduction problem in a closed cavity. The numerical examples illustrate the capability of the proposed method to accurately estimate both the value function and its associated uncertainty, while extending classical Kalman-based temporal-difference learning to a broader class of stochastic systems.
[LG-22] Good Practice Guide for quantifying uncertainties for machine learning models applied to photoplethysmography signals
链接: https://arxiv.org/abs/2607.19999
作者: P. Harris,C. Bench,M. Rinkevičius,V. Marozas,L. Coquelin,A. Thompson,M. Nandi,U. Hackstein,P.J. Aston
类目: Machine Learning (cs.LG)
*备注:
Abstract:This Good Practice Guide presents work done in the QUMPHY project (Uncertainty quantification for machine learning models applied to photoplethysmography signals) that considered both machine learning and uncertainty quantification for problems which used photoplethysmography (PPG) signals from wearable devices as input. It provides high-level guidance on what types of machine learning model might be used and how different models compare when applied to both regression and classification tasks. It provides guidance on the implementation of different methods for uncertainty quantification, covering both model-dependent and model-independent techniques, and on the validation of the results provided by those methods. It also describes six benchmark problems together with pointers to different benchmark datasets for each problem. Software is described that can assist practitioners in implementing the methods described herein and there is a brief consideration of ethical issues. It concludes with a summary and recommendations.
[LG-23] Diffusion ReRoll: Revisable Denoising for Robotic Sequential Prediction
链接: https://arxiv.org/abs/2607.19919
作者: Seonsoo Kim,Seongil Hong,Jun-Gill Kang
类目: Robotics (cs.RO); Machine Learning (cs.LG)
*备注: Project Page: this https URL
Abstract:We propose Diffusion ReRoll, a diffusion-based framework for robotic sequential prediction that enables revisable denoising over horizons. Existing diffusion-based sequence predictors typically perform a single monotonic denoising process. In contrast, Diffusion ReRoll selectively re-noises regions that have become locally stable while the remaining regions continue denoising, so the re-noised regions can be refined again using context from the rest of the horizon. This structured re-noising enables iterative cross-horizon revision, allowing earlier and later segments to revise one another, while maintaining local consistency. We evaluate Diffusion ReRoll against full-sequence diffusion and causal denoising based on Diffusion Forcing across long-horizon planning, policy learning, and unified video-action modeling. On OGBench PointMaze and AntMaze, Diffusion ReRoll achieves relative gains in average success rate of 21% over Diffusion Forcing in matched guidance-based planning and 23% over Diffuser in matched goal-inpainting. In diffusion-policy-style action prediction, Diffusion ReRoll improves average success by 56.5% relative to Diffusion Policy across different prediction horizons and history lengths on the LIBERO-10 multi-task benchmark. In unified video-action prediction, Diffusion ReRoll improves policy and inverse dynamics performance, especially under out-of-distribution evaluation, and achieves the best action-video consistency. These results support structured re-noising as an effective mechanism for revisable robotic sequence generation.
[LG-24] Nonlinear Bias-Compensated Adaptive Filter and Its Application for Time-Series Prediction
链接: https://arxiv.org/abs/2607.19902
作者: Yi Peng,Haiquan Zhao,Jinhui Hu
类目: Machine Learning (cs.LG); Audio and Speech Processing (eess.AS)
*备注:
Abstract:Most existing nonlinear adaptive filtering algorithms only account for output noise, neglecting the fact that input noise is also prevalent in practice. Although the recently proposed bias-compensated kernel least mean square (BCKLMS) algorithm addresses input noise in the nonlinear errors-in-variables (EIV) model, it still suffers from two major limitations. First, the use of a fixed-size dictionary restricts network growth but also prevents it from fully capturing the characteristics of the input signal. Second, as an least mean square (LMS) based algorithm, it exhibits poor robustness in the presence of non-Gaussian noise in the output signal. To overcome these issues, this paper proposes the random Fourier bias-compensated filter under general adaptive function (RFFBCGA) algorithm. Within the random Fourier feature based bias-compensated (RFFBC) framework, the proposed algorithm not only maintains a fixed network structure and effectively mitigates input noise interference through the BC term, but also achieves improved characterization of the input signal. Moreover, by leveraging the flexible form of the general adaptive (GA) function, the algorithm’s robustness across various noise scenarios is further enhanced. Extensive simulations, including real-world time series prediction tasks, demonstrate the superiority of the proposed method.
[LG-25] Local Causal Structure Learning in the Presence of Latent Variables and Selection Bias
链接: https://arxiv.org/abs/2607.19866
作者: Zheng Li,Hao Zhang,Ruxin Wang,Ruichu Cai,Kun Zhang,Feng Xie
类目: Machine Learning (cs.LG)
*备注:
Abstract:Discovering the direct causes and effects of a target variable from observational data is a fundamental problem in causal discovery, with broad applications in domains such as gene regulatory analysis and biomedical research. Existing causal discovery methods either learn a global causal structure, which incurs substantial computational cost, or assume the absence of latent variables and selection bias, assumptions that are often violated in real-world settings. Motivated by these challenges, we study local causal structure learning in the presence of latent variables and selection bias. Specifically, we first characterize a local region that enables target-specific causal discovery without recovering the entire global structure. We then establish a theoretical bridge between causal information learned from the observed distribution induced on this local region and the corresponding information in the global causal structure. Building on these foundations, we propose LoCaLS, a local causal structure learning algorithm that is sound and complete under standard assumptions and identifies the same direct causes and effects of a target variable as those identifiable by global causal discovery methods, while allowing for latent variables and selection bias. Extensive experiments on random and real-world structures demonstrate that the proposed method consistently achieves higher structural accuracy than existing local methods while requiring substantially less computational effort than state-of-the-art global methods. Furthermore, applications to two real-world gene expression datasets reveal biologically plausible target-specific causal structures, demonstrating its practical applicability in large-scale biological data analysis.
[LG-26] Adversarial Frontiers: Minimum-Norm Attack Ensembles for Robustness Evaluation
链接: https://arxiv.org/abs/2607.19855
作者: Luca Scionis,Luca Melis,Maura Pintor,Fabio Brau,Ambra Demontis,Giorgio Fumera,Fabio Roli,Battista Biggio
类目: Machine Learning (cs.LG); Cryptography and Security (cs.CR)
*备注:
Abstract:Adversarial robustness is commonly evaluated with predefined attack ensembles, such as AutoAttack, at a single perturbation budget \varepsilon and on a selective choice of perturbation norms. We argue this formulation is fundamentally limited. First, robustness–perturbation curves may intersect or decay at different rates across models, making single- \varepsilon rankings unstable. Second, current ensembles provide no evidence of optimality, leaving an unknown gap to worst-case performance. Third, fixed attack configurations provide no systematic control over the trade-off between attack strength and evaluation cost. To address these limitations, we introduce a unified evaluation framework based on a comprehensive pool of minimum-norm attacks and robustness–perturbation curves across \ell_0 , \ell_1 , \ell_2 and \ell_\infty norms. We define the attack frontier as the worst-case robustness estimate the attack pool produces against a model. We then formalize evaluation as a frontier-approximation problem, constructing minimum-norm attack ensembles, optimized subsets of the comprehensive pool, that approach the frontier under a controllable query budget, with larger budgets monotonically tightening the estimate. Furthermore, we define the defense frontier as the maximum robustness across the model set at each perturbation size. We finally propose the Defense Optimality Index to rank defenses by their gap to the defense frontier, providing a ranking without selecting a reference \varepsilon . On CIFAR-10 and ImageNet, our ensembles match or exceed AutoAttack on most defenses at every budget tier, at fixed and controllable query cost, offering practitioners a query-controlled, curve-based alternative to fixed- \varepsilon evaluation.
[LG-27] Asymptotically Optimal Regret for Reinforcement Learning without Horizon Dependence
链接: https://arxiv.org/abs/2607.19854
作者: Runlong Zhou,Zihan Zhang,Maryam Fazel,Simon S. Du
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: 78 pages
Abstract:We study horizon-free regret minimization for finite-horizon time-homogeneous tabular Markov decision processes with S states, A actions, horizon H , and per-trajectory total reward bounded by 1 . We propose a new algorithm and prove a regret upper bound [\tilde O(\sqrtSAK+S^8A^3)] with failure probability \delta , where K is the number of episodes and \tilde O(\cdot) hides \mathsfpoly\log(S,A,K,1/\delta) . Thus, the regret is H -free and asymptotically optimal, matching the contextual-bandit lower bound \Omega(\sqrtSAK) up to logarithmic factors. This completely removes the \log H dependence from the previous \tilde O(\sqrtSAK\log H+S^2A\log H) guarantee of Zhang et al. (2021), and drastically improves the prior best horizon-free regret \tilde O(\sqrtS^9A^3K) of Zhang et al. (2022) asymptotically. The main technical difficulty is that the optimal value functions \V_h^_h=1^H are time-inhomogeneous even though the transition kernel is time-homogeneous. A direct union bound over all value functions typically incurs an additional \min\log H,S\ factor. We avoid this factor by (i) exploiting the monotonicity of V_h^ in h and (ii) non-trivially projecting the value functions onto an S -dimensional grid. Our analysis relies on three additional ingredients. First, we introduce a horizon-truncation argument that enables reward-based exploration and removes the cost of a separate reward-free exploration phase. Second, we design a cutting bonus that preserves both optimism and the monotonicity needed for planning. Third, we prove a new bound on total deviation for time-homogeneous MDPs, which controls the clipped variance terms in the cutting bonus with adjustable polynomial dependence on S and without any dependence on H . Together, these tools yield an asymptotically optimal horizon-free regret guarantee. Comments: 78 pages Subjects: Machine Learning (cs.LG); Machine Learning (stat.ML) Cite as: arXiv:2607.19854 [cs.LG] (or arXiv:2607.19854v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.19854 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Runlong Zhou [view email] [v1] Wed, 22 Jul 2026 07:42:19 UTC (111 KB) Full-text links: Access Paper: View a PDF of the paper titled Asymptotically Optimal Regret for Reinforcement Learning without Horizon Dependence, by Runlong Zhou and 3 other authorsView PDFHTML (experimental)TeX Source view license Current browse context: cs.LG prev | next new | recent | 2026-07 Change to browse by: cs stat stat.ML 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-28] A Structure-Adaptive Random Feature Method for High-Dimensional Elliptic PDEs
链接: https://arxiv.org/abs/2607.19786
作者: Jiale Linghu,Hao Dong,Yangshuai Wang
类目: Numerical Analysis (math.NA); Machine Learning (cs.LG)
*备注:
Abstract:Random-feature methods reduce high-dimensional elliptic PDE collocation to linear coefficient problems, but full-dimensional trial spaces overlook lower-dimensional structure. We introduce the Hierarchical Analysis-of-Variance Random Feature Method (HA-RFM), which selects coordinate blocks using closed Sobol indices of the PDE residual, identifies oblique low-rank features from fitted-predictor gradients, and couples all retained features in one regularized least-squares solve. Under structural and stability hypotheses, we establish an L^2 error bound that links solution and residual truncation to finite-width approximation and regularized finite-sample fitting, and we derive guarantees for width and structure recovery. The resulting width is polynomial in the dimension at fixed interaction order, with dimension-independent higher-order contributions under uniform structural control. Residual screening achieves exact recovery of the prescribed three-pair support, while fitted-predictor gradients recover oblique directions through dimension 50 . In random-ridge tests, less than 1% additional width reduces errors by factors of 14 - 39 over coordinate blocks and 34 - 100 over equal-width full-dimensional RFM. Semilinear computations extend HA-RFM through dimension 100 , while dense and distributed interactions delineate the coordinate families required for broader structure.
[LG-29] AlphaRoute: Large Language Models as Semantic Optimizers for Multi-Objective Routing
链接: https://arxiv.org/abs/2607.19768
作者: Kabir Murjani,Mishri Bhavsar,Manish I. Patel,Jonti Talukdar
类目: Machine Learning (cs.LG); Hardware Architecture (cs.AR)
*备注: 7 pages, 5 figures. Accepted for publication in the IEEE International Conference on LLM-Aided Design, 2026, Stanford University, Stanford, CA, USA. Code available at this https URL
Abstract:Very Large Scale Integration (VLSI) global routing is an NP-hard combinatorial optimization problem requiring signal net assignment across capacity-constrained 3D grids while minimizing congestion, wirelength, and via transitions. Because traditional heuristics rely on static penalty schedules that fail on complex congestion topologies, we present AlphaRoute: a multi-objective adaptive search framework reformulating rip-up and reroute (RR) into a dynamic optimization system. We introduce SHAP-based overflow decomposition to isolate per-net congestion, driving targeted subgraph extraction via 3D Dijkstra maze routing and an adaptive PathFinder policy. Crucially, AlphaRoute employs Large Language Models (LLMs) as semantic policy optimizers. Bounded by a deterministic knowledge graph, the LLMs interpret congestion metrics to dynamically adjust penalty parameters. Evaluated on ISPD 2025 benchmarks, AlphaRoute reduces overflow by 98.6% on MEMPOOL. On the constrained ARIANE design, we achieve an overflow of 146,109 (a 29.8x reduction in overflow over the state of the art), yielding a penalized score of S_orig = 0.0538 versus the State-of-the-art (SOTA) 1.780. These results demonstrate that superior algorithmic search geometry can overcome the latency of interpreted Python implementations.
