本篇博文主要内容为 2026-09-03 从Arxiv.org论文网站获取的最新论文列表，自动更新，按照NLP、CV、ML、AI、IR、MA六个大方向区分。
说明：每日论文数据从Arxiv.org获取，每天早上12:30左右定时自动更新。
提示: 当天未及时更新，有可能是Arxiv当日未有新的论文发布，也有可能是脚本出错。尽可能会在当天修复。
目录
概览 (2026-09-03)
今日共更新632篇论文,其中:
- 自然语言处理共102篇(Computation and Language (cs.CL))
- 人工智能共164篇(Artificial Intelligence (cs.AI))
- 计算机视觉共136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16篇(Multiagent Systems (cs.MA))
- 信息检索共41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] Post-Training Language Models for Gold-Medal Performance in Coding Competitions
【速读】：该论文旨在解决大语言模型在高难度竞赛编程（如国际信息学奥林匹克竞赛，IOI）中推理能力不足的问题，尤其针对复杂算法问题的求解准确率与效率瓶颈。其核心解决方案在于构建一个端到端的专用化训练流程，包含大规模题目筛选、合成推理轨迹生成、监督微调（SFT）及强化学习（RL）等关键步骤，并引入一种基于反馈的测试时计算策略GenCorrect，通过迭代生成、评估与优化多种解法路径，显著提升解题质量。实验表明，经过后训练与GenCorrect增强后，Nemotron-3-Nano-CC在IOI 2025中的得分从130提升至468，超越金牌线（438.3），而更强大的Nemotron-3-Ultra-CC系统在模拟真实竞赛环境下的得分达到535.4，首次超越最高人类得分（498.27），验证了该方法在真实竞赛场景下的卓越性能。
链接: https://arxiv.org/abs/2609.02849
作者: Aleksander Ficek,Sean Narenthiran,Mehrzad Samadi,Somshubra Majumdar,Boris Ginsburg
机构: 未知
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Multiagent Systems (cs.MA); Software Engineering (cs.SE)
备注:
Abstract:Competitive programming has become a key test of large language model reasoning, with international competitions such as IOI and ICPC representing its most challenging settings. We present an end-to-end specialization pipeline combining large-scale problem curation, synthetic reasoning traces, supervised fine-tuning (SFT), and reinforcement learning (RL). Using 22,000 curated problems, we train Nemotron-3-Nano-CC (30B-A3B) with SFT and RL and Nemotron-3-Ultra-CC (550B-A55B) with SFT alone. We further introduce GenCorrect, a feedback-driven test-time compute strategy that iteratively generates, evaluates, and refines diverse solutions. On IOI 2025, Nano-CC improves from 130 points to 291 after post-training and to 468 with GenCorrect, exceeding the gold threshold of 438.3 while Ultra-CC reaches 502. Guided by these results, we develop a competition-specific Ultra-CC system and evaluate it prospectively during IOI 2026. Under the same time, internet-access, and submission constraints as human contestants, it scores 535.4 out of 600, exceeding both the gold threshold of 361.12 and the top human score of 498.27. To our knowledge, this is the first AI system to outscore the highest-scoring human contestant on an IOI problem set.
[MA-1] Collective creativity in hybrid societies
【速读】：该论文试图解决的核心问题是：在生成式AI广泛应用于文化创作的背景下，如何准确理解其对创造力的影响，尤其是在当前学术界对AI是否丰富或削弱文化存在分歧的背景下。论文指出，这种分歧源于将“新颖性”（novelty，单个作品的属性）与“多样性”（diversity，群体整体的属性）混淆。其解决方案的关键在于重新界定创造力的本质——主张在生成式AI语境下，创造力应被视为由人与算法共同构成的混合集体（hybrid collectives）的属性，而非个体的独立特质。研究发现，尽管AI辅助可显著提升个体产出的新颖性，但可能抑制整体多样性；然而，这种结果并非不可避免。由于人类与模型具有互补性的搜索机制，混合群体能够同时实现更高的新颖性与更优的多样性，且机器生成的创新方案可被融入并持续存在于人类文化中。因此，最终决定创造性成果的关键因素在于群体组成结构：参与主体的类型、比例及其连接方式。问题的焦点已从“AI是否促进或破坏创造力”转变为“何种人机混合模式可在保持个体创新优势的同时维持集体多样性”。
链接: https://arxiv.org/abs/2609.02620
作者: Mason Youngblood,Katie Mudd,Manuel Anglada-Tort,Cameron Jones,Elena Miu,Diana Omigie,Margaret Schedel
机构: 未知
类目: Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Multiagent Systems (cs.MA)
备注:
Abstract:Generative AI is changing how cultural artifacts are created and circulated, and with it our understanding of creativity itself. Researchers disagree about whether these tools enrich or impoverish culture, and we argue that much of that disagreement comes from conflating two distinct components of creativity: novelty, a property of single artifacts, and diversity, a property of populations. We argue further that creativity in the context of generative AI is best understood as a property of hybrid collectives, or populations of interacting people and algorithms, rather than of individuals. AI-assisted ideation reliably raises the novelty of individual output while narrowing diversity in the aggregate, but this is not an inevitable consequence of putting machines in the loop. Because humans and models search in complementary ways, mixed groups can outperform and out-diversify groups of either kind alone, and machine-discovered solutions can enter human culture and persist there. What decides the outcome is composition: which agents are present, in what proportion, and how they are connected. The question is no longer whether AI helps or harms creativity, but which mixtures let individual gains accumulate without eroding collective diversity.
[MA-2] Competitive Market Behavior of LLM s
【速读】：该论文旨在解决生成式 AI（Generative AI）作为经济主体在人类设计的市场机制中是否具备有效参与能力的问题，尤其关注这些机制在面对由大语言模型（Large Language Models, LLMs）驱动的代理时能否实现预期的资源配置效率。其核心问题在于验证LLM代理与基础市场机制（如双拍卖机制）之间的兼容性，即此类代理是否能够促成高效市场均衡。解决方案的关键在于通过复现经典的经济实验，将人类被试替换为LLM代理，系统性地评估其在双拍卖环境中的行为表现。研究发现，由LLM代理组成的市场收敛至均衡的速度显著迟缓甚至无法收敛，导致资源配置效率低于人类市场；进一步分析表明，不同模型家族及市场角色间存在显著的行为异质性，并且通过对链式思维（Chain-of-Thought, CoT）轨迹的词汇学分析揭示：代理执行交易决策的时机往往伴随着从策略性考量向紧迫性判断的转变。该研究公开了测试框架，为后续对LLM代理市场行为的评估提供了可复用工具。
链接: https://arxiv.org/abs/2609.02580
作者: Pawel Struski,Jakub Swistak,Inez Okulska,Przemyslaw Biecek
机构: 未知
类目: Multiagent Systems (cs.MA); Artificial Intelligence (cs.AI); General Economics (econ.GN)
备注:
Abstract:Large language models (LLMs) are increasingly deployed as economic agents, yet there is little evidence whether LLM agents are suited for participating in market mechanisms designed for humans, and whether these mechanisms deliver desired outcomes when faced with LLM agents. We address this question by replicating seminal economic experiments, replacing human subjects with LLM agents. We place agents in a double auction environment, which is a widely-used market mechanism. We check whether such a market is able to deliver an efficient allocation of resources, thereby testing a novel dimension of alignment of LLM agents – their compatibility with a fundamental market mechanism. We find that markets populated by LLM agents exhibit slower or no convergence towards market equilibrium, thus providing less efficient allocations than markets populated by humans. We then analyze agents’ individual trading decisions and find substantial heterogeneity both across model families and market roles. We also run a lexical analysis of Chain-of-Thought (CoT) traces generated by the agents. We find that the decision to execute a trade rather than continue incrementally adjusting prices is associated with a shift from strategic considerations toward urgency. We publicly release our testing framework, which can be used for future evaluations.
[MA-3] Codebook Agent : Amortized Topology Design for LLM Multi-Agent Systems
【速读】：该论文旨在解决大语言模型（Large Language Model, LLM）多智能体系统中通信拓扑结构动态适应查询以提升准确率与效率的问题。现有方法将此问题建模为条件图生成任务，即通过变分、自回归或扩散解码器在 N×N 邻接矩阵空间中搜索候选拓扑，并依赖图网络代理基于效用和结构成本（如边数）对样本进行排序。然而，作者指出该范式存在根本性偏差：实验表明，经过奖励筛选后的拓扑最终坍缩为约六种固定结构，即使代码本容量从8增至64也未显著改善多样性；边数与实际令牌消耗呈负相关（皮尔逊相关系数 r≈−0.4），说明稀疏化反而增加推理开销；此外，在主流基准测试中智能体共享同一配置（agent profile）时，基于消息传递的评分器具有邻接无关性，无法有效区分候选拓扑。针对上述问题，本文提出Codebook Agent——一种向量量化自编码器将成功拓扑压缩至一个与查询无关的16项代码本；通过奖励加权的MLP将查询嵌入映射为代码分布；并采用一个读取展平邻接矩阵的MLP代理，在单次批量前向传播中联合回归真实效用与任务归一化的令牌成本，对解码出的候选拓扑进行重排序。该方案摒弃了迭代搜索与测试时的消息传递机制，实现了六项基准上最高的平均准确率（84.6，优于最强基线的83.0），拓扑生成耗时仅2.4毫秒，且减少21.9%–33.2%的LLM令牌使用量。其核心创新在于将动态拓扑生成转化为可高效优化的代码本选择与端到端评分机制，克服了传统方法在表达能力、效率与可区分性方面的局限。
链接: https://arxiv.org/abs/2609.02264
作者: Jinxi Yu,Yubei Li,Eric Hanchen Jiang,Zhi Zhang,Dong Liu,Wenxiao Zhao,Levina Li,Kai-Wei Chang,Ying Nian Wu
机构: 未知
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Multiagent Systems (cs.MA)
备注:
Abstract:Adapting the communication topology of an LLM multi-agent system to each query improves both accuracy and efficiency, yet current designers treat this as conditional graph generation: a variational, autoregressive, or diffusion decoder searches the N \times N adjacency space, and a graph-network proxy trained on utility and a structural cost such as edge count ranks the sampled candidates. We argue that this formulation is misaligned with the problem. Empirically, topologies that survive a reward filter collapse to about six distinct graphs even when the codebook capacity grows from 8 to 64; edge count is negatively correlated with measured token consumption (Pearson r \approx -0.4 ), so sparsifying the graph makes inference more expensive; and a message-passing scorer over agent-profile nodes is adjacency-invariant whenever agents share a profile—the default configuration of published benchmarks—so it cannot rank candidates at all in that regime. These three facts motivate Codebook Agent: a vector-quantized autoencoder compresses successful topologies into a query-independent 16-entry codebook; a reward-weighted MLP maps the query embedding to a distribution over codes; and an MLP proxy that reads the flattened adjacency, regressed on measured utility and per-task normalized token cost, reranks the top decoded candidates in a single batched forward pass. With no iterative search and no message passing at test time, Codebook Agent is the most accurate method on all six benchmarks we compare (84.6 average against 83.0 for the strongest prior designer), emits a topology in 2.4 ms, and uses 21.9–33.2% fewer LLM tokens.
[MA-4] RideSkill: A Hierarchical Algorithm for Generalized Ride Sharing with LLM -Driven Automatic Evolution
【速读】：该论文旨在解决共享出行系统中多乘客共乘（ride-sharing）的复杂调度问题，核心挑战在于如何在不确定且动态变化的环境下，高效地将具有不同起终点（OD）对的订单进行打包并分配给车辆，同时应对状态空间与动作空间因车辆共享而指数级扩大的难题。现有基于多智能体强化学习（MARL）的方法存在泛化能力差、目标迁移性低及大规模系统中训练困难（如维度灾难）等问题；尽管近期有研究尝试引入大语言模型（LLM），但多数方法无法支持车辆共享，且依赖高频调用LLM进行推理，难以满足实时部署需求。本文提出RideSkill，一种分层式共享出行调度框架，其关键在于利用LLM辅助的自动化算法设计机制，通过一个可学习的技能库（skill repository）、组合器（combiner）和再定位器（repositioner）实现高效、自适应的调度决策。其中，组合器从技能库中为每辆车动态分配合适技能以适应不同场景与平台目标，再定位器则通过序列化移动空闲车辆至新兴需求区域，避免车辆冲突。所有模块均通过基于LLM的自动演化方法训练完成，无需在部署阶段调用LLM，从而显著提升实时性能，解决了现有方法在可扩展性、实时性与通用性方面的瓶颈。
链接: https://arxiv.org/abs/2609.02250
作者: Zijian Zhao,Sen Li,Xialiang Tong,Mingxuan Yuan
机构: The Hong Kong University of Science and Technology (香港科技大学); Noah’s Ark Lab, Huawei (华为诺亚方舟实验室)
类目: Multiagent Systems (cs.MA); Computation and Language (cs.CL); Emerging Technologies (cs.ET); Machine Learning (cs.LG)
备注:
Abstract:Ride-sharing, which allows multiple passengers with different origin-destination (OD) pairs to share a single vehicle, is a challenging operational problem, as it requires orders with different OD pairs to be efficiently bundled and assigned to vehicles under uncertain and varying scenarios. Although multi-agent reinforcement learning (MARL) solutions have achieved promising performance, they suffer from limited generalization (adapting to different environmental scenarios), low transferability (adapting to different platform objectives), and training difficulties in large-scale systems, such as the curse of dimensionality. Recently, motivated by the scaling of large language models (LLMs), several works have incorporated LLMs into ride-hailing systems, either by employing LLMs directly as decision-making agents or using them for automatic algorithm design. However, none of these approaches support vehicle sharing, which complicates the problem by expanding both the state and action spaces exponentially. Moreover, most of them require frequent LLM calls at inference time, making them infeasible for real-time deployment. To address these issues, we propose RideSkill, a hierarchical method for ride-sharing that leverages LLM-assisted automatic algorithmic design. RideSkill consists of a combiner that assigns appropriate skills to each vehicle from a learned skill repository, enabling adaptive dispatch under varying scenarios and objectives, and a repositioner that sequentially relocates idle vehicles to emerging regions, avoiding conflicts among vehicles. Crucially, the skill repository, combiner, and repositioner are all trained by an LLM-based automatic evolutionary method, eliminating the need for LLM calls during deployment and thus ensuring high real-time performance.
[MA-5] Propose to Learn Learn to Propose: Evaluability-Aware Assistance under Bounded Rationality
【速读】：该论文旨在解决现有生成式AI助手在协作过程中忽视用户评估能力有限性的问题，即传统方法假设用户能够可靠地评估任意提案，但在实际应用中由于认知资源受限（bounded rationality），用户对复杂或模糊提案的评估往往不可靠。其核心挑战在于如何在保证提案质量的同时，主动设计具有可评估性的提案，以有效探测并学习用户的隐含偏好与评估约束。解决方案的关键在于提出一种“可评估性感知的提案规划”（evaluability-aware proposal planning）框架——ProSE（Hidden-Parameter Sequential Assistance Problem），将提案同时视为任务干预手段和对用户潜在偏好的探测工具。该框架通过引入KL正则化的有界理性二元响应模型，使接受提案的决策在价值收益与评估代价之间进行权衡，从而显式建模用户评估能力的边界。分析表明，高接受概率的提案未必是最具信息量的探测信号，因此单纯追求可接受性的规划策略会系统性表现不佳。为此，作者提出\textscProSE-Plan，一种深度为2的贝叶斯自适应规划器，通过联合评估不同响应可能性及其引发的后验信念更新来评分提案。在控制图模拟实验中，当评估成本成为瓶颈时，\textscProSE-Plan显著优于无评估意识及短视基线方法，且探针-采纳消融实验验证了其能识别出简单方法忽略的高信息量提案。研究结果揭示了用户可评估性作为智能辅助规划中的关键维度，与生成质量、偏好推断共同构成多维优化目标。
链接: https://arxiv.org/abs/2609.02242
作者: Yifan Zhu,Sammie Katt,Samuel Kaski
机构: ELLIS Institute Finland; Department of Computer Science, Aalto University, Finland; Department of Computer Science, University of Manchester, United Kingdom
类目: Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC); Multiagent Systems (cs.MA)
备注: 28 pages, 7 figures, 8 tables. Under review
Abstract:AI assistants often collaborate by proposing candidate edits, plans, or designs that users evaluate before adoption. Existing assistance methods focus on proposal quality or user-goal inference, often assuming that the user can reliably evaluate any proposal, which can fail in practice because of bounded rationality. We study evaluability-aware proposal planning, where proposals serve both as task interventions and as probes for learning latent preferences and evaluation constraints, where the resulting belief updates then guide later proposals. We formalise this setting as ProSE, a hidden-parameter sequential assistance problem, and instantiate it with a KL-regularised bounded-rational binary response model in which acceptance trades off value gain against a distance-dependent evaluability penalty. Analysing the planning consequence of this likelihood reveals that likely accepted proposals and informative probes need not coincide, which explains why planners that only pursue acceptance systematically underperform. We operationalise ProSE with \textscProSE-Plan, a depth-2 Bayes-adaptive planner that scores proposals by possible responses and response-induced posterior beliefs. In controlled graph simulations, \textscProSE-Plan improves over evaluability-unaware and myopic baselines when evaluation cost is the bottleneck, and a probe-commit ablation confirms that our approach selects informative proposals that simpler methods miss. Our results thus identify user evaluability as a planning-relevant dimension of AI assistance, complementary to generation quality and preference inference.
[MA-6] ClaimReceipt: Verifying Evidence Sufficiency and Coverag e in Agent Evaluations NEURIPS2026
【速读】：该论文旨在解决智能体评估中两个关键的证据性问题：一是报告的声明是否可从保留的证据中重新计算（充分性），二是保留的记录是否覆盖了承诺的实验集合（覆盖范围）。传统日志和哈希链接的记录无法可靠回答这两个问题。为此，论文提出ClaimReceipt（声明凭证）机制，这是一种与声明相关的凭证规范及选择性验证器，能够将类型化的交易证据绑定至签名的实验清单，并针对每个声明返回PASS、INVALID或INCONCLUSIVE的结果。通过在实现前冻结规范（SHA-256 18d109…b81），在1,392条历史买卖记录上，CR-2验证器成功复现了全部五项人工标注的审计结论，精确重放了600条确定性记录和792条生成后记录，验证了13个声明字段组在测试扰动下的非冗余性，并在11个语义故障案例中均正确返回预期结果，且无误报（0/8假阳性）。随后开展的前瞻性CR-3阶段进一步验证：30项任务在推理前被承诺，终端凭证被签名并链式连接，私有证据加密供审计。完整证据下获得覆盖与会计验证通过；若仅隐藏一个终端凭证，则返回INCONCLUSIVE_COVERAGE；而若所有私有证据均未开放，则覆盖与协议验证仍成立，但经济声明变为不可判定，与预先注册的预测完全一致。实验表明，凭证注入仅增加0.021%模型推理时间及每笔交易9.9 KB开销。此外，规范可读性探测显示，即使对作者自身而言，冻结后的规范仍不够清晰明确。因此，论文强调，声明验证不仅需要充分的证据，还需一个已承诺的实验宇宙，以使遗漏行为得以显现。
链接: https://arxiv.org/abs/2609.01992
作者: Peiying Zhu,Sidi Chang
机构: 未知
类目: Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR); Multiagent Systems (cs.MA)
备注: Submitted to Who Verifies the Agents? Toward Reliable Agent Development (NeurIPS 2026 workshop). 8 pages, 1 figure, 7 tables
Abstract:Agent evaluations face two distinct evidentiary questions: whether a reported claim is recomputable from retained evidence (sufficiency), and whether the retained records cover the committed experiment set (coverage). Generic logs and hash-linked transcripts answer neither reliably. We introduce ClaimReceipt, a claim-relative receipt specification and selective verifier that binds typed transaction evidence to a signed experiment manifest and returns PASS, INVALID, or INCONCLUSIVE per claim. We freeze the specification before implementation (SHA-256 18d109…b81). On 1,392 historical buyer–seller records, a CR-2 verifier reproduces all five manually labeled audit verdicts, exactly replays 600 deterministic and 792 post-generation records, makes every one of 13 declared field groups non-redundant under tested ablations, and returns the expected result on 11/11 semantic faults with 0/8 false positives. We then run a separate prospective CR-3 epoch: 30 assignments are committed before inference, terminal receipts are signed and chained, and private evidence is encrypted for an auditor. Complete evidence yields coverage and accounting PASS; withholding one terminal receipt returns INCONCLUSIVE_COVERAGE, while withholding all private openings preserves coverage and protocol verification but makes economic claims inconclusive, exactly matching a preregistered prediction. Receipt instrumentation adds 0.021% of model-inference time and 9.9 KB per transaction. A specification-legibility probe indicates that our own frozen specification is not yet unambiguous to an independent reader. Claim verification therefore requires both claim-sufficient evidence and a committed universe against which omissions become visible.
[MA-7] When Agents Implement Systems: A Case Study in Defects Detection and Evaluation Rigor
【速读】：该论文旨在解决大语言模型（Large Language Model, LLM）编码代理在系统级需求实现中的行为缺乏实证刻画的问题，具体聚焦于数据系统构建中涉及的模式设计、异步编排、配置正确性以及检索-过滤权衡等关键挑战。其核心解决方案在于通过一个端到端的案例研究，评估单一LLM代理在预设架构约束下自主完成多组件数据系统实现的能力，包括代码实现、缺陷诊断与修复，以及开放性交互设计选择。研究发现，在单次会话中代理引入了五类违反约束的缺陷，并通过不同检测方法识别。此外，针对架构中明确指定的检索-过滤权衡问题，研究在公开的HotpotQA基准上进行评估，将实体识别阶段替换为使用黄金标注的实体集合以规避模型访问限制，采用标准召回率作为评价指标。结果表明，经过图结构识别的实体集过滤后，检索召回率在预算为3时即达上限；而未过滤的搜索在最大预算10时仅能恢复全部必要证据的69%，且该差距在所有预算水平下均显著（符号检验p < 0.0001）。研究最后讨论了代理自主性的边界，指出其在部分决策中虽表现出自主性，但存在未对引发修正的回归测试进行重新验证的情况，揭示了当前代理在自我验证与持续改进方面的局限。
链接: https://arxiv.org/abs/2609.01985
作者: Phanindra Reddy Madduru
机构: Amazon.com(亚马逊)
类目: Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注: 4 pages
Abstract:As LLM coding agents increasingly perform end-to-end engineering work, we lack empirical characterization of how they behave on systems-level requirements: schema design, async orchestration, configuration correctness, and retrieval-filtering trade-offs. We present a case study of one such agent implementing a multi-component data system against a detailed pre-existing specification. Storage technologies, schema, entity-resolution algorithm, and retrieval-filtering strategy were fixed in advance; the agent autonomy was in the implementation, in diagnosing and fixing defects it introduced, and in interaction-design choices left open. Over a single session, we catalog five such defects, categorized by constraint violated and detection method. We further evaluate, on the public HotpotQA benchmark, the one retrieval trade-off specified in that architecture: restricting candidates to a graph-identified entity set before ranking versus unfiltered search. We substitute the benchmark gold evidence labels for entity identification, since we lacked LLM access to run that stage, and report standard recall rather than the benchmark own accuracy metrics. Across retrieval budgets from 1 to 10 and 100 questions against a pooled corpus of 2994 paragraphs, filtered recall reaches its ceiling by a budget of 3, expected once candidates are restricted to the gold paragraphs themselves, while unfiltered search recovers all required evidence only 69 percent of the time even at a budget of 10, a gap that holds at every budget tested, with sign test p less than 0.0001. We close with a discussion of where the agent autonomy succeeded versus required correction, including one instance where a claimed performance fix was never re-measured on the regression that motivated it.
[MA-8] Bonded Recourse for Smart-Contract Settlement of Compensable Agent Side Effects
【速读】：该论文旨在解决自主代理（autonomous agent）在跨组织边界执行工具操作时，因允许但失败的操作所导致的残余损害（residual harm）无法补偿的问题。现有方案如授权（authorization）与本地回滚（local compensation）仅能覆盖事前准入与运行时回滚，却无法处理已发生但未成功操作所带来的持续性负面影响。其解决方案的关键在于设计并实现Recourse协议——一种基于智能合约的可补偿副作用结算机制，将每项被批准的操作绑定于作用范围（scope）、恢复策略（recovery）、证据（evidence）、赔付（payout）及抵押物（collateral）五个维度。Recourse通过分离事前资格审查与事后客观可结算性，引入类型化收据（typed receipts），支持在乐观预言机（optimistic-oracle）挑战模式下对客观残余索赔进行可计算验证；而主观或不完整索赔则交由ERC-792仲裁机制处理或直接排除。系统已在Base Sepolia链上部署，并开发了与Postgres、Git及云兼容本地沙箱的适配器，通过确定性测试框架、沙箱日志、对抗性扫描和基于属性的模糊测试进行评估。结果表明，在对比仅依赖授权或本地补偿的基线系统时，带有抵押保障的覆盖机制显著减少了未补偿损害，且链上层提供了中立托管、公开挑战、非合作支付以及跨组织信任假设下的可移植历史记录能力。
链接: https://arxiv.org/abs/2609.01939
作者: Laurent Bindschaedler,Quentin Botha,Christoph Siebenbrunner
机构: Max Planck Institute for Software Systems (马普所软件系统研究所); Vienna University of Economics and Business (维也纳经济大学)
类目: Cryptography and Security (cs.CR); Distributed, Parallel, and Cluster Computing (cs.DC); Multiagent Systems (cs.MA)
备注: 9 pages, 3 figures, 8 tables. Accepted at BCCA 2026 (IEEE International Conference on Blockchain Computing and Applications)
Abstract:Autonomous agent runtimes execute tool actions that mutate databases, repositories, and cloud services across organizational boundaries. Authorization and local compensation cover pre-action admission and in-runtime rollback, but neither settles the residual harm left after a permitted action fails. We design Recourse, a smart-contract settlement protocol for compensable agent side effects that binds each admitted action to scope, recovery, evidence, payout, and collateral. Recourse separates ex ante eligibility from ex post objective settleability: typed receipts make objective residual claims computable under an optimistic-oracle challenge pattern, while subjective or incomplete claims route to ERC-792 arbitration or exclusion. We implement the contract suite, deploy it on Base Sepolia, build adapters against Postgres, Git, and cloud-compatible local sandboxes, and evaluate the system on a deterministic harness, sandbox traces, adversarial sweeps, and property-based fuzzing. Against authorization-only and local-compensation baselines, bonded coverage cuts uncompensated harm. The on-chain tier supplies neutral custody, public challenge, non-cooperative payout, and portable history under cross-organizational trust assumptions.