[LG-30] Koopman Dreamer: Spectrally Constrained Latent Dynamics for Stable World-Model Imagination
链接: https://arxiv.org/abs/2607.19719
作者: Jiaqi Li,Xinglong Zhang,Haibin Xie,Yixing Lan,Wei Pan,Xin Xu
类目: Machine Learning (cs.LG); Robotics (cs.RO)
*备注: 20 pages, 13 figures, 11 tables
Abstract:Latent world models improve sample efficiency in continuous control by optimizing policies over imagined latent trajectories, but common neural transitions offer limited direct control over modal persistence and error accumulation in long rollouts. We propose Koopman Dreamer, a Dreamer-style world model with a spectrally constrained deterministic latent dynamics core. Its Koopman-inspired backbone uses two-dimensional rotation–scaling blocks with bounded radii to represent damping, rotation, and near-periodic modes. Linear and low-rank bilinear action terms capture global and state-dependent control effects, while stochastic-state modulation supplies local correction information. To reduce the mismatch between posterior-conditioned training and prior-only imagination, the model combines posterior-conditioned EMA teacher targets with one-step consistency, multi-step rollout, and open-loop observation-prediction objectives. We further derive a multi-step rollout-error bound that separates amplification by the spectral backbone and bilinear interaction from the additive effects of stochastic-state mismatch and modeling residuals, clarifying the trade-off between error attenuation and long-term information retention. Experimental results on proprioceptive continuous-control tasks from the DeepMind Control Suite and UAV-LiDAR autonomous navigation demonstrate that Koopman Dreamer improves the stability of long-horizon latent rollouts and achieves stronger closed-loop control performance on tasks that rely on high-quality multi-step imagination.
[LG-31] How Fast Can Reward Models Score? A Systems Study of C and PyTorch Inference Runtimes for RLHF
链接: https://arxiv.org/abs/2607.19712
作者: Venkata Naga Sai Vishnu Rohit Pulipaka,Anish Katta,Deva Rohit Reddy Peddireddy
类目: Machine Learning (cs.LG)
*备注:
Abstract:In RLHF pipelines, reward scoring blocks policy updates. Slow scoring bottlenecks the entire loop, since no update runs until every rollout gets a score. And yet most setups just default to PyTorch eager mode or this http URL, no one checks if that’s actually fastest. Scoring itself is small. Rollout generation eats far more of a typical RLHF step. But scoring and generation fight over the same CPU and GPU resources, so a faster scoring engine doesn’t shrink step time on its own. It mainly frees up capacity generation can use instead. We built a native C++ inference engine on ONNX Runtime. First step: confirm correctness. Output matched the PyTorch reference to 5.7 x 10^-6 on CPU and 4.2 x 10^-3 on GPU, close enough to trust. Then we tested it against PyTorch eager mode, this http URL, and FastAPI, on both CPU and GPU. CPU was decisive. Our engine beat every baseline, confidence intervals didn’t even overlap. GPU gave a different view: we beat PyTorch and FastAPI, but this http URL came out ahead. Further testing traced the speedup to ONNX Runtime itself, not C++ as a language. And batching strategy mattered more than either the language or the runtime choice, more than we expected. The results are from repeated, independent runs, since single runs just aren’t reliable enough to trust.
[LG-32] Efficient Clustering with Provable Guardrails for LLM Inference at Scale ICML
链接: https://arxiv.org/abs/2607.19704
作者: Longshaokan Wang,Wai Tsang Keung,Punit Ghodasara,Roman Wang,Ali Dashti,Francesc Moreno-Noguer
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: Accepted for presentation at ICML HiLD workshop 2026 (non-archival)
Abstract:Scaling LLM-based applications to millions of users is bottlenecked by the inference cost and latency of modern foundation models. A natural fix is to cluster the inputs and call the LLM only on cluster representatives, letting other members inherit the output – but this is only safe if each member is measurably close to its representative. Existing clustering methods do not offer such per-sample quality control at scale: none jointly guarantee a minimal within-cluster similarity, exact matching of categorical attributes, and scalability to tens of millions of samples. We propose a two-stage algorithm that generates initial clusters with Mini-batch K-Means, then greedily selects representatives within each initial cluster – a step equivalent to the Johnson-Chvatal heuristic for Set Cover over alpha-balls in embedding space. The algorithm enforces the similarity and attribute guardrails exactly by construction, and runs in O(nd + n^2 d/K) time and O(nd + n^2/K^2) memory for n samples, feature dimension d , and K initial clusters – linear in n when K grows proportionally with n . We provide benchmarks against common clustering methods on internal and public datasets: our method not only delivers per-sample guardrails but also runs 10-1000x faster and scales to data sizes where most standard methods become intractable. Deployed on 38 million customers for a persona-based recommender, the clustering method cut downstream cost and latency by 50-fold while preserving personalization and unblocked the production launch.
[LG-33] Leverag ing ECRAM for Edge Continual Learning
链接: https://arxiv.org/abs/2607.19661
作者: Nabila Tasnim,Haoran Liu,Qing Cao,Saugata Ghose
类目: Hardware Architecture (cs.AR); Emerging Technologies (cs.ET); Machine Learning (cs.LG)
*备注: extended abstract of AICS 2025 poster
Abstract:Several edge computing platforms, such as autonomous vehicles and smart sensing devices, need to adapt to dynamic environments in real time by learning from new data in the field. Continual learning has emerged as a promising solution for edge training, by incorporating techniques that successfully combine a highly summarized version of previously trained data (to avoid catastrophic forgetting) with recently sensed data. However, as is the case with other ML algorithms, continual learning generates significant data movement between general-purpose CPUs/GPUs and memory, impacting the suitability of continual learning for edge platforms. In-memory computing (IMC; also known as processing-using-memory) can curtail this waste and make continual learning feasible at the edge, but it faces two unique challenges: (1) IMC architectures make use of noisy computation operations that significantly harm training accuracy; and (2) IMC architectures have poor and often incomplete support for resource-efficient training. To address these challenges, we propose CLASP (the Continual Learning Acceleration System Platform), which to our knowledge is the first end-to-end system with IMC acceleration for continual learning. The hardware and software of CLASP are co-designed to support a wide range of continual learning algorithms, through software-visible assembly-level instructions that can be incorporated without constraints into ML-based algorithms. CLASP is designed around a back-end-of-line (BEOL) compatible ECRAM device that we fabricate, which can overcome the challenges of IMC-based training using other emerging memory devices. We show that CLASP with ECRAM approaches the accuracy of in-GPU training, while delivering a speedup of 67x and energy savings of 132x for learning without forgetting and experience replay using MNIST. Comments: extended abstract of AICS 2025 poster Subjects: Hardware Architecture (cs.AR); Emerging Technologies (cs.ET); Machine Learning (cs.LG) Cite as: arXiv:2607.19661 [cs.AR] (or arXiv:2607.19661v1 [cs.AR] for this version) https://doi.org/10.48550/arXiv.2607.19661 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-34] Expert-Guided Forecast Editing for Time-Series Foundation Models
链接: https://arxiv.org/abs/2607.19659
作者: Hung Le,Minh Hoang Nguyen,Manh Nguyen,Huu Hiep Nguyen,Dai Do
类目: Machine Learning (cs.LG)
*备注: preprint 34 pages
Abstract:Time-series foundation models can forecast across heterogeneous domains without task-specific training, but their forecasts are fixed once produced and cannot directly incorporate task-specific expert feedback. We study expert-guided forecast editing: a frozen foundation model generates candidate future trajectories, and an expensive expert evaluator scores them to guide forecast revision. Under a tight query budget, two natural strategies sit at opposite ends: best-of- N purely exploits the foundation model’s predictive distribution, while optimization approaches mostly explore the forecast horizon as an unstructured high-dimensional vector. Each extreme is individually sub-optimal. We introduce \textbfDEFT, an expert-guided forecast editing framework that balances the two by first exploiting the foundation model’s predictive samples in a decomposed trend–seasonal space, then exploring around them via component-wise refinement. DEFT queries the expert only on complete trajectories, then reuses scores for the trend and seasonal components that appeared in the queried recombinations. This lets each expert query provide structured component-level feedback while keeping the foundation model frozen. We compare DEFT against direct search approaches, including best-of- N , cross-entropy methods, and Bayesian optimization, under matched expert-query budgets. Across two forecasting benchmarks consisting of 78 datasets, three time-series foundation models, four feedback types, and seven query budgets, DEFT consistently improves the effectiveness of expert guidance. A molecular-dynamics case study further suggests that the same principle extends to more physically grounded feedback, supporting the hypothesis that sparse test-time guidance should be spent balancing prior exploitation with structured exploration.
[LG-35] HypEMBER: Hypernetwork-based Ensemble for Robust Policy Learning of Parametrized Dynamical Systems
链接: https://arxiv.org/abs/2607.19628
作者: Nicolò Botteghi,Gabriele Pascali,Urban Fasel,Andrea Manzoni
类目: Machine Learning (cs.LG)
*备注:
Abstract:In this work we investigate reinforcement learning (RL) as a framework for the robust control of parametrized dynamical systems in presence of measurements and model uncertainties. High-dimensional state spaces, expensive numerical solvers, the partial knowledge of the governing equations, and the dependence on physical parameters that may be uncertain or difficult to estimate accurately, make the use of standard RL approaches computationally unfeasible. Indeed, lack of robustness and poor generalization across parameter variations are further amplified in presence of noisy or incomplete measurements, ultimately hampering control performance. To address these challenges, we introduce HypEMBER, a novel RL framework based on the combination of hypernetworks and ensemble learning. In the proposed approach, both the policy and value functions are represented through hypernetworks that generate the weights of the underlying models conditioned on the physical parameters of the system, thereby enabling parametric generalization across different dynamical regimes. In addition, an ensemble of policy and value approximators is employed to quantify epistemic uncertainty, leading to improved exploration strategies and enhanced robustness during and after training. The performance of the proposed framework is assessed on two representative parametrized control problems: (i) the one-dimensional Kuramoto-Sivashinsky equation and (ii) a particle-navigation task in a two-dimensional time-dependent gyre flow, focusing on robustness with respect to measurement noise and parameter misspecification. Numerical results demonstrate that HypEMBER consistently improves training stability and sample efficiency, while achieving superior robustness to uncertainties affecting both the system dynamics and the available observations, in comparison with state-of-the-art RL methods.
[LG-36] From Bit-Position Sensitivity to Unequal Error Protection for DNN Inference Memory MICRO2026 MICRO
链接: https://arxiv.org/abs/2607.19623
作者: Muhammad Husnain Mubarik,Karthik Mohan Kumar,Pedro Antonio Pena,Keshavan Varadarajan,Kunal Tyagi
类目: Hardware Architecture (cs.AR); Machine Learning (cs.LG)
*备注: Accepted to appear at the 59th IEEE/ACM International Symposium on Microarchitecture (MICRO 2026). 14 pages
Abstract:We characterize per-bit-position fault sensitivity in ML inference across 16 workloads – spanning transformer-based models and attention-free CNNs – and across three floating-point formats. Our central empirical finding is a sharp bit-sensitivity transition: flipping any of the least-significant fraction bits up to a data-type-specific threshold, Xsafe, degrades task metrics by less than 1% under deterministic single-bit stress tests. Sensitivity rises through the upper fraction bits and spikes at the exponent-mantissa boundary, where a single-bit flip causes catastrophic collapse. Because low-order bits are largely inconsequential while high-order and exponent bits are critical, uniform SECDED protection – which guards every bit equally at 12.5% storage overhead – is unnecessarily conservative. We derive per-data-type Xsafe floors (FP16: 6, BF16: 4, FP32: 15) and workload-aware tiers that widen the unprotected region for resilient model classes, raising ECC savings to 37.5-62.5% without retraining. Text-conditioned diffusion models dictate the conservative floor; vision encoders, NLU models, and resilient LLMs tolerate wider bypass regions. These floors and tiers drive an Unequal Error Protection (UEP) codec with per-cacheline data-type tags and a dual-partition SRAM architecture for ML accelerators. Validation across 870+ fault-injection runs confirms selective protection holds under contiguous 2- and 3-bit upsets. The codec reduces ECC area by 27.8% relative to uniform SECDED; dual-voltage operation of the non-critical partition lowers gross BF16 read energy by about 17%, with a roughly 4% dual-partition macro-area overhead.
[LG-37] he Mechanism Matters: When Knowledge Graphs Help Reinforcement Learning
链接: https://arxiv.org/abs/2607.19616
作者: Mohammed Sameer Syed
类目: Machine Learning (cs.LG)
*备注:
Abstract:Knowledge graphs (KGs) are widely used to inject prior knowledge into reinforcement learning (RL), yet the literature is dominated by single-domain, positive-result method papers, so we lack a systematic account of when KG structure helps an agent, when it is neutral, and when it hurts. We conduct a controlled study that independently varies the RL task, the injection mechanism (state features, action masking, or potential-based reward shaping), and KG quality. Using a synthetic, fully controllable KG over MiniGrid environments, we report three findings. First, on compositional sparse-reward tasks structured KG guidance improves sample efficiency and solve reliability (70% to 97% of seeds), and a shuffle control that permutes the KG’s edges while preserving their count collapses the benefit toward baseline (masking p=0.0001; shaping p=0.006), so the gain is structural rather than generic regularization. Second, KG value scales with the amount of task-relevant knowledge the graph contains. Third, and most consequential, safety depends on the mechanism: soft, optimality-preserving injection benefits from correct knowledge and harmlessly ignores incorrect knowledge, whereas hard masking is brittle, forbidding essential actions when the KG is incomplete or corrupted and making a wrong KG worse than none. A UMLS-derived clinical case study on sepsis management under offline RL is a careful null, underscoring that benefits require task structure the chosen mechanism can exploit. Our results give practitioners concrete guidance on how, and how much, to trust a KG when using it to guide RL.
[LG-38] CRB-Driven Beamforming and Trajectory Optimization for UAV-assisted ISAC System
链接: https://arxiv.org/abs/2607.19609
作者: Yi Yang,Qianqian Zhang,Huaxia Wang
类目: Information Theory (cs.IT); Machine Learning (cs.LG); Robotics (cs.RO)
*备注: 7 pages, accepted by 2026 IEEE 104th Vehicular Technology Conference (VTC2026-Fall)
Abstract:In this paper, we study an unmanned aerial vehicle (UAV)-assisted integrated sensing and communication (ISAC) system, where a UAV enhances the sensing capability of a base station (BS) towards a target while ensuring reliable communication towards a downlink user. This architecture is practically attractive for future wireless networks due to the UAV’s controllable mobility and adaptive sensing coverage in wireless environments. The sensing performance is characterized by the average Cramér-Rao bound (CRB), which quantifies the minimum variance of the unbiased angle-of-arrival estimation. To enhance the sensing performance, the UAV trajectory and beamforming parameters are jointly optimized under power and mobility constraints, while satisfying communication requirements to the downlink user. To address the resulting non-convex problem, we employ null-space projection for beamforming design and adopt deep reinforcement learning for the trajectory optimization over a discrete-time scale. In each time slot, beamforming is optimized based on the channel state information to improve CRB performance while mitigating interference between the BS and the communication user. Simulation results demonstrate that the proposed method significantly reduces the time-averaged CRB by over 10%, compared with the ISAC system without UAV assistance, and also achieves a higher sensing accuracy than both the fixed-UAV-trajectory and the maximum-ratio-transmission-based beamforming benchmarks.
[LG-39] End-to-End Differential Privacy in Training Deep Neural Network Classifiers
链接: https://arxiv.org/abs/2607.19580
作者: Huaiyuan Rao,Calvin Hawkins,Alexander Benvenuti,Matthew Hale
类目: Machine Learning (cs.LG); Cryptography and Security (cs.CR)
*备注: 32 pages, 10 tables, 3 figures
Abstract:Differentially private machine learning enables model training on sensitive data while ensuring that individual data is unlikely to be recoverable from the parameters of the resulting model. However, existing work often privatizes both training inputs and their labels, and these protections may be conservative when labels are public or can be safely made public. Therefore, in this work we propose a novel private training framework that instead privatizes training inputs while keeping labels public. We consider neural networks with softmax output layers, and thus the mapping from training inputs to the output of the softmax layer is a mapping onto the unit simplex. We randomize softmax outputs during training by applying the Dirichlet mechanism to enforce differential privacy for the training inputs, hence the ``end-to-end’’ label. Because training data is reused across multiple training epochs, we use the notion of \Renyi differential privacy to formulate tight bounds on the strength of privacy provided by the Dirichlet mechanism across repeated uses. We show empirically that we attain new state-of-the-art accuracy when training from scratch on CIFAR10, MNIST, MedMNIST, FashionMNIST, and SVHN across all privacy budgets evaluated. Notably, when implementing (\epsilon, \delta) -differential privacy with \delta=10^-5 , we improve the prior state-of-the-art accuracy from 78.37% to 88.17% at \epsilon=4 on CIFAR10, and our approach has 82.96% accuracy even for \epsilon=1 , which significantly outperforms prior work.
[LG-40] Agent -Centric Animal Pose Forecasting
链接: https://arxiv.org/abs/2607.19548
作者: Eyrun Eyjolfsdottir,Kristin Branson
类目: Machine Learning (cs.LG)
*备注:
Abstract:Understanding animal behavior at an algorithmic level – what animals attend to, how they form internal models and plans, and how this maps to action – remains a central challenge in neuroscience and ethology. Data-driven generative models offer a path toward this understanding. We introduce a framework for training agent-centric autoregressive models of animal behavior from tracked pose, applicable to single animals and to groups in which each agent senses and responds to its conspecifics. Our models input egocentric sensory observations and output egocentric movements, mirroring the biological constraint that animals observe and act on the world from their own reference frame. Social behavior emerges from agents independently sensing and responding to one another. This agent-centric formulation requires managing many parallel representations of the same data, along with ML-specific transformations like discretization. We release a general-purpose library focused on the composable sequences of operations that translate between these representations. We show that trained models capture the distribution of social behavior in groups of courting Drosophila, and our library includes quantitative tools for measuring fit. We demonstrate how the library supports systematic comparison across input and output representations and that it adapts straightforwardly to a new domain.
[LG-41] he C-index illusion: discrimination without calibration in published survival models
链接: https://arxiv.org/abs/2607.19526
作者: Rafael da Silva,Danilo Alvares
类目: Machine Learning (cs.LG)
*备注: 34 pages, 6 figures, 18 tables
Abstract:“Stop Chasing the C-index when Evaluating Survival Analysis Models” (ICML 2026, Spotlight) argued normatively, on synthetic data, that evaluating survival models by discrimination alone, i.e. the concordance index, produces systematically misleading model comparisons, because the metric ignores calibration and time-dependent accuracy. Whether this matters for real, published, non-clinical models has not been tested. We reproduce three published survival-ML models across three structurally distinct domains (hard-drive failure, peer-to-peer credit default, and user disengagement on digital platforms), validate our evaluation instrument against the anchor paper’s own synthetic experiment, and test five pre-registered hypotheses under a Holm-corrected family-wise error rate. Three of five reject. A model that reproduces the published discrimination almost exactly (C = 0.9595 vs. 0.958 reported) fails a formal calibration test at p = 2.6e-136; a broad feature-ablation search finds no single attribute responsible for this discrimination, so the calibration failure is not an artifact of a trivial shortcut. A lender’s estimated default risk is biased upward by roughly two percentage points, growing to nearly four points in the riskiest segment, when loan prepayment is treated as non-informative censoring rather than as a competing risk. A platform’s churn model shows probability estimates that degrade with the prediction horizon even as its global discrimination stays within the pre-registered C-index band. A direct test of whether metric choice inverts which model is preferred does not reject, though with limited power given only two to three models per domain; the failure mode we document is better characterized as misplaced confidence in a chosen model than as choice of the wrong one. We release a reusable, pre-registered evaluation harness with full code and a single annotated notebook.
[LG-42] Do Sheaf Neural Networks Use Holonomy? A Measure–Intervene–Control Study
链接: https://arxiv.org/abs/2607.19514
作者: Ankit Grover,Rémi Bourgerie
类目: Machine Learning (cs.LG)
*备注:
Abstract:Geometric architectures are often justified by internal mechanisms such as rotations, yet task performance alone cannot show whether those mechanisms drive predictions. Using sheaf neural networks (SNNs) as a testbed, we introduce the first basis-independent measurement of trained triangle-loop products, separating rotation, stalk-space area, and orientation. In a custom high-homophily GraphUniverse regime, Neural Sheaf Propagation (NSP) increases the triangle-weighted mean two-dimensional SO(2) loop rotation from 0.010 to 0.388 radians for triangle counting, while the community-detection comparison ends at 0.029 radians. Across the training-set-size experiment, replacing all learned SO(2) transports with identities sharply increases test error, establishing post-training sensitivity to the complete learned connection. However, a graph-summary ridge predictor is more accurate, diagonal maps also improve, and fixed-degree graphs develop increasing rotation without outperforming the training-mean predictor. This measure-intervene-control study separates geometric change, connection sensitivity, and evidence for triangle-specific computation.
[LG-43] otal Variation Distance Estimation in Autoregressive Models
链接: https://arxiv.org/abs/2607.19510
作者: Eric Price,Kevin Tian,Zhiyang Xun,Yusong Zhu
类目: Machine Learning (cs.LG); Data Structures and Algorithms (cs.DS); Methodology (stat.ME); Machine Learning (stat.ML)
*备注: 39 pages, 11 figures, code is available at this https URL
Abstract:Modern LLM deployments use a number of implementation choices and inference optimizations (e.g., batching, custom kernels, and quantization) on top of fixed weights, so two engines serving “the same model” can produce meaningfully different distributions. We study the problem of estimating the total variation (TV) distance between two length- n autoregressive distributions to additive error \varepsilon , under three access models. (1) Under sample access, we use \widetildeO(n^2 K/\varepsilon^2) queries, where K is the maximum support of the next-token distribution. This improves upon the \widetildeO(n^3 m/\varepsilon^5) -query estimator of Meel et al. (2025), where m \geq K is the total size of the token alphabet. (2) Under logit access, we use O(n/\varepsilon^2) queries, and this is tight. (3) Under noisy logit access, we smoothly interpolate between the above two guarantees: if probability values are given to relative error \sigma , we use \widetildeO((n+n^2\sigma^2)/\varepsilon^2) queries. We complement our theoretical results with an empirical evaluation of our algorithms, for example measuring the distance between SGLang and vLLM serving identical weights. Our experiments highlight the robustness and practicality of estimating the total variation distance, which remains estimable where the KL divergence is infinite. Our code is available at this https URL.
[LG-44] Generating Bearing Vibration Signals at User-Specified Fault Probabilities Using PR-GAN and Counterfactual Methods
链接: https://arxiv.org/abs/2607.19455
作者: Seyed Mohammadreza Alavi,Ardeshir Shojaeinasab,Reza Jalayer,Masoud Jalayer,Behnam Bahrak
类目: Machine Learning (cs.LG)
*备注:
Abstract:In bearing vibration datasets, most samples receive predicted fault probabilities close to 0 or 1, while samples with intermediate (gray-zone) probabilities are rare. Such borderline samples are important because they reflect conditions in which maintenance decisions may require additional inspection or a conservative response and are useful for studying decision boundaries. To address this scarcity, this paper proposes and compares two approaches that generate vibration signals whose predicted fault probability matches a target probability of 0.25, 0.50, or 0.75. We use the average output of a heterogeneous ensemble classifier with different architectures and random initializations as a fixed, gradient-accessible probability oracle. The first, training-based approach, Probability-Regularized Generative Adversarial Network (PR-GAN), extends Wasserstein Generative Adversarial Network with Gradient Penalty (WGAN-GP) and edits a real signal through a residual generator while pushing the classifier output toward the target probability. The second is a training-free, per-sample Wachter-style counterfactual (CF) procedure that directly optimizes each input signal to reach the target probability while remaining close to the source signal. We evaluate both methods on the Case Western Reserve University (CWRU) and Paderborn bearing datasets using mean absolute target-probability error, time-domain total variation, and frequency-domain log power spectral density (log-PSD) differences. Across all settings, CF reaches the target with a mean absolute probability error of 0.005-0.008 and a within-tolerance success rate of 1.000 on retained samples, whereas PR-GAN’s mean error is 0.046-0.059 with success rates between 0.501 and 0.680. CF therefore steers the probability more reliably and requires smaller average L1 changes, whereas PR-GAN has a lower reported runtime in most settings.
[LG-45] Marine Engine Fault Dataset: Open-Access Data under Controlled Reference and Fault Scenario Conditions
链接: https://arxiv.org/abs/2607.19444
作者: Ahmad BahooToroody,Oleksiy Bondarenko,Mohammad Mahdi Abaei,Niki Yoichi,Enrico Zio
类目: Machine Learning (cs.LG); Signal Processing (eess.SP); Systems and Control (eess.SY)
*备注: submitted to a journal and currently is under review, 34 pages, 13 Figures
Abstract:Open-access datasets for marine-engine predictive maintenance remain scarce, particularly those from controlled fault experiments with documented operating conditions, subsystem-level interventions and system-level measurements. This work presents the Marine Engine Fault Dataset, an openly available dataset from a turbocharged, intercooled three-cylinder marine diesel engine operated on a testbed under both reference and fault-scenario conditions. The experimental campaign combined a reference-performance program across the 30-90% load range with scenario-based tests in which abnormal conditions were introduced after stabilized fault-free operation, enabling controlled comparison between baseline and fault-affected behaviour. Five anomaly classes were implemented through physical interventions affecting major engine subsystems: cooling-water pump cavitation, compressor air-filter clogging, air-cooler fouling, injection-valve nozzle clogging and turbine degradation induced through increased exhaust-side restriction. The released data comprise multi-sensor time-series of operating, thermal, pressure, flow and combustion-related variables, with a separate reference-performance record and metadata for structured reuse. Technical validation shows that the reference measurements remain physically coherent across the operating range and that the imposed anomalies produce interpretable response patterns consistent with the affected subsystems, including progressively distinguishable behaviour where different severities were implemented. By combining controlled fault realization, multi-load operation and system-level measurements within a real marine-engine platform, the dataset provides a well-documented benchmark for anomaly detection, fault diagnosis, degradation modelling and related condition-monitoring studies in maritime machinery.
[LG-46] Adaptive Multi-Expert Graph Transformer for Interpretable EEG-Based Diagnostics
链接: https://arxiv.org/abs/2607.19429
作者: Maryam Rahimimovassagh,Md Elias Hossain,Ivan Garibay,Niloofar Yousefi
类目: Machine Learning (cs.LG)
*备注:
Abstract:Electroencephalographic (EEG) abnormalities arise from dynamic changes in neural synchrony across spatial and temporal scales, yet many computational approaches reduce these dynamics to static features. We present a Spatial Multi-Expert Graph Transformer that models each EEG recording as a sequence of dynamic functional connectivity graphs. Time-resolved connectivity is estimated using the weighted Phase Lag Index (wPLI), and hierarchical graph encoding aggregates information from electrode to regional and global levels. A multi-expert transformer architecture enables subtype-aware reasoning, with a gating mechanism adaptively fusing expert outputs for global abnormality prediction. Experiments on the TUAB dataset show competitive abnormal EEG detection performance and demonstrate the potential of dynamic graph modeling with adaptive expert fusion for interpretable, subtype-aware spatial–temporal analysis.