[MA-9] Agent Flight Recorder: Tamper-Evident Audit Trails with On-Chain Anchoring for Long-Horizon Tool-Using Agents
【速读】：该论文旨在解决长时序智能体（long-horizon agents）在执行数千次连续操作过程中，因序列性失败（sequential failures）导致的可追溯性缺失问题，尤其针对诸如代码代理误删生产数据库或提示注入（prompt injection）跨代理传播等高风险事件中所引发的因果关系模糊、责任归属不清及第三方不可抵赖验证难题。其核心解决方案是提出“智能体飞行记录器”（Agent Flight Recorder），通过将每个智能体动作结构化为包含意图（intent）、执行（execution）到来源（provenance）等八个语义字段的标准化事件，并利用哈希链（hash chaining）与默克尔分批（Merkle batching）技术实现篡改检测与紧凑的包含证明。对于跨组织争议场景，系统采用周期性链上锚定（on-chain anchoring）机制，将每个周期的根哈希（epoch root）上链，仅存储32字节根值与回指指针，不暴露事件内容，确保任何持有披露载荷和默克尔证明的验证者均可独立验证，无需预先信任中介。实验表明，完整系统在合成工作负载下仅引入约48微秒/事件的中位延迟与512字节/事件的开销，每10万事件成本为2.30美元（100事件周期），且对编辑、删除、重排序及分叉篡改均实现100%检出率与零误报；相比传统文本搜索，结构化取证查询在护栏（guardrail）与委托（delegation）查询上的精确率分别达1.0，显著优于未结构化搜索的0.013与0.077。
链接: https://arxiv.org/abs/2609.01931
作者: Laurent Bindschaedler,Quentin Botha,Christoph Siebenbrunner
机构: Max Planck Institute for Software Systems (马普所软件系统研究所); Research Institute for Cryptoeconomics, Vienna University of Economics and Business (维也纳经济与商业大学密码经济学研究院)
类目: Cryptography and Security (cs.CR); Multiagent Systems (cs.MA)
备注: 9 pages, 1 figure, 4 tables. Accepted at BCCA 2026 (IEEE International Conference on Blockchain Computing and Applications)
Abstract:Long-horizon agents execute thousands of actions, resulting in sequential failures rather than isolated errors. When a coding agent deletes a production database or a prompt injection spreads across agents, the incident raises questions of causality, authority, and non-repudiable third-party verification. The Agent Flight Recorder captures each agent action as a structured, canonically serialized event binding eight semantic fields from intent through execution to provenance. Hash chaining and Merkle batching provide tamper evidence and compact inclusion proofs. For cross-organizational disputes where no party’s infrastructure qualifies as neutral ground, periodic on-chain anchoring of epoch roots lets any verifier with the disclosed payload and Merkle proof check the record independently, without pre-agreeing on a trusted intermediary. The on-chain footprint is minimal: each anchor stores a 32-byte epoch root and a back-pointer, and no event content touches the chain. We evaluate the system across five cumulative ablation configurations on synthetic agent workloads. The full system adds ~48 microseconds median per-event latency and 512 bytes per event. L2 anchoring costs 2.30 per 100K events at 100-event epochs. The full integrity stack detects edit, delete, reorder, and fork tampering at 100% with zero false positives. Structured forensic queries achieve 1.0 precision on guardrail and delegation lookups where unstructured text search yields 0.013 and 0.077 respectively.
[MA-10] Epistemic Sybil Resistance: Multiplying AI Agents Without Multiplying Evidence
【速读】：该论文旨在解决多智能体生成式AI系统在集体推理过程中因智能体间存在隐含的证据共源性（epistemic Sybil问题）而导致的可信度误判问题。其核心挑战在于：多个看似独立的报告可能源自同一证据源，而真正独立的证据却可能生成高度相似的报告，导致仅依赖报告内容进行聚合的机制无法区分“重复”与“独立佐证”，从而引发后验概率校准偏差。解决方案的关键在于识别并建模报告间的隐含证据依赖关系，而非简单依据报告数量、智能体数量或文本相似性进行聚合。研究提出通过共享根节点的高斯模型揭示共源性不等同于完全冗余，并实证表明重复提取可逼近源级信息上限，但由共享基础模型引发的提取误差相关性会进一步降低该上限。实验在超过2万次受控的大语言模型智能体调用中验证了这一理论：当证据根数固定而报告数从1增至32时，朴素后验覆盖率由0.940骤降至0.263；反之，当报告数固定而证据根数增至16时，不同聚合器表现趋于无统计差异。此外，实测显示智能体重复提取误差高度相关（γ_cal = 0.719），引入考虑误差相关性的聚合机制可有效恢复校准。研究还通过控制变量分离表示相似性与证据来源，发现前者对聚类数量推断的影响远大于后者。因此，集体推理必须追踪证据的溯源路径与依赖结构，而非依赖报告或智能体的表层特征。
链接: https://arxiv.org/abs/2609.01873
作者: Marc Bara
机构: Universitat Oberta de Catalunya (UOC), Barcelona, Spain
类目: Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注: 23 pages, 13 figures. Code and data: this https URL
Abstract:Multi-agent AI systems improve inference by spawning agents and synthesizing reports. But another agent is not another observation: apparently independent reports may descend from the same evidence, and genuinely independent evidence can produce nearly identical reports. We formalize this as an epistemic Sybil problem. A report Z is an epistemic Sybil extension relative to reports R when I(Theta; Z | R) = 0. No report-only aggregator can generally distinguish replication from independent corroboration: identical reports can warrant different posteriors under unobserved ancestry. A Gaussian shared-root model shows common ancestry does not imply complete redundancy. Repeated extraction adds information toward a source-level ceiling, and correlated extraction errors, which a shared base model can induce among independent agents, lower that ceiling further. We test these predictions with more than 20,000 controlled LLM-agent report and extraction calls on synthetic evidentiary documents. Holding one evidence root fixed while report multiplicity rises from 1 to 32 collapses naive posterior coverage from 0.940 to 0.263. Holding report count fixed while evidence-root multiplicity rises from 1 to 16 closes the gap, and the aggregators are statistically indistinguishable at k = 16. The agent’s replicate extraction errors are correlated (gamma_cal = 0.719, estimated out of sample), and a correlated-extraction aggregator restores calibration accordingly. A controlled manipulation isolates representation similarity from evidential ancestry. It changes a report-space deduplication mechanism’s mean inferred cluster count by 1.425 (95% CI [1.363, 1.485]), whereas a fourfold change in true ancestry changes it by only 0.040 ([-0.045, 0.120]). Collective inference should therefore track evidential ancestry and dependence, not agent or report multiplicity or similarity.
[MA-11] ArcticSwarm: Deferring Early Consensus in Long-Horizon Multi-Agent Research
【速读】：该论文旨在解决多智能体系统（Multi-agent Systems）在开放性、长时程科研任务中缺乏可靠验证机制（verifier）时，难以有效开展深度研究的问题。传统方法依赖多数投票或自一致（self-consistency）作为代理验证器，但在无验证器的场景下，平行智能体易重复探索相同证据，且因共享部分发现导致搜索过早收敛至早期候选解，限制了探索广度。其解决方案的关键在于提出 ArcticSwarm 架构，通过将证据收集（evidence gathering）与证据整合（evidence integration）分离：子智能体将发现发布至共享公告板，同时采用门控隔离（gated isolation）机制，使选定的搜索任务可保留自身先验信息，避免过早达成共识；并在三个承诺边界（commitment boundaries）实施结构化审查，仅允许高置信度候选方案被传播。实验表明，该设计显著提升了搜索多样性与最终性能，在 BrowseComp-Plus 全集上使用 Qwen 3.5-27B 模型达到 82.6% 的准确率，优于基线模型（70.6%），并在实时网页检索任务（live-web BrowseComp）中以 GPT-5 达到 73.6%，大幅超越报告的供应商系统（54.9%）和 MiroFlow（63.4%）。结果证明，限制证据收集阶段的同行读取并强化假设传播前的承诺边界，是提升长时程多智能体深度研究能力的核心。
链接: https://arxiv.org/abs/2609.01870
作者: Soyoung Yoon,Boyi Liu,Yite Wang,Ruofan Wu,Canwen Xu,Nikki Lijing Kuang,Seung-won Hwang,Yuxiong He,Zhewei Yao
机构: Seoul National University (首尔国立大学); Snowflake AI Research (Snowflake人工智能研究)
类目: Multiagent Systems (cs.MA)
备注:
Abstract:Multi-agent systems have shown strong performance in domains with reliable verifiers such as coding, where multi-parallel candidate generation selected by a verifier is effective. However, such pipelines would not generalize to open-ended, long-horizon research tasks without a verifier. While majority voting or self-consistency is often used to reach consensus as a proxy verifier, parallel agents repeatedly explore the same evidence, while access to peers’ partial findings cause search to converge on an early candidate before alternatives are tested. We present ArcticSwarm, a multi-agent research architecture that separates evidence gathering from evidence integration. Subagents publish findings to a shared bulletin board, while gated isolation lets selected search tasks maintain their own prior, preventing early consensus. Structured review at three commitment boundaries enforce only confident candidates to be propagated. As a result, ArcticSwarm reaches 82.6% on the full BrowseComp-Plus set with the open-weight Qwen 3.5-27B model, compared with 78.8% without gated isolation and 74.5% additionally with structured review disabled, outperforming aligned baseline MiroFlow runs (70.6%). Extending to live-web BrowseComp, ArcticSwarm reaches 73.6% with GPT-5, which is well above the reported provider system (54.9%) and MiroFlow (63.4%). Overall, the results show that restricting peer reads during evidence gathering and strengthening commitment boundaries before a hypothesis is shared can broaden search and improve long-horizon multi-agent deep research.
[MA-12] Differential Games for Compositional Handling of Competing Control Tasks
【速读】：该论文旨在解决单智能体多目标动态系统中因控制目标间存在动态冲突而导致全局加权方法不可行的问题，尤其在大规模系统中，传统基于统一权重的优化策略难以有效平衡多个相互竞争的目标。其核心解决方案是提出一种基于微分博弈（differential games）的“分而治之”（Divide and Conquer）控制设计方法：将每个控制目标关联至一个虚拟输入，并构建一个非合作型、有限或无限时域的微分博弈框架，其中各代表玩家分别优化针对自身目标、系统全状态及其它虚拟输入的虚拟代价函数，同时考虑其他参与者的最优策略。通过求解该博弈的纳什均衡（Nash Equilibrium），合成出一种复合控制器，实现对多个冲突目标的稳定协调。该方法为控制工程师提供了直观且模块化的参数调优机制，适用于复杂动态系统的分阶段设计流程。论文进一步给出了连续与离散时间系统的严格数学推导，并开发了一个开源Python工具包，用于求解无限时域微分博弈中出现的耦合代数黎卡提方程（Coupled Algebraic Riccati Equations），并通过倒立摆小车和分层控制四旋翼两个基准案例验证了该方法在轨迹跟踪精度与多目标鲁棒调节性能方面显著优于经典线性二次型调节器（LQR）。
链接: https://arxiv.org/abs/2609.01838
作者: Joshua Shay Kricheli
机构: 未知
类目: Multiagent Systems (cs.MA)
备注: Thesis submitted in partial fulfillment of the requirements for the Master of Sciences degree, The Department of Computer Science, The Faculty of Natural Sciences, Ben-Gurion University of the Negev, Israel
Abstract:We introduce a novel Divide and Conquer control design methodology leveraging differential games in single-agent, multi-objective dynamical systems. The proposed framework associates each control objective with a virtual input and establishes a non-cooperative, finite or infinite horizon differential game among representative players. Each player optimizes a distinct virtual cost function tailored to its specific goal, the full system state, and the other virtual inputs, while accounting for the remaining players’ optimal policies. By establishing a Nash Equilibrium for this game, we synthesize a composite controller that achieves a stable balance across competing objectives, providing control engineers with an intuitive and modular framework for parameter re-tuning throughout the design cycle. We provide formal mathematical derivations for both continuous-time and discrete-time dynamical systems, targeting large-scale single-agent applications where complex, dynamically conflicting control objectives make global weighting intractable. To demonstrate the methodology, we developed an open-source Python package implementing a novel numerical algorithm for solving Coupled Algebraic Riccati Equations arising in infinite-horizon differential games. We evaluate the approach on two benchmark case studies: an inverted pendulum on a cart and a non-linear hierarchically controlled quadrotor. The resulting closed-loop performance is compared against the classical Linear Quadratic Regulator (LQR) across various transient and steady-state control metrics, demonstrating superior trajectory tracking and robust multi-objective regulation.
[MA-13] Agents That Model Agents : Five Principles Toward a Theory of Mind for 6G Networks
【速读】：该论文旨在解决未来6G网络中基于大语言模型（Large Language Model, LLM）代理的无线接入网（Radio Access Network, RAN）管理架构所面临的认知一致性与幻觉传播风险问题。当前架构假设代理间通信传递的是客观事实，但实际消息本质上是发送方推理过程的主观痕迹——即使语法正确，也可能蕴含由生成式AI产生的幻觉（hallucination），进而引发不可见于协议验证的级联故障。其核心解决方案在于引入心智理论（Theory of Mind, ToM），要求接收方在行动前对对方信念及其应有信念进行建模，从而识别并抑制非理性信息传播。通过将代理间的交互建模为细胞层（cellular sheaf）上的认知通道，提出一个统一的鲁棒多智能体系统框架，并衍生出五项设计原则：（i）消息是发送方隐含推理的证据；（ii）信任为连续的认知信噪比（cognitive Signal-to-Noise Ratio, SNR），即断言精度与偏离建模信念程度之比；（iii）全网一致性与抗幻觉传染性可通过层的拉普拉斯算子计算；（iv）信念建模需严格限制在两级以节省计算资源并避免互信息衰减；（v）可信容量受限于任务目标对齐度，而非链路带宽。基于本地部署的10亿参数电信语言模型的信令风暴实验验证了该框架的有效性：认知SNR可识别出被多数邻居错误支持的幻觉节点，发散门将错误个体排名高于正确者；唯有两层深度的心智理论能恢复正确决策；而谱间隙（spectral gap）决定了拓扑结构是否能在近实时预算内达成一致性。
链接: https://arxiv.org/abs/2609.01779
作者: Hatim Chergui,Carolina Fernández-Martínez,Mehdi Bennis,Merouane Debbah
机构: 未知
类目: Networking and Internet Architecture (cs.NI); Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注: 8 pages, 3 figures, 3 tables
Abstract:Future 6G networks will rely on Large Language Model (LLM) agents to manage the Radio Access Network (RAN). However, current architectures assume inter-agent messages convey objective facts. A message is instead a \emphtrace of the sender’s reasoning: it carries a subjective conclusion, so a syntactically valid report can propagate an AI hallucination and trigger a cascading outage invisible to protocol validation. Reading such a trace requires a Theory of Mind (ToM)—before acting, the receiver must model what the peer believes, and what a peer in that position should have believed. Modeling these interactions as cognitive channels on a cellular sheaf, we obtain a unified framework for resilient multi-agent systems, from which five design principles emerge: (i) a message is evidence of the sender’s hidden reasoning; (ii) trust is a continuous cognitive Signal-to-Noise Ratio (SNR)—asserted precision over deviation from the modeled peer belief; (iii) network-wide consistency and resistance to hallucination contagion are computable via the sheaf’s Laplacian; (iv) peer-modeling must halt at exactly two levels to conserve compute and survive mutual information decay; and (v) credible capacity is bounded by operational goal alignment, not link bandwidth. A signaling-storm study on locally deployed 1B-parameter telecom language models validates it: cognitive SNR isolates a hallucinating peer that three of its four neighbors agree with, where a divergence gate ranks every wrong peer above the right one; only depth two ToM recovers the correct action; and the spectral gap decides whether a topology reaches consistency inside the near-real-time budget.
[MA-14] Harness Engineering in LLM Tool Use via Agent -Native Reusable Tool Primitives
【速读】：该论文旨在解决大语言模型（LLM）在调用外部工具时面临的两大核心问题：一是由于工具输出类型与API接口模式不兼容，导致多步、多轮推理过程脆弱且难以持续；二是当工具库规模庞大时，模型性能显著下降。其解决方案的关键在于提出工具原语（Tool Primitives），即以自然语言作为工具调用的统一接口，将每个工具封装为具备内部模式解析与执行能力的LLM接口，从而实现工具间的自然语言通信，支持嵌套和多轮调用。在此基础上，构建了ToolFace——一个包含25,519个函数的中心化工具仓库，允许模型在推理时动态检索相关工具，避免在上下文中枚举原始API模式，显著降低上下文负担。为进一步保障复杂场景下工具调用的可靠性，论文还提出了HEART框架，一个基于代理原生（Agent-native）、可复用工具原语的工程化调度系统，由规划器（Planner）、路由器（Router）和验证器（Verifier）组成，协同完成动态工具规划、多步执行与反馈驱动的容错恢复。实验表明，HEART在多个基准测试中平均优于监督微调（SFT）模型10%，并超越GPT-5.4、Claude-4.6-Sonnet和Gemini-3.1-Pro约6%，同时将API调用成本降低高达85%；在50个真实任务中实现84%的任务完成率，是三款前沿商业模型平均成功率（22%）的3.8倍。
链接: https://arxiv.org/abs/2609.01736
作者: Haibo Jin,Suijin Wang,Xucheng Yu,Haojing Luo,Haohan Wang
机构: University of Illinois at Urbana-Champaign (伊利诺伊大学香槟分校); Starc Institute
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG); Multiagent Systems (cs.MA)
备注: 21 pages
Abstract:Large language models (LLMs) augmented with external tools have demonstrated remarkable capability in solving complex real-world tasks. However, existing approaches suffer from two key challenges: brittle multi-step and multi-turn reasoning caused by incompatible tool output types and API schemas, and performance degradation under large tool catalogues. To address these, we introduce \textbfTool Primitives, a design that replaces rigid API schema-based invocation with natural language as the interface for tool calling, where each tool is wrapped with an LLM interface that handles schema resolution and execution internally, enabling natural inter-tool communication for nested and multi-turn tool calling. Building on Tool Primitives, we host \textbfToolFace, a centralized repository of 25,519 functions from which LLMs dynamically retrieve only the relevant tools at inference time, eliminating the need to enumerate raw API schemas in context. To orchestrate Tool Primitives and ToolFace reliably in complex settings, we further propose \textbfHEART, a \textbfHarness \textbfEngineering framework via \textbfAgent-native, \textbfReusable \textbfTool Primitives, comprising a Planner, Router, and Verifier that jointly support dynamic tool invocation planning, multi-step execution, and feedback-driven recovery. Experiments on five benchmarks demonstrate that HEART outperforms SFT-based models by 10% on average and surpasses GPT-5.4, Claude-4.6-Sonnet, and Gemini-3.1-Pro by 6% on average while reducing API cost by up to 85% . On 50 real-world tasks, HEART achieves 84% task completion, 3.8\times the average of three frontier commercial models ( 22% ). Comments: 21 pages Subjects: Software Engineering (cs.SE); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG); Multiagent Systems (cs.MA) Cite as: arXiv:2609.01736 [cs.SE] (or arXiv:2609.01736v1 [cs.SE] for this version) https://doi.org/10.48550/arXiv.2609.01736 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[MA-15] PRISM: An Agent ic Multi-Model Architecture for Proactive Safety in Autonomous Transportation Systems
【速读】：该论文旨在解决当前自动驾驶与智能交通系统在复杂城市环境中对弱势道路使用者（Vulnerable Road Users, VRUs）安全保护不足的问题，尤其针对现有高级驾驶辅助系统（ADAS）普遍采用的“反应式”风险应对机制——即仅在危险事件发生后才启动干预——所导致的滞后性与局限性。其核心挑战在于如何实现从被动避撞向主动、持续的风险管理转变，以有效降低日益增长的VRU伤亡率。为此，论文提出PRISM（Proactive Risk Intelligence and Safety Management）架构，其关键创新在于构建了一个基于多模型协同的代理式（agentic）安全体系，通过逆向碰撞概率建模将传统的二元碰撞分类器转化为可解释的动态安全评分；系统集成轨迹动力学、环境风险与VRU交互三类专用模型，并由融合强化学习、上下文记忆与特征级归因的推理层进行实时协调。该架构支持分层级的安全干预策略（从静默监控到紧急告警），并能根据实时场景动态调整安全阈值，突破了传统规则系统静态阈值的限制。在三个自然驾驶数据集共1,296个场景中的验证表明，系统平均安全得分为68/100，77.6%的场景被识别为需预警，近事故率3.8%，其中11%触发干预或应急响应，且特征归因一致指出轨迹风险与VRU接近度为首要安全影响因素，充分证明其在密集城市环境中实现可解释、前瞻性交通安全管理的有效性。
链接: https://arxiv.org/abs/2609.01623
作者: Joyjit Roy,Samaresh Kumar Singh,Sushanta Das
机构: American Center for Mobility(美国汽车中心); Independent Researcher(独立研究员)
类目: Multiagent Systems (cs.MA); Computer Vision and Pattern Recognition (cs.CV); Emerging Technologies (cs.ET); Machine Learning (cs.LG); Physics and Society (physics.soc-ph)
备注: 15 pages, 7 figures, 8 tables. Accepted for presentation at ASCE2027: The Infrastructure and Engineering Experience, Philadelphia, PA, March 1-5, 2027. Extends SafeDriver-IQ ( arXiv:2603.14841 )
Abstract:Autonomous and intelligent transportation systems operate in complex urban environments where safety depends on interactions among vehicle behavior, environmental conditions, and vulnerable road users (VRUs) such as pedestrians and cyclists. Most advanced driver assistance systems (ADAS) employ reactive mechanisms that activate only after hazards have emerged, a critical limitation underscored by rising VRU fatalities in the United States. This study introduces PRISM (Proactive Risk Intelligence and Safety Management), an agentic multi-model safety architecture that transitions from reactive crash avoidance to proactive, continuous risk management. PRISM employs inverse crash-probability modeling to convert binary crash classifiers into dynamic, interpretable safety scores. Three specialized models addressing trajectory kinematics, environmental risk, and VRU interaction operate concurrently, coordinated by a reasoning layer incorporating reinforcement learning, contextual memory, and feature-level attribution. The system provides graduated safety interventions across four tiers, from silent monitoring to emergency alerts. Unlike rule-based systems with static thresholds, PRISM dynamically adjusts safety parameters in real time. Validated across 1,296 scenarios from three naturalistic driving datasets without dataset-specific retraining, the system yielded a mean safety score of 68 out of 100, classified 77.6% of scenarios as advisory, and flagged a near-miss rate of 3.8%, with 11% of scenarios escalating to intervention or emergency response. Feature attribution consistently identified trajectory risk and VRU proximity as primary safety factors. PRISM provides a unified, interpretable framework for proactive transportation safety with emphasis on VRU risk reduction in dense urban environments. Comments: 15 pages, 7 figures, 8 tables. Accepted for presentation at ASCE2027: The Infrastructure and Engineering Experience, Philadelphia, PA, March 1-5, 2027. Extends SafeDriver-IQ (arXiv:2603.14841) Subjects: Multiagent Systems (cs.MA); Computer Vision and Pattern Recognition (cs.CV); Emerging Technologies (cs.ET); Machine Learning (cs.LG); Physics and Society (physics.soc-ph) Cite as: arXiv:2609.01623 [cs.MA] (or arXiv:2609.01623v1 [cs.MA] for this version) https://doi.org/10.48550/arXiv.2609.01623 Focus to learn more arXiv-issued DOI via DataCite
自然语言处理
[NLP-0] User Feedback Provides a Unique Signal that LLM s Can not Detect
【速读】： 该论文旨在解决当前大型语言模型（Large Language Models, LLMs）在利用用户交互反馈进行优化时面临的有效性质疑问题。尽管已有研究认为用户反馈存在固有的噪声，难以有效用于模型改进，但本文通过实证表明，用户反馈实际上是一种高度可操作的学习信号，其“无效性”源于现有评估范式中存在的系统性偏差。解决方案的关键在于：通过构建具有明确真实答案的合成数据与自然场景下的真实数据相结合，验证反馈在不同情境下的有效性；并揭示评估偏差的根源——当模型仅依赖反馈成功修复问题时，大语言模型评判者（LLM judges）往往无法识别真正修正后的输出，反而偏好表现较差的基线输出。这一发现揭示了评估机制本身对反馈价值的误判，为更准确地利用用户反馈提供了理论依据和方法路径。
链接: https://arxiv.org/abs/2609.02859
作者: Shachar Don-Yehiya,Leshem Choshen,Omri Abend
机构: The Hebrew University of Jerusalem (耶路撒冷希伯来大学); IBM Research (IBM 研究院); MIT (麻省理工学院); MIT-IBM Watson AI Lab (麻省理工学院-IBM 沃森人工智能实验室)
类目: Computation and Language (cs.CL)
备注:
Abstract:Harnessing naturally occurring feedback from user interactions offers a promising learning signal for Large Language Models (LLMs). However, recent studies suggest this feedback is inherently noisy and difficult to leverage effectively. We challenge this conception by demonstrating that user feedback is a highly actionable signal for improvement, and that its perceived ineffectiveness stems from a systematic bias in current evaluation paradigms. To isolate the usefulness of feedback, we construct synthetic data with a definitive ground truth, alongside naturalistic data to validate that our findings hold in real-world scenarios. By comparing model revisions generated with and without access to feedback across both settings, we show that feedback-informed revisions resolve targeted issues at significantly higher rates than baseline revisions. Finally, we expose the root of the evaluation bias: when a model successfully fixes an issue exclusively due to feedback, LLM judges frequently fail to identify the genuinely corrected response, systematically preferring inferior baseline outputs instead.
[NLP-1] DiscoSign: Discourse-Aware Text to Sign Language Gloss Translation EMNLP2026
【速读】： 该论文旨在解决传统手语处理系统仅在句子层面进行翻译，忽视了对理解手语至关重要的篇章现象这一关键问题。其解决方案的核心在于提出一个基于语言学研究的模块化大语言模型（LLM）框架——DiscoSign，专门针对三个核心篇章现象进行建模：（i）空间指代消解（spatial coreference resolution），确保实体在篇章中保持一致的空间位置；（ii）问答小句（Question-Answer Clauses, QACs），即具有特定篇章功能的伪分裂结构；（iii）概念-词素一致性（concept-gloss consistency），保障英文概念与美国手语（ASL）手势之间稳定映射。为克服传统翻译评估指标无法衡量篇章层面质量的局限，研究引入一套新型评估指标，分别量化各篇章连贯维度的表现。实验结果表明，相较于仅依赖句子级别的翻译方法，该框架显著提升了空间一致性与实体追踪能力，同时保持了与现有单句翻译相当的高质量表现，首次建立了系统性的篇章级文本到手语词素翻译框架及其配套评估方法。
链接: https://arxiv.org/abs/2609.02796
作者: Vasileios Baltatzis,Mert Inan,Connor Gillis,Raja Kushalnagar,Lorna Quandt,Leah Findlater,Colin Lea
机构: Apple(苹果); Northeastern University (东北大学); Gallaudet University (加劳德特大学)
类目: Computation and Language (cs.CL)
备注: Accepted at EMNLP 2026 Main Conference
Abstract:Sign language processing systems have traditionally operated at the sentence level, ignoring critical discourse phenomena fundamental to sign language comprehension. We introduce DiscoSign, a computational approach for discourse-aware text to sign language gloss translation grounded in linguistic research. We address three key phenomena within our modular Large Language Model (LLM)-based translation framework: (i) spatial coreference resolution, where entities maintain consistent spatial locations throughout discourse; (ii) Question-Answer Clauses (QACs), pseudocleft structures serving specific discourse functions; and (iii) concept-gloss consistency, ensuring stable mappings between English concepts and American Sign Language (ASL) signs. Traditional translation metrics fail to capture discourse-level quality, so we introduce a suite of novel evaluation metrics designed to assess each dimension of discourse coherence addressed by our framework. Experiments on sentence-level and discourse-level datasets show that our approach for discourse-aware processing significantly improves spatial consistency and entity tracking relative to sentence-only translation, while maintaining competitive single-sentence gloss translation quality. Our work establishes the first systematic framework for discourse-level text to sign language gloss translation with corresponding evaluation methodology.
[NLP-2] EarlyEval: Cheaper Agent Evaluation via Early Outcome Prediction
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）智能体评估成本高昂的问题，尤其在迭代开发过程中，前沿模型在代理基准测试上的单次运行成本可达数百至数千美元，且需反复执行。现有方法主要依赖基准蒸馏以减少评估任务数量，但未降低每个保留任务的执行成本。本文提出“早期结果预测”（early outcome prediction）这一互补性效率提升策略，核心在于通过分析智能体的中间行为，在任务完成前即预判其最终结果，从而提前终止无望的任务执行。其关键创新在于构建名为EarlyEval的轻量级框架，利用LightGBM训练成功与失败两类分类器，基于行为特征、文本内容及参考解等多维度输入，当任一分类器达到校准后的置信度阈值时立即终止智能体运行，每步开销可忽略不计。在SWE-bench Verified、TerminalBench和Toolathlon三个基准上，EarlyEval可减少13%–26%的智能体步骤，最多节省44.1%的输入令牌与29.4%的输出令牌，同时保持89%–97%的预测准确率，对单个智能体的解决率影响仅平均下降1–2个百分点。
链接: https://arxiv.org/abs/2609.02783
作者: Yuling Shi,Zhensu Sun,Junsen Dong,Chengcheng Wan,David Lo,Xiaodong Gu
机构: Shanghai Jiao Tong University (上海交通大学); Singapore Management University (新加坡管理大学); East China Normal University (华东师范大学); Shanghai Innovation Institute (上海创新研究院)
类目: Computation and Language (cs.CL)
备注: Code and data available at this https URL
Abstract:Evaluating LLM agents is essential for guiding their development, yet it has grown prohibitively expensive: a single pass of a frontier model over an agentic benchmark can cost hundreds to thousands of dollars, a price paid repeatedly across iterative development cycles. Prior efforts, centered on benchmark distillation, reduce the number of evaluation tasks but leave the cost of executing each retained task untouched. In this work, we introduce early outcome prediction, a complementary axis of efficiency that instead cuts cost within each task. Our key insight is that an agent’s final outcome is often evident from its intermediate behavior well before execution completes. We instantiate this idea in EarlyEval, a lightweight framework that trains a pair of LightGBM success and failure classifiers over behavioral, textual, and reference-solution features, and halts an agent run the moment either classifier crosses a calibrated confidence threshold, adding negligible per-step overhead. Across three benchmarks, SWE-bench Verified, TerminalBench, and Toolathlon, EarlyEval can eliminate 13%-26% of agent steps and up to 44.1% input tokens and 29.4% output tokens at 89%-97% prediction accuracy, while perturbing per-agent resolve rates by only one to two percentage points on average.
[NLP-3] ShallowStream: Index Shallow then Answer Deep for Streaming Video Understanding
【速读】： 该论文旨在解决多模态大语言模型（Multimodal Large Language Models, MLLMs）在处理连续视频流时计算开销过大这一核心问题，尤其针对现有方法普遍忽视模型深度维度导致的重复全深度预填充（prefill）所引发的高延迟与键值缓存（KV cache）线性膨胀问题。其解决方案的关键在于提出ShallowStream框架，通过利用MLLM的浅层（shallow layers）实现帧编码与检索索引构建的并行化：在流式处理过程中，始终维护一个基于浅层KV缓存的轻量级索引；在查询阶段，借助浅层生成的注意力分数对上下文帧进行评分，并采用兼顾多样性的选择策略，高效精准地召回相关证据。该设计显著降低了每帧预填充延迟和端到端响应时间，相较最强基线方法分别提升52.1倍和11.9倍，同时保持相当的性能水平。
链接: https://arxiv.org/abs/2609.02780
作者: Jitai Hao,Ke Yang,Qiang Huang,Jun Yu
机构: 未知
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注: Work in Progress
Abstract:Streaming video understanding is a critical capability for real-world applications, including embodied intelligence, autonomous driving, industrial monitoring, surveillance and early warning, and wearable assistants. However, processing continuous video streams with multimodal large language models (MLLMs) is computationally expensive. Existing efforts have explored reducing streaming overhead through visual token pruning, token merging, quantization, on-demand frame retrieval, and context offloading. However, most existing methods overlook the dimension of model depth. Repeatedly executing full-depth MLLM prefill over incoming frames is prohibitively expensive, incurring substantial computational overhead and causing the KV cache to grow at a rate directly proportional to the prefill depth. To address these challenges, we propose ShallowStream, a novel framework that leverages the shallow layers of an MLLM to simultaneously perform frame encoding and retrieval index building. During stream processing, ShallowStream maintains an always-on lightweight index using the KV cache of shallow layers. During query-time answering, we leverage the attention scores generated by the shallow layers to score context frames and employ a diversity-aware selection strategy to retrieve precise and comprehensive evidence. ShallowStream achieves performance on par with the strongest existing streaming methods, while reducing per-frame prefill latency and 10-second end-to-end latency by up to 52.1x and 11.9x, respectively. Our code is available at this https URL.
[NLP-4] HyperStyler: Low-resource Authorship Style Transfer via Context-aware Style Navigation and Hypernetworks EMNLP2026
【速读】： 该论文旨在解决低资源作者风格迁移（Low-resource Authorship Style Transfer, LAST）中的核心挑战：在仅有少量目标作者参考样本的情况下，如何同时实现高风格保真度与原文语义的精确保留。现有方法通常将多样化的参考文本压缩为单一静态的作者嵌入表示，导致上下文相关的风格差异被平均化；同时依赖隐含表示进行风格控制，造成风格与内容的耦合。本文提出的HyperStyler架构通过解耦风格选择与风格实现两个阶段来突破上述瓶颈：其中，风格导航器（Stylo-navigator）联合建模源文本上下文与目标作者参考样本，以预测动态的风格坐标；风格超网络（Stylo-hypernet）则通过动态参数调制而非隐藏状态注入的方式实现风格表达，从而实现更灵活、解耦的风格控制。实验结果表明，HyperStyler在Reddit、Blog和News数据集上均显著优于现有方法，包括基于大语言模型（LLM）的方法，并具备出色的跨领域泛化能力；尤其值得注意的是，其仅需比T5-large多出2.4%的参数量，且推理速度超过LLMs 1.8倍，展现出卓越的效率与性能平衡。
链接: https://arxiv.org/abs/2609.02772
作者: Jongkyung Shin,Minguk Jeon,Chanwoo Park,Chiehyeon Lim
机构: UNIST(韩国科学技术院); POSTECH(浦项科技大学); POSCO Holdings Inc.(浦项控股公司)
类目: Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026 (Main)
Abstract:Low-resource authorship style transfer (LAST) aims to rewrite text into the style of an arbitrary target author using only a few reference examples while preserving the original meaning. Existing methods often struggle to achieve both high style fidelity and semantic preservation because they compress diverse references into a single static author embedding, which averages out context-dependent stylistic variation, and rely on hidden representations for style control, which entangle style with content. We propose HyperStyler, a novel architecture that decouples LAST into style selection and style realization. Stylo-navigator predicts style coordinates by jointly modeling the source context and target-author references, and Stylo-hypernet realizes them via dynamic parameter modulation instead of hidden-state injection. Our experiments on Reddit, Blog, and News datasets demonstrate that HyperStyler consistently outperforms prior methods including LLM-based approaches and generalizes robustly across domains. Notably, HyperStyler achieves superior performance with as few as 2.4% additional parameters over T5-large, while being over 1.8x faster than LLMs at inference.
[NLP-5] From Reweighting to Rewriting: Unlocking the Intervention Effects of Influential Samples in Training Data Attribution
【速读】： 该论文旨在解决训练数据归属（Training Data Attribution, TDA）中干预价值评估的局限性问题，即现有方法在选择关键训练样本后，采用传统重加权（reweighting）策略难以实现显著且持续的行为改变。其核心挑战在于：尽管影响函数（Influence Functions, IF）能有效识别对模型行为具有高影响力的训练样本，但基于重加权的干预手段往往无法充分释放这些样本的潜在干预潜力。为此，论文提出了一种影响引导的响应重写（influence-guided response rewriting） 方法，其关键创新在于：利用IF识别出具有高影响力的目标样本，并在保持指令不变的前提下，通过替换其输出响应为与目标行为一致或对立的监督信号（behavior-aligned/opposed supervision），从而实现更有效的干预。实验结果表明，在四款开源大语言模型上，响应重写相较于重加权能够产生更强、更持久且可双向调控的行为变化；进一步分析显示，由IF选出的样本在重写策略下展现出显著更高的干预杠杆效应，且行为改变高度集中于目标相关维度。该研究揭示了影响估计所捕捉的局部重加权效应与被识别样本实际具备的广泛干预潜力之间的本质差异，强调应发展更具干预意识的TDA评估范式。
链接: https://arxiv.org/abs/2609.02771
作者: Yuzhang Luo,Chenpeng Wang,Jianhui Chen,Liangming Pan
机构: Peking University (北京大学); YIXIN (亿心); Beijing Academy of Artificial Intelligence (北京人工智能研究院)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Training data attribution (TDA) aims to identify training examples that shape model behavior, but its intervention value depends on both which examples are selected and how they are modified. Influence functions (IF) estimate behavioral changes under infinitesimal reweighting, yet IF-selected examples often show limited advantages over random selection under conventional weight-based interventions. This raises the question of whether influential examples lack intervention value or whether reweighting fails to realize their behavioral this http URL introduce influence-guided response rewriting, which uses IF to identify intervention targets and replaces their responses with behavior-aligned or behavior-opposed supervision while keeping instructions fixed. Across four open-weight LLMs, we compare rewriting and reweighting on the same influence-selected examples using epistemic abstention as our primary testbed. Response rewriting produces stronger, more persistent, and bidirectional behavioral shifts, while reweighting the same examples yields weak and inconsistent effects. Further analyses show that influence-selected examples provide greater rewriting leverage than alternative selectors, with changes remaining concentrated on target-relevant behaviors. The same qualitative contrast extends to safety refusal. These results distinguish the local reweighting effects captured by influence estimates from the broader intervention leverage of the examples they identify, motivating intervention-aware evaluation of TDA methods.
[NLP-6] Untangling the Mechanisms of Misleading Context in Medical Question Answering ML4H2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在医疗问答中因误导性上下文（misleading context）导致判断偏差的问题。其核心挑战在于：尽管模型具备专家级表现，但其推理过程易受虚假证据或单纯断言等误导性线索影响，进而产生错误决策。解决方案的关键在于揭示并监控此类误导对模型推理机制的影响——研究发现，模型对“断言”类误导线索的敏感性远高于“虚构证据”，且前者在推理过程中更隐蔽、披露率更低（仅7–90%），而后者虽被广泛披露（81–98%），但其影响路径不同：虚构证据早期介入并逐步累积影响，而断言则在推理末期直接扭转结论。通过分析开放模型的完整推理轨迹（reasoning trace），研究构建了一个基于大语言模型的监测器（LLM monitor），可在5%假阳性率下识别出78%的被误导决策，显著优于仅依赖输出响应的检测方法（最高仅32%）。因此，关键突破在于：只有通过可访问的、完整的推理轨迹，才能有效识别最隐蔽且最具危害性的误导性断言，而这一能力目前被前沿模型提供商所限制，构成当前生成式医疗系统可信性与可解释性的重大瓶颈。
链接: https://arxiv.org/abs/2609.02754
作者: Robin Linzmayer,Noémie Elhadad
机构: Columbia University (哥伦比亚大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 25 pages, 10 figures. Submitted to ML4H 2026
Abstract:Large language models now answer medical questions with expert-level performance. However, the context these systems act on can be misleading, and misleading context can corrupt a model’s medical judgment. To understand how misleading context corrupts this judgment, we examine the model’s susceptibility to the context, disclosure of it, mechanism of corrupted reasoning, and monitorability of the decision. On the medical reasoning subset of MedMisBench, a clinician-reviewed question-answering benchmark of 8,627 questions, we inject two types of misleading context cues, fabricated evidence and a bare assertion. We test three reasoning models, two that expose their full reasoning trace and one frontier model that exposes only its response. All three are more susceptible to the assertion than to the fabricated evidence, adopting the asserted answer 10 to 27 points more often. The misleading cues are disclosed in 81 to 98% of traces but only 7 to 90% of responses, and the assertion is disclosed less often than evidence based cues. Resampling from reasoning traces without disclosure shows the two cues corrupt reasoning differently, evidence entering early and accumulating while the assertion redirects the conclusion near its end. An LLM monitor catches 78% of corrupted decisions at 5% false positives when reading an open model’s trace with guidance, against at most 32% from any response. The misleading context that models are most susceptible to is disclosed least, and was caught reliably only from an open reasoning trace, which frontier providers withhold.
[NLP-7] Repo-To-Skill: Distilling GitHub Repositories Into AI4AI Skills
【速读】： 该论文旨在解决当前自主研究代理在执行机器学习（Machine Learning, ML）研究任务时，因缺乏可复用的“操作性知识”（operational knowledge）而导致效率低下与性能受限的问题。操作性知识指将理论方法转化为实际有效执行的能力，这类知识虽广泛存在于开源库和论文中，但以人类可读的形式存在，体积庞大且难以直接加载至任务执行流程中。为突破此瓶颈，论文提出一种基于技能（skill）驱动的研究代理DisCo，其核心解决方案在于通过两种互补的提炼机制：一是任务无关型提炼，将领域内广泛使用的开源仓库中的知识压缩为通用、可验证的技能；二是任务导向型提炼，针对具体研究任务生成专用技能。前者构建了包含5000余项经验证技能的AREX-Skill Library，覆盖20个研究领域与178个能力类别。实验表明，在固定模型架构（GPT-5.5）、研究框架与执行预算的前提下，引入技能后的代理在MLE-bench、PaperBench、FrontierCS和PassNet等基准上分别提升134.3%、34.4%、9.2%和14.0%，显著优于无技能版本。这一性能跃升源于在固定系统中注入经过提炼的操作上下文，实现了对领域知识的高效复用与精准调用。
链接: https://arxiv.org/abs/2609.02749
作者: Jianlyu Chen,Yuyang Hu,Hongjin Qian,Jiawei Liu,Wenqing Wei,Xiaolong Chen,Defu Lian,Zhicheng Dou,Chaozhuo Li,Qiwei Ye,Zheng Liu
机构: Beijing Academy of Artificial Intelligence(北京人工智能学院); University of Science and Technology of China(中国科学技术大学); Renmin University of China(中国人民大学); Hong Kong Polytechnic University(香港理工大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 48 pages, 3 figures
Abstract:Autonomous agents are beginning to carry out machine-learning (ML) research end to end. These agents combine a model backbone with a harness for planning, execution, memory, and verification, but this architecture still leaves domain-specific know-how outside the agent. We call this missing layer operational knowledge, the know-how that separates knowing a method from making it work. That knowledge is not absent from the field. It appears in repositories and papers, but in forms written for human readers and too large to load during a task. Once distilled into compact, verified skills, this knowledge can be reused across tasks rather than rediscovered during each run. We present DisCo, a skill-powered research agent that creates skills and uses them during research. Its distillation runs in two complementary forms: task-agnostic, condensing the field’s widely used repositories into reusable skills, and task-oriented, producing the skills a concrete task calls for. The former, applied across the open ecosystem, yields the AREX-Skill Library, with 5,000+ verified skills distilled from 1,000 widely used ML repositories and organized into 20 areas and 178 capability families. With the GPT-5.5 backbone, research harness, and downstream execution budget held fixed, the skill-equipped research agent scores 134.3% higher on MLE-bench, 34.4% higher on PaperBench, 9.2% higher on FrontierCS, and 14.0% higher on PassNet than the same agent without skills. These gains come from adding distilled operating context under that fixed setup. Comments: 48 pages, 3 figures Subjects: Artificial Intelligence (cs.AI); Computation and Language (cs.CL) Cite as: arXiv:2609.02749 [cs.AI] (or arXiv:2609.02749v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2609.02749 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-8] Language Models Can Control Their Own Attention
【速读】： 该论文旨在解决大语言模型在长上下文生成中因全局注意力机制导致的计算效率低下问题：尽管模型仅关注上下文中的少数关键片段，却仍需遍历整个键值缓存（KV cache）进行注意力计算，造成显著的冗余开销。现有方法通过轻量级代理分数预筛选相关令牌以缓解此问题，但仍存在每步O(N)的时间复杂度。本文提出一种内在式解决方案——声明式注意力（Declarative Attention, DA），其核心思想是利用模型自身在思维链（chain-of-thought）中对相关信息位置的隐含认知能力，主动“声明”其在生成过程中所需关注的上下文区域。DA将生成过程划分为三种模式：全局（全上下文）、聚焦（特定区域）和局部（仅近期输出），推理引擎据此解析这些声明并跳过大部分不必要的KV缓存读取。在15个长上下文任务的零样本评估中，基于现成模型（Gemma-4-31B、Qwen-3.6-27B）的DA显著减少了解码过程中的总注意力访问量（分别降低52.0%和31.1%），同时仅带来微小的性能下降（1.27个百分点、2.75个百分点），且随着模型规模增大，误差进一步缩小。该方法开辟了稀疏注意力的新维度，为未来基于训练优化的更高效策略提供了探索空间。
链接: https://arxiv.org/abs/2609.02737
作者: Namgyu Ho,Huzama Ahmad,Woosung Koh,Se-Young Yun,Tal Schuster,Cicero Nogueira dos Santos
机构: KAIST AI(韩国科学技术院人工智能); Google DeepMind(谷歌深度思维)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Language models spend most of their attention on a small fraction of context, yet they read the entire KV cache to find the few tokens that matter. If the user asks about a previous detail in a 1M-token conversation, global attention layers must scan the full context to generate each token of the reply. A prominent approach mitigates this cost by pre-selecting relevant tokens via lightweight proxy scores, but this extrinsic scoring still incurs O(N) per step. We take an intrinsic approach motivated by the simple question: wouldn’t the model already know which parts of the context are relevant? To this end, we introduce Declarative Attention (DA), a protocol that elicits the model to declare where it needs to attend within its chain-of-thought, partitioning generation into three modes: global (full context), focus (a specific region), and local (recent output only). The inference engine parses these declarations like tool calls and skips most of the KV cache read. Under zero-shot evaluation across 15 long-context tasks, DA on off-the-shelf models (Gemma-4-31B, Qwen-3.6-27B) significantly reduces total attended tokens during decoding (52.0%, 31.1%) with modest accuracy drops (1.27pp, 2.75pp) that shrink with model scale. DA unlocks a new axis of sparse attention, with further potential under training-based methods that future work can explore.
[NLP-9] Choosing a PEFT Variant for Per-Patient Dysarthric ASR: A Single-Speaker Case Study on Two ASR Bases
【速读】： 该论文旨在解决针对特定患者（特别是存在严重构音障碍的中风后患者）的自动语音识别（ASR）系统在个性化适配中的性能优化问题，重点在于评估参数高效微调（PEFT）方法在个体化、单说话人场景下的有效性与实用性。其核心挑战在于如何在保持低计算与存储开销的前提下，实现对重度构音障碍语音的高精度识别。解决方案的关键在于系统性比较七种基于低秩适配器（LoRA家族）的方法（包括LoRA、QLoRA、AdaLoRA、DoRA、LoHA、VeRA和VB-LoRA），并验证注意力投影层适配器在提升词错误率（CER）方面的显著优势。研究发现，在两个主流模型基座（Whisper-large-v3与多语言Qwen3-ASR-1.7B）上，尽管全量微调表现最优，但仅需115 MB参数量的LoRA适配器（同时适配前馈网络块）即可达到接近全微调的效果（误差仅0.66个百分点），且存储成本仅为全微调的约3.7%。此外，实证表明4比特（NF4）QLoRA在该规模下并未带来预期的内存节省反而性能下降，而其他变体如LoHA虽有一定增益，但未超越标准LoRA。最终选择更简单、经济的LoRA作为推荐方案，并通过6点注册网格实验揭示约5分钟患者录音可捕获超过45%的零样本到30分钟的性能提升，为临床部署提供了关键的轻量化训练策略。
链接: https://arxiv.org/abs/2609.02735
作者: Bernard Muller,László Tóth,LaVonne Roberts
机构: The Scott-Morgan Foundation (斯科特-摩根基金会); University of Szeged (塞格德大学)
类目: Computation and Language (cs.CL); Sound (cs.SD)
备注: 2 figures. Submitted to Speech Communication (Elsevier)
Abstract:Per-patient adapters are the preferred production architecture for dysarthric automatic speech recognition (ASR), yet parameter-efficient fine-tuning (PEFT) variants have not been compared in the speaker-dependent, per-patient regime. We present a single-speaker case study comparing seven LoRA-family methods (LoRA, QLoRA, AdaLoRA, DoRA, LoHA, VeRA, VB-LoRA) on two production bases (Whisper-large-v3 with Hungarian fine-tuning, and a multilingual Qwen3-ASR-1.7B checkpoint) for one post-stroke Hungarian male speaker (S1, 409 utterances; severe dysarthria on auditory-perceptual clinical assessment). Attention-projection adapters substantially improve CER on both bases. Across three seeds, a paired bootstrap detects no significant LoRA-DoRA difference (p0.5; 13.86/13.90 % CER on Whisper, 28.10/28.33 % on Qwen3-ASR), so we adopt the simpler, cheaper LoRA. Real 4-bit (NF4) QLoRA is worse on every seed and both bases (14.56/30.09 % CER) with no memory saving at this scale, and LoHA, VeRA, VB-LoRA and AdaLoRA do not reach the LoRA family, though LoHA still gives an 18.6 % relative CER reduction on Whisper. On the same base, full fine-tuning is more accurate (11.43 % CER), but a 115 MB LoRA that also adapts the feed-forward blocks reaches within 0.66 pp of it at approximately 3.7 % of the per-patient storage. A 6-point enrollment grid shows about 5 min of patient audio captures 45.6 % of the zero-shot-to-30-min CER reduction, with further gains at 10 and 30 min (caveat: one speaker, one language, severe post-stroke dysarthria). Training scripts and recipes will be released, source-available under a research-use licence, on publication.
[NLP-10] CORAL: An LLM -Native Harness for Production Recommender Systems RECSYS’26
【速读】： 该论文旨在解决生产环境推荐系统在持续演化过程中性能衰退的问题，即随着内容、用户行为及上游模型的动态变化，传统依赖人工工程师通过在线实验（A/B测试）进行优化的方法存在响应滞后、工程成本高、覆盖范围有限等瓶颈，难以实现对推荐系统中检索、排序与服务策略的持续迭代。其解决方案的关键在于提出CORAL（Constraint-Optimized Recommender via an Agentic Loop），一种基于大语言模型（LLM）原生构建的闭环智能体系统：该系统以持续循环的方式运行，每轮周期中，智能体观测实时运营信号，结合历史决策与结果的记忆进行推理，并调用工具（包括数值优化器以确保变更在固定资源预算内）来动态重构推荐策略；其优化过程被建模为一个部分可观测、非平稳且受约束的优化问题，策略通过上下文学习不断改进，无需参数更新。在两个大规模社交平台上的A/B实验验证表明，该框架可在不增加服务成本的前提下提升用户参与度，或在不降低参与度的情况下降低服务成本，有效跨越“参与度-效率”权衡边界，且性能随迭代次数提升，证明单一智能体闭环可自动化完成以往需人类算法工程师在明确约束下执行的持续优化任务。
链接: https://arxiv.org/abs/2609.02730
作者: Muhammad Rafay Azhar,Yuhang Zhou,Gilbert Jiang,Yuchen Wang,Rahul Sharma,Matthew DeSousa,Jiayi Liu,Xin Guo,Lizhu Zhang,Xiangjun Fan
机构: Meta AI(元宇宙人工智能)
类目: Computation and Language (cs.CL)
备注: Accepted by RecSys '26 OARS Workshop
Abstract:Production recommender systems shape what billions of people see, and sustaining their performance requires continual optimization: as content, user behavior, and upstream models shift, the choices governing retrieval, ranking, and serving must be revisited. Traditionally, human engineers test such changes through online experiments–a slow, reactive process limited by engineering effort, leaving parts of the system unrevised as conditions change. Although large language models have been applied to ranking, user modeling, and offline model development, few systems place an agent in a continual closed loop that acts on a live recommender and learns from the measured effects of its decisions. We present CORAL (Constraint-Optimized Recommender via an Agentic Loop), an LLM-native harness that closes this loop: each cycle, the agent observes operating signals, reasons over a memory of past decisions and outcomes, and invokes tools–including a numerical optimizer that keeps changes within a fixed operating budget–to reconfigure the recommender, with measured outcomes informing the next cycle. We formulate this as a partially observed, non-stationary, constrained optimization problem in which the policy improves in context, without parameter updates, from its prior actions. Across two large-scale social platforms, evaluated with A/B experiments, the same harness improves engagement at no additional serving cost on one and reduces serving cost without degrading engagement on the other, spanning the engagement-efficiency frontier. Performance improves as the loop iterates, suggesting that a single agentic loop can automate continual optimization work traditionally performed by human algorithm engineers under explicit guardrails.