[LG-47] Reward-Aware Population Scaling of Evolutionary Strategies in LLM Fine-Tuning ICML2026
链接: https://arxiv.org/abs/2607.19408
作者: Sung Cho,Gyubin Han
类目: Machine Learning (cs.LG)
*备注: 15 pages, 2 figures. Accepted at the 4th HiLD (High-dimensional Learning Dynamics) Workshop, ICML 2026
Abstract:Using Evolutionary Strategies (ES) for fine-tuning large language models is attractive because it is memory-efficient, parallel, and compatible with black-box or discrete rewards. Yet its population-size conclusions conflict sharply: fine-tuning with cross-entropy (CE) reward succeeds with N=1 , while binary-reward training often needs N \approx 30 . We show this gap is largely about reward design and normalization, not population size. In the capable-model regime we study, z-score advantage normalization can cause N=2 to fail. Disabling normalization lets binary-reward ES with N=2 improve on GSM8K and TREC across capable models spanning 0.5B-7B, where the normalized variant collapses or degrades. This small- N risk is set by reward granularity: binary accuracy reward induces a zero-advantage probability q that depends in closed form on base accuracy, batch size, and intra-pair correctness correlation; a zero-training probe on Qwen2.5-Instruct/GSM8K matches the formula with mean absolute error 0.020 across 12 configurations and finds the availability threshold N_\mathrmavail to be small in this capable-model regime. The implication is not that N=2 is universally sufficient, but that small-population failure in capable-model binary ES can be an implementation artifact rather than an intrinsic population limit.
[LG-48] NMR Elucidation as an Agent ic Search Problem Not a Modeling Problem
链接: https://arxiv.org/abs/2607.19406
作者: Irina Espejo Morales,Damon Hinz,Marvin Alberts,Geraud Krawezik,Haewon Jeong,Shirley Ho
类目: Machine Learning (cs.LG)
*备注:
Abstract:Structural elucidation from Nuclear Magnetic Resonance (NMR) data remains a fundamental bottleneck across chemistry, materials science, and biology. We demonstrate that an agentic AI system can perform this task at a level comparable to graduate-level chemistry students. Instead of training a model to directly map spectra to structures, we build a single autonomous agent, backed by a frozen LLM, that interacts with a curated environment with access to domain-specific processing tools, validation checks, tabulated chemical shifts, and instructions that outline the stepwise nature of a chemist’s thinking process. On the Alberts dataset, our agent elucidates structures with a top-1 accuracy of 71%, comparable to the performance of graduate students at 66% top-1 accuracy. On the van Bramer and AstraZeneca datasets, our agent achieved 80% and 20% top-1 accuracy respectively, outperforming zero-shot end-to-end deep learning models which were trained on large datasets of simulated spectra. These results show that reframing NMR elucidation as an LLM-guided constrained search, rather than a modeling task, yields substantial gains and suggests a path toward multi-step orchestration frameworks that integrate a variety of tools, models, and domain knowledge to assist in automating spectroscopic analysis.
[LG-49] Reproducing Recurrent Transformers: The CoTFormer
链接: https://arxiv.org/abs/2607.19405
作者: Aras Kavuncu,Bryan Vullo,Alberto Berni
类目: Machine Learning (cs.LG)
*备注: 7 pages, 3 figures, reproducibility report submitted to University of Southampton
Abstract:The CoTFormer architecture formalizes Chain-of-Thought as a form of recurrent latent computation, preserving intermediate states as attendable representations to mimic explicit reasoning traces. In this work, we evaluate CoTFormer and its structural variants across perplexity and compute efficiency metrics. Furthermore, we extend evaluation to controlled algorithmic settings to determine whether this recurrent framework improves out-of-distribution generalisation on inductive reasoning tasks.
[LG-50] Predictive single cell foundation model for gene regulation and aging with privacy-preserving tabular learning
链接: https://arxiv.org/abs/2607.19400
作者: Jiayuan Ding,Jianhui Lin,Ziyang Miao,Nils Mechtel,Shiyu Jiang,Yixin Wang,Zhaoyu Fang,Jorge D. Martin-Rufino,Chen Weng,Reuben Saunders,Weize Xu,Jonathan S. Weissman,Min Li,Jiliang Tang,Wei Ouyang,Yuancheng Ryan Lu,Xiaojie Qiu
类目: Machine Learning (cs.LG); Genomics (q-bio.GN)
*备注: 98 pages, 4 main figures, and 15 supplementary figures
Abstract:Pre-trained foundation models (FMs) have begun transforming single-cell genomics, but scaling them raises privacy concerns. Moreover, unlike text data, single-cell data is unordered and exhibits a unique tabular structure that current single-cell FMs overlook. We introduce Tabula, a privacy-preserving FM designed with federated learning (FL) that explicitly models the tabular structure of single-cell data. To deploy Tabula, we further developed Chiron, a decentralized AI agent-enabled platform for collaborative training across institutions without sharing raw data. Beyond strong performance across downstream benchmarks, Tabula reveals combinatorial regulatory logic across diverse biological systems, including hematopoiesis, pancreatic endogenesis, neurogenesis, and cardiogenesis. Using a new scRNA-seq dataset of paired young and aged human fibroblasts, Tabula nominates rejuvenation factors through age- and identity score-guided in silico prioritization, outperforming conventional approaches. Thus, Tabula represents an important advance in single-cell foundation modeling by integrating tabular learning with FL, paving the way toward privacy-preserving virtual cells for human health.
[LG-51] Memory Merge DQN: Sensitivity Weighted Target Updates for Stable Value Learning
链接: https://arxiv.org/abs/2607.19397
作者: Adrian Ly,Richard Dazeley,Peter Vamplew,Sunil Aryal,Francisco Cruz
类目: Machine Learning (cs.LG)
*备注:
Abstract:Deep Q-networks use target networks to stabilise bootstrapped value learning, but the standard hard copy update also introduces a tradeoff. Holding the target network fixed, improves short term stability, yet each hard update abruptly replaces the target parameters with the newest online network and discards recent parameter history. This can produce sudden changes in the bootstrap target and may remove value function structure that remains useful later in training. This paper introduces Memory Merge DQN, a target network update mechanism that maintains a short memory of recent historical online network copies and constructs the target network by merging network parameters based on the Q-value sensitivity rather than copying only the newest online network. Memory Merge gives greater influence to parameters that remain locally important for current Q-value behaviour, while using a recency prior to keep the merged target close to the latest online parameters. The method is inspired by Fisher Weight Model Merging, but uses Q-value sensitivity rather than Fisher information as the weighting signal. This paper evaluates Memory Merge DQN on Atari environments against DQN, Averaged DQN, DQN with layer normalisation, and PQN (with gradient clipping). The results show that Memory Merge DQN is highly competitive and it achieves the largest number of first place final performance results among the evaluated methods, beats DQN, Averaged DQN, and PQN (with gradient clipping), and produces substantial gains in several games where preserving useful value-function parameters appears beneficial. These findings suggest that selectively merging recent parameter weights and history can improve the stability and final performance of DQN agents, and that target network design is an important mechanism for preserving useful value function structure during long horizon value learning.
[LG-52] Predicting Groundwater Arsenic Concentrations Using Graph Neural Networks AAAI’26
链接: https://arxiv.org/abs/2607.19392
作者: William Xing,Stephanie Yang,Aarush Bandemegal,Anushree Misra,Ananya Kalapatapu,Brennan Lagasse,Kevin Zhu
类目: Machine Learning (cs.LG)
*备注: 7 pages, 5 figures, 1 table, AAAI '26 AI4ES workshop acceptance
Abstract:Arsenic contamination in groundwater presents a longstanding public health crisis in the United States, especially for households depending on private wells. Accurate and spatially informed prediction of arsenic concentration is vital to identify high-risk areas and focus mitigation efforts. However, there is a lack of generalizable models for representing continuous variation in arsenic concentrations across regions. In this work, we pose arsenic prediction as a regression task and construct a spatially integrated dataset to aggregate over 74,000 arsenic samples from the Water Quality Portal (WQP), Mineral Resources Data System (MRDS), and Gridded National Soil Survey Geographic Database (gNATSGO). Specifically, we use a variety of techniques including kNearest Neighbors (k-NN) and Geographic Information Systems (GIS) to join arsenic measurement points from across the United States by location. Building on this dataset, we evaluate a diverse suite of machine learning models, including tree-based ensemble approaches, multilayer perceptrons, and spatially aware graph neural networks (GNN). Our findings show that while gradient-boosted trees are still considered state-of-the-art in the field of tabular data, GNNs are able to further account for spatial dependence to match or outperform the results of gradient-boosted trees. These results demonstrate that graph-based and spatially informed learning can enhance environmental prediction and provide a foundation for improved groundwater risk mapping and monitoring.
[LG-53] he Orthogonalized Read Is a Removable Training Scaffold for Recurrent Memory
链接: https://arxiv.org/abs/2607.19390
作者: Keston Aquino-Michaels
类目: Machine Learning (cs.LG)
*备注: 18 pages, 8 figures. Code, per-seed results, and checkpoints: this https URL
Abstract:A recent report finds that orthogonalizing the mLSTM memory matrix at read time (five Newton-Schulz iterations, trained through) substantially improves noisy associative recall. The effect replicates, but it is not a memory improvement. Training on this task is a long chance plateau followed by a sharp escape, and the orthogonalized read acts by re-conditioning the learning problem during the plateau. Three properties establish this. It must be self-consistent: an exact recursive least-squares read (the Mesa layer) reproduces it, while straight-through halves, delta-rule writes, frozen random keys, and plain normalization all fail. It is uniform: across a learning-rate x hardness grid it multiplies the escape hazard roughly six-fold with no detectable hardness dependence, widening the workable learning-rate corridor that narrows for the baseline. And it is removable: applied to failed models at inference it rescues none, and annealed away on an escape-triggered schedule it leaves numerically stock mLSTMs at full accuracy. Much of the published gain needs no architecture at all – solved-rate at a fixed budget measures escape hazard, which follows a heat/noise law (learning-rate elasticity +3.0, gradient-noise elasticity -1.65) under which the original vocab-96 result is a large-batch noise condition rather than a capacity one. Decoding the memory state directly shows failed models carry roughly half their associations in linearly recoverable form: the plateau is a readout failure over half-written storage. Two conclusions travel beyond the intervention: recall benchmarks used for architecture selection partly measure trainability, and the system is a fully instrumented model organism of “emergence,” in which a sharp behavioral threshold demonstrably arises from a censored metric over gradually accumulating structure.
[LG-54] Neural Operator Surrogates for Two-Dimensional Neutron Flux Estimation
链接: https://arxiv.org/abs/2607.19388
作者: Japan K. Patel,Barry D. Ganapol,Anthony Magliari,Matthew C. Schmidt,Todd A. Wareing
类目: Machine Learning (cs.LG)
*备注: to be submitted to ANS Winter 2026
Abstract:This work extends our one-dimensional single-sweep neural-operator studies to two dimensions. We consider one-group transport with isotropic scattering. As in the one-dimensional work, we use Fourier neural operators (FNOs) to approximate the high-fidelity scalar flux. Additionally, we also investigate U-shaped neural operators (UNOs) in this study. We consider three surrogates. The first two map the material and source fields directly to the flux, one using an FNO and one using a UNO. The third is an FNO that additionally takes the scalar flux after one source iteration, the single-sweep approximation, as an input. Each case is solved to high fidelity with a verified discrete-ordinates solver, and an average relative L_2 error norm is used to characterize the quality of the inferred maps. We train every surrogate over three random seeds so that differences between them can be assessed against run-to-run variability. Two questions guide the study: whether the single-sweep input improves accuracy over the direct maps, and whether training on the logarithm of the flux improves accuracy in the strongly attenuated regions relevant to shielding.
[LG-55] STN-TGAT: Top-K Portfolio Construction via Prior-Guided Graph Attention with Learnable Soft-Threshold Sparsification
链接: https://arxiv.org/abs/2607.19385
作者: Haoran Guo,Yutong Lu,Li Zhang
类目: Machine Learning (cs.LG)
*备注:
Abstract:This paper tackles the problem of stock ranking and portfolio construction under realistic investment settings by jointly modeling temporal dynamics and cross-sectional dependencies. We propose the Soft-Threshold NMI-prior Transformer Graph Attention Network (STN-TGAT), which integrates a temporal Transformer with a Graph Attention Network to capture long-horizon sequential patterns and dynamic inter-stock relationships. An NMI-based prior graph combined with a soft-threshold sparsification mechanism enhances structural robustness by mitigating noisy correlations while preserving informative connections. The portfolio formation process incorporates practical considerations, including Top-5 selection within the Top-50 S\P 500 constituents, explicit weight allocation, and transaction cost adjustment, thereby aligning the evaluation with real-world trading conditions. Empirical results on real-world data demonstrate that STN-TGAT consistently outperforms benchmark models from predictive accuracy and investment profitability measured by portfolio returns. These findings suggest that combining decision-aligned training with adaptive relational modeling provides a coherent and practically effective framework for data-driven portfolio construction.
[LG-56] SUM: Unified Geometric Surgery on Spatio-Temporal Adaptation Vectors for Federated Class Incremental Learning ECCV2026
链接: https://arxiv.org/abs/2607.19384
作者: Jaeik Kim,Jaeyoung Do
类目: Machine Learning (cs.LG)
*备注: Accepted to ECCV 2026
Abstract:Real-world intelligent systems often require both distributed collaboration across data-isolated clients and continual adaptation to evolving tasks. This setting naturally gives rise to Federated Class Incremental Learning (FCIL), which combines Federated Learning (FL) and Continual Learning (CL). However, their combination introduces two coupled sources of interference: spatial interference from heterogeneous clients and temporal interference from sequential tasks, jointly leading to Spatial-Temporal Catastrophic Forgetting (ST-CF). Existing approaches typically address spatial and temporal interference with separate mechanisms, often incurring additional client-side computation or communication, while leaving directional interactions among updates during aggregation unregulated. In this paper, we reinterpret FCIL as a unified multi-task learning problem, where both client and task updates are represented as adaptation vectors in a shared parameter space. Based on this view, we propose Unified Geometric Surgery on Spatio-Temporal Adaptation Vectors (SUM), a purely server-side framework that performs geometric surgery on adaptation vectors during aggregation. Spatial SUM mitigates client-level interference within each round, while causal online temporal SUM removes cross-task interference over time without additional client-side computation, communication, or memory beyond standard federated training. Empirically, SUM achieves up to 22% improvement over prior FCIL methods across diverse vision and language benchmarks while remaining robust to unreliable clients and maintaining computational efficiency.