[NLP-11] Door-in-the-Face Requests and Refusal Behaviour in Large Language Models
【速读】： 该论文旨在探究“门在脸前”（door-in-the-face）这一人类心理影响策略是否适用于大语言模型。在人类行为中，先提出一个被拒绝的大请求，随后跟进较小的请求，可显著提升后者被接受的概率。研究通过在九个来自三家厂商的生产级模型上进行实验，测试当模型拒绝一个较大请求后，再提出相同主题的较小请求时，其合规率是否高于直接提问。结果表明，该策略的效果因模型而异：在Anthropic的前沿模型Opus 5上，该技术有效，小请求的响应率为65.8%，远高于直接提问的29.3%；而在OpenAI与Google的前沿模型以及Haiku 4.5上，该策略反而降低合规率15.5至23.0个百分点。控制实验进一步验证，这种效应源于对同一主题请求的拒绝本身，而非无关话题的拒绝，说明“让步”行为在所有模型中均具影响，但模型对自身拒绝行为的后续反应存在家族差异。此外，该策略无法迁移至公开基准数据集中的拒绝案例。研究发现，决定“撤回策略”能否生效的关键因素在于请求内容的本质——将265个被拒绝的指令类请求重构为同主题的解释性请求后，其中263例成功消除拒绝。因此，人类影响力技巧在语言模型中的适用性具有高度模型家族特异性，需逐模型分析。
链接: https://arxiv.org/abs/2609.02707
作者: Til Jordan
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 28 pages (9 pages of content plus references and appendix), 5 figures, 9 tables. Preprint, under review
Abstract:Does the door-in-the-face technique work on language models? In humans, a large request that is refused makes a smaller follow-up request more likely to be granted. We test this on nine production models from three providers: each model refuses a large request, then receives a smaller version of the same request, and we compare its compliance with asking directly. The answer depends on the model. On Anthropic’s frontier models the technique works: Opus 5 answers the smaller request 65.8% of the time after refusing the larger one, against 29.3% when asked directly. On the frontier models of OpenAI and Google, and on Haiku 4.5, it backfires, lowering compliance by 15.5 to 23.0 points. A control locates the effect: a refused large request on an unrelated topic does less than the related one on all nine models, so the concession itself matters everywhere, while the reaction to having just refused something differs by model family. The technique does not transfer to refusals drawn from public benchmarks. What decides whether a retreat can work is what the request asks for: rewriting 265 refused requests for usable instructions into requests for explanations of the same topic removed the refusal in 263 cases. Human influence techniques port to language models one model family at a time.
[NLP-12] race as State: Reasoning Traces as Conditional States for Long-Context Transformers
【速读】： 该论文旨在解决因果性Transformer模型在处理长上下文推理任务时存在的根本性瓶颈：即当前模型按顺序（自回归）处理信息，而某些任务状态需依赖后续信息才能确定，导致早期推理阶段无法利用最终获得的关键状态信息。为形式化这一不匹配问题，作者引入“条件状态更新任务”进行分析，发现对于因果状态更新处理器而言，在输入末尾提供条件信息可能导致最坏情况下内存需求呈指数级增长，而在开头提供则可显著降低开销。针对此问题，论文提出“迹作为状态”（Trace as State）的解决方案，其核心在于将先前推理过程中收集到的推理轨迹（reasoning traces）作为任务状态的文本代理，并在重新读取长上下文前将其置于输入序列的起始位置，从而允许早期推断结果指导后续的重读过程。与作为对照的“迹附加”（Trace Append）方法——将相同轨迹置于上下文末尾——相比，该方案在三个模型和三个长上下文数据集上的27组实验中，有26组表现更优；例如在GraphWalks Parents任务中，DeepSeek V4 Pro Preview的精确匹配率从初始的29.2%提升至81.8%，而GLM-5.2更是实现从66.4%到100.0%的突破性提升。实验表明，尽管保持了原始因果Transformer结构，仅通过将推理迹前置即可显著增强长上下文推理能力，关键在于重构信息流顺序以弥合因果处理与任务状态依赖之间的鸿沟。
链接: https://arxiv.org/abs/2609.02702
作者: Xu Zou,Jie Tang
机构: Z.ai; Tsinghua University (清华大学)
类目: Computation and Language (cs.CL)
备注: preprint
Abstract:Transformers process information causally, but long-context reasoning may depend on task state discovered only later. We formalize this mismatch through conditional state update tasks. For causal state update processors, providing the condition first can require exponentially less memory in the worst case than providing it last. Motivated by this principle, we introduce Trace as State. We use collected reasoning traces as a textual proxy for task state and place it before the long-context block on a fresh pass, allowing information derived previously to guide rereading. We conduct extensive experiments on Trace as State and Trace Append, a matched control that uses the same task state proxy but put it after the context. Across three models and three long-context datasets, Trace as State outperforms Trace Append in 26 of 27 reported combinations of model, task, and metric. On GraphWalks Parents, exact match lifts DeepSeek V4 Pro Preview from 29.2% on the initial pass and 43.0% with Trace Appendto 81.8% with Trace as State, and from 66.4% and 83.2% to 100.0% for GLM-5.2. These results show that placing traces before the context can improve long-context reasoning while retaining the causal transformer structure. Comments: preprint Subjects: Computation and Language (cs.CL) Cite as: arXiv:2609.02702 [cs.CL] (or arXiv:2609.02702v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2609.02702 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-13] DKL: Decoupled Knowledge Learning for Instruction-Tuned Language Models
【速读】： 该论文旨在解决生成式 AI（Generative AI）在使用检索增强生成（RAG）技术时，因检索失败或不完整而导致的事实性错误（即幻觉）问题。现有方法如微调（fine-tuning）虽能提升知识注入效果，但需大量合成问答对；而基于扩展预训练（EPT）的方法虽避免了合成数据生成，却会损害指令微调语言模型（Instruct LLM）的指令遵循能力，进而需要额外的指令微调（IFT），但后者成本高昂且依赖可用的指令微调语料库。为克服上述局限，本文提出一种轻量级解决方案——解耦知识学习（DKL, Decoupled Knowledge Learning）。其核心在于：不对指令微调语言模型直接进行扩展预训练，而是将其对应的基座语言模型（base LLM）进行扩展预训练以注入新知识，随后通过模型合并（model merging）将知识权重迁移至指令微调语言模型中，从而在不破坏其指令遵循能力的前提下实现知识注入。实验表明，该方法在检索失败场景下将RAG准确率从54.17%提升至79.26%，且显著减少所需训练数据量，优于已有方法。
链接: https://arxiv.org/abs/2609.02685
作者: Kushagra Bhushan,Meghanadh Pulivarthi,Sai Krishna Reddy Sathi,Gaurav Pandey,Sonam Gupta,Vineet Kumar,Jaydeep Sen,Yatin Nandwani,Sachindra Joshi,Dinesh Raghu
机构: IBM; Indian Institute of Technology, Madras(印度理工学院马德拉斯分校)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 20 pages, 4 figures, 15 tables
Abstract:RAG has become the de facto method for incorporating new, corpus-specific knowledge into an instruction following LLM (Instruct LLM). Although RAG-based prompting improves factual grounding, it fails when retrieval is incorrect or incomplete, leading to hallucinations. Finetuning methods such as RAFT and PA-RAG enhance RAG by injecting new knowledge into the model’s parameters, but require generating a massive amount of synthetic QA that covers the entire corpus. Extended Pre-Training (EPT) on the text corpus avoids the need for comprehensive synthetic data generation but compromises an Instruct LLM’s instruction-following capabilities, necessitating instruction fine-tuning (IFT) after pre-training. However, IFT is costly and may be infeasible due to the unavailability of an instruction-tuning corpus. In this work, we propose DKL-Decoupled Knowledge Learning for Instruction-Tuned Language Models. Instead of doing EPT on the Instruct LLM, DKL performs EPT on its corresponding base LLM to infuse new knowledge. These knowledge infused weights are then merged with the Instruct LLM, imparting new knowledge without affecting their instruction-following capabilities. DKL is a lightweight method that avoids expensive instruction fine-tuning and relies on model merging to infuse the new knowledge into the Instruct LLM without destroying its instruction following capabilities. Empirical results show that DKL improves RAG accuracy from 54.17 to 79.26 on retrieval failure cases, while outperforming prior approaches with substantially less training data.
[NLP-14] From Tokens to Semantics: Leverag ing Complementary Signals for Hallucination Detection in Black-Box LLM s
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在面向公众或高风险应用场景中，因生成内容存在幻觉（fabrication）而带来的用户与机构风险，同时避免因误报（false alarms）过度消耗有限的人工审核资源。其核心挑战在于：当缺乏可信上下文或参考文档时，如何有效识别模型输出中的错误。解决方案的关键在于利用黑盒模型API可访问的两类信号——语义熵（semantic entropy，衡量采样响应间语义分歧程度）与基于词元概率的不确定性（token-level uncertainty），并设计融合机制以互补二者缺陷。具体而言，提出TopK方法通过聚合多轮采样结果中的词元级信号增强不确定性检测；引入混合方法CoCoA，结合目标响应的不确定性与语义差异性；进一步提出两种监督学习方法：Gated通过门控机制将单一语义簇情形交由聚合词元特征分类器处理，Stacked则联合学习语义不确定性与更广泛的词元特征。实验评估覆盖七项基准（含五项公开数据集及两项自建数据集），涵盖四种语言模型，在不同假阳性率预算（1%–15%）下验证性能鲁棒性，并分析生成策略与校准敏感性及数据集特性的影响。结果表明，无监督方法如TopK和CoCoA虽无需标注数据但需精细调参，而监督方法Stacked在近半数场景中表现最优，揭示不存在普适最优方案，强调了根据实际需求进行方法选择与配置的重要性。
链接: https://arxiv.org/abs/2609.02679
作者: Urja Pawar,Rajitha Ramanayake,Owen O’Neill,Nabeel Kemal,Abhishek Mandal,Houssem Chatbri,Christopher Martin
机构: BNY
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:When LLMs support public-facing or high-stakes workflows, missed fabrications can harm users and institutions, while false alarms consume limited human-review capacity. When no trusted context or reference document is available, we study two signals accessible through black-box model APIs: semantic entropy, which measures disagreement among sampled response meanings, and uncertainty derived from token log-probabilities. Their failure modes can be complementary: semantic entropy becomes uninformative when responses form one semantic cluster, while token uncertainty can miss consistently confident errors. We extend token-based uncertainty detection by aggregating token-level signals across sampled responses through our TopK method, evaluate the hybrid CoCoA method, which combines target-response uncertainty with semantic dissimilarity, and propose and study two supervised methods: Gated, which routes single-cluster cases to an aggregated-token-feature classifier, and Stacked, which learns jointly from semantic uncertainty and broader token features. We evaluate seven benchmarks, including five public benchmarks (four text datasets and multimodal handwritten-cheque extraction) and two constructed benchmarks (Financial Summaries and Long-Text QA), using four language models. In our evaluation across models and datasets, Stacked gave the best performance in nearly half of the cases, while TopK and CoCoA remain competitive without supervised training labels, although their thresholds require careful calibration. No method is universally strongest. We therefore evaluate performance at false-positive-rate budgets from 1% to 15%, assess their sensitivity to generation and calibration choices, and examine variation across dataset characteristics.
[NLP-15] oHC: Orthogonal Hyper-Connections on SO(4) via Quaternions
【速读】： 该论文旨在解决超连接（Hyper-Connections, HC）架构中因残差流间混合矩阵无约束而导致的训练不稳定性问题。传统HC使用可学习的n×n残差矩阵在每层对n条并行残差流进行混合，但若该矩阵未加限制，其缩放因子会随层数累积放大，引发梯度爆炸或消失。为解决此问题，已有方法如流形约束超连接（Manifold-constrained Hyper-Connections, mHC）将矩阵限制于双随机矩阵集，从而将最大缩放因子控制在1以内，防止残差流被放大；然而该约束仅限制上界，下界无约束，导致混合过程持续压缩各流间的差异，使残差流随深度趋同，丧失多样性。针对此缺陷，本文提出正交超连接（Orthogonal Hyper-Connections, oHC），将残差矩阵约束于旋转群SO(n)，确保混合操作在所有方向上既不放大也不衰减残差流，从而维持训练稳定性并保留流间差异。在实际实现中，针对近期HC模型常用的四流结构，oHC采用单位四元数对SO(4)进行显式参数化，无需额外参数，以固定符号加法模式替代迭代投影，计算效率高于mHC。实验表明，oHC在多种下游任务中均显著优于单流残差基线、mHC及恒等矩阵约束的iHC，验证了其有效性与优越性。
链接: https://arxiv.org/abs/2609.02672
作者: Haoqiang Guo,Xuyi Chen,Bo Ke,Yishu Lei,Ziyang Xu,Shikun Feng,Ximen,Wenhan Luo
机构: 未知
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Hyper-Connections (HC) replace the single residual stream of a Transformer with n parallel ones, mixing them at every layer with a learned n \times n residual matrix. Leaving that matrix unconstrained places no limit on the factor by which the mixing step rescales the residual streams, and that factor compounds across layers, which destabilizes training. Manifold-constrained Hyper-Connections (mHC) address this by restricting the matrix to the doubly stochastic matrices. That caps the factor at one, so the mixing can no longer amplify any direction, but nothing bounds it from below. We prove that inside this set the mixing step can reduce the norm of the residual streams only by shrinking the differences between the streams, while their mean is left unchanged; and since the reduction accumulates over layers, the streams grow more alike and their diversity is spent with depth. We therefore propose Orthogonal Hyper-Connections (oHC), restricting the residual matrix to the rotation group SO(n) , so that the mixing step can neither amplify nor attenuate the residual streams in any direction, which keeps training stable and no longer forces the differences between the streams to contract. Specifically, at the four streams used by recent HC models we parameterize the group in closed form by a pair of unit quaternions, which adds no parameters, replaces the iterative projection with a fixed pattern of signed additions, and can be constructed faster than mHC. We evaluate oHC across a comprehensive set of downstream tasks, where it outperforms the single-stream residual baseline, mHC and iHC, which fixes the residual matrix to the identity.
[NLP-16] WinoQueer-NL: Assessing Bias in Dutch Language Models toward LGBTQ Identities
【速读】： 该论文旨在解决荷兰语语言模型中针对酷儿群体（queer）的隐性偏见缺乏系统评估的问题，尤其针对英语语料库基准（如WinoQueer）在荷兰语文化语境下的适用性不足。其核心解决方案在于构建一个经过文化与语言适配的荷兰语酷儿偏见评估数据集，基于英文WinoQueer基准设计，包含成对的刻板印象与反刻板印象句子，并通过43名荷兰酷儿参与者的在线调查验证其文化相关性，识别出22项新发现的偏见。最终发布的数据集包含42,906条句子，采用日志似然比（log-likelihood ratio）作为偏见度量指标，对多种荷兰语专用及多语言模型（包括掩码语言模型MLM和自回归语言模型ARLM）进行评估。关键发现表明，尽管整体平均偏见得分接近中性（约50%），但模型在不同酷儿身份上的表现存在显著差异：部分模型对跨性别身份的刻板印象句选择高达97%的频率，而对同性恋相关表述仅6%，且跨性别与非二元性别相关表述始终呈现最高偏见得分。研究强调了基于本土文化语境构建评估数据集的重要性，以精准识别并缓解对边缘化群体造成不成比例影响的语言模型偏见。
链接: https://arxiv.org/abs/2609.02651
作者: Jiska Beuk,Gerasimos Spanakis
机构: Maastricht University (马斯特里赫特大学)
类目: Computation and Language (cs.CL)
备注: under review, dataset available via this https URL
Abstract:While English language models have been widely examined for anti-queer bias, Dutch models remain understudied. To address this gap, we developed a culturally and linguistically adapted Dutch dataset based on the English WinoQueer benchmark, containing pairs of stereotypical and counter-stereotypical sentences. To validate and expand it, we conducted an online survey with 43 Dutch queer participants, confirming 145 of 171 stereotypes as culturally relevant and identifying 22 new biases through free-text responses. The final released dataset, comprising 42,906 sentences, was evaluated using a range of Dutch-specific and multilingual models, including both masked language models (MLMs) and autoregressive language models (ARLMs), with bias measured via a score comparing log-likelihoods of stereotypical versus counter-stereotypical sentences. While the mean bias score across models appeared neutral (~50%), closer analysis revealed significant disparities: some models favored stereotypical sentences up to 97% of the time for transgender identities, but only 6% of the time for gay-related pairs, with transgender and non-binary identities consistently receiving the highest bias scores. Our findings highlight the importance of culturally grounded datasets for evaluating and mitigating biases that disproportionately impact marginalized groups in Dutch language models.
[NLP-17] Loom: Weaving Diagnostic Strands into Free-Text Consensus via Embedding-Space Reweighting EMNLP2026
【速读】： 该论文旨在解决在真实工业场景中部署自然语言处理（NLP）系统时，如何将大量噪声大、相互冲突的开放形式文本假设（open-form hypotheses）有效聚合为可靠共识的核心挑战。现有方法存在明显局限：单一的大语言模型（Large Language Model, LLM）代理虽具备强大的表达能力，适用于根因分析（Root Cause Analysis, RCA）等任务，但受限于上下文长度、幻觉加剧以及推理延迟过高；而传统弱监督方法虽具统计严谨性，却仅适用于离散类别，难以应对开放式推理需求。本文提出Loom——一种生成式共识框架，其核心创新在于通过将模块化启发式规则（诊断模板，动态填充事件特定实体、时间与指标）生成的开放形式假设投影至连续嵌入空间，并采用基于迭代质心重加权的算法化解析冲突信号，最终以共识权重驱动轻量级LLM的合成步骤。该方案在OpenRCA基准测试中实现了精度-效率的帕累托最优：在Bank和Market-2上达到顶尖自主代理性能，在Market-1和Telecom上略有差距，但所有数据集均仅需单次LLM调用（相比基线快约26倍，使用8B参数合成器时提速约33倍）。关键解决方案在于将生成式推理与确定性共识机制相结合，既保留了生成式AI（Generative AI）的灵活性，又通过可解释的共识过程增强了领域专家（Subject Matter Experts, SMEs）的信任感。
链接: https://arxiv.org/abs/2609.02649
作者: Ron Begleiter,Katya Egert Berg,Gilad Saban,Gil Shabat
机构: NVIDIA(英伟达); Tel Aviv, Israel
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: Accepted to EMNLP 2026
Abstract:Aggregating noisy, conflicting textual hypotheses into a reliable consensus is a fundamental challenge when deploying NLP systems in real-world industrial settings. While monolithic Large Language Model (LLM) agents offer unbounded expressivity for tasks like Root Cause Analysis (RCA), they suffer from context limits, compounding hallucinations, and prohibitive inference latency. Traditional weak supervision offers statistical rigor but is mathematically restricted to discrete classes. We present Loom, a generative consensus framework deployed for real-world RCA that bridges these paradigms. Loom aggregates open-form hypotheses emitted by modular heuristics (diagnostic templates dynamically populated with episode-specific entities, times, and metrics) by projecting them into a continuous embedding space, and resolves conflicting signals with an iterative centroid-based reweighting algorithm. The resulting consensus weights ground a single lightweight LLM synthesis step. Evaluated on the OpenRCA benchmark, Loom occupies the accuracy–efficiency Pareto frontier: it matches a state-of-the-art autonomous agent on Bank and Market-2 and trails on Market-1 and Telecom, while using a single LLM call per incident on all four datasets ( \sim 26 \times faster; \sim 33 \times with an 8B-parameter synthesizer). We discuss our deployment experience, highlighting lessons learned regarding the trade-offs between agentic depth and inference latency, negative results in redundancy detection, and how deterministic consensus fosters trust among Subject Matter Experts~(SMEs).
[NLP-18] aRA: Training-Aware Low-Rank Adaptation Initialization EMNLP2026
【速读】： 该论文旨在解决低秩适配（LoRA）在参数高效微调（PEFT）中因低秩分解导致的信息瓶颈问题，其性能对初始化高度敏感。现有方法虽通过利用预训练权重、激活或梯度的主成分来构建高质量初始化，但未能直接考虑全秩模型的训练动态。本文提出一种面向训练过程的低秩适配初始化方法——训练感知低秩适配初始化（TaRA），其核心在于使低秩因子所诱导的梯度在训练初期尽可能逼近对应全秩权重矩阵的真实梯度，从而提升初始阶段的梯度保真度。该方法基于数学推导实现，计算开销极小，在多种复杂微调任务中均显著优于现有最先进方法，展现出简单、鲁棒且可扩展的优势，为高效LoRA初始化提供了新的解决方案。
链接: https://arxiv.org/abs/2609.02639
作者: Taehyeon Kim,Eunhyeok Park
机构: Pohang University of Science and Technology (POSTECH)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Accepted to the EMNLP 2026 Main Conference
Abstract:Low-Rank Adaptation (LoRA) has become a de facto standard for parameter-efficient fine-tuning (PEFT), yet its performance is highly sensitive to initialization due to the information bottleneck imposed by low-rank decomposition. Existing approaches attempt to construct high-quality LoRA initializations by exploiting principal components of pretrained weights, activations, or gradients. However, these methods do not directly account for the training dynamics of the full-rank model. In this paper, we propose Training-aware Low-Rank Adaptation Initialization (TaRA), a method that initializes LoRA such that the gradients induced by the low-rank factors closely approximate the gradient of the corresponding full-rank weight matrix. Derived from a mathematical formulation, TaRA improves gradient fidelity at the start of training while introducing negligible computational overhead. Across diverse and challenging fine-tuning tasks, TaRA consistently outperforms prior state-of-the-art methods, establishing a simple, robust, and scalable solution for effective LoRA initialization.
[NLP-19] Scalable Direction-Following TTS via Voice Impression-Guided Pseudo Triplet Construction INTERSPEECH2026
【速读】： 该论文旨在解决语音合成中“遵循表演指令”的问题，即在保持说话人身份和语言内容不变的前提下，根据给定的表演方向（如“更激动”或“更缓慢”）生成语调或风格上有所调整的新语音。其核心挑战在于缺乏真实标注的三元组数据（参考语音、方向描述、修改后语音），难以训练模型准确理解并实现风格迁移。为此，作者提出一种可扩展的伪三元组（pseudo-triplet）构建流程，通过可控印象风格的生成式语音合成模型生成具有受控风格差异的语音，并利用大语言模型（LLM）从估计的印象差异中自动生成自然语言方向描述，从而构建大规模伪三元组数据。实验表明，仅使用伪三元组即可实现稳定的说话人保真度风格修改；结合真实录制数据后，方向对齐效果进一步提升，同时维持良好的说话人相似性。该方法有效缓解了高质量标注数据稀缺的问题，为灵活、可控的语音风格编辑提供了可行方案。
链接: https://arxiv.org/abs/2609.02623
作者: Kenichi Fujita,Yusuke Ijima
机构: 未知
类目: ound (cs.SD); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 5 pages,4 figures, Accepted to INTERSPEECH 2026
Abstract:Voice actors often re-read the same script while modifying their delivery in response to performance directions. We study this setting as direction-following TTS, where a system generates a new utterance that reflects a given direction relative to a reference utterance while preserving speaker identity and linguistic content. A key challenge is the lack of training data capturing such relative modifications. To address this, we propose a scalable pseudo-triplet construction pipeline that generates~(reference utterance, direction text, modified utterance) triplets. It generates controlled style variations using an impression-controllable TTS model and uses an LLM to produce natural language directions from estimated impression differences. Experimental results demonstrate that pseudo-triplets alone enable stable speaker-preserving modification, and that combining pseudo and recorded data further improves direction alignment while maintaining speaker similarity. Audio examples are available on our demo page this https URL
[NLP-20] Predictors of Loneliness in Older Adults Using Multimodal Analysis of Speech and Language
【速读】： 该论文旨在解决老年人孤独感（loneliness）在自然对话情境中缺乏可扩展、客观的检测方法这一关键公共卫生问题。现有手段难以在不依赖主观报告的情况下，有效识别孤独情绪的言语与语音特征。其解决方案的关键在于构建一个融合语言与声学特征的多模态分析框架：通过结合心理语言学词典、n-gram模型与主题建模等文本特征，以及音高、语调、音量等声学特征，系统性地分析310名老年人在半结构化电话访谈中的语言行为。研究发现，孤独感水平与否定表达、负面语调及冲突相关语言显著正相关，而较低孤独感则与社交提及、动机性词汇及情感丰富性呈负相关；更重要的是，多模态模型（相关系数r = 0.298）显著优于仅基于文本或仅基于音频的模型，表明言语内容与语音表现共同构成孤独感的复合表征。研究支持将基于语音的分析作为心理评估的辅助工具，用于早期识别情绪性孤独，而非独立诊断手段。
链接: https://arxiv.org/abs/2609.02606
作者: Vinmay Khandode,Sai Karthik Kosuri,Neil K. R. Sehgal,Adam Greene,Elif Alpoge,Elana Duffy,Matthew Lee Smith,Thomas K.M. Cudjoe,Sharath Chandra Guntuku
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:Loneliness is a critical public health issue among older adults, linked to higher risks of depression, cognitive decline, and mortality. Scalable, objective methods for its detection remain limited, particularly in natural conversational contexts. We analyzed speech and language markers of loneliness in 310 older adults using semi-structured telephone interviews to help understand how they process feeling lonely and how their language differs at different levels of feeling loneliness. Our multimodal framework combined linguistic features (psycholinguistic dictionaries, n-grams, and topic models) with acoustic features (pitch, tone, loudness) to examine associations with self-reported loneliness scores. Both predefined and data-driven methods captured patterns in verbal content and vocal delivery. Higher loneliness was associated with negations(r = 0.11), negative tone(r = 0.12), and conflict-related language. Lower loneliness was linked to social references(r = -0.18), motivational drives(r = -0.11), and emotional richness in speech(r = -0.12). We also found that the multimodal model (r = 0.298) outperforms the text-only and audio-only models. Findings suggest that loneliness manifests through both linguistic and acoustic cues, supporting the potential of speech-based analysis in psychological assessments and as an early indicator of emotional loneliness when used alongside existing assessments, rather than as standalone diagnostic tools.