[LG-57] Air Quality Arena: A Large-Scale Multi-Region Ground Monitoring Dataset and Benchmark for Air Quality Forecasting with Time-Series Foundation Models
链接: https://arxiv.org/abs/2607.19381
作者: Rishi Bharadwaj,Manik Gupta,Pandarasamy Arjunan
类目: Machine Learning (cs.LG)
*备注: 20 pages
Abstract:Air pollution causes an estimated 7.9 million premature deaths annually, making accurate forecasting a critical public health priority. Machine learning is increasingly being applied to forecast air pollution levels, yet existing benchmarks remain narrow in both geographic scope and pollutant coverage, and fail to evaluate the latest generation of time series foundation models (TSFMs) on real world, large scale data. We present Air Quality Arena (AQA), a large scale multi-country and multi-pollutant dataset (AQA-Data) and benchmark (AQA-Bench) to address this gap. AQA covers 6 major pollutants over a three year period across 7 diverse countries and 4 continents, with more than 14,000 station-pollutant series, aiming to provide a comprehensive benchmark for air quality tasks. We benchmark this dataset across 11 leading time series foundation models and classical baselines to assess performance on short-term air quality forecasting. Our results demonstrate that TSFMs are effective zero-shot forecasters and consistently outperform classical baselines, with our top-performing model employing a cross-modal architecture that leverages a vision foundation model for time series forecasting. AQA is publicly released at this http URL
[LG-58] CruiseBench: A Real-Flight-Aligned N-CMAPSS Benchmark for Engine RUL Prediction
链接: https://arxiv.org/abs/2607.19380
作者: Pu Cheng,Qiang Miao
类目: Machine Learning (cs.LG)
*备注: 20 pages, 7 figures
Abstract:Remaining useful life (RUL) prediction estimates how long an engine can continue safe operation and is central to maintenance planning. N-CMAPSS extends C-MAPSS by simulating run-to-failure aero-engine trajectories using recorded real-flight profiles and retaining complete within-flight time series rather than cycle-level snapshots. However, this added realism reduces evaluation control because full-flight records increase data volume and entangle degradation cues with operating-regime variation, complicating preprocessing choices and direct comparisons of RUL modeling performance. To mitigate this issue, this paper proposes CruiseBench, a cruise-stage RUL benchmark derived from N-CMAPSS. It introduces CPM-N-CMAPSS (Cruising-Period Mask for N-CMAPSS), a mask artifact that stores cycle-local cruising intervals identified by the common-altitude method for the nine accessible subdatasets. CruiseBench applies a fixed protocol to the masked rows, using scenario descriptors and measured sensors as inputs while excluding virtual sensors, health parameters, and auxiliary metadata from the feature tensor, preserving native-resolution windows, and applying dataset-wise RUL caps. Experiments with LSTM, GRU, TCN, and TSMixer provide baseline results for this setting. Under CruiseBench-eta5-W256-S10, TSMixer obtains the lowest average RMSE, 3.4\pm1.71 , and Saxena score, (2.50\pm2.99)\times 10^4 . Ablation studies show that flight-stage selection, temporal downscaling method, and RUL-cap threshold affect reported results. With its fixed cruise-stage protocol, CruiseBench provides a reproducible sub-benchmark for controlled RUL model comparison and CPM-N-CMAPSS provides a stage-specific data foundation for future transfer-learning and domain-adaptation studies.
[LG-59] Bayesian Wind Tunnels for Model Selection
链接: https://arxiv.org/abs/2607.19379
作者: Siddhartha R Dalal,Vishal Misra,Abhay Parekh
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: 27 pages
Abstract:Prior work has shown that transformers can perform exact Bayesian filtering within a fixed hypothesis class. Can they also perform Bayesian model selection – identifying the correct hypothesis class from data? We introduce model-selection Bayesian wind tunnels: controlled environments where ground-truth posteriors over hypothesis classes are available in closed form. Using fixed-point-free involutions – whose defining property f(f(x))=x is purely relational – a 2.8M-parameter transformer achieves 0.01-bit entropy agreement with the Bayesian optimum (3 seeds), with both integer tokens and opaque symbols whose meanings change every episode. This extends to non-nested comparisons: involutions vs. 3-cycles (where neither class is a subset of the other) achieve class-posterior MAE under 0.001, demonstrating genuine model selection beyond simplicity/subset bias. We then identify a sharp perceptual access condition: when the discriminative statistic requires arithmetic – modular addition (rotations) or multiplication (f(x)=cx mod p) – model selection succeeds with integer tokens but fails completely with opaque symbols, and this boundary persists under 112x scaling (2.8M to 316M parameters). A stationarity control confirms the operative factor: opaque tokens with a fixed relabeling succeed (0.009-bit MAE), showing that stable semantics, not integer identity, enable circuit compilation. Header subtask diagnostics localize the failure to the composition of header inversion with arithmetic rather than header parsing itself. Probing frontier LLMs on the same tasks shows qualitative Bayesian behavior but a large calibration gap (~55x), measured through lossy probes and therefore directional rather than exact. Comments: 27 pages Subjects: Machine Learning (cs.LG); Machine Learning (stat.ML) ACMclasses: I.2.6; G.3 Cite as: arXiv:2607.19379 [cs.LG] (or arXiv:2607.19379v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.19379 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Vishal Misra [view email] [v1] Wed, 1 Jul 2026 04:05:08 UTC (124 KB) Full-text links: Access Paper: View a PDF of the paper titled Bayesian Wind Tunnels for Model Selection, by Siddhartha R Dalal and Vishal Misra and Abhay ParekhView PDFHTML (experimental)TeX Source view license Current browse context: cs.LG prev | next new | recent | 2026-07 Change to browse by: cs stat stat.ML 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-60] Reliability-Aware Hard–Soft Physics-Informed Neural Networks for Robust Learning of Challenging Partial Differential Equations
链接: https://arxiv.org/abs/2607.19377
作者: Duc Tien Nguyen,Hang Tran,Trinh Minh Tuan,Nguyen Duc Manh,Dinh Gia Ninh
类目: Numerical Analysis (math.NA); Machine Learning (cs.LG); Fluid Dynamics (physics.flu-dyn)
*备注:
Abstract:Physics-informed neural networks (PINNs) provide a mesh-free framework for solving partial differential equations, but their training is often affected by loss imbalance, optimization stiffness, and difficulty in capturing localized or multi-mode solution structures. Hard-soft PINNs (HSPINN) alleviate part of this difficulty by embedding Dirichlet or periodic constraints directly into the trial space, but the resulting fixed admissible representation can still be poorly conditioned for sharp or heterogeneous residual fields. This paper proposes a reliability-aware hard-soft PINN (RA-HSPINN) that preserves exact embedded constraints while introducing a bounded learnable reliability field to modulate the interior representation. The method combines this reliability-aware ansatz with inverse-EMA global loss balancing and lightweight regularization, while retaining the standard mean-square residual form. The reliability field is a numerical modulation variable, not a physical parameter or calibrated probability. RA-HSPINN is evaluated on nonlinear Burgers equations, periodic convection, a mixed-boundary Poisson problem, and a mixed first-order Poisson system. Compared with HSPINN, it reduces the relative error by 98.65% for sharp-gradient Burgers, 72.42% for Burgers data with noisy and incompatible initial conditions, 61.18% for smooth periodic convection, 60.02% for localized periodic convection, 29.36% for mixed-boundary Poisson, and 82.17% for a multi-mode mixed first-order Poisson system. The results show that reliability-aware modulation is most beneficial when hard-soft trial spaces are admissible but difficult to optimize, especially in localized, unreliable-data, and multi-mode PDE regimes.
[LG-61] owards Torque-Driven Reinforcement Learning for Quadruped Locomotion
链接: https://arxiv.org/abs/2607.18365
作者: Jordan Dowdy,Jean Chagas Vaz
类目: Robotics (cs.RO); Machine Learning (cs.LG); Systems and Control (eess.SY)
*备注: 6 pages, 4 figures. Accepted manuscript. Published in the 2026 IEEE/SICE International Symposium on System Integration (SII), pp. 1259-1264
Abstract:Reinforcement learning (RL) for legged robots is advancing locomotion, demonstrating its ability to adapt to new and challenging terrain. Traditionally, these RL locomotion frameworks are position-based, making the policy less adaptable to terrain types and requiring state estimation techniques in the observation space, i.e., linear velocity. Moreover, these RL frameworks often use small, lightweight quadrupeds that are limited in their viability for high-complexity tasks due to hardware constraints. This work explores an RL torque control framework for heavyweight high-torque quadrupeds. The RL framework in this paper can traverse rough terrain and effectively track a desired linear velocity without requiring knowledge of the agent’s current velocity. Using Nvidia’s Isaac Sim and Isaac Lab, simulation results of the RL torque control policy are shown on the Unitree B1 quadruped, achieving speeds of 3.5 m/s and 1.5 rad/s. In addition, the quadruped can walk up and down stairs without the aid of an exteroceptive sensor.
[LG-62] Isaac Sim-to-Real: Reinforcement Learning based Locomotion for Quadrupeds
链接: https://arxiv.org/abs/2607.18135
作者: Jordan Dowdy,Jean Chagas Vaz
类目: Robotics (cs.RO); Machine Learning (cs.LG)
*备注: 6 pages, 5 figures. Accepted manuscript. Published in the 2025 IEEE 21st International Conference on Automation Science and Engineering (CASE), pp. 2194-2199
Abstract:Learning-based approaches to locomotion have risen in popularity in recent years, showing the capability for complex legged locomotion and whole-body control. Reinforcement learning (RL), the primary learning-based approach for locomotion, often utilizes a high-performance simulation tool, providing a controlled and efficient training and development environment. However, policies that perform well in simulation frequently encounter unexpected challenges when deployed on a physical system, known as the sim-to-real gap. This work presents a robust RL locomotion framework capable of whole-body control. The proposed RL framework utilizes Nvidia’s new set of simulation tools, Isaac Sim, and its companion RL framework, Isaac Lab, for training, achieving a zero-shot sim-to-real policy. The performance of our policy is validated on physical hardware using the Unitree Go1, with experimental results showing similar velocity tracking performance to the quadruped’s integrated controller, with a greater ability to recover from large disturbances, and achieve linear velocities of 2.0 m/s and angular velocities of 1.8 rad/s.
[LG-63] Lipschitzian SLLNs for random functions
链接: https://arxiv.org/abs/2607.20411
作者: Lai Tian,Johannes O. Royset
类目: Optimization and Control (math.OC); Machine Learning (cs.LG); Statistics Theory (math.ST)
*备注: 30 pages
Abstract:We prove strong laws of large numbers for locally Lipschitz functions in the Lipschitz pseudometric. Our results hold under either a topological or a model-theoretic condition, with the latter encompassing functions jointly definable in o-minimal structures but extending substantially beyond this class. Applications include uniform convergence of limiting and Clarke subdifferentials and finite-sample identification of solutions. Consequently, we identify broad classes of functions for which the failure phenomena revealed by our previous negative results [Tian and Royset, arXiv:2511.16568, 2025] do not occur.
[LG-64] Statevector-Referenced Geometry Survival of a Four-Qubit ZZ Quantum Kernel on IBM Quantum Hardware: A Fixed-Subset Diagnostic Across Three Execution Configurations
链接: https://arxiv.org/abs/2607.20377
作者: Rostyslav Sipakov
类目: Quantum Physics (quant-ph); Machine Learning (cs.LG)
*备注: 3 supplementary notes. Code and data available at Zenodo: this https URL . Preprint also available at this http URL : this https URL
Abstract:Quantum-kernel methods encode a dataset’s geometry in a Gram matrix, so learning claims on hardware kernels assume the intended geometry survives execution. We measure that survival for one frozen four-qubit ZZ feature-map kernel on N=24 real indoor air-quality windows, reconstructed on ibm_fez (1024 shots per circuit) under baseline, dynamical decoupling alone, and gate twirling alone, each a single non-interleaved job. Every configuration returned a complete, finite, positive-semidefinite Gram matrix and preserved the centered statevector geometry to a substantial but incomplete descriptive degree (full-matrix centered kernel alignment, CKA, 0.933-0.989). Gate twirling was most faithful on every reported geometry axis, with the only jackknife-resolved improvement over baseline (persisted Spearman, mean absolute error, and full-matrix CKA diagnostics); dynamical decoupling alone was not separated from baseline at the frozen-window scale. Residual hardware distortion, not finite sampling, dominates the discrepancy. Yet fidelity and label alignment were reversed: the most faithful configuration had the lowest centered kernel-target alignment, which sits at or below label-permutation references for statevector and hardware alike. We read the small hardware uplift as a normalization property of the non-affine distortion, not captured signal. These are descriptive results for single jobs on one backend, not causal mitigation-efficacy estimates; no quantum-advantage, hardware-classifier-superiority, or forecasting claim is made. Implementation fidelity and task relevance are distinct axes; hardware quantum machine-learning studies should report both.