[NLP-21] When Persona Attributes Improve Population Alignment in Large Language Models
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在预测调查问卷中人类参与者回答时表现不一致的问题，尤其是现有研究中关于角色提示（persona prompting）效果混杂且缺乏明确规律的现象。其核心挑战在于如何有效选择用于构建角色描述的属性（persona attributes），以提升模型生成响应与真实人类响应之间的相关性。论文提出，调查问题中人类回答的固有变异程度（observed human response variation）可能是导致当前性能波动的关键解释因素。为此，研究系统比较了多种属性选择方法在不同社会调查任务中的表现，覆盖两个国家、四种通用社会调查、六种主流LLM及每项调查下的二十个预测任务。结果表明，角色提示的有效性取决于问题的人类响应变异性以及属性选择策略的适配性，从而为判断角色提示在调查预测中的适用场景提供了实证依据，并揭示了不同属性选择方法的相对有效性，推动了对生成式人工智能（Generative AI）在社会科学预测应用中可控性和可解释性的理解。
链接: https://arxiv.org/abs/2609.02526
作者: Leon Fröhling,Jens Rupprecht,Markus Strohmaier,Claudia Wagner
机构: GESIS – Leibniz Institute for the Social Sciences(德国社会科学研究机构); University of Mannheim(曼海姆大学); Complexity Science Hub(复杂科学中心); RWTH Aachen University(亚琛工业大学)
类目: Computation and Language (cs.CL); Computers and Society (cs.CY)
备注: 45 pages, 15 figures
Abstract:Large Language Models (LLMs) are increasingly used to predict the responses of human participants in survey panels. Towards that goal, persona prompting has recently emerged as a technique to inform and align large pretrained language models. Persona prompting refers to the practice of using short textual descriptions of ‘personas’ in prompts to steer the LLM’s generations. Personas describe individuals through different attributes such as their socio-demographics, attitudes, or behaviors, with the aim of aligning LLMs to produce responses that correlate with the corresponding human responses. Yet, recent work has produced mixed and partly conflicting results of persona prompting without clear patterns of success and failure. Among the few consistent findings is that the selection of persona attributes matters, and that using more attributes does not necessarily lead to better performance. It remains unclear how different attribute selection methods perform and how to choose among them. In this paper, we propose that observed human response variation of a survey question is a potential explanation for the mixed performance observed so far. In addition, we compare the performance of persona prompting associated with different methods for selecting persona attributes. We evaluate these methods on four different (general) social surveys across two countries, six LLMs, and twenty prediction tasks per survey. Our work helps to identify when persona prompting can be expected to be useful in survey prediction tasks, and provides new insights on the effectiveness of different attribute selection methods for LLM-based survey prediction using persona prompting.
[NLP-22] Debias-SparseGPT : Bias-Aware Pruning for Large Language Models EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在采用权重稀疏化（weight sparsification）等模型压缩技术（如剪枝和量化）后，模型偏差被放大的问题。现有方法如SparseGPT虽能实现高效部署，但其输出对提示中的角色线索（persona cues）敏感，导致生成内容存在显著偏差。为应对这一挑战，论文提出一种后训练剪枝方法Debias-SparseGPT，其核心创新在于引入基于人口统计学对比输入的二阶表示去偏机制（representational debiasing with a second-order term over demographically contrasting inputs），在剪枝过程中主动抑制模型对特定群体的偏见。实验结果表明，在多种生成式LLM及不同稀疏度设置（25%、50%、结构化2:4稀疏）下，Debias-SparseGPT均能有效降低剪枝引发的偏差，同时保持模型困惑度（perplexity）和零样本准确率（zero-shot accuracy）不变；尤其在最严格的2:4结构化稀疏条件下，通过向校准集注入长上下文、内容丰富的样本，进一步提升了下游任务性能与公平性。总体而言，Debias-SparseGPT在不牺牲稀疏模型计算效率的前提下，显著优化了模型偏差-性能权衡。
链接: https://arxiv.org/abs/2609.02496
作者: Irina Proskurina,Guillaume Metzler,Antoine Gourru,Julien Velcin
机构: Laboratoire Hubert Curien, UMR CNRS 5516, Saint-Étienne, France(圣埃蒂安, 法国); Université Claude Bernard Lyon 1, Université Lumière Lyon 2, ERIC(克莱蒙-贝尔纳里昂第一大学, 里昂第二大学, 经济与管理研究机构); École Centrale de Lyon, LIRIS, CNRS UMR 5205(里昂中央理工学院, 利里斯实验室, 国家科学研究中心5205联合研究单位)
类目: Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026, Code: this https URL
Abstract:Model compression techniques such as pruning and quantization facilitate the efficient deployment and acceleration of Large Language Models (LLMs). However, recent studies show that weight sparsification methods, such as SparseGPT, can amplify existing biases in models, with outputs varying significantly depending on persona cues in the prompt. In this paper, we introduce Debias-SparseGPT, a post-training pruning method incorporating representational debiasing using a second-order term defined over demographically contrasting inputs. We perform empirical validation of our method over a wide range of generative LLMs. Across models and sparsity regimes (25%, 50%, and structured 2:4 sparsity), Debias-SparseGPT consistently reduces pruning-induced bias compared to SparseGPT while preserving model perplexity and zero-shot accuracy. Under the most restrictive 2:4 structured sparsity pattern, which most aggressively degrades model quality, augmenting the calibration set with long-context, content-rich examples further improves both downstream performance and fairness. Overall, Debias-SparseGPT advances the bias-performance trade-off while preserving the computational efficiency of sparse models.
[NLP-23] How LLM s Build Fictional Worlds: Setting and Narrative Space in AI-Generated Creative Storytelling
【速读】： 该论文旨在解决生成式 AI 在构建故事世界（storyworld）时与人类创作在叙事空间分布上的差异问题，尤其聚焦于“设定”（setting）这一可量化的维度。其解决方案的关键在于通过微调的 BERT 分类器，将叙事空间划分为五种类型——“行动空间”（action space）、“感知空间”（perceived space）、“视觉空间”（visual space）、“描述性空间”（descriptive space）和“无空间”（no space），并以此对 GPT 4.1、LlaMA 3.3、Mistral 3.2 和 Gemma 3 等模型生成的英文与德文故事，以及来自 Project Gutenberg 的人类创作小说进行系统比较。研究发现，人类作品主要依赖“行动空间”，强调角色与环境的具身互动，而大语言模型则系统性地过度生成“感知空间”，更侧重氛围与情感的营造。这种差异在叙事时间上保持稳定，且表现出模型特异性和语言敏感性特征，揭示了生成式 AI 在世界建构策略上的根本性偏离。
链接: https://arxiv.org/abs/2609.02482
作者: Katrin Rohrbacher,Björn Nieth,Emmanuelle Salin,Bjoern Eskofier,Michaela Mahlberg
机构: FAU Erlangen-Nürnberg(弗莱堡大学埃尔朗根-诺伊堡分校); Munich Center for Machine Learning (慕尼黑机器学习中心); LMU München(慕尼黑大学); Institute of AI for Health, Helmholtz Zentrum München(亥姆霍兹慕尼黑研究所人工智能健康研究所); University of Birmingham(伯明翰大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:In this paper, we analyze how Large Language Models (LLMs) employ worldbuilding strategies, focusing on setting as one measurable dimension of storyworld construction. We compare 1,000 AI-generated stories per model in English and German with human-authored fiction from Project Gutenberg. Building on prior work, we operationalize setting through five types of narrative space: “action”, “perceived,” “visual,” “descriptive” and “no space”, identified using fine-tuned BERT classifiers for German and English. We generate narratives using GPT 4.1, LlaMA 3.3, Mistral 3.2, and Gemma 3 and compare their spatial distributions to a human-authored baseline. We find that human-authored texts predominantly employ “action space,” grounding narratives in embodied character-environment interaction, whereas LLMs systematically overproduce “perceived space,” emphasizing atmosphere and affect. This divergence remains stable across narrative time. Overall, our findings show that LLMs exhibit worldbuilding patterns that differ consistently from human-authored fiction in ways that are both model-specific and language-sensitive.
[NLP-24] Prag Align: Feedback-Guided Prag matic Alignment for Controlled Synthetic Dialogue Generation
【速读】： 该论文旨在解决生成式对话在隐私受限服务场景中应用时，如何有效保持对话的交际意图、情感意义与自然对话流的问题。现有方法在单次生成（one-shot generation）或重复生成（repeated generation）中难以同时满足多维度质量要求，尤其在情感对齐方面表现不佳。其解决方案的核心是提出PragAlign框架，采用“生成—评估—修订”反馈引导循环机制，基于大语言模型（LLM）的评估器对意图一致性、情感一致性、连贯性、流畅性和整体质量进行多维度评分，并提供针对性反馈以支持最多三轮精细化优化。实验表明，该框架在800组对话规范下实现了99.50%的评估者定义通过率，显著优于单次生成（72.25%）和无结构反馈的重复生成（95.88%）。分析显示，迭代尝试主要提升整体生成成功率，而结构化反馈则关键作用于最后一公里的多约束协同满足，尤其是情感对齐的改善最为显著。尽管人工评估验证了意图表达与对话流的可识别性，但情感适切性仍存在主观性强、稳定性不足的问题，表明情感实现与独立人类感知质量仍是待突破的技术挑战。
链接: https://arxiv.org/abs/2609.02480
作者: Smitha Muthya Sudheendra,Jaideep Srivastava
机构: University of Minnesota, Twin Cities(明尼苏达大学双城分校)
类目: Computation and Language (cs.CL)
备注:
Abstract:Synthetic dialogue generation can support research in privacy-restricted service settings, but generated conversations must preserve communicative intent, affective meaning, and natural dialogue flow. We introduce PragAlign, a feedback-guided framework for controlled synthetic dialogue generation conditioned on service context, target intent, and target emotion, with auxiliary trait-style controls. PragAlign uses a generate–evaluate–revise loop in which an LLM-based evaluator scores intent alignment, emotion alignment, coherence, fluency, and aggregate quality, then provides criterion-specific feedback for up to three refinement rounds. On 800 matched dialogue specifications, PragAlign achieves 99.50% evaluator-defined acceptance, compared with 72.25% for one-shot generation and 95.88% for repeated generation without structured feedback. This indicates that repeated attempts account for much of the gain over one-shot generation, while structured feedback primarily improves last-mile multi-constraint satisfaction rather than broad average quality. Refinement gains are concentrated in emotion alignment, which is also the dominant failure mode in ablations. A separate human evaluation of 1,200 generated dialogues shows that intent expression and dialogue flow are highly recognizable to annotators, while emotion appropriateness is less stable and more subjective. These results support PragAlign as a quality-control framework for improving evaluator-defined communicative constraint satisfaction, while showing that affective realization and independent human-perceived quality remain open challenges.
[NLP-25] Learning to Fuse LLM s with Ontology Rankers for Rare-Disease Diagnosis
【速读】： 该论文旨在解决生成式AI在罕见病诊断中缺乏可追溯证据链的问题，即尽管大语言模型（Large Language Models, LLMs）能够基于患者表型描述生成差异性诊断，但其预测结果缺乏与疾病候选项对应的清晰证据支持。为此，论文提出一种基于行为的融合模型（behavior-based fusion model），其关键在于不替代传统本体排序器（ontology ranker），而是通过综合分析两种系统的排名列表、两者之间的共识程度以及每个候选疾病背后的本体支持度，动态学习在具体病例中对各系统应予的信任权重。该方法保留了本体排序器的结构化证据优势，同时利用LLMs的诊断能力提升整体性能。研究通过消除基准测试集中的文献泄露路径（test-set leakage pathway），验证了融合模型在Phenopacket Store和RAMEDIS数据集上分别将Phenomizer Recall@1提升7.86和20.18个百分点；即使仅使用其他7个开源LLM训练的融合模型，在接入DeepSeek-V4-Flash API后仍实现5.19个百分点的召回率增益，且无需重新训练。更重要的是，90.8%的正确融合诊断仍保留可审查的候选级本体证据，证明该方案能够在增强诊断能力的同时维持诊断工具的可解释性与可信度。
链接: https://arxiv.org/abs/2609.02473
作者: Zhaoyang Jiang,Zhizhong Fu,Yunsoo Kim,Zicheng Li,Xuanqi Peng,Fei Teng,Jiacong Mi,Honghan Wu
机构: University of Glasgow(格拉斯哥大学); University of Electronic Science and Technology of China(电子科技大学); Shanghai Jiao Tong University School of Medicine(上海交通大学医学院)
类目: Computation and Language (cs.CL)
备注:
Abstract:Ontology rankers remain useful for rare-disease diagnosis because each candidate can be traced to matched patient phenotypes. Large language models (LLMs) can generate differential diagnoses from the same patient description, but their predictions lack an equally clear evidence trail. Rather than asking which system should replace the other, we ask whether an LLM can improve the ranker without giving up its evidence. Our behavior-based fusion model examines the two ranked lists, their agreement, and the ontology support behind each candidate, and learns how much to rely on each system for the individual case. Before comparison, we remove a documented test-set leakage pathway caused by benchmark cases and ontology annotations being derived from the same publications. Across eight open LLMs, fusion improves Phenomizer Recall@1 by 7.86 percentage points on Phenopacket Store and 20.18 points on RAMEDIS. When paired with DeepSeek-V4-Flash through an API, a fusion model trained only on the other LLMs improves Recall@1 from 0.1657 to 0.2176, a 5.19-point gain, without retraining. For 90.8% of correct fused diagnoses, the disease retains candidate-level ontology evidence that can be inspected. These results show that LLMs can strengthen an established diagnostic tool without discarding the structured evidence that makes it useful.
[NLP-26] Scalable Kronecker-Fisher Approximation: Efficient Hessian Analysis for Billion-Parameter Language Models Compression
【速读】： 该论文旨在解决在千亿参数大模型中进行海森矩阵（Hessian）分析时，因完整计算和存储费舍尔信息矩阵（Fisher matrix）不可行而导致的可扩展性问题。其核心挑战在于如何在不显式存储整个费舍尔矩阵的前提下，有效捕捉跨层交互作用并实现对模型敏感性的精确评估。解决方案的关键在于提出一种基于克罗内克积（Kronecker-based）的可扩展近似方法，该方法通过低秩结构建模跨层相关性，从而在保持理论严谨性的同时显著降低计算与内存开销。实验结果表明，该方法能够准确识别出模型中对扰动最敏感的组件——如值投影层（value projection layers）表现出最高的敏感度和最强的跨层相关性，并在量化、稀疏化、层间破坏及灾后微调等任务中展现出与性能退化和恢复能力高度相关的预测能力。这一框架为大型模型的脆弱性分析提供了实用且理论支撑的工具，推动了混合精度分配、分层稀疏性设计以及跨层甚至单个权重组的自适应低秩分解等新型压缩与优化策略的发展。
链接: https://arxiv.org/abs/2609.02451
作者: Viacheslav Yusupov,Daria Cherniuk,Evgeny Frolov
机构: 未知
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:In this paper, we propose a scalable Kronecker-based approximation that captures cross-layer interactions without storing the entire Fisher matrix, enabling practical Hessian analysis for billion-parameter networks where full computation is infeasible. Our approach reveals consistent vulnerability patterns: value projection layers exhibit the highest sensitivity and strongest cross-layer correlations across multiple model families, while other components exhibit architecture-specific behaviors. Through extensive experiments on quantization, sparsification, inter-layer corruption, and post-corruption fine-tuning, we demonstrate that our approximation strongly correlates with both performance degradation and recovery. Our framework provides a practical, theoretically grounded tool for identifying fragile components in large models, opening new avenues for guided compression and optimization strategies, such as mixed-precision allocation, layer-wise sparsity, and adaptive low-rank decomposition across layers and even individual weight groups.
[NLP-27] When Decodability Is Not Enough: Logical Validity Representations Behavioral Dissociation and Causal Tests in Language Models
【速读】： 该论文旨在解决大语言模型在逻辑推理任务中看似具备推理能力，但其内部表征与行为输出之间是否存在一致性的问题。具体而言，研究关注的是：尽管模型在行为层面的表现接近随机水平，其隐藏状态中是否仍编码了逻辑有效性（logical validity）的信息。解决方案的关键在于通过设计匹配的“有效-无效”前提-结论对，在多种推理类型、语义领域、模板结构及难度层级下系统评估模型的内部表征。研究发现，逻辑有效性信息在隐藏状态中几乎可被完美解码，并且在未见模板、领域和推理家族条件下仍保持强可解码性；即使在行为错误的例子中，有效性信息也高度可解码。然而，通过留一法消融实验揭示出该解码能力存在明显局限，且基于探测方向的干预仅表现出微弱且非特异性的效果。这表明，逻辑有效性的内部表征、行为表达以及因果使用是相互独立的过程——有效性信息可在模型内部被强解码，却未必稳定地反映在输出行为中。
链接: https://arxiv.org/abs/2609.02438
作者: Smitha Muthya Sudheendra,Jaideep Srivastava
机构: University of Minnesota, Twin Cities(明尼苏达大学双城分校)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Large language models can look capable of logical reasoning, but correct or incorrect answers alone tell us little about what the model represents internally. We study logical verification in five open-weight transformer models using matched valid–invalid premise–claim pairs that vary across inference families, semantic domains, templates, and difficulty levels. Despite near-chance behavioral performance, logical validity is often almost perfectly decodable from hidden states and remains strongly decodable under held-out templates, domains, and inference families. Validity also remains highly decodable on behaviorally incorrect examples in the conditions where correctness-conditioned evaluation is well defined. At the same time, exhaustive leave-one-out tests reveal clear limits to this generalization, and interventions along probe-derived validity directions have only weak, nonspecific effects compared with random controls. Our results suggest that representing validity, expressing it in behavior, and using it causally are distinct. Validity related information can be strongly decodable from a model’s hidden states without being reliably expressed in its output.
[NLP-28] UTP-Bench: Uncertainty-aware Travel Planning Benchmark EMNLP2026
【速读】： 该论文旨在解决现有大语言模型（LLM）在自动化行程规划中忽视现实世界不确定性的问题。传统旅行规划基准如TravelPlanner和TripCraft假设环境为确定性，仅评估静态约束满足情况，未能考量交通延误、人流波动等随机扰动对行程可行性的影响。为此，本文提出UTP-Bench1——一个面向不确定性感知的旅行规划大规模基准数据集，整合了覆盖印度504个城市的实际旅行数据，包括景点、餐厅、住宿及多模态交通网络，并引入从主要城市采集的实证延迟分布与客流密度模式，以真实模拟运行中的不确定性扰动。其解决方案的关键在于构建一套可量化评估行程鲁棒性的三元评价指标体系：缓冲时间充分性评分（Buffer Adequacy Score, BAS）、客流敏感时间评分（Crowd-Aware Timing Score, CATS）与交通延迟吸收能力评分（Transport Delay Absorption Score, TDAS），从而系统评估生成行程在面对动态干扰时的稳定性与适应性。实验表明，当前主流大语言模型（如GPT-5、Qwen3、Mistral和Phi-4）生成的行程在时间缓冲策略、延迟感知交通调度及客流敏感规划方面与人工制定行程存在显著差距，凸显了现有模型在应对真实复杂场景下的不足。
链接: https://arxiv.org/abs/2609.02421
作者: Etcharla Revanth Rao,Priyanshu Karmakar,Shubhojit Mallick,Manish Gupta,Shreya Ghosh,Abhik Jana
机构: Microsoft(微软); IIT Bhubaneswar(印度理工学院布巴内斯瓦尔)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 34 pages, 12 figures, 16 Tables, EMNLP 2026
Abstract:Large Language Models (LLMs) have recently demonstrated strong capabilities in automated travel itinerary generation. However, real- world travel planning is inherently uncertain: transportation delays, crowd fluctuations, and unexpected stochastic delays frequently inval- idate otherwise feasible schedules. Existing benchmarks like TravelPlanner and TripCraft assume deterministic environments, evaluating only static constraint satisfaction and ignoring whether generated plans remain robust when such uncertainties arise. To address this limitation, we introduce UTP-Bench1 , a large-scale benchmark for uncertainty-aware travel planning. The dataset integrates real-world travel data spanning 504 cities of India, including attractions, restau- rants, accommodations, and multi-modal trans- portation networks. To model realistic disrup- tions, UTP-Bench incorporates empirical delay distributions and crowd-density patterns col- lected from major cities, enabling evaluation of travel plans under stochastic conditions. We further propose three evaluation metrics, namely Buffer Adequacy Score (BAS), Crowd- Aware Timing Score (CATS), and Transport Delay Absorption Score (TDAS), which quan- tify the ability of generated itineraries to main- tain robustness against transit delays and crowd variability. Experiments with state-of-the-art LLMs like GPT-5, Qwen3, Mistral and Phi-4 re- veal substantial gaps between model-generated and human-authored plans, particularly in tem- poral buffering, delay-aware transportation scheduling, and crowd-sensitive planning.
[NLP-29] Before the Script Set the Stage: How Worldview Simulation Amplifies Psychologically Grounded Persuasion in Multi-Turn Jailbreaking EMNLP2026
【速读】： 该论文旨在解决多轮对话中恶意意图通过对话过程分散传播所导致的安全漏洞问题，现有方法未能揭示驱动模型脆弱性的具体对话机制。其解决方案的关键在于提出BLUEPRINT安全评估框架，该框架将影响策略空间与情境上下文模块解耦：一方面采用基于18个理论驱动的影响因子的因子化社交影响策略空间，另一方面引入WORLDVIEWSIM这一跨轮次的情境上下文模块，并结合蒙特卡洛树搜索（Monte Carlo Tree Search）在四轮对话轨迹中优化各轮次的影响因子组合。实验表明，BLUEPRINT在六种前沿模型上实现了接近天花板的攻击成功率（ASR），且平均查询次数仅为2.46，显著低于现有方法。进一步分析揭示了抗性模型的特定脆弱性模式——不同模型对不同的影响因子及策略转移敏感，但均表现出共同的恢复路径：转向具体、可执行的任务表述能有效规避强硬拒绝状态。消融实验确认，操作性提示最为关键，增益框架异常有效，而部分合法性诉求甚至可能产生反效果。研究结果表明，构建鲁棒的多轮安全机制不仅需监控有害内容本身，更需关注对话状态如何使不当请求呈现出具体且局部可执行的表征。
链接: https://arxiv.org/abs/2609.02414
作者: Siyu Chen,Haoran Wang,Xiaojian Li,Yao Huang,Yinpeng Dong,Wei Xu
机构: Shanghai Qi Zhi Institute (上海奇智研究院); Tsinghua University (清华大学); Fangcun AI (方村AI); Institute for Interdisciplinary Information Sciences, Tsinghua University (清华大学交叉信息研究院)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 19 pages, 7 figures. Accepted to Findings of EMNLP 2026
Abstract:Multi-turn jailbreak attacks demonstrate that harmful intent can be distributed across dialogue, yet existing methods obscure what conversational mechanisms drive vulnerability. We introduce BLUEPRINT, a safety-evaluation framework separating a factorized social-influence strategy space from WORLDVIEWSIM, a cross-turn situational context module. Monte Carlo Tree Search optimizes turn-level combinations of 18 theory-grounded influence factors across a four-turn trajectory. Across six frontier models, BLUEPRINT achieves near-ceiling ASR on major open-weight and proprietary models, while requiring the fewest average queries (2.46). The resulting trajectories further reveal model-specific vulnerability among resistant targets: each responds to distinct influence factors and strategy transitions, yet all share a common recovery pathway-shifting toward concrete, executable task framing consistently escapes hard-refusal states. Ablations confirm operational cues matter most: making requests actionable has the largest impact, gain framing is unusually potent, and some legitimacy appeals can backfire. These findings suggest robust multi-turn safety requires monitoring not only harmful content, but also how dialogue state makes unsafe requests appear concrete and locally executable.
[NLP-30] Improving Health Literacy through Lay Summarization of Radiological Reports: An Evaluation of BioNER and Retrieval-Augmented Generation
【速读】： 该论文旨在解决医学影像报告（radiology reports）因专业术语密集而难以被患者理解的问题，尤其针对当前患者依赖公开大型语言模型（LLM）解读报告所引发的事实性错误与幻觉风险。其核心解决方案是通过结合命名实体识别（Named Entity Recognition, NER）与检索增强生成（Retrieval-Augmented Generation, RAG）技术，提升自动生成通俗化摘要（lay summary）的质量、事实一致性及可读性。研究发现，基于NER的临床关键信息提取能持续改善摘要的可读性和整体质量，而单独使用RAG反而可能引入无关检索内容导致幻觉；当RAG与NER协同使用时，在少样本场景下性能下降，但在微调（fine-tuned）设置下可显著提升可读性。最终，经过微调的BioBART模型结合NER实现了最佳综合表现，表明以实体感知为导向的信息提取是生成患者友好型摘要的关键驱动力。
链接: https://arxiv.org/abs/2609.02396
作者: Egecan Çelik Evgin,İlknur Karadeniz,Olcay Taner Yıldız
机构: Özyeğin University (Özyeğin大学); Galatasaray University (加拉塔萨雷大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Radiology reports are written primarily for clinicians, and their specialized terminology often makes them difficult for patients to interpret. As a result, many patients turn to publicly available Large Language Models (LLMs) to help explain their reports, despite well-documented risks of factual inaccuracies and hallucinations. Automated lay-summary generation has emerged as a promising alternative, yet the effectiveness of retrieval-enhanced and clinically informed approaches for radiology-specific communication remains underexplored. This study investigates the extent to which Retrieval-Augmented Generation (RAG) and Named Entity Recognition (NER) improve the quality, factual consistency, and readability of automatically generated lay summaries compared with standard LLM-based generation. We develop a framework combining NER-based extraction of clinically relevant findings with a RAG mechanism for contextual grounding, evaluated across few-shot and fine-tuned variants of two models (Qwen, BioBART). Results show that NER consistently improves readability and overall quality, while RAG alone offers no benefit and can introduce hallucinations from irrelevant retrieved terms. Combining RAG with NER degrades performance in few-shot settings but improves readability when fine-tuned. Fine-tuned BioBART with NER achieves the best overall performance, highlighting entity-aware extraction as the primary driver of improved patient-friendly summaries.
[NLP-31] PolERo: Studying Political Evasion in Romanian EMNLP2026
【速读】： 该论文旨在解决跨语言背景下政治回避（political evasion）识别的泛化问题，即现有基于英文数据集构建的响应清晰度与回避策略分类框架在非英语语境下是否具备可迁移性。其核心挑战在于：如何在不同语言和政治文化背景下有效识别政治人物在回答敏感问题时所采用的微妙回避策略，尤其是依赖语用线索的模糊性回避类型。解决方案的关键在于构建首个罗马尼亚总统官方发言记录的多标签标注数据集PolERo（包含3,574对问答样本），并系统评估多种分类方法在跨语言条件下的表现，包括基于TF-IDF的基线模型、微调编码器模型、提出的滑动窗口编码器架构以及零样本/少样本大语言模型（LLM）提示方法；同时通过联合双语训练和机器翻译增强数据，探究跨语言迁移机制。实验表明，微调编码器表现优异，跨语言迁移具有不对称性，而涉及语用线索的模糊回避类别仍是各类模型共同面临的最大难题。
链接: https://arxiv.org/abs/2609.02391
作者: Gabriel Stefan,Sergiu Nisioi
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted to EMNLP 2026 Main Conference
Abstract:Political evasion refers to responses that engage with a question while withholding the requested information. Recent NLP work frames political evasion as a classification task using a two-level taxonomy of response clarity and fine-grained evasion strategies. Existing work on response clarity and evasion classification is limited to English, leaving open whether the taxonomy and model behavior transfer across languages and political contexts. We introduce PolERo, a dataset of 3,574 human-annotated question-answer pairs extracted from official transcripts of five Romanian presidents. We evaluate multiple classification approaches on both datasets under matched conditions, including TF-IDF baselines, fine-tuned encoder models, a proposed sliding-window encoder, and zero/few-shot LLM prompting. We study cross-lingual transfer through joint bilingual training and machine-translation-based data augmentation. Our results indicate that fine-tuned encoders are competitive, cross-lingual transfer is asymmetric, and ambivalent evasion categories involving pragmatic cues remain the main challenge across all model families.
[NLP-32] NE-R1: Enhancing Named Entity Recognition Model via Reinforcement Learning EMNLP2026
【速读】： 该论文旨在解决大语言模型在命名实体识别（NER）任务中对长尾及领域特定实体识别能力不足的问题，其核心挑战在于模型参数化知识的局限性。为应对这一问题，作者提出NE-R1框架，其关键解决方案是设计一种“按需检索”（retrieval-on-demand）机制，并通过两阶段训练方法实现：首先进行多任务指令微调初始化，随后采用基于思维链（CoT）的端到端强化学习（RL）优化。为合理权衡模型内部参数知识与外部知识的使用，该框架引入一个多维度奖励函数，综合考虑识别准确率与检索收益。实验表明，NE-R1在多个基准测试上均达到先进水平，在域内评估中平均F1得分提升2.52%，零样本跨域评估中提升1.18%。
链接: https://arxiv.org/abs/2609.02366
作者: Meixuan Chen,Hehan Li,Ruizhi Zhao,Xin Lu,peizhi xu,Liwei Qian,LI Meifang,shuanglong li,Hanmeng Liu,Xin Pei,Yanbiao Ma
机构: Baidu, Inc., Beijing, China; Shenzhen University, Shenzhen, China; Shenzhen Graduate School, Peking University; Gaoling School of Artificial Intelligence, Renmin University of China, Beijing, China; Beijing Key Laboratory of Research on Large Models and Intelligent Governance; Engineering Research Center of Next-Generation Intelligent Search and Recommendation, MOE
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: EMNLP2026
Abstract:Named Entity Recognition (NER) has achieved substantial progress since the advent of large language models (LLMs). Nevertheless, the recognition of long-tail and domain-specific entities remains challenging due to the deficiency in parametric knowledge. Retrieval-augmented generation (RAG) offers a promising remedy by injecting external knowledge, but it also introduces noise and unnecessary cost when dealing with familiar cases. In this paper, we propose NE-R1, a novel framework for adaptive retrieval-augmented NER. We design a “retrieval-on-demand” mechanism for NER. Then we integrate it into models by a two-stage training method: (1) multi-task instruction tuning initialization; (2) end-to-end RL optimization with CoT. To achieve reasonable selection between parameterized and external knowledge, we design a multi-dimensional reward considering both accuracy and retrieval benefit. NE-R1 achieves state-of-the-art performance on various benchmarks, with an average F1 score gain of 2.52% in in-domain evaluation and 1.18% in zero-shot cross-domain evaluation.
[NLP-33] SonicCaps: Large-Scale Diverse and Fine-Grained Captioning for Improved Audio-Retrieval
【速读】： 该论文旨在解决现有音频描述数据集存在的语义多样性不足、描述内容泛化且缺乏声学细节，以及音频与文本之间一对一映射无法反映听觉感知固有模糊性的问题。其解决方案的关键在于构建大规模、高多样性的音频描述数据集SonicCaps，通过条件于音频和文本的多模态大语言模型（Qwen3-Omni）生成约1500万条与70万段音频片段配对的描述文本；为显式提升多样性，采用结构化提示工程与少量样本生成策略，每段音频生成约24条不同视角的描述，涵盖主描述、改写变体（如冗余度与风格差异）及语义标签。人类评估表明，SonicCaps在描述性与精确性方面显著优于现有数据集，且与质量评价高度相关。进一步地，基于多描述采样策略在SonicCaps上训练的CLAP模型，在音频检索与零样本分类任务中展现出更强的泛化能力，覆盖公共与商业基准。研究团队已将SonicCaps及两个专用CLAP模型发布于Hugging Face平台。
链接: https://arxiv.org/abs/2609.02343
作者: Zineb Lahrichi,Marc Ferras,Gaël Richard,Geoffroy Peeters
机构: Sony CTC(索尼中国研究中心); LTCI, Telecom Paris, Institut polytechnique de Paris(电信学院，巴黎理工学院)
类目: ound (cs.SD); Computation and Language (cs.CL); Multimedia (cs.MM); Audio and Speech Processing (eess.AS)
备注:
Abstract:Recent advances in audio-language modeling have been driven by large-scale audio captioning datasets. However, existing datasets remain limited by low semantic diversity, generic descriptions lacking acoustic details, and one-to-one audio-caption mappings that poorly reflect the inherent ambiguity of auditory perception. We introduce SonicCaps, a large-scale audio captioning dataset comprising ~15M captions paired with ~700k audio clips, generated using a multi-modal large language model (Qwen3-Omni) conditioned on both audio and text. To explicitly promote diversity, we generate around 24 captions per audio via structured prompt engineering and few- shot generation, spanning main descriptions, rephrased variants (verbosity, style) and semantic tags. Human evaluation shows that SonicCaps is rated significantly higher than existing captioning datasets, with fine-grained analyses indicating that our captions are perceived as more descriptive and precise, which strongly correlates with quality judgments. Finally, training CLAP models on SonicCaps with a multi-caption sampling strategy consistently improves audio retrieval and zero-shot classification, with stronger generalization across public and commercial benchmarks. We release both SonicCaps and two specialized CLAP models on hugging face: this https URL.
[NLP-34] SALA: Semantic-Aware Logical Alignment for Complex Reasoning in In-Context Learning EMNLP2026
【速读】： 该论文旨在解决复杂推理任务中基于上下文学习（In-Context Learning, ICL）时演示样本选择不准确的问题。传统基于表面相似性的检索方法无法捕捉问题求解的深层逻辑，而现有基于逻辑的方法依赖预定义的推理步骤和严格的匹配规则，难以适应灵活多样的推理过程。为此，本文提出一种语义感知的逻辑对齐框架SALA（Semantic-Aware Logical Alignment），其核心在于无需依赖固定规则库，而是自动学习与特定任务相关的推理操作，并将这些操作嵌入连续的语义空间中，利用动态时间规整（Dynamic Time Warping, DTW）实现推理序列的柔性对齐。该方法在保持高度可解释性的同时，实现了对推理逻辑的软匹配，有效提升了演示样本的选择质量。实验在四个推理基准和三类大语言模型上的结果表明，SALA显著优于现有方法，且分析验证了推理操作归纳与逻辑语义对齐的关键作用。
链接: https://arxiv.org/abs/2609.02336
作者: Zhao Ji,Wenqing Chen,Zhixuan Chu,Jianxing Yu,Jingping Liu,Shanhe Zhao,Zibin Zheng
机构: Sun Yat-sen University (中山大学); Zhuhai Key Laboratory of Trusted Large Language Models (珠海市可信大语言模型重点实验室); Zhejiang University (浙江大学); School of Artificial Intelligence, Sun Yat-sen University (中山大学人工智能学院); Key Laboratory of Sustainable Tourism Smart Assessment Technology, Ministry of Culture and Tourism (文化和旅游部可持续旅游智能评估技术重点实验室); Merchants Union Consumer Finance Company Limited (招商联消费金融有限公司)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Accepted for publication in Findings of EMNLP 2026
Abstract:Effective in-context learning (ICL) for complex reasoning relies on selecting the right demonstrations. Traditional retrieval methods based on surface similarity fail to capture the underlying problem-solving logic. Recent logic-based methods address this by matching predefined reasoning steps, but the rigid rules and exact-match criteria is improper to handle flexible or diverse reasoning processes. To address the problem, we propose SALA, a Semantic-Aware Logical Alignment framework. Instead of relying on a fixed inventory, SALA automatically learns task-specific reasoning operations. It then embeds these operations into a continuous semantic space and uses dynamic time warping (DTW) to align the reasoning sequences. This approach allows for soft, flexible matching of reasoning logic while remaining highly interpretable. Experiments across four reasoning benchmarks and three LLMs demonstrate that SALA outperforms existing demonstration selection methods. Further analysis confirms the roles of the operation induction and the logical semantic alignment.
[NLP-35] DiffIE: Diffusion-based Open Information Extraction
【速读】： 该论文旨在解决开放信息抽取（Open Information Extraction, OpenIE）中多关系三元组共存导致的多输出问题。传统神经方法通常采用自回归生成或固定槽位预测，前者虽灵活但效率低且易冗余，后者虽高效却将抽取预算与训练阶段耦合，缺乏灵活性。本文提出DIFFIE，其核心创新在于将条件离散扩散过程的随机性本身作为抽取机制：通过对每个词的角色标签进行独立的逆向扩散轨迹采样，生成候选三元组池，再通过宽松匹配进行聚类并排序输出。该方法在推理阶段可自由调节候选池大小与输出数量，实现了抽取预算与训练过程的解耦，并将测试时计算资源作为可调参数。实验表明，DIFFIE在CaRB (1-1)任务上同时取得新的F1和AUC最优结果，在BenchIE上超越最强规则系统ClausIE，且在标准CaRB与WiRe57评估中保持竞争力，综合四项基准测试得分最高。消融实验进一步验证了均匀离散扩散优于吸收态扩散，且非扩散型随机标签器无法复现其性能增益，表明扩散过程中的随机性是处理具有多个有效输出的结构化预测任务的有效机制。
链接: https://arxiv.org/abs/2609.02315
作者: Konstantin Fedorov,Valentin Malykh
机构: Matrosov Institute for System Dynamics and Control Theory, SB RAS(系统动力学与控制论研究所, 俄罗斯科学院西伯利亚分院); AI Talent Hub, ITMO University(人工智能人才中心, 赫尔岑国立教育大学); MWS AI; Trusted AI Research Center, RAS(可信人工智能研究中心, 俄罗斯科学院); IITU University(国际工业大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:A single sentence often expresses multiple valid relational triplets, which makes Open Information Extraction (OpenIE) fundamentally a multi-output task. Existing neural systems handle this by autoregressive generation, which is flexible but slow and prone to redundancy, or by fixed-slot prediction, which is efficient but couples the extraction budget to training. We introduce DIFFIE which instead treats the stochasticity of conditional discrete diffusion as the extraction mechanism itself: independent reverse-diffusion trajectories over per-token role tags produce a pool of candidate triplets, which are clustered under lenient matching and ranked to form the output. Both the pool size and the number of returned extractions are inference-time choices, decoupling the extraction budget from training and exposing test-time compute as a tunable axis. DIFFIE achieves the new state of the art in CaRB (1-1) both F1 and AUC, and outperforms the strongest rule-based system (ClausIE) in BenchIE; it also remains competitive in standard CaRB and WiRe57 evaluations, giving the best average score among systems that report all four benchmarks. Ablations show that uniform discrete diffusion outperforms absorbing state diffusion in our setting, and that a matched non-diffusion stochastic tagger does not reproduce its gains. Our results indicate that diffusion stochasticity is an effective mechanism for structured prediction tasks with multiple valid outputs.
[NLP-36] Efficient GUI Agents : A Systems Survey of Observation Memory Action and Runtime Optimization EMNLP2026
【速读】： 该论文旨在解决当前图形用户界面（GUI）智能体在实际部署中面临的效率瓶颈问题。尽管现有研究主要以任务成功率作为核心评价指标，但作者指出，实际应用的可行性更依赖于多维度的效率表现，包括观察效率、上下文与内存效率、动作效率以及规划器端与系统级效率。其解决方案的关键在于从端到端系统视角重构评估框架，系统性地分析并优化上述四类效率维度。通过深度文献挖掘与引用链扩展，研究发现近年来的进展集中于若干核心机制：采用选择性读取而非全上下文摄入以降低计算负担；基于全局至局部的视觉资源分配策略提升感知效率；引入可恢复的内存机制替代原始历史重放以减少存储开销；设计验证感知的控制逻辑以增强决策鲁棒性；以及构建支持在GUI与非GUI执行模式间动态切换的混合运行时架构。这些机制共同推动了高效GUI智能体的发展，但仍面临验证器成本透明化、跨基准可比性不足，以及在真实延迟与隐私约束下观察、记忆与执行层的协同设计等关键开放问题。
链接: https://arxiv.org/abs/2609.02309
作者: Bizhe Bai,Jiakang Yuan,Hongming Wu,Xinyue Wang,Jie Ren,Siyao Chen,Yuchen Ya,Fan Bai,Pai Peng,Huafeng Qin,Tao Chen
机构: Fudan University (复旦大学); Shanghai Innovation Institute (上海创新研究院); Chongqing Technology and Business University (重庆工商大学)
类目: Computation and Language (cs.CL)
备注: Accept at Grounding Language Models: Learning Faithfully and Efficiently @ EMNLP 2026
Abstract:GUI agents increasingly operate across websites, mobile apps, and desktop environments, yet the field still reports progress primarily through task success. We argue that practical deployment depends equally on efficiency: how much context, computation, action budget, and runtime overhead an agent consumes while succeeding. This survey studies efficient GUI agents through an end-to-end systems lens that preserves the current technical axes of observation efficiency, context and memory efficiency, action efficiency, and planner-side/system efficiency. For each subsection, we expand the seed literature through targeted search plus backward and forward citation chaining, then synthesize the dominant mechanisms, reported efficiency signals, and new overheads they introduce. Across the literature, recent progress converges on a small set of recurring ideas: selective reading instead of full-context ingestion, global-to-local visual allocation, recoverable memory rather than raw history replay, verification-aware control, and hybrid runtimes that can switch between GUI and non-GUI execution. We conclude by identifying the main open problems, including honest accounting of verifier cost, cross-benchmark comparability, and co-design of observation, memory, and execution layers under real latency and privacy constraints.
[NLP-37] Improving Evaluation Realism with Inference-Time Compute and Deployment Scaffolds NEURIPS2026
【速读】： 该论文旨在解决对齐评估中的“评估意识”（evaluation awareness）问题：具备能力的模型能够识别自身处于测试而非实际部署状态，从而导致安全评估结论的可信度下降。为应对这一挑战，论文提出两项关键技术。其一为“批判优化”（critique refinement），即在每次模拟器动作生成时增加推理阶段的计算开销：模拟器生成多个候选动作，通过目标模型实例提供的反馈对这些动作进行迭代优化，以提升其与真实部署场景的相似性，并选取最接近真实部署的候选动作继续评估。其二为“DISH”（Deployment-Imitating SWE-Agent Harness），通过构建一个代理框架将目标模型包裹其中，缩小编码场景下模拟环境与真实部署环境之间的差距。实验表明，两项技术具有可组合性，联合应用时带来的真实感提升显著优于单独使用任一方法。研究结果表明，自动化手段可有效增强对齐评估的真实性，且相较于延长审计时长，更高效地利用额外计算资源。
链接: https://arxiv.org/abs/2609.02302
作者: Axel Ahlqvist,Richard Guan,Juan-Pablo Rivera,Adeline Kassler,Dmitrii Troitskii,Alexandra Souly,Kai Fronsdal,Robert Kirk,John Hughes
机构: Meridian Visiting Researcher Programme; Cambridge Boston Alignment Initiative; UK AI Security Institute; Meridian Labs; Anthropic
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 70 pages, 43 figures, 4 tables (13 figures in the main text). Under review at NeurIPS 2026. Code: this https URL and this https URL ; reproduction assets: this https URL
Abstract:A core obstacle to alignment evaluation is evaluation awareness: capable models can tell when they are being tested rather than deployed, weakening the conclusions a safety evaluation can support. We present two techniques that make simulated alignment evaluations harder to distinguish from real deployments. Our first technique, critique refinement, spends additional inference-time compute on each simulator action: the simulator generates multiple candidate actions, refines them using feedback from an instance of the target model on how to make them more realistic, and continues the evaluation with the most deployment-like candidate. Our second technique, DISH (Deployment-Imitating SWE-Agent Harness), wraps the target in an agent harness, reducing the gap between simulated and real deployment environments in coding settings. We test the techniques on multiple target models and find that they compose: applying both yields larger realism gains than either alone. Our results show that automated approaches can improve the realism of alignment evaluations, and that these improvements use additional compute more effectively than making the audits longer.
[NLP-38] SCX Router: Streaming Zero-Shot Model Selection with a Decoder-KV Classifier and a Real-World Task Ontology
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在实际应用中因推理端点在质量、成本、延迟、上下文支持、工具使用、领域专长及推理行为等方面存在显著异质性，导致难以通过人工启发式方法实现稳定且高效的“速度-成本-质量”权衡问题。其核心解决方案是提出一种轻量级的GLiClass-based路由机制（\router），该机制无需自回归生成即可为每个推理时的模型标签分配适应性评分。关键创新在于：采用0.6B参数的释放检查点，结合Qwen3解码器与浅层双向评分器，通过解码器-KV执行路径实现在会话中保持纯文本键值缓存，仅对新增对话轮次进行编码，并在不持久化缓存的前提下评估候选标签令牌；同时，该模型可预测任务类型、难度、推理模式、预期输出长度并支持自定义零样本标签。研究构建了一个包含23个家族、115种任务类型、345个可路由子类型、1,173个合成示例及30个独立领域的任务本体，基于此生成了15万条验证器评分的任务和1.5万条开放式任务，用于训练Qwen3解码器，并显式分离请求预测与任务级策略（如资格、成本、缓存复用、安全性和主权等）。在六个LiveBench子集上的实验表明，该路由器性能优于平均候选模型，在选定的1,000任务子集上达到0.707的综合顶1得分，优于最强固定模型的0.696，且在不同基准下均表现出显著增益。
链接: https://arxiv.org/abs/2609.02292
作者: Ihor Stepanov,Aleksandr Smechov,Mykhailo Shtopko,Dmytro Vodianytskyi,Oleksandr Lukashov
机构: Knowledgator; SCX.ai Holdings Limited
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 20 pages, 10 tables, 6 figures
Abstract:The rapid proliferation of large language models (LLMs) and the growing diversity of their applications presents a unique optimization opportunity: selecting the right model for the task, while optimizing for speed, cost, and quality at a per-task level. However, inference endpoints can vary widely in quality, price, latency, context support, tool use, domain expertise, and reasoning behavior. This heterogeneity makes manual heuristics difficult to maintain and unlikely to achieve consistently favorable speed–cost–quality trade-offs on their own. We introduce \router, a lightweight GLiClass-based router that assigns a suitability score to each inference-time model label without autoregressive generation. The released 0.6B-parameter checkpoint combines a Qwen3 decoder with a shallow bidirectional scorer. Its decoder-KV execution path preserves a text-only key–value cache across a session, encodes only new dialogue turns, and evaluates transient candidate-label tokens without adding them to the persistent cache. The same checkpoint also predicts task type, difficulty, reasoning mode, and expected output length, and supports custom zero-shot labels. For task generation, we construct a task ontology with 23 families, 115 task types, 345 routable subtypes, 1,173 synthetic examples, and an orthogonal axis of 30 domains. Using this structure, we generate 150,000 verifier-scored tasks and 15,000 open-ended tasks. We then train the Qwen3 decoder on these tasks, while explicitly separating learned request prediction from per-task policies for attributes such as eligibility, cost, cache reuse, safety, and sovereignty. Across six LiveBench subsets, the router outperforms the mean candidate; on the selected 1,000-task subset, it achieves an aggregate top-1 score of 0.707 versus 0.696 for the strongest fixed model, with benchmark-dependent gains.
[NLP-39] Entangled Representations Amplify Collateral Damage in Unlearning
【速读】： 该论文旨在解决生成式人工智能（Generative AI）中因表征纠缠（representational entanglement）导致的遗忘困难问题，即当不同知识领域（如生物学与非生物学知识）在神经网络中共享表征结构时，难以实现精准的局部遗忘而不损害保留知识。其解决方案的关键在于通过选择性梯度掩码（Selective Gradient Masking, SGTM）技术，构建一组参数规模为254M的英文维基百科语言模型，并系统性地控制生物学与非生物学知识之间的解耦程度，从而在受控实验条件下验证表征纠缠对遗忘性能的影响。实验结果表明，在固定遗忘水平下，解耦程度更高的模型在三种标准遗忘方法中均表现出显著更优的“保留-遗忘”权衡：其中两种方法下保留代价降低约4倍，第三种方法下降低1.3倍。由于该干预仅改变模型结构而未调整数据或遗忘算法，这一发现为解释性研究领域长期怀疑的论断提供了直接证据——表征纠缠是导致遗忘过程中附带损伤（collateral damage）的重要原因。该实验设计亦可推广至检验其他来自解释性研究的结构性假设。
链接: https://arxiv.org/abs/2609.02285
作者: Evžen Wybitul,Tim G. J. Rudner,Christian Schroeder de Witt
机构: University of Oxford(牛津大学); University of Toronto(多伦多大学)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注:
Abstract:A long-held intuition in interpretability research is that representational entanglement, the sharing of structure between knowledge domains in a neural network, makes unlearning harder. While the intuition is widespread, it has never been directly tested in a controlled experiment. We present a way to do so: by repurposing Selective Gradient Masking (SGTM), we train a suite of six 254M-parameter language models on English Wikipedia with graded levels of disentanglement between biology and non-biology knowledge. Applying three standard unlearning methods to every model in the suite, we find that more disentangled models consistently achieve better retain-forget trade-offs: at a fixed level of forgetting, the most disentangled models incur roughly 4\times lower retain cost under two of the three methods, and 1.3\times lower under the third. Because our intervention changes only the model, not the data or the unlearning algorithm, this is direct evidence that representational entanglement is one of the causes of collateral damage in unlearning, as interpretability researchers have long suspected. A similar design could be used to test other structural claims from interpretability.
[NLP-40] Do Large Language Models Capture the Diversity in their Training Data?
【速读】： 该论文旨在解决生成式模型（Generative AI）在生成输出时所表现出的条件多样性不足问题，即模型生成结果的条件熵显著低于其训练数据中的条件熵，表明模型未能充分捕捉训练数据中潜在的多样化输出。其核心解决方案在于提出一种后处理校正机制，通过为每个输入生成多个候选输出，并利用基于冯诺依曼熵（von Neumann entropy）的矩阵熵投影进行重加权，从而在保持与原始模型分布相近的前提下提升条件多样性。该方法的关键在于证明了矩阵形式条件熵函数的凹性，使得熵约束下的投影优化问题具有凸性，进而可采用可扩展的镜面下降（mirror-descent）算法高效求解。该研究建立了一个信息论框架，用于量化并缓解现代生成模型与其训练数据之间的系统性条件多样性差距。
链接: https://arxiv.org/abs/2609.02275
作者: Youqi Wu,Farzan Farnia
机构: The Chinese University of Hong Kong (香港中文大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Large language models are trained to model conditional distributions over text, yet it remains inadequately understood whether they capture the full diversity of plausible outputs present in their training data. We study this question through an information-theoretic lens by comparing the conditional entropy of model-generated outputs with that of the corresponding training data. Given paired input-output samples, we use conditional entropy and its matrix-based analogue based on von Neumann entropy to measure output variability beyond what is explained by the conditioning input, without requiring multiple reference outputs for the same prompt. Across LLM families with publicly available training data, including OLMo, Pythia, and GPT-Neo, we consistently find that model-generated outputs exhibit lower conditional entropy than their training data, across different model scales, sequence lengths, and decoding strategies. We observe a similar conditional diversity gap beyond language modeling, including class-conditioned ImageNet generators and text-conditioned models trained on MS-COCO. To address this gap, we propose a post-hoc correction mechanism that generates multiple outputs for each input and reweights them through a matrix-entropy projection, increasing conditional diversity while remaining close to the original model distribution. We prove the concavity of the matrix-based conditional entropy functional, which makes the resulting entropy-constrained projection a convex optimization problem, and develop a scalable mirror-descent algorithm for its implementation. Our results reveal a systematic conditional diversity gap between modern generative models and their training data, and provide an information-theoretic framework for measuring and mitigating this gap.