[LG-65] Label-Free Finite-Volume-Residual Training of Attention Graph Neural Networks for Coupled Thermo-Fluid Fields
链接: https://arxiv.org/abs/2607.20321
作者: Tianyu Li,Zhiwei Cao,Qingang Zhang,Ruihang Wang,Binyang Song,Yonggang Wen
类目: Fluid Dynamics (physics.flu-dyn); Machine Learning (cs.LG)
*备注: 51 pages, 27 figures
Abstract:Neural surrogates are widely used in scientific machine learning for fast prediction of three-dimensional (3D) thermo-fluid fields. However, generating training data using conventional numerical solvers often incurs substantial computational and storage costs. We propose to train an attention graph neural network by minimizing the finite-volume method (FVM) residuals of the governing equations. These residuals are evaluated directly on the mesh, requiring no labeled data. We evaluate the trained surrogates against computational fluid dynamics (CFD) references and a data-supervised baseline across four scenarios. On the two steady-state benchmarks, the FVM-loss model achieves an all-field normalized root-mean-square error (nRMSE) of 2.3-2.8%. It demonstrates close agreement with the CFD references, including the buoyancy-energy coupling. On the two parametric transient cases, the FVM-loss model outperforms the supervised baseline in terms of accuracy, while avoiding the data-generation cost entirely. These results indicate that the FVM loss can provide a practical training signal for neural surrogates and reduce the model development cost.
[LG-66] Adaptive deep nonparametric regression from dependent data under covariate shift
链接: https://arxiv.org/abs/2607.20309
作者: William Kengne,Ehud Mossa Ockegna
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:Covariate shift often occurs because, in many real applications, the source and the target observations may be generated from different distributions. In this case, the standard metric under the source distribution is not appropriate. This paper considers deep neural network estimators for nonparametric quantile and Huber regression under covariate shift and from dependent observations. We deal with a generalized Bernstein-type inequality that is satisfied by many classical models, including i.i.d. observations, \phi -mixing, strong mixing, and \mathcalC -mixing processes. To perform the covariate shift phenomenon, we propose a sparse-penalized deep neural network (SPDNN) estimator that takes into account the discrepancy between the source and target distributions of the data. When the density ratio (between the source and target distributions of the covariate) is unknown, a two steps pre-training procedure is carried out: the first step is devoted to the construction of a least squares SPDNN estimator of the density ratio; which is used in the second step to perform a pre-training reweighted SPDNN estimator of the regression function. For both the quantile and the Huber regression, non-asymptotic error bounds of the proposed SPDNN estimators are established in the class of Hölder smooth functions. These estimators can adaptively attain (up to a logarithmic factor) the minimax optimal convergence rate from i.i.d. data as well as from several classical time series models.
[LG-67] Adaptive Bayesian Online Learning via Expert Aggregation
链接: https://arxiv.org/abs/2607.20239
作者: Jungbin Jun,Ilsang Ohn
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:Bayesian online learning promises uncertainty-aware prediction on data streams, but its performance hinges on inferential choices, including learning rates, prior distributions and variational families, which are usually fixed before seeing the stream. We address this by treating Bayesian update rules as experts and aggregating the Bayesian experts according to sequential predictive losses. We prove that the resulting aggregate competes with the best expert in hindsight at an aggregation cost determined by how each expert’s per-round performance is evaluated. We instantiate the framework in online conformal inference and Gaussian process regression. The conformal inference application yields a smoothed Bayesian counterpart of adaptive conformal inference with long-run randomized coverage, while the Gaussian process application gives an oracle inequality in cumulative predictive Kullback-Leibler risk and adaptation to unknown Hölder smoothness up to logarithmic factors. Experiments show that the aggregate tracks strong experts without oracle expert selection.
[LG-68] Dynamical and Optimization Trade-offs of Levi–Civita Coordinates for Learned Close-Encounter Dynamics
链接: https://arxiv.org/abs/2607.20235
作者: Abhishek Shankar
类目: Computational Physics (physics.comp-ph); Earth and Planetary Astrophysics (astro-ph.EP); Machine Learning (cs.LG)
*备注: 16 pages, 4 figures
Abstract:Classical regularization removes the binary-collision singularity from the Kepler problem, but its value as a representation for learned Hamiltonian dynamics has not been systematically isolated. We compare Cartesian and planar Levi–Civita formulations of a perturbed Kepler system with a smooth quadrupole potential. With the perturbation supplied analytically, a Levi–Civita Hamiltonian splitting holds the maximum relative energy error near 2.1\times10^-5 through eccentricity e=0.99 , while the Cartesian splitting becomes unstable. This advantage persists at matched physical horizon and force-evaluation budget, where the regularized baseline is 3\times10^-5 , about 4.7 – 8.3 orders of magnitude below the Cartesian arm depending on eccentricity. In held-out high-eccentricity tests with matched sampling, regularized models produce finite rollouts in 40/40 runs versus 0/40 for Cartesian. However, the fixed-shell construction supplies the regularized model with the exact initial orbit energy, and survival still carries \mathcalO(1) energy error. Four neural residual objectives fail to approach the analytic result. Exact-feature controls show that the regularized residual is a four-monomial degree-6 polynomial that a direct least-squares solve fits to the baseline. The remaining exact-feature gap is due to severe raw-basis ill-conditioning: orthogonalization restores baseline fitting for L-BFGS in two iterations. Small MLPs remain at \mathcalO(1) rollout error even after gauge symmetrization. Levi–Civita coordinates therefore improve dynamical conditioning while worsening raw-basis optimization conditioning; accurate neural residual learning remains unresolved. This is a controlled falsification-plus-trade-off study, not a solution to learned close-encounter dynamics.
[LG-69] Statistical Inference for Rank Allocation in Low-Rank Adaptation
链接: https://arxiv.org/abs/2607.20205
作者: Yihang Gao,Vincent Y. F. Tan
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Statistics Theory (math.ST)
*备注: 77 pages
Abstract:Low-rank adaptation (LoRA) has become a widely used parameter-efficient fine-tuning method for large language models. Since different modules and layers may contribute unequally to downstream adaptation, allocating rank resources under a fixed parameter budget is an important problem for balancing efficiency, expressiveness, and generalization. Existing adaptive rank methods address this problem mainly through carefully designed importance scores constructed from gradient-derived sensitivity and uncertainty measures, without an explicit statistical interpretation. In this paper, we formulate LoRA rank allocation as a statistical hypothesis testing problem and propose StatLoRA, a statistical inference-based rank allocation method. StatLoRA associates each LoRA component with a test statistic and uses estimated p-values to determine which components should be retained or pruned under a prescribed rank budget. The proposed testing procedure is supported by our central limit theory for stochastic optimizer trajectories. In particular, we establish asymptotic normality for a broad class of commonly used optimizers in deep learning, including AdamW, and derive the corresponding asymptotic distributions for the proposed component scores used in hypothesis testing. We evaluate StatLoRA on LoRA fine-tuning of DeBERTaV3-base, BART-Large, and Qwen2.5-7B across natural language understanding, natural language generation, and question answering tasks. Experiments show that StatLoRA achieves comparable or better performance than vanilla LoRA, AdaLoRA, and IGU-LoRA under matched rank budgets. Sensitivity analyses and empirical diagnostics further support the stability of the proposed hypothesis-testing-based allocation rule and provide empirical evidence for the asymptotic theory of component scores.
[LG-70] Hard Guarantees at a Measured Price: Entropy-Stable Learned Finite Volumes for Compressible Flow
链接: https://arxiv.org/abs/2607.20171
作者: Denis Gueyffier(ONERA – Institut Polytechnique de Paris)
类目: Fluid Dynamics (physics.flu-dyn); Machine Learning (cs.LG); Numerical Analysis (math.NA)
*备注: 15 pages, 7 figures, 3 tables
Abstract:Learned solvers for compressible flow are usually compared to classical methods at equal mesh resolution rather than at equal computational cost, and they typically offer no guarantee that their solutions remain physically admissible. We present a learned finite volume scheme for the two-dimensional Euler equations on unstructured meshes, admissible by construction and with an entropy-stable interior flux. We evaluate it under protocols fixed before any computation: frozen thresholds, falsification clauses, negative controls, a factor decomposition of the learned components, and an iso-cost comparison against the refined classical baseline. The decomposition produced the central result: the guarantee machinery alone, with both learned heads switched off (the unlearned skeleton), is the strongest scheme at equal mesh on every periodic case. At equal wall-clock cost the picture inverts into a map. Learning pays robustly only on the wall case whose boundary-condition type it never saw (10.8%). Its periodic gains flip sign with the evaluation draw (+10% on one held-out case, -12% on the hardest). The skeleton is the only method whose iso-cost gain never changes sign, at a measured overhead of 1.74x per step. The guaranteed variant completes 36 of 36 rollouts, Mach extrapolation and unseen wall included, with zero negativity events. We fix the guaranteed scheme’s one remaining out-of-distribution weakness, Mach extrapolation, at inference time: with scale-invariant network inputs, a specific-entropy floor, and no retraining, the corrected arm overtakes the unconstrained arm on one Mach case, cuts its deficit on the other by a third, passes the skeleton on the unseen wall, and keeps the guarantee. A spatial gate closes the loop: activating the heads only near the walls beats both the skeleton and the corrected arm, and transfers unchanged to a second wall geometry.
[LG-71] Plausibility-Driven Prioritization of Candidate Biomedical Annotations
链接: https://arxiv.org/abs/2607.20163
作者: Emanuele Cavalleri,Miad Alavinezhad,Dario Malchiodi,Marco Mesiti
类目: Quantitative Methods (q-bio.QM); Databases (cs.DB); Machine Learning (cs.LG)
*备注:
Abstract:The rapid growth of biomedical knowledge has made the validation of automatically generated biological annotations a major bottleneck in biomedical curation. While computational methods can rapidly produce large numbers of candidate annotations, determining which are biologically valid still requires costly expert review. Prioritizing these candidates before manual curation has therefore become a fundamental challenge. Machine learning techniques can support this process by exploiting biomedical knowledge graphs (bioKGs), which capture biological entities and their functional associations. In this work, we propose a framework that leverages bioKGs to estimate the plausibility of candidate annotations and guide expert curation. Starting from knowledge graph embeddings, we train relation-specific binary classifiers using a community-based negative sampling strategy to obtain reliable confidence estimates. We then introduce a family of plausibility measures that combine classifier confidence, classifier reliability, and the semantic context provided by alternative relationships involving the same pair of biological entities. Unlike conventional confidence estimation, the proposed approach explicitly accounts for multiple biologically meaningful relations that may coexist between the same entities. Experimental results on five large bioKGs demonstrate that the proposed negative sampling strategy consistently improves classifier robustness, increasing balanced accuracy by an average of 5.8%. Moreover, the plausibility measures outperform classifier confidence alone, enabling more effective prioritization of candidate annotations for expert review. Overall, our results show that the use of bioKGs improves the efficiency of AI-assisted biomedical curation while preserving expert control over the final annotation assessment.
[LG-72] Directional Kernel Mean Difference: A Fast Signed Statistic for Univariate Distribution Comparison
链接: https://arxiv.org/abs/2607.20119
作者: Shijie Zhong,Jiangfeng Fu
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:We introduce the Directional Kernel Mean Difference (DKMD), a signed statistic for univariate distribution comparison that preserves the direction of distributional shifts. Unlike the squared Maximum Mean Discrepancy (MMD), which discards directional information by squaring the RKHS distance, DKMD integrates the difference of kernel mean embeddings against a fixed odd weighting function. This construction yields three structural properties: antisymmetry, immunity to symmetric distributional differences, and directional monotonicity under stochastic dominance. We derive a data-driven Riemann estimator that ensures asymptotic consistency with the continuous formulation, strictly preserving the theoretical guarantees of the signed statistic in empirical evaluations. To overcome the quadratic computational cost of kernel methods, we develop an O(N \log N) prefix–suffix scanning algorithm that exploits the total order of the real line while requiring only O(N) memory. Experiments on synthetic benchmarks demonstrate that DKMD correctly isolates directional shifts from symmetric perturbations, remains robust to heavy-tailed outliers that can flip the sign of the mean difference, and scales to millions of samples in seconds.
[LG-73] Non–negative matrix factorization using the textitR package textsfnnmf
链接: https://arxiv.org/abs/2607.20084
作者: Volkan Sevinç,Nikolas Kontemeniotis,Theodoros Perdikis,Michail Tsagris
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:Non–negative matrix factorization (NMF) has become an established dimensionality reduction technique for extracting latent structures from non–negative data and has found widespread applications in fields such as bioinformatics, text mining, image analysis, and recommender systems. As the popularity of NMF has increased, numerous \textitR packages implementing different optimization strategies and computational frameworks have been developed. Despite their widespread availability, comprehensive evaluations of these implementations under real–world data conditions remain limited. Consequently, researchers often lack objective guidance when selecting an appropriate package for practical applications. This study introduces a new \textitR package for NMF and offers asystematic performance comparison with two widely available \textitR packages for NMF analysis. Rather than relying on simulated datasets, the evaluation is conducted using real–world data to better reflect the complexity, heterogeneity, and noise characteristics encountered in practical analytical settings. The packages are assessed using a consistent experimental framework, with emphasis on computational efficiency, convergence behavior, reconstruction accuracy, memory utilization, and the stability of the resulting matrix factorization.
[LG-74] Antigen-specific Antibody Multi-modal Foundation Model for Functional Antibody Design
链接: https://arxiv.org/abs/2607.20057
作者: Xiaoliang Shi,Zichen Wang,Runze Ma,Zhongyue Zhang,Shuangjia Zheng
类目: Biomolecules (q-bio.BM); Machine Learning (cs.LG)
*备注:
Abstract:Antibodies are essential proteins that play a central role in immune recognition by binding specific antigen molecules. Although recent protein language models have enabled progress in single-chain protein modeling and generation, they often fall short in antigen-specific antibody design, where effective modeling requires explicit pairing between antibody and antigen, particularly at the epitope level. To address these limitations, we introduce AAMFM, an Antigen-specific Antibody Multimodal Foundation Model that learns unified representations of antibody sequences and structures conditioned on antigen context. AAMFM incorporates rich antigen information including geometric interfaces and epitope annotations via a cross-modal adapter, enabling joint modeling of antibody-antigen interactions in a shared latent space. To further guide the model toward functional relevance, we fine-tune AAMFM using Calibrated Direct Preference Optimization (Cal-DPO), leveraging preference signals extracted from a strong structural prior to align learning with binding-specific objectives. Extensive experiments demonstrate that AAMFM achieves state-of-the-art performance in functional antibody design, revealing its potential for antigen-specific antibody engineering. Our code is available at this https URL.
[LG-75] PN-QNN: Harnessing Physical Noise as a Native Regularizer in Photonic Hybrid Quantum Neural Networks
链接: https://arxiv.org/abs/2607.20045
作者: Farah Elnakhal,Alberto Marchisio,Nouhaila Innan,Gabriel Falcao,Muhammad Shafique
类目: Quantum Physics (quant-ph); Machine Learning (cs.LG)
*备注: Accepted at the IEEE International Conference on Quantum Computing and Engineering (QCE), 2026
Abstract:Physical noise in near-term quantum hardware is usually treated as a nuisance to suppress. We ask whether it can instead act as a hardware-native regularizer for photonic hybrid quantum-classical neural networks (PHQCNNs), analogous to noise-injection regularization in classical deep learning. Using Quandela’s Perceval simulator and the MerLin framework, we build PHQCNNs for Iris, Digits, and MNIST and inject Perceval’s seven-parameter physical noise model directly into training. A genetic algorithm searches the six continuous noise dimensions and 1 boolean parameter to find, per dataset, the configuration maximizing validation accuracy, compared against a noiseless baseline across five seeds. GA-tuned noise yields modest accuracy gains on Iris (+0.82pp) and Digits (+1.45pp), but a clear degradation on MNIST (-1.21pp). Per-parameter sweeps show that no individual noise parameter is consistently beneficial, motivating the joint search, while a second-order loss expansion shows that physical noise induces a Tikhonov-like regularization term whose effect is dataset-dependent. Physical photonic noise can thus act as a free regularizer, but not universally.
[LG-76] Hypothesis-and-Refinement Learning of Organic Structures from Multimodal Spectroscopic Data
链接: https://arxiv.org/abs/2607.19816
作者: Chengchun Liu,Zhiyuan Yan,Li Yuan,Hao Li,Boxuan Zhao,Yonghong Tian,Bartosz A. Grzybowski,Fanyang Mo
类目: Chemical Physics (physics.chem-ph); Machine Learning (cs.LG); Computational Physics (physics.comp-ph); Data Analysis, Statistics and Probability (physics.data-an)
*备注: 6 figures
Abstract:Determining molecular structures from spectroscopic data remains fundamentally challenging because the inverse problem is intrinsically underdetermined: individual spectra are sparse, low-dimensional, and encode only partial structural evidence relative to the vast space of possible molecules. We address this challenge by formulating automated structure elucidation as a scalable hypothesis-refinement paradigm that tightly integrates spectral evidence with large-scale molecular priors. To supply structure-resolving NMR signals for multimodal learning, we construct \textbfQM9SPIN, a DFT-derived dataset comprising diverse 1D and 2D spectra, including J-coupling, DEPT experiments, and explicit spin–spin interactions. On this foundation, we introduce \textbfSpectroMol, a spectrum-to-structure model that proposes chemically valid molecular hypotheses conditioned on multimodal spectral inputs. Complementarily, we develop \textbfMS-Mol2Mol, a high-resolution mass-constrained molecular generator that integrates molecular formula, exact mass, and degree of unsaturation within a conditional generative prior trained on 400 million molecules, ensuring global compositional consistency and chemically realistic refinement. The integrated system achieves 93.8% top-1 accuracy on the simulated benchmark, adapts effectively from simulated to experimental spectra with limited experimental fine-tuning, and further improves experimental predictions through mass-guided refinement, establishing a scalable route toward automated, data-driven organic structure elucidation.
[LG-77] A Multiclass Quantum Aligned Centroid Kernel
链接: https://arxiv.org/abs/2607.19782
作者: Kilian Tscharke,Pascal Debus
类目: Quantum Physics (quant-ph); Machine Learning (cs.LG)
*备注:
Abstract:Kernel methods are powerful tools in machine learning but commonly used full-Gram kernels face three key limitations: (1) quadratic scaling with training set size; (2) the use of fixed, non-trainable kernels; and (3) the absence of an intrinsic formulation for multiclass classification. We present McQuack, a trainable quantum kernel method for multiclass problems that achieves linear scaling in the number of training samples. This is accomplished by replacing the full training-set Gram matrix with a trainable sample-to-(class-centroid) fidelity matrix. We evaluate the model in simulation and on 124 qubits of two IBM devices, across more than 150 datasets. In simulation, McQuack outperforms existing “pure” quantum baselines, while results from hardware inference – obtained without training – achieve performance similar to an RBF kernel. Finally, we study the trainability of the model and observe no evidence of barren plateaus in our experiments with up to 13 qubits, and highlight the importance of parameter initialization for successful optimization.
[LG-78] Machine Can Automatically Discover Parametric Functions to Model HEP Data
链接: https://arxiv.org/abs/2607.19750
作者: Ho Fung Tsoi,Dylan Rankin,Cecile Caillol,Miles Cranmer,Sridhara Dasu,Javier Duarte,Philip Harris,Elliot Lipeles
类目: High Energy Physics - Experiment (hep-ex); Machine Learning (cs.LG)
*备注: 6 pages. to be presented at ICHEP 2026
Abstract:In HEP data analyses, finding an adequate function to model binned data has largely relied on a manual process: guess a functional form by intuition, fit, examine, then repeat until successful. We show that this iterative process can be automated by a machine using symbolic regression, which performs a data-driven search over function space without requiring prior knowledge of what an adequate function should look like. We present the SymbolFit package, which pairs symbolic regression with uncertainty modeling to target HEP analysis use cases, and demonstrate it on the CMS and ATLAS Run 2 dijet spectra: 560 independent seeded runs across seven simple fit configurations generated over 1000 functions fitting the spectra with \chi^2/\textNDF\approx 1 , and 111 of the runs rediscovered the very dijet and UA2 functions used in published dijet searches.
[LG-79] Domain-Adapted Power Curve for Cross-Farm Applications
链接: https://arxiv.org/abs/2607.19744
作者: Ahmadreza Chokhachian,V. Roshan Joseph,Yu Ding
类目: Applications (stat.AP); Machine Learning (cs.LG)
*备注: Submitted to Renewable Energy
Abstract:The wind energy industry relies on accurate power curve models to make power forecast, evaluate turbine performance, quantify upgrade, or support site-planning decisions. In this paper, we focus on site-planning power curves, i.e., we investigate how power curve models trained using turbine data on an operating wind farm can be transferred to a new, undeveloped farm. The traditional wisdom in the wind energy literature relies on distance, layout, or terrain characteristics for making cross-farm power curve transfer. Through the lens of domain adaptation, we propose a more reliable transfer learning approach for cross-farm power curve modeling. In the cross-farm applications, a domain is specified by the temporal environmental variates and spatial terrain variables. Domain adaptation is to find a capable similarity metric to adapt the domain on the new farm to that on the existing farm. Empirical results show that our domain adapted power curve consistently outperforms competing approaches by an appreciable margin for site-planning power predictions.
[LG-80] Data-Poisoning Audits for Causal Effect Estimation
链接: https://arxiv.org/abs/2607.19692
作者: Kwangho Kim
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Methodology (stat.ME)
*备注:
Abstract:Observational causal analyses increasingly pool records across sites, vendors, and collection systems, creating vulnerability to append-only attacks in which plausible records are strategically selected to alter a reported treatment effect. We develop a data-poisoning audit for augmented inverse-probability-weighted estimation. The analyst specifies a finite catalog of feasible records, an append budget, and nested source capacities, and the adversary selects a feasible subset to maximize movement in a prespecified direction. With preprocessing and nuisance fits held fixed, we propose a greedy scan that computes the exact finite-sample worst-case movement at every append budget. To account for nuisance refitting, we go on to derive a total-influence score combining each record’s direct contribution with its effect through the propensity and outcome models. We further obtain a conservative finite-budget bound for the fully refitted estimate. Extensive simulations validate the exact result and show that total influence improves local refit prediction, while multisite and public-data analyses demonstrate material sensitivity at small append budgets. By translating adversarial data-composition risk into movement curves and critical budgets, the framework supports more reliable causal reporting and the design of source-level safeguards.
[LG-81] Optimal Recalibration of an Online Predictor
链接: https://arxiv.org/abs/2607.19689
作者: Lunjia Hu,Kevin Tian,Chutong Yang
类目: Machine Learning (stat.ML); Data Structures and Algorithms (cs.DS); Machine Learning (cs.LG)
*备注:
Abstract:We study the problem of recalibrating an online predictor [KE17, OKS24]: given an arbitrary “hint” sequence of forecasts, the learner must output new predictions that are calibrated while incurring small excess error relative to the original forecasts, under a proper loss. We give an online algorithm that achieves (\varepsilon, \varepsilon^2) -recalibration for Lipschitz proper losses in T \approx \varepsilon^-3 rounds, using an imbalanced extension of the recent simultaneous Blackwell approachability reduction framework of [HTY26]. We show that this tradeoff is optimal by proving a matching lower bound for recalibrating against the squared loss. We also prove a companion \mathcalK_2 -recalibration theorem that obtains the same tradeoffs up to a logarithmic factor. As our main application, we show how our recalibration algorithms can be combined with the online refinement method of [FH23] to obtain simultaneous \varepsilon -calibration and \varepsilon^2 -calibeating for smooth proper losses at the same asymptotic rate, improving upon prior works that achieved these properties separately or with a worse \varepsilon dependence. In particular, the \mathcalK_2 variant answers a question of [CHJL26] on simultaneously achieving near-optimal calibeating and calibration rates. We also derive extensions to settings with multiple hint sequences. Finally, we empirically evaluate our algorithms on a classification dataset undergoing distribution shift. Subjects: Machine Learning (stat.ML); Data Structures and Algorithms (cs.DS); Machine Learning (cs.LG) Cite as: arXiv:2607.19689 [stat.ML] (or arXiv:2607.19689v1 [stat.ML] for this version) https://doi.org/10.48550/arXiv.2607.19689 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-82] Nuclear Quantum Effects as a Denoising Problem
链接: https://arxiv.org/abs/2607.19680
作者: Weizhou Wang,Jonathan Weare,Aaron R. Dinner
类目: Chemical Physics (physics.chem-ph); Statistical Mechanics (cond-mat.stat-mech); Machine Learning (cs.LG); Computational Physics (physics.comp-ph); Quantum Physics (quant-ph)
*备注: 9 pages, 3 figures
Abstract:Nuclear quantum effects are rigorously captured by imaginary-time path integrals, which map the quantum Boltzmann distribution onto a ring polymer of classical replicas. Yet the nuclear masses, the coupling to the environment, and the boundary conditions of the path remain hard-wired in the simulation or the trained model, even though this quantum context enters the path measure only through a quadratic action known in closed form. Here we show that a denoiser trained on classical Boltzmann statistics alone, composed at sampling time with an analytic Gaussian component carrying the entire quantum context, yields the quantum Boltzmann distribution of the nuclei. Such a composition exists and is exact whenever the training noise does not exceed the intrinsic quantum uncertainty of the target ensemble, and it is invariant across all quantum contexts admitted by this bound. We show exact transfer across temperature, isotopic mass, dissipation strength, and the boundary conditions of the path in theory and in numerical experiments, without retraining. The last yields the end-to-end displacement and momentum distributions of a tagged nucleus from open imaginary-time paths. The same invariance extends in principle to the permuted boundary conditions of bosonic exchange, with the identical denoiser. In this view, the noise of generative modeling and the quantum fluctuations of the nuclei are two faces of the same quadratic structure.
[LG-83] A Deep Learning Framework for Predicting Solar EUV Irradiance During Significant Flares
链接: https://arxiv.org/abs/2607.19597
作者: Sathvik Soman,Jason T. L. Wang,Haimin Wang,Haodi Jiang
类目: olar and Stellar Astrophysics (astro-ph.SR); Instrumentation and Methods for Astrophysics (astro-ph.IM); Machine Learning (cs.LG)
*备注: 8 pages, 9 figures
Abstract:We present FlareEUV, a multimodal deep learning framework for predicting daily extreme ultraviolet (EUV) irradiance at 6.5 nm over three consecutive days during significant solar flares, using multi-instrument observations from NASA’s Solar Dynamics Observatory (SDO). We consider 33 significant flares in the period between 2011 and 2014 in Solar Cycle 24. The SDO observations include 13 co-aligned full-disk images, comprising eight AIA EUV/UV and five HMI magnetic/continuum products. FlareEUV learns the relationship between magnetic structure and coronal emission from the raw imaging data using a lightweight attention-based architecture. Our experimental results demonstrate the good performance of FlareEUV in short-term EUV irradiance forecasting during the significant flares and its superiority over baseline methods.