[NLP-41] CoMerge: Conflict-Driven Preference Optimization for Multi-Task Model Merging EMNLP2026
【速读】： 该论文旨在解决多任务大语言模型（LLM）在模型合并过程中因参数干扰导致性能下降的问题，尤其关注现有方法未能有效利用朴素合并策略暴露的退化行为。其核心解决方案是提出一种冲突驱动的偏好优化框架（CoMerge），将模型合并重构为偏好优化问题。关键在于采用自监督的冲突驱动策略，利用朴素合并方法（如任务算术）产生的缺陷作为硬负样本，构建无需外部标注的偏好对，并通过偏好优化微调轻量级、张量级别的合并系数，从而在缓解参数空间冲突的同时保留各任务特异性能力。实验表明，CoMerge在MergeBench上实现0.9968的平均归一化性能，显著优于所有评估的数据无关与数据依赖型合并基线；在Llama-3.1-8B-Instruct上，其在指令遵循和安全性等敏感任务上表现显著提升，且仅优化1,445个标量系数即达到接近全参数微调的效果。
链接: https://arxiv.org/abs/2609.02273
作者: Mingjie Zheng,Zihao Chen,Wenqing Chen,Weile Yuan,Zhixuan Chu,Jianxing Yu,Zibin Zheng
机构: Sun Yat-sen University (中山大学); HiThink Research (杭州慧思科技); Zhejiang University (浙江大学); Ministry of Culture and Tourism (文化和旅游部)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Accepted for publication at the EMNLP 2026 Main Conference
Abstract:Model merging provides an efficient paradigm for constructing multi-task large language models (LLMs) without full model retraining, yet it remains challenged by parameter interference. While existing methods aim to preserve the capabilities of individual expert models and mitigate interference, they generally do not directly learn from the potentially degraded behaviors exposed by naive merging. In this paper, we propose a conflict-driven preference optimization framework for model merging (CoMerge), which reformulates model merging as a preference optimization problem. The approach utilizes a self-supervised, conflict-driven strategy that leverages the defects of naive merging methods (e.g., task arithmetic) as hard negative samples to construct preference pairs without external annotations. By applying preference optimization to refine lightweight, tensor-wise merging coefficients, CoMerge enables the model to mitigate parameter-space conflicts while preserving task-specific capabilities. Extensive experiments show that CoMerge achieves an average normalized performance of 0.9968 on MergeBench, outperforming all evaluated data-free and data-driven model-merging baselines. Furthermore, on Llama-3.1-8B-Instruct, CoMerge yields marked improvements on conflict-sensitive tasks such as instruction following and safety, while remaining highly competitive with full-parameter fine-tuning despite optimizing only 1,445 scalar coefficients.
[NLP-42] PaperCompiler: Faithful Paper-to-Code Generation via Repository-Level Specification Compilation
【速读】： 该论文旨在解决科研论文到代码仓库（repository-level）实现之间忠实映射的难题，核心挑战在于：研究论文通常以高层次描述方法、隐含实现假设，且要求生成的代码仓库需保持方法逻辑一致性、评估协议完整性和跨文件结构统一性。现有论文到代码代理（paper-to-code agents）的中间输出多为自由格式的计划或摘要，易被下游编码代理忽略、误读或压缩，导致算法简化与代码结构不一致。其解决方案的关键是提出PaperCompiler框架，该框架将基于论文的证据转化为显式的、可执行的代码仓库级实现规范（implementation specifications），在保留原始来源可追溯性的前提下，明确区分论文支持、推断得出、外部委托及未决信息，并编码非退化性要求、所有权分配、跨文件依赖关系和文件级约束。在此编译后规范的指导下进行代码仓库生成，同时保留对论文未规定局部工程决策的灵活性。实验表明，PaperCompiler在Paper2CodeBench基准上显著优于基线模型，参考基准保真度相对提升13.8%（从3.64增至4.15），高严重性评估批评率由13.2%降至6.1%，验证了其在提升生成代码质量与一致性方面的有效性。
链接: https://arxiv.org/abs/2609.02272
作者: Yunhao Liu,Hong Phuc Pham,Jaehong Yoon
机构: NTU Singapore(南洋理工大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
备注: 9 pages
Abstract:Faithfully translating research papers into repository-level implementations remains challenging because papers often describe methods at a high level, leave implementation assumptions implicit, and require generated repositories to preserve method logic, evaluation protocols, and cross-file consistency. Despite recent advances in paper-to-code agents, their intermediate outputs are often presented as free-form plans or summaries that downstream coding agents may ignore, reinterpret, or compress, leading to algorithmic simplification and inconsistent repository structure. To address these challenges, we introduce PaperCompiler, a paper-to-code generation framework that compiles paper-grounded evidence into explicit repository-level implementation specifications. PaperCompiler grounds implementation-relevant evidence while preserving source provenance and distinguishing paper-supported, inferred, externally delegated, and unresolved information. The resulting specifications encode non-degradation requirements, ownership assignments, cross-file dependencies, and file-level constraints. Repository generation proceeds under these compiled specifications while retaining flexibility over local engineering choices not fixed by the paper. PaperCompiler outperforms strong baselines on Paper2CodeBench, achieving a 13.8% relative improvement in reference-based fidelity (from 3.64 to 4.15) and reducing high-severity evaluator critiques (from 13.2% to 6.1%).
[NLP-43] From Detection to Characterization: A Large-Scale Study of Rag ebait on Japanese X
【速读】： 该论文旨在解决在线内容中“愤怒诱饵”（ragebait）的可靠大规模检测与系统性分析不足的问题，从而阻碍了对其传播范围、社会影响及治理策略的理解。其解决方案的关键在于构建一个基于大语言模型（LLM）辅助标注的高质量标注数据集，并训练多种日语语言模型以实现对ragebait的有效识别；最终采用集成分类器在大规模日语推文数据集上进行分析，揭示了ragebait在政治、社会争议议题中的高发性及其传播速度更快、引发更多负面情绪反应（如愤怒、恐惧、厌恶、悲伤与惊讶）等特征，为理解并应对网络情绪化内容提供了可量化的技术基础与实证依据。
链接: https://arxiv.org/abs/2609.02262
作者: Zhiyang Qi,Kazuhiro Ito,Jinghui Chen,Hibiki Nakamura,Zhangxuan Chen,Erina Murata,Masaki Chujyo,Fujio Toriumi
机构: The University of Tokyo (东京大学)
类目: ocial and Information Networks (cs.SI); Computation and Language (cs.CL)
备注: Accepted at WI-IAT 2026. This is the pre-camera-ready version
Abstract:Ragebait refers to online content intentionally designed to provoke anger or outrage and thereby increase attention and engagement. However, reliable large-scale detection and systematic analysis of ragebait remain limited, hindering efforts to understand its prevalence, impact, and mitigation. This study aims to develop an effective ragebait detection framework and to clarify the characteristics of ragebait at scale, providing a basis for understanding and mitigating emotionally provocative content online. We constructed a labeled dataset with the assistance of a large language model (LLM) and trained several Japanese language models for ragebait detection. The resulting ensemble classifier was then applied to a large-scale dataset of Japanese-language posts on X. Our analysis shows that ragebait is more prevalent in politically and socially contentious topics, including politics, discrimination, public health, and interpersonal conflict. Ragebait posts also spread faster and receive more negative reactions than non-ragebait posts, particularly anger, fear, disgust, sadness, and surprise. These findings demonstrate the utility of the proposed detector and provide a large-scale characterization of ragebait in Japanese online discourse.
[NLP-44] APEx: Distillation of Agent Procedural Experience for Adaptive Deep Research Question Answering
【速读】： 该论文旨在解决大语言模型在处理复杂、长时程任务时，因缺乏有效经验学习机制而导致的持续优化瓶颈问题。现有方法要么依赖冗长的任务特定交互轨迹，增加决策负担；要么通过固定方式提炼过程性技能，但其与下游策略适应脱节，难以实现动态优化。为此，论文提出APEx（Adaptive Procedural Experience）框架，其核心在于构建分层的经验利用机制：将交互历史组织为实例级轨迹记忆与类别级过程性技能，并通过执行器（Executor）、蒸馏器（Distiller）与规划器（Planner）构成闭环架构进行协同优化。该框架采用三阶段交替的广义奖励策略优化（GRPO）训练范式，实现基于奖励引导的技能蒸馏，而非静态提示生成。在测试阶段，蒸馏所得技能作为过程先验，通过技能引导的测试时强化学习（test-time reinforcement learning）实现规划器的在线适应，结合技能对齐正则化，在无需真实标签的情况下完成自提升并抑制策略漂移。实验在7个基准上验证了APEx的优越性，性能超越GPT-5.4达14.7分，优于最强的记忆增强基线3.0分。
链接: https://arxiv.org/abs/2609.02253
作者: Jie Ding,Rui Sun,Xinyuan Zhang,Zeyu Zhang,Xin Liu
机构: University of Science and Technology of China(中国科学技术大学); ByteDance Inc.(字节跳动公司); Chinese Academy of Sciences(中国科学院)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Deep research agents augment large language models with external tools to answer complex, long-horizon questions through multi-turn reasoning. Learning from prior experience is crucial for continual improvement, yet existing methods either retrieve verbose task-specific traces that burden decision-making, or distill procedural skills that remain decoupled from downstream policy adaptation. We propose APEx, a hierarchical experience utilization framework that organizes interaction history into instance-level trajectory memories and category-level procedural skills, and couples them through a closed-loop architecture of Executor, Distiller, and Planner. The three modules are optimized via a three-stage alternating GRPO training paradigm, enabling reward-guided skill distillation rather than fixed-prompt generation. At test time, distilled skills serve as procedural priors for online Planner adaptation through skill-guided test-time reinforcement learning, allowing ground-truth-free self-improvement with skill-alignment regularization to prevent policy drift. Experiments on 7 benchmarks demonstrate that APEx achieves state-of-the-art performance, surpassing GPT-5.4 by 14.7 points and the strongest memory-augmented baseline by 3.0 points.
[NLP-45] LeakageBench: Document-Level Leakage Risk for Redacting Personally Identifiable Information in Document Images
【速读】： 该论文旨在解决真实场景下文档图像（如扫描件、截图和PDF渲染图）中个人身份信息（PII）擦除的可靠性问题，其核心挑战在于：现有PII基准多聚焦于文本层面，未能充分评估文档级泄露风险——即使仅遗漏一个敏感标识符，整页仍可能构成泄露隐患。为此，研究提出LeakageBench，一个包含500张文档图像与11,954个符合通用数据保护条例（GDPR）标准的PII标注的基准数据集，涵盖直接标识符、关联键及上下文再识别表面等多类敏感信息。通过实体级F1、群体级泄露率与文档级泄露率等多维度指标，系统评估了通用OCR流水线、商业及任务适配型OCR依赖检测器以及无OCR视觉-语言模型的表现。结果显示，尽管代码解释器（Code Interpreter）可将GPT-5.5的定位F1从0.090提升至0.249，但关键的页面级泄露率仍高达0.968，表明当前方法在提升检测精度的同时，仍难以实现高召回率且空间精准的文档级安全擦除。因此，解决方案的关键在于构建兼具高召回率与空间对齐能力的文档级红蓝对抗评估框架，以推动生成式AI与工具链协同下的鲁棒性PII擦除技术发展。
链接: https://arxiv.org/abs/2609.02207
作者: Vishnu Prasad Vijaya Kumar,Santhosh Venkatesh,Ivan P. Yamshchikov
机构: Center for Artificial Intelligence and Robotics (CAIRO) Technical University of Applied Sciences Würzburg-Schweinfurt (THWS), Würzburg, Germany; DataX, Frankfurt am Main, Germany
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注:
Abstract:Real-world personally identifiable information (PII) redaction often operates on document images—scans, screenshots, and PDF renderings—where OCR errors, layout structure, and visual noise determine whether sensitive information is actually removed. Existing PII benchmarks are mostly text-centric and do not measure document-level redaction risk: a page remains unsafe if even one identifier is missed. We introduce LeakageBench, a challenge set of 500 document images with 11,954 GDPR-aligned PII annotations spanning direct identifiers, linkage keys, and contextual re-identification surfaces. We evaluate generic OCR pipelines, commercial and task-adapted OCR-dependent detectors, and OCR-free vision-language models using entity-level F1, group-wise leakage, and document-level leakage metrics. Code Interpreter raises GPT-5.5 localization F1 from 0.090 to 0.249, but critical page-level leakage remains 0.968. These results show that stronger detection and tool assistance improve localization without making most pages safe for release. LeakageBench provides a diagnostic benchmark for high-recall, spatially grounded PII redaction in document images.
[NLP-46] Breadth Beats Depth: Improving GCG-Based Jailbreak Optimization with Breadth-Oriented Suffix Search
【速读】： 该论文旨在解决基于优化的越狱攻击（jailbreak attack）方法中，现有贪婪坐标梯度（Greedy Coordinate Gradient, GCG）技术因依赖平均对抗损失与深度贪婪搜索而导致的局限性问题。具体而言，此类方法容易过度关注易于触发的攻击行为，而忽视潜在更具潜力的后缀空间区域，从而影响攻击的有效性与泛化能力。其解决方案的关键在于提出一种名为BOSS（Breadth-Oriented Suffix Search）的即插即用框架，通过引入尾部聚焦对抗损失（Tail-Focused Adversarial Loss, TFAL）、标准源模型损失以及行为覆盖度（behavior coverage）作为多维度评估指标，实现对终止后缀的更优选择；同时，采用多短路径并行探索策略，并对有前景的后缀进行选择性延续，从而在保持高攻击成功率的同时显著降低优化时间。
链接: https://arxiv.org/abs/2609.02172
作者: Shiliang Xiao,Jingsong Wei,Yuzhi Liang,Yufan Zheng,Xia Li,Qiliang Lin
机构: Guangdong University of Foreign Studies(广东外语外贸大学); School of Information Science and Technology(信息科学与技术学院)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Optimization-based jailbreak attacks such as Greedy Coordinate Gradient (GCG) achieve strong effectiveness and transferability by optimizing adversarial suffixes on white-box source models. However, existing GCG-based methods rely on averaged adversarial loss and deep greedy search, which can over-emphasize easy-to-jailbreak behaviors and overlook promising regions of the suffix space. We propose BOSS, a plug-and-play framework that improves GCG-based jailbreak optimization through breadth-oriented suffix search. BOSS uses Tail-Focused Adversarial Loss (TFAL), standard source loss, and behavior coverage to select terminal suffixes, then explores multiple short trajectories and selectively continues promising suffixes. Experiments on public benchmarks show that BOSS improves attack success rates across multiple GCG-based methods while reducing optimization time.
[NLP-47] Do Cantonese-Adapted Language Models Better Predict Cantonese Reading? A Cross-Model Eye-Tracking Evaluation
【速读】： 该论文旨在解决语言多样性背景下，特定方言（粤语）训练对生成式语言模型在心理语言学预测能力上是否具有提升作用的问题，尤其关注粤语与普通话之间在自然语言处理（NLP）模型适应性上的差异。当前对于粤语专用训练能否有效增强模型对人类阅读预期的刻画仍存在争议。其解决方案的关键在于通过真实自然的粤语眼动数据，系统比较同一模型家族内不同粤语适配版本在信息论指标上的表现，具体包括词汇意外度（lexical surprisal）、词性意外度（POS surprisal）、目标词前熵（entropy before the target）以及熵减少量（entropy reduction）。研究发现，经过更深入粤语持续预训练与指令微调的CantoneseLLM-7B在多数指标（尤其是词汇意外度及联合四维模型）上表现最优，表明更充分的粤语特化训练可显著提升模型对人类阅读行为的预测拟合度；然而，不同信息论指标对模型排名的影响存在差异，例如熵减少量则偏好原始模型CKIP GPT-2 Tiny，提示模型性能评估需结合具体度量方式综合考量。
链接: https://arxiv.org/abs/2609.02163
作者: Ziqi Zhang,Emmanuele Chersoni,Mohammad Momenian
机构: The Hong Kong Polytechnic University(香港理工大学); The University of Hong Kong(香港大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Information-theoretic measures derived from autoregressive language models are widely used to characterize the expectations that shape human reading, but whether language-variety-specific training improves such psycholinguistic alignment remains unclear. This question is still open for Cantonese, where recent NLP evaluations reported mixed benefits from Cantonese-specific training relative to Mandarin-oriented or general-purpose models. Using naturalistic Cantonese eye-tracking data, we compare two within-family adaptation contrasts: CKIP GPT-2 Tiny versus its lightly Cantonese-adapted JED351 derivative, and Qwen2.5-7B versus CantoneseLLM-7B, which underwent substantially more extensive Cantonese continued pretraining and instruction tuning. From each model, we derive lexical surprisal, POS surprisal, entropy before the target, and entropy reduction. Lexical surprisal and the joint four-metric model consistently favor CantoneseLLM-7B, followed by Qwen2.5-7B, CKIP, and JED351, whereas entropy reduction favors CKIP. These results suggest that more extensive Cantonese-specific training can be associated with stronger predictive fit, while model rankings also depend on the information-theoretic measure being evaluated.
[NLP-48] OBJECTION! Lawyer Agents Mitigate Guilty Bias in Legal Judgment Prediction EMNLP2026
【速读】： 该论文旨在解决法律判决预测（Legal Judgment Prediction, LJP）模型中存在的“有罪偏见”（Guilty Bias）问题，即模型在训练过程中过度依赖检方陈述，导致对无罪结果的预测能力严重不足，尤其在推理阶段无法有效缓解此类偏差。其核心挑战在于现有方法虽通过三步推理结构或合成无罪数据提升整体准确率，但未能在推理阶段真正实现对有罪假设的批判性检验。本文提出OBJECTION——一种基于对抗性辩护律师代理（Adversarial Lawyer Agent）的推理时干预机制，将该代理嵌入到“行为构成、违法性、可责性”的三步法律推理流程中，在每一阶段主动引入法律辩护论点，以挑战模型对被告有罪的预设。该方案的关键创新在于实现了动态、阶段性的反向论证，推动模型进行更符合“无罪推定”原则的实质性法律推理。为全面评估效果，研究构建了包含3400个真实案例的“自然无罪”（Natural Innocent）数据集，克服了以往合成数据的局限性。实验结果表明，OBJECTION将错误有罪率（False Guilty Rate, FGR）从当前最优基线的82.93%大幅降至16.69%，显著提升了模型在无罪判定上的公平性与可靠性，标志着法律人工智能向契合司法伦理的重要进展。
链接: https://arxiv.org/abs/2609.02158
作者: Jaehoon Jeong,Jay-Yoon Lee
机构: Seoul National University (首尔国立大学); Korean National Police Agency (韩国国家警察局)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted to EMNLP 2026 Main Conference. Dataset: this https URL
Abstract:Legal Judgment Prediction (LJP) models are typically trained on documents that describe facts from a prosecutorial perspective. Existing datasets further exhibit severe label imbalance toward guilty outcomes. Consequently, these models suffer from “Guilty Bias”, blindly accepting the prosecution’s narrative as objective truth. Previous studies employing three-step reasoning structures or training on synthetically generated innocence data improve overall accuracy, but they still fail to mitigate bias at inference time. In this paper, we introduce OBJECTION, an inference-time pipeline that integrates an Adversarial Lawyer Agent into each 3-step reasoning of offense, unlawfulness, and culpability. Unlike generic critics, our agent actively challenges the model’s presumptions of guilt by injecting legal defense arguments at each reasoning stage. To thoroughly evaluate this, we present a new “Natural Innocent” dataset including 3.4k real-world cases, overcoming the limitations of synthetic innocence benchmarks. Test results show that OBJECTION drastically reduces the False Guilty Rate (FGR) from 82.93% (SOTA baseline) to 16.69%, proving its capability to perform substantive legal reasoning. This work denotes a key progress toward aligning Legal AI with the presumption of innocence. Comments: Accepted to EMNLP 2026 Main Conference. Dataset: this https URL Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI) Cite as: arXiv:2609.02158 [cs.CL] (or arXiv:2609.02158v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2609.02158 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-49] A Layered Taxonomy for Chinese Learner Grammatical Error Annotation
【速读】： 该论文旨在解决中文学习者写作中语法错误标注的一致性与语言学意义不足的问题。其核心解决方案是提出一种分层标注体系，将计算型中文语法错误修正（CGEC）与教学导向的错误分析相衔接：首先在字符和标点层面识别拼写错误，并通过编辑操作类型与子类型进行标注；对于其他错误，则采用包含编辑操作、语言领域及词性三个层级的核心标签，并可选地引入汉语特有的扩展维度，如体貌、情态、比较、论元结构及补语等。该标注体系基于CGEC资源、学习者错误分类体系及现代汉语语法理论构建，通过自动提取的MuCGEC编辑数据进行覆盖度分析，并借助五种大语言模型对样本数据的初步一致性测试进行验证，结果表明该分层方法具有可行性，但部分类别边界仍需进一步优化。
链接: https://arxiv.org/abs/2609.02153
作者: Mengyang Qiu,Jungyeul Park
机构: Saint Elizabeth University (圣伊丽莎白大学); KAIST (韩国科学技术院)
类目: Computation and Language (cs.CL)
备注:
Abstract:Grammatical error annotation in Chinese learner writing requires labels that are both consistent and linguistically meaningful. This paper proposes a layered scheme linking computational Chinese grammatical error correction (CGEC) with pedagogical error analysis. The scheme first identifies character- and punctuation-level orthographic errors, labeling them by edit operation and subtype. Other errors receive a three-layer core label combining edit operation, linguistic domain, and part of speech, with optional Chinese-specific extensions for aspect, modality, comparison, argument structure, and complements. Drawing on CGEC resources, learner-error taxonomies, and Mandarin grammar, the taxonomy is evaluated through a coverage analysis of automatically extracted MuCGEC edits and a preliminary consistency study in which five large language models apply it to a sample. The results support the layered approach while identifying category boundaries requiring further refinement.
[NLP-50] EmoStance: Response-Side Affective-Orientation Control for Empathetic Response Generation via Emoji Weak Supervision EMNLP2026
【速读】： 该论文旨在解决情感回应生成中模型难以准确把握回应方情感立场（response-side affective-orientation）的问题，即在生成回应时不仅需决定“说什么”，还需根据对话中前言者的共情状态合理调整回应的情感态度。其核心解决方案是利用多标注者提供的表情符号（emoji）分布作为弱情感-态度证据（weak affective-attitudinal evidence），而非将其视为输出标签或黄金标准，从而构建一个潜在的控制空间，以操作性地逼近听者立场（listener stance）。为此，研究构建了EmojiDialogue数据集，该数据集在EmpatheticDialogues基础上扩展了话语级的表情符号投票与置信度评分，并提出EmoStance模型：该模型首先建模说话方的情感表达，再基于对话上下文与发言者角色预测软性回应倾向，并通过连续前缀嵌入（continuous prefix embeddings）对冻结的指令微调大语言模型（instruction-tuned LLM）进行可控引导。在包含20名标注者、800次判断的盲评实验中，EmoStance实现了62.2%的决定性胜率，在上下文特异性与感知响应性方面提升显著，且与外部知识方法具有互补性。
链接: https://arxiv.org/abs/2609.02133
作者: Ziyuan Jin,Yuxuan Ge,Zheng Tian
机构: ShanghaiTech University (上海科技大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Accepted to the Main Conference of EMNLP 2026
Abstract:Empathetic response generation requires models to decide not only what to say, but also how to respond to the previous speaker’s affective situation. We formulate this as response-side affective-orientation control and use multi-annotator emoji distributions as weak affective–attitudinal evidence, rather than as output symbols or gold labels, to induce a latent control space that operationally approximates listener stance. We construct EmojiDialogue, an utterance-level extension of EmpatheticDialogues with emoji votes and confidence scores, and propose EmoStance, which models source-side affective expression, predicts a soft response-side orientation from dialogue context and speaker roles, and steers a frozen instruction-tuned LLM through continuous prefix embeddings. In blind pairwise evaluation with 20 annotators and 800 judgments, EmoStance achieves a 62.2% decisive win rate, with the clearest gains in contextual specificity and perceived responsiveness, while remaining complementary to external-knowledge methods. Code, annotation metadata, and reconstruction scripts are available in our GitHub repository: this https URL.
[NLP-51] C3T: Counterfactual Causal Reasoning for Sentiment Shifts in Social-Media Conversation Trees EMNLP2026
【速读】： 该论文旨在解决社交媒体中谣言传播对话树内情感动态演变的因果推理问题，即在多层级回复结构中，用户对信息修正、证据呈现或攻击性言论等话语行为如何引发情感变化。其核心挑战在于准确识别导致情感转移的关键干预因素，并实现对因果路径的可解释建模。解决方案的关键在于提出一种新型框架——C³T（Counterfactual Causal Conversation Transformer），该模型基于公开的谣言对话数据集构建了包含细粒度情感标签、父子节点情感偏移标签、校准的多标签干预标记及显式因果源标注的因果情感推理层（CaSiRe）。C³T通过联合预测节点情感与情感偏移，学习稀疏祖先归因，并利用反事实干预嵌入的开关机制进行潜在结果估计，从而支持对具体话语行为（如否认/更正、证据引用、毒性攻击）的因果效应分析。实验表明，该方法在事件级划分下显著提升泛化鲁棒性与归因准确性，揭示了否认/更正和证据可降低下游负面情绪，而毒性则加剧负面情绪，且相较于仅依赖大语言模型提示的基线，结构感知的反事实建模在因果推断可靠性上更具优势。
链接: https://arxiv.org/abs/2609.02131
作者: S M Rafiuddin,Atriya Sen
机构: Oklahoma State University (奥克拉荷马州立大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 23 pages, 3 figures, 7 tables; accepted to the EMNLP 2026 Main Conference
Abstract:Sentiment in social-media threads does not only vary across posts; it shifts as users react to claims, corrections, evidence, and hostility within a branching reply tree. We study why sentiment changes in rumor-centric conversation trees by treating discourse moves (e.g., denial/correction, evidence/link, toxicity/attack) as candidate interventions and asking (i) what sentiment a reply expresses, (ii) whether the sentiment shifts relative to its parent, and (iii) which prior message most plausibly drove the reply’s sentiment. To support this setting, we introduce CaSiRe, a causal sentiment reasoning layer over public rumor conversation datasets that adds post-level sentiment labels, induced parent-child shift labels, calibrated multi-label intervention tags, and explicitly annotated causal-source labels. We then propose C ^3 T (Counterfactual Causal Conversation Transformer), a thread-structured temporal model that jointly predicts node sentiment and shifts, learns sparse ancestor attribution, and supports counterfactual queries by forcing conversational intervention embeddings on or off to estimate potential outcomes. Under an event-level split, C ^3 T improves out-of-event robustness and attribution over text-only, graph-based, and temporal baselines, and yields interpretable model-based effects: denials/corrections and evidence reduce downstream negativity, while toxicity increases it. We also benchmark open-weight LLM prompting baselines and find that added conversational context helps, but attribution remains less reliable, motivating structure-aware counterfactual modeling for social-media analysis.