[LG-84] Machine-learned syndrome post-selection for reliable quantum error correction
链接: https://arxiv.org/abs/2607.19563
作者: Tobias Haug,Askery Canabarro,Leandro Aolita
类目: Quantum Physics (quant-ph); Machine Learning (cs.LG)
*备注: 12 pages, 7 figures
Abstract:Quantum error correction can be enhanced by post-selecting out runs that are likely to produce a logical failure, but the most accurate measures for that require costly decoder-level information. We introduce a practical, decoder-agnostic post-selection method that learns directly from syndrome data. The method trains a supervised classifier to distinguish between syndromes from low- and high-noise regimes, and then uses the classifier’s output as an abort score for new runs, without requiring logical-error labels, correction operators, or code-specific likelihood calculations. We validate the approach in three complementary settings: circuit-level simulations of the Gross bivariate-bicycle code, code-capacity simulations of the surface code, and experimental logical magic-state distillation data from the QuEra neutral-atom processor. In the Gross and surface codes, learned syndrome post-selection reduces the conditional logical error rate at a fixed acceptance rate, with performance comparable to syndrome-weight filtering. For the surface code, the learned classifier reveals a post-selection transition distinct from the conventional decoding threshold. In the experimental data, the machine-learning score outperforms syndrome-weight post-selection and, when combined with logical-gap filtering, improves the output fidelity beyond using the logical gap alone. These results show that syndrome-only learning provides a scalable and hardware-compatible route to improving the reliability of quantum error correction.
[LG-85] Online Optimization of Difference-of-Convex Compositions with Smooth Mappings
链接: https://arxiv.org/abs/2607.19553
作者: Jingwei Ji,Jong-Shi Pang,Renyuan Xu
类目: Optimization and Control (math.OC); Machine Learning (cs.LG)
*备注:
Abstract:We study online optimization for a broad class of structured non-convex non-smooth problems where each loss is a composition of a difference-of-convex function with a smooth mapping, and the feasible region is defined by constraint functions of the same kind. We propose a time-smoothed proximal linear algorithm and a local-regret measure based on a proximal residual mapping. We show that this residual is a proper stationarity measure for the original problem: its fixed-point condition implies first-order stationarity. Our analysis relies on a tangent-cone characterization for a feasible region described by composite difference-of-convex constraints, which is of independent interest and allows each update to be computed via a convex optimization oracle, despite the non-convexity of the problem. We establish a local-regret bound and a bound on the total number of inner convex subproblems. We also derive an error bound connecting the proximal residual to the distance to stationarity, providing a quantitative certificate of approximate stationarity. Subjects: Optimization and Control (math.OC); Machine Learning (cs.LG) Cite as: arXiv:2607.19553 [math.OC] (or arXiv:2607.19553v1 [math.OC] for this version) https://doi.org/10.48550/arXiv.2607.19553 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Jingwei Ji [view email] [v1] Tue, 21 Jul 2026 20:13:40 UTC (75 KB) Full-text links: Access Paper: View a PDF of the paper titled Online Optimization of Difference-of-Convex Compositions with Smooth Mappings, by Jingwei Ji and 2 other authorsView PDFHTML (experimental)TeX Source view license Current browse context: math.OC prev | next new | recent | 2026-07 Change to browse by: cs cs.LG math 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
[LG-86] RELTA-SGLD: Relative-Growth Localized Taming for Nonconvex Stochastic-Gradient Langevin Learning
链接: https://arxiv.org/abs/2607.19544
作者: Yiwei Zhou,Ziheng Chen
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Optimization and Control (math.OC)
*备注: 40 pages, 1 figure, 4 tables, Code available at: this https URL
Abstract:We introduce RELTA-SGLD, a taming scheme that stabilizes superlinear stochastic-gradient updates while reducing unnecessary suppression of the original learning drift. A threshold determines where the taming turns on, while a relative-growth principle derived from the one-step Lyapunov stability condition determines the required taming strength. Together, they produce a lighter \lambda -scale denominator and preserve a nonvanishing far-tail return. As a consequence, we prove polynomial moment stability and first-order stationary accuracy in both W_1 and W_2 for nonconvex SGLD with superlinearly growing stochastic-gradient oracles, improving the corresponding half-order and quarter-order bounds for comparable stochastic-gradient tamed schemes. On Fashion-MNIST under active stabilization pressure, RELTA improves the mean learning metrics over both untamed SGLD and TUSLA and remains competitive with a tuned AdamW reference. In an ordinary-training regime, its lighter localized denominator reduces unnecessary perturbation of the original update and maintains nearly untamed learning dynamics.
[LG-87] Equilibrium Causal Games: Separation Identification and the Identifiability of Cyclic Latent States
链接: https://arxiv.org/abs/2607.19531
作者: Faraz Dadgostari,Neda Nazemi
类目: Optimization and Control (math.OC); Machine Learning (cs.LG); Systems and Control (eess.SY)
*备注: Working Paper
Abstract:Power grids, markets, and interacting populations, settle into feedback driven equilibria observed through unknown sensors. Our Equilibrium Causal Game (ECG) joins a game to its cyclic causal model, hidden inputs, sensor map, and rules for interventions and equilibrium selection; interventions edit declared objects and recompute equilibrium. Under stated conditions, ECG-separation is sound but incomplete in our examples. Back-door/half-trek routes identify observed queries. Yet for an untouched rotationally symmetric Gaussian block, second moments determine only a source-frame rotation, across which distinct-variable effects generically change. Unknown sensing creates a separate ambiguity. In passive stable linear models without self-effects, unknown wiring and full-rank unknown sensing leave B completely unidentified for d\ge2 . Under LiNG, non-Gaussianity removes the source rotation; mechanism interventions separate sensing from interactions. With unknown support, invariant sensing, aligned responses, and well-posed single-target interventions identify (H,B) up to declared equivalence. Of d targets, d-1 suffice exactly when the sole untargeted node directly parents all others; otherwise d are needed. Acquisition probes are excluded; known wiring gives no universal count. With nonlinear sensing, isotropic Gaussian source blocks admit hidden twists within and across blocks in labelled environments preserving required radial laws. Conversely, under stated positivity, informative one-block changes, rank, and irreducibility conditions, the finest independent source-block representation is identified within the stated alternative class up to block permutation and blockwise coordinate changes, but not downstream mechanisms or the sensor/interaction split. Together, these results show which causal conclusions equilibrium data support and which require targeted experiments.
[LG-88] Boltzmann-Expected Molecular Design with Decoupled Annealing Flows
链接: https://arxiv.org/abs/2607.19519
作者: Selma Moqvist,Richard Beckmann,Ross Irwin,Rocío Mercado,Simon Olsson
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:Most 3D properties relevant to molecular design, including free energies and shape descriptors, are \textitexpectations over the Boltzmann distribution over 3D configurations of a molecular graph. However, existing property-guided generative models tie each property to a single structure, ignoring the underlying ensemble. We recast 3D molecular design as \textbfBoltzmann-expected design and realise it with \textbfDECAF (Decoupled Annealing Flows), which factorise the joint distribution over graphs and coordinates into two conditional flow models: a graph-conditioned flow p(x\mid\mathcalG) , acting as a \textitBoltzmann emulator , and a coordinate-conditioned flow p(\mathcalG\mid x) , proposing new graphs from 3D information. By alternating the two flows, DECAF optimises molecular graphs with a simulated-annealing acceptance rule whose scoring function is evaluated on ensembles drawn from p(x\mid\mathcalG) , making ensemble statistics, not single-conformer properties, the design target. The resulting loop requires no retraining to change objectives. On GEOM-Drugs, we show that ensemble-aware optimisation produces graphs whose mean radius of gyration and solvent-accessible surface area consistently shift toward targets, while single-conformer optimisation degrades on larger drug-like molecules where Boltzmann distributions are broadest. DECAF extends to multi-objective trade-offs and, uniquely among 3D generative models, to \textbfhigher-moment design : jointly optimising an ensemble property’s variance and skewness to produce flexible molecules biased to a prescribed conformational regime: we verify the conformational distributions of these higher-moment designs with all-atom MD simulations.
[LG-89] nsor Network Machine Learning for Wildfire Susceptibility Mapping: from Grokking Dynamics to Quantum Mixedness of Class Representations
链接: https://arxiv.org/abs/2607.19503
作者: Domenico Pomarico,Alessandra Costantino,Gabriel Ramirez Sanchez,Loredana Bellantuono,Davide D’ Alò,Mario Elia,Alessandro Fania,Francesco Giordano,Niloofar Kheirkhahan,Raffaele Lafortezza,Ester Pantaleo,Sabina Tangaro,Roberto Bellotti,Alfonso Monaco,Nicola Amoroso
类目: Physics and Society (physics.soc-ph); Machine Learning (cs.LG); Data Analysis, Statistics and Probability (physics.data-an)
*备注:
Abstract:A quantum-inspired tensor network framework for wildfire susceptibility classification in the Gargano region is introduced, leveraging AlphaEarth embeddings and Matrix Product State models. The approach combines scalable geospatial representations with an interpretable quantum mask, enabling both binary and multiclass classification of wildfire susceptibility. Beyond predictive performance, the study reveals a pronounced grokking transition in the binary case and provides a detailed analysis of inter-class confusion in the multiclass setting. By introducing level-resolved mixedness diagnostics based on reduced density matrices, we show that the MPS classifier naturally encodes a hierarchy of class distinguishability, with non-adjacent categories becoming more separable than neighboring ones. These results demonstrate that tensor network models not only achieve competitive classification accuracy but also offer a physically grounded framework to quantify and interpret class separability in complex environmental datasets.
[LG-90] A Bayesian Framework for Built-in Input Dimension Reduction for Gaussian Process Modeling
链接: https://arxiv.org/abs/2607.19498
作者: Eric Herrison Gyamfi,Emily L. Kang,Bledar A. Konomi,Guang Lin
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Probability (math.PR); Applications (stat.AP); Computation (stat.CO); Methodology (stat.ME)
*备注:
Abstract:Gaussian process (GP) modeling is widely used in computational science and engineering. However, fitting a GP to high-dimensional inputs remains challenging due to the curse of dimensionality. While various methods have been proposed to reduce input dimensionality, they typically follow a two-stage approach, performing dimension reduction and GP fitting separately. We introduce a Bayesian framework that seamlessly integrates dimensionality reduction with GP modeling and inference. Our approach, built on a hierarchical Bayesian model with priors on the Stiefel manifold, enforces orthonormality on the projection matrix and enables posterior inference via Hamiltonian Monte Carlo with geodesic flow. Additionally, we extend this framework by incorporating Deep Gaussian Processes (DGP) with built-in dimension reduction, providing a more flexible and powerful tool for complex datasets. Through extensive numerical studies, we demonstrate that while the proposed Bayesian method incurs higher computational costs, it improves predictive performance and uncertainty quantification, providing a principled and robust alternative to existing methods.
[LG-91] rend strength predicts when generative foundation models win: a power-controlled benchmark a mechanism and an actionable selection rule
链接: https://arxiv.org/abs/2607.19383
作者: Ahmed Cherif
类目: Applications (stat.AP); Machine Learning (cs.LG)
*备注: 12 pages, 6 figures. Code and per-forecast data released
Abstract:Pretrained generative foundation models cast forecasting as conditional generation from a learned predictive distribution and forecast unseen series zero-shot. We establish three results that turn their reported success into an actionable, mechanistic understanding. First (a positive benchmark result): on a power-controlled study of 1728 rolling-origin forecasts over 36 series from 19 datasets spanning the full range of STL trend strength (F_T in [0.17, 1.00]), a zero-shot Chronos model significantly outperforms four strong classical baselines – drift, seasonal-naive, Theta, and additive Holt-Winters/ETS – with the best mean MASE (1.187 vs. Theta 1.337, ETS 1.656; Friedman chi^2 = 46.08, p = 8.75e-09; Holm-corrected Wilcoxon p = 0.015 against every baseline; a Nemenyi critical difference separating it from the classical pack). Second (a novel, quantified mechanism): a controlled synthetic experiment with a known trend-generating process shows why – and reveals that the win does not come from better trend extrapolation. When the true trend is linear, damped, or exponential, additive ETS tracks the slope (slope-tracking ratio 1.02, 1.34, 0.98) whereas Chronos systematically under-extrapolates, behaving as a trend-shrinkage estimator (ratio 0.80, 0.49, 0.36). Third (an actionable selection rule): because the advantage is a shrinkage effect, it is predictable from trend strength alone – the generative model wins 78% of low-trend series but only 44% of high-trend ones, and its edge over ETS is significant on the low-trend stratum (0.982 vs. 1.671, p = 0.002) yet a tie on the high-trend stratum (p = 0.18). Trend strength, computable before forecasting from the training context alone, is therefore a practical a-priori indicator of when to deploy a foundation model. We additionally document a calibration shortfall (80% intervals cover 0.77).
[LG-92] Refnd: Preventing Data Leakage in Relational Datasets
链接: https://arxiv.org/abs/2607.19376
作者: Anthony Lavertu,Jacob Cote,Jacques Corbeil,Sophie Gobeil,Pascal Germain
类目: Quantitative Methods (q-bio.QM); Machine Learning (cs.LG)
*备注:
Abstract:Machine learning models trained on biochemical data are routinely evaluated using splits that fail to account for relational structure, causing information leakage and over-optimistic performance estimates. Existing splitting methods lack theoretical grounding and scale at best quadratically. We introduce the Relational Generative Process (RGP), a mathematical formalization explaining why relational structure arises in biochemical datasets, and Refnd, a splitting algorithm that leverages a proximity graph computed in loglinear time using Hierarchical Navigable Small World (HNSW). We validate on an antimicrobial peptide dataset, showing that Refnd splits yield lower but more realistic evaluation performance than traditional splits. Refnd is applicable to any dataset arising from an RGP such as protein sequences and structures, small molecules, and nucleotide sequences, and is openly available as a Rust accelerated Python package: pip install refnd.
附件下载