[NLP-52] AI agents reshape consensus formation in human groups
【速读】： 该论文旨在解决在人机混合群体中，随着生成式 AI（Generative AI）代理比例增加，共识形成机制如何演变的问题。其核心挑战在于揭示人机协同情境下群体规范的生成路径、内容特征及其合法性感知的变化规律。解决方案的关键在于通过设计一个协作描述游戏实验，系统性地调控大语言模型（LLM）代理在群体中的比例，并观察其对共识形成过程的影响。研究发现，存在三种典型的共识形成范式：低比例时由人类主导形成具象、整体且基于真实世界类比的共识；中等比例时因认知差异导致收敛受阻；高比例时则转向由代理主导的抽象、信息密度较低且具有几何分割特征的共识。机制层面，代理影响力源于表达空间中的共享语言先验及其表达选择的稳定性，而人类虽初始排斥与AI交互，但最终在从众压力下逐渐趋同。该研究揭示了代理比例与透明度是影响群体规范生成、内容属性及社会合法性的关键设计变量。
链接: https://arxiv.org/abs/2609.02122
作者: Lin Chen,Ziyi Liu,Xia Hu,Yong Li
机构: Northeastern University (东北大学); Tsinghua University (清华大学); Beijing National Research Center for Information Science and Technology (北京信息科学与技术国家研究中心); Shanghai Artificial Intelligence Laboratory (上海人工智能实验室)
类目: Computation and Language (cs.CL); Computers and Society (cs.CY); Social and Information Networks (cs.SI)
备注:
Abstract:As large language model (LLM) agents shift from tools to participants in human groups, a fundamental question for collective behavior is how their growing presence reshapes consensus formation. Here we study mixed human-AI groups in a collaborative description game, in which shared conventions emerge through repeated rounds of random pairwise communication. Varying the proportions of LLM agents, we identify three distinct regimes of consensus formation: low agent proportions facilitate human-led consensus, intermediate proportions disrupt convergence, and high proportions restore strong consensus while shifting it toward agent-led conventions. Crucially, these regimes differ not only in the strength of convergence, but also in the semantic grounding and communicative form of the resulting consensus: human-led consensus is more concrete, holistic, and grounded in shared real-world analogies, whereas agent-led consensus is more abstract, less information-dense, and more geometrically segmented. Mechanistically, agent influence arises from a shared linguistic prior that places agents near one another in the expression space, combined with relatively stable expression choices across rounds; humans initially resist adopting expressions from partners perceived as AI but gradually yield to conformity pressure. These findings provide evidence that AI composition can shape the emergence, content, and perceived legitimacy of group norms, making agent proportion and transparency important design variables for human-AI systems.
[NLP-53] xt2ql: Multi-Target Natural Language Querying via a Language-Agnostic Intermediate Representation
【速读】： 该论文旨在解决自然语言到数据库查询接口（NL2DB）长期存在的三大结构性问题：仅限于关系型SQL的查询目标、查询时对大语言模型（LLM）推理的无条件依赖，以及生成查询在语义错误时缺乏运行时反馈信号。其解决方案的关键在于提出一个语言无关的中间表示（QueryIR）与可插拔渲染器架构，通过单一七阶段检测流水线同时支持SQL和GraphQL目标。系统提供零LLM的确定性模式，在无需API调用的情况下实现100%执行准确率，中位延迟仅为3.2毫秒；同时，每个生成查询均携带基于加性信号模型计算的运行时置信度分数（范围[0.15, 0.97]）。在Spider和BIRD基准上随机选取的50个查询样本评估显示，基于LLM的模式达到62-70%的精确匹配率和84-91%的执行准确率，而确定性模式在全部100个测试用例中均实现100%执行准确率且无解析错误。消融实验表明，模式感知提示（schema-aware prompting）是提升准确率的核心因素，在两个基准上相较无模式基线带来+18.4个百分点的精确匹配率提升。该框架已开源，采用Apache 2.0许可协议。
链接: https://arxiv.org/abs/2609.02115
作者: Ritesh Kumar
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Databases (cs.DB)
备注:
Abstract:Natural language interfaces to databases have traditionally suffered from three structural limitations: exclusive targeting of relational SQL, unconditional dependence on large language model (LLM) inference at query time, and absence of any runtime signal when generated queries are semantically incorrect. This paper presents text2ql, an open-source Python framework that addresses all three limitations through a language-agnostic Intermediate Representation (QueryIR) and a pluggable renderer architecture. A single seven-stage detection pipeline serves both SQL and GraphQL targets; a zero-LLM deterministic mode delivers 100% execution accuracy at a median latency of 3.2 ms with no API cost; and every generated query carries a runtime confidence score in [0.15, 0.97] computed from an additive signal model. Evaluated on 50-query random samples from the Spider and BIRD benchmarks (indicative results; full-set evaluation is planned), the LLM-backed mode achieves 62-70% exact match and 84-91% execution accuracy; the deterministic mode achieves 100% execution accuracy with zero parse errors across all 100 test cases. An ablation study isolates schema-aware prompting as the dominant accuracy lever, contributing +18.4 percentage points of exact-match gain over the schema-free baseline on both benchmarks. text2ql is publicly available at this https URL under the Apache 2.0 license.
[NLP-54] Predict Dont Iterate: Efficient Adaptive-Length Infilling for Diffusion Language Models EMNLP2026
【速读】： 该论文旨在解决生成式语言模型在填充任务（infilling）中因生成跨度长度固定而带来的局限性，尤其是现有扩散语言模型（Diffusion Language Models, DLMs）在处理动态长度填充时存在的两大问题：一是对初始长度设定高度敏感，需预先指定长度以启动搜索过程，导致结果易受初始值影响；二是推理效率低下，现有方法通过在生成过程中插入长度调整操作或多次执行去噪置信度搜索来寻找合适长度，引入大量额外前向传播，显著增加计算开销。针对上述问题，本文提出PILL（Probing-based Infilling with preset-Length-free decoding），其核心创新在于无需预设初始长度即可实现高效推理，通过基于探测的机制直接确定最优生成长度，避免了重复搜索和复杂操作，从而大幅减少前向传播次数。实验表明，在涵盖五种不同家族、架构与训练策略的DLM模型上，于八个填充基准测试中，PILL相比最强基线平均提升4.8%的通过率（代码任务）和6.0 BLEU-2（文本任务），同时推理速度提升1.82倍，显著提升了生成效率与质量。
链接: https://arxiv.org/abs/2609.02108
作者: Haobo Xu,Sirui Chen,Yuanchen Bei,Lingjie Chen,Yuchen Yan,Dongqi Fu,Jingrui He,Hanghang Tong
机构: University of Illinois at Urbana-Champaign (伊利诺伊大学香槟分校); Meta
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted at EMNLP 2026 (Main Conference)
Abstract:Diffusion language models (DLMs) have emerged as a promising alternative to the auto-regressive paradigm. With bidirectional attention and any-order generation, DLMs naturally fit infilling tasks, which require generating a middle span conditioned on both the prefix and the suffix. However, infilling is sensitive to the length of the span, while DLMs require the length to be fixed before generation. Although prior studies extend DLMs to dynamic lengths, they still suffer from two limitations. (i) Sensitivity to initial length. These methods require a preset length to initialize the search and are highly sensitive to this initial length, often yielding suboptimal results. (ii) Inference inefficiency. They either insert length-changing operations during generation or repeatedly search for an appropriate length using multi-step denoising confidence, both of which introduce substantial extra forward passes and computational cost. Therefore, we propose PILL (Probing-based InfiLling with preset-Length-free decoding), an efficient infilling method for DLMs that requires no preset initial length and adds far fewer extra forward passes than baselines, substantially reducing inference time. Experiments show that, across five DLMs spanning different families, architectures, and training recipes on eight infilling benchmarks, PILL improves over the strongest baseline by +4.8 average pass rate on code and +6.0 BLEU-2 on text, while running 1.82x faster than that baseline. The code is available at this https URL.
[NLP-55] MASkills: Continual Skills Optimization for Multi-Agent LLM Systems
【速读】： 该论文旨在解决基于大语言模型（LLM）的多智能体系统在复杂任务中持续学习与优化的问题，尤其是现有自省方法所构建的经验记忆难以有效调用、精炼与扩展的瓶颈。其核心挑战在于如何将交互经验转化为可复用、可进化且可规模化的能力单元。为此，论文提出MASkills框架，其关键创新在于引入“智能体技能”（agent skills）作为可操作的学习单元——即结构化的程序性知识，明确描述了行动时机、执行方式及资源/工具调用策略。解决方案的核心是构建一个集成技能条件信用分配、分层信用聚合与动量平滑优化的智能体优化流程，使智能体技能库能够通过精炼、归纳、整合与剪枝实现持续演化。实验在HotpotQA、LoCoMo和GAIA等多个基准上验证了该框架在多智能体任务中的有效性。
链接: https://arxiv.org/abs/2609.02094
作者: Huaiyuan Yao,Xiaoou Liu,Charles Fleming,Tianlong Chen,Hua Wei
机构: Arizona State University (亚利桑那州立大学); Cisco Research (思科研究院); University of North Carolina at Chapel Hill (北卡罗来纳大学教堂山分校)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 14 pages, 4 figures
Abstract:LLM-based multi-agent systems have shown strong performance on complex tasks, yet continual improvement from interaction experience remains challenging. Existing self-reflection methods build experience memories, but memories are mostly hard to invoke, refine, or scale, while agent skills offer a more actionable unit: structured procedural knowledge that specifies when to act, how to act, and which resources or tools to use. We introduce MASkills, a continual learning framework that optimizes multi-agent LLM systems through agent skills. MASkills presents a new agent-optimization pipeline that integrates skill-conditioned credit assignment, hierarchical credit aggregation, and momentum-smoothed optimization, enabling agent skill libraries to evolve through refinement, induction, consolidation, and pruning. Experiments on HotpotQA, LoCoMo, and GAIA demonstrate the effectiveness of MASkills across multiple agentic tasks. Our code is available at this https URL
[NLP-56] Selective Knowledge Edit Reversal via Gated Singular Vector Shrinkage EMNLP2026
【速读】： 该论文旨在解决大语言模型中由参数修改引发的恶意知识编辑所带来的安全风险问题，核心挑战在于如何实现对特定编辑内容的精准逆向操作，同时保留其他有益的编辑结果。现有方法多采用全局移除策略，易导致部分有效编辑被误删。为此，本文提出一种基于谱分解的选择性逆向框架，其关键在于假设每次知识编辑在权重矩阵的主子空间中以稀疏方式编码，进而通过分析编辑后权重矩阵的主奇异子空间，定位对编辑敏感的特征成分。实验表明，该方法可在多种场景下有效实现目标知识的定向逆向，同时保持无关编辑内容不变。研究结果验证了不同编辑在主导奇异分量中可被区分的可行性，为实现编辑特异性组件定位与语言模型修复提供了新的思路，证明了选择性谱逆向在精细化知识管理中的潜力。
链接: https://arxiv.org/abs/2609.02091
作者: Weifeng Jiang,Ruirui Chen,Qianren Mao,Junnan Liu,Qili Zhang,Kwok-Yan Lam
机构: Nanyang Technological University (南洋理工大学); Singapore; Digital Trust Centre (数字信任中心); Institute of Advanced Intelligence (先进智能研究所); Agency for Science, Technology and Research (科技研究局); Zhongguancun Laboratory (中关村实验室); Monash University (蒙纳士大学)
类目: Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026 Findings
Abstract:Knowledge editing provides an efficient way to update factual knowledge in large language models. However, malicious edits may introduce safety risks, making it necessary to reverse undesirable editing effects. Existing reversal methods for parameter-modifying edits mainly focus on global removal, which may also erase beneficial edits that should be preserved. In this paper, we study selective reversal of edited knowledge, where the goal is to reverse targeted edited facts while preserving the remaining edited facts. Based on the hypothesis that each edit is sparsely encoded within the dominant subspace of the edited matrix, we propose a spectral-based reversal framework that locates edit-sensitive components within the dominant singular subspace of edited weights. Experiments across multiple settings demonstrate the effectiveness of our method in reversing selected edits while preserving unrelated edited facts. These results suggest that different edits are sparsely encoded within dominant singular components and can be separable when the number of edits is moderate, making selective spectral reversal a promising direction for locating edit-specific components and repairing edited language models.
[NLP-57] IDEEA: training-free Input-Dependent stEEring via Activation cluster matching EMNLP2026
【速读】： 该论文旨在解决现有无需训练的模型对齐方法（如生成式AI中的提示工程或激活向量调优）普遍存在的输入无关性缺陷问题。传统方法通常为所有输入共享单一方向向量进行对齐，忽略了不同输入在激活空间中分布差异及对应最优对齐方向的多样性，这限制了对齐效果的精准性。其核心解决方案是提出一种全新的、无需训练的输入依赖型对齐框架——IDEEA（Input-Dependent stEEring via Activation cluster matching），其关键在于：针对每个注意力头，基于正负样本激活支持集进行聚类分析，并通过求解一个最优匹配问题，构建一组以激活簇条件为依据的、指向目标概念的方向向量集合；在推理阶段，根据当前输入的激活特征动态选择最匹配的簇条件方向进行对齐。该方法揭示了表征空间中同一概念可能分布在多个独立子区域，而非单一连续区域，从而实现更精确且保留原始输入语义的对齐。实验表明，IDEEA在TruthfulQA上的“真实性×信息量”指标平均提升9.9%（最高达23.5%），显著优于最佳的输入无关基线方法。
链接: https://arxiv.org/abs/2609.02089
作者: Zheng Wang,Muchen Li,Renjie Liao,Yan Leng
机构: University of British Columbia (不列颠哥伦比亚大学); Vector Institute for AI (向量人工智能研究所); Canada CIFAR AI Chair (加拿大魁北克人工智能主席); University of Texas at Austin (德克萨斯大学奥斯汀分校)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: Accepted to EMNLP 2026 Findings
Abstract:Steering aligns large language models (LLMs) by injecting a bias into selected activations at inference time, offering a far cheaper alternative to weight-update methods such as supervised fine-tuning or reinforcement learning. However, most existing training-free steering methods are input-independent: a single direction is fitted once and shared across all inputs. This is fundamentally limiting as different inputs occupy different regions of the activation space and admit different optimal steering directions toward the same target concept, much as the gradient with respect to a fixed loss varies from input to input. We close this gap with IDEEA (Input-Dependent stEEring via Activation cluster matching), a training-free framework for input-dependent steering. IDEEA clusters the positive and negative activation supports per attention head, and solves an optimal-matching problem to construct a set of cluster-conditional directions, all about the target concept. At inference time, it picks from this pool of directions and uses the one that best matches the input’s own activation for steering. IDEEA aligns the model toward the target concept while preserving the input’s original representation, evidence that activations encoding a concept occupy several distinct sub-regions of the representation space rather than a single one. IDEEA improves the truth \times info rate in TruthfulQA by an average of 9.9% (up to 23.5%) over the best input-independent baseline.
[NLP-58] XMerge: Cross-Axis Selection and Reconstructive Layer Merging for LLM Depth Compression NEURIPS2026
【速读】： 该论文旨在解决大模型深度压缩（depth compression）中普遍存在的性能退化问题，即现有方法在移除完整Transformer层时往往导致模型质量显著下降，且性能损失在不同模型间呈现不可预测的波动。其核心解决方案是提出一种无需任务标签或端到端微调的后训练压缩方法XMerge，其关键在于两个协同机制：一是跨轴选择（cross-axis selection），通过识别隐藏状态变化幅度与角度均较低的层间区块以确定可安全移除的层；二是局部边界重建（local boundary reconstruction），对相邻保留的层进行重拟合，使其输出逼近原始双层组合的输出。该方法不引入额外的架构变更或推理时参数，保持了标准服务架构的兼容性。实验表明，在7个Llama和Qwen骨干模型（0.5B–8B）、5个基线方法及3种压缩程度下，XMerge在最激进的移除策略（k=4）中表现最优，于CORE（22项任务聚合）和MMLU基准上分别在6/7和5/7模型上领先，且未出现显著困惑度上升。统计分析显示其性能优势具有高度置信度，且在所有14个（模型，压缩策略）组合中均未发生崩溃，零样本与上下文学习场景下均位居前列，同时在首个校准探测中表现最佳。消融实验证明局部重建贡献主要增益，而跨轴融合在选择轴冲突时仍具辅助作用，其额外计算开销可通过数万次请求后的单标记解码效率提升得以回收。
链接: https://arxiv.org/abs/2609.02083
作者: Jundong Hu,Shekar Ramachandran
机构: PayPal AI(贝宝人工智能)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注: Preprint. Under review at a NeurIPS 2026 workshop. 21 pages total, 5 figures, 25 tables
Abstract:Removing complete transformer layers preserves a standard serving architecture, but existing depth-compression methods can lose substantial quality, and the loss varies unpredictably across models. We introduce XMerge, a post-training method with two components. Cross-axis selection identifies a block with low relative-magnitude and angular hidden-state change, and local boundary reconstruction re-fits the adjacent surviving block to match the original two-block output. XMerge uses no task labels or end-to-end fine-tuning, and it introduces neither architectural changes nor additional inference-time parameters. Across seven Llama and Qwen backbones (0.5B-8B), five published baselines, and three layer-reduction levels, its advantage over baselines is largest at the most aggressive removal: at k=4 it ranks first on six of seven backbones on CORE (a 22-task aggregate) and, separately, on six of seven on MMLU (five of seven on both at once), while avoiding the large perplexity increases of several competing operators. In a task-level bootstrap, the 95% confidence intervals for the three largest CORE margins exclude zero; the remaining margins are consistent with ties. Across the 14 (model, regime) cells it is also the only evaluated operator that never collapses, ranking top-2 in both zero-shot and in-context regimes; on a first calibration probe (one backbone) it is the best-calibrated operator. Ablations show that local reconstruction provides most of the gain, while cross-axis fusion helps when the two selection axes disagree. The additional construction cost is recovered through per-token decode savings after roughly tens of thousands of requests.
[NLP-59] ransfer Safety Awareness for Cross-Modal Safety Drift in Multimodal Large Language Models EMNLP
【速读】： 该论文旨在解决多模态大语言模型（Multimodal Large Language Models, MLLMs）中存在的跨模态安全漂移（cross-modal safety drift）问题，即当一个看似无害的文本查询在视觉图像的语境下可能隐含有害意图时，模型的安全响应率显著下降。其核心挑战在于，模型对视觉输入中的风险线索关注度不足，且难以有效触发拒绝机制。解决方案的关键在于提出一种轻量级的方向修正方法——安全意识表征迁移（Safety-awareness Representation Transfer, SRT），通过将源自不安全文本处理的安全信号迁移至视觉模态，增强模型对跨模态潜在危害的敏感性。SRT在保持预训练多模态模型主干（frozen MLLM backbone）不变的前提下，实现了对多种跨模态场景下安全性的有效提升，同时维持了模型的原有功能性能。
链接: https://arxiv.org/abs/2609.02082
作者: Tianqi Xiao,Shiyao Cui,Minghao Zhang,Junxiao Yang,Renmiao Chen
机构: The Conversational AI (CoAI) group, DCST, Tsinghua University(清华大学数据科学与技术研究院对话人工智能团队); Northwestern Polytechnical University(西北工业大学); School of Cyberspace Security, Beijing University of Posts and Telecommunications(北京邮电大学网络空间安全学院)
类目: Multimedia (cs.MM); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Cryptography and Security (cs.CR)
备注: EMNLP Findings
Abstract:Visual modality enhances the capabilities of multimodal large language models (MLLMs) but also introduces a safety concern: a benign textual query may convey harmful intent when grounded in a visual image. We term this cross-modal safety drift and our pilot studies show that the safety response rate for such requests is substantially lower than that for requests containing explicitly unsafe text. This paper aims to systematically study this issue. First, we conduct an empirical analysis to identify representative unsafe response patterns. Building on these, we interpret model representations and attentions, revealing that visually risky cues receive limited attention and weakly trigger refusal. Motivated by the observation that safety signals from unsafe text processing can be transferred, we propose safety-awareness representation transfer (SRT), a lightweight direction-refinement method that mitigates cross-modal safety drift with a frozen MLLM backbone. Experiments across multiple benchmarks and models show that SRT effectively improves safety in diverse cross-modal settings while preserving utility. Code is available at this https URL.
[NLP-60] HyGRAIL: Cost-Aware and Evidence-Grounded Scientific Hypothesis Discovery over Knowledge Graphs
【速读】： 该论文旨在解决科学知识图谱（Scientific Knowledge Graph, SKG）中存在大量缺失的类型化链接（typed links）的问题，这些缺失链接可能代表潜在的科学假说，如材料与应用之间的未探索关联。然而，科学假说发现面临的核心挑战在于：真实假说在候选对中极为稀疏，导致现有方法难以高效且可靠地识别有效假设——图神经网络（Graph Neural Networks, GNNs）虽计算高效但对模糊案例可靠性不足，而大型语言模型（Large Language Models, LLMs）虽具备丰富知识但计算成本过高，且缺乏对图结构的天然依赖性。为此，论文提出HyGRAIL，一种兼顾成本意识与证据可解释性的框架，其核心解决方案在于采用异质GNN初步筛选与LLM深度验证相结合的两阶段机制：首先利用GNN对候选假说进行评分并识别出“验证校准的模糊区域”，仅将图结构不确定性高的案例路由至LLM进行进一步审查；随后，针对每个被路由的假说，从知识图谱中检索节点级关联和多跳关系路径，并通过模板化或基于LLM的自然化方法将其转化为自然语言形式的结构化证据；最终由一个基于自然化证据与验证选择的决策标准进行判断。实验结果表明，在MatKG数据集上，HyGRAIL达到0.429的最高F1分数，相较于最强基线提升0.242，较纯GNN基线提升0.322；同时，GNN预筛使LLM调用率平均降低54.36%。消融实验证明，所检索的图谱证据对于可靠验证至关重要，且紧凑、双面支撑的证据形式比单纯增加检索量更具有效性。
链接: https://arxiv.org/abs/2609.02056
作者: Yihang Sun,Zhihan Zhu,Zhiyuan Jiang,Jingyi Ge,Zixuan Li,Jiaxuan You
机构: University of Illinois Urbana-Champaign(伊利诺伊大学厄本那-香槟分校)
类目: Computation and Language (cs.CL)
备注:
Abstract:Scientific knowledge graphs organize entities and relations extracted from scientific literature, but they remain inherently incomplete. Missing typed links in such graphs can therefore represent plausible scientific hypotheses, such as unexplored associations between materials and applications. However, scientific hypothesis discovery is challenging because true discoveries are extremely sparse among typed candidate pairs: graph neural networks (GNNs) are efficient but unreliable for ambiguous cases, while large language models (LLMs) are knowledgeable but too costly to apply exhaustively and are not naturally grounded in graph structures. We propose HyGRAIL, a cost-aware and evidence-grounded framework that combines heterogeneous GNN triage with LLM-based hypothesis review. HyGRAIL first uses a GNN to score candidate hypotheses and identify a validation-calibrated ambiguous region, routing only graph-uncertain cases to LLM review. For each routed hypothesis, HyGRAIL retrieves node-level associations and multi-hop relational paths from the knowledge graph (KG), then converts this structured evidence into natural language through template-based or LLM-based naturalization. An LLM review agent finally judges each hard hypothesis using the naturalized evidence and validation-selected decision criteria. On MatKG, HyGRAIL achieves the best F1 score of 0.429, improving over the strongest prior baseline by 0.242 F1 points and over the GNN-only baseline by 0.322. Meanwhile, GNN triage reduces the LLM call rate by 54.36% on average. Ablation studies further show that retrieved graph evidence is crucial for reliable hypothesis verification and that compact, two-sided evidence is more effective than simply increasing retrieval quantity.
[NLP-61] Privacy Washing: Detecting Internal Contradictions in Privacy Policies
【速读】： 该论文旨在解决隐私政策中普遍存在但未被系统识别的“隐私洗白”（privacy washing）问题，即政策文本内部承诺与实际数据处理实践之间存在矛盾。其核心挑战在于如何自动识别并验证这些隐性不一致。解决方案的关键在于构建一个四阶段自动化分析流水线：首先通过自然语言处理提取关键声明，继而利用兼容性过滤与自然语言推理（Natural Language Inference, NLI）筛选潜在矛盾，再经由多模型大语言模型（LLM）面板进行联合验证，最终通过主题分析确认矛盾模式。该方法以三模型LLM面板的多数投票机制作为矛盾判定标准，确保结果具备可复现性与一致性。实验结果表明，在2015年和2026年的两组隐私政策样本中，第三方数据共享相关的矛盾是主要类型，且在长达11年的跨度中呈现显著重复性，暗示此类矛盾源于政策结构设计的共性而非单纯的故意欺骗。尽管存在未经过人工专家验证导致精度未知、以及不同阶段滤波配置差异带来的可比性限制等局限，该方法仍成功揭示了隐私政策中系统性矛盾的存在及其稳定性特征。
链接: https://arxiv.org/abs/2609.02055
作者: Thomas Brackin
机构: Varitas(Varitas); Seattle, WA, USA
类目: Computers and Society (cs.CY); Computation and Language (cs.CL)
备注:
Abstract:Privacy policies may contain internal contradictions in which commitments are undermined by practices documented elsewhere in the same policy. We operationalize this phenomenon, privacy washing, through a four-stage pipeline: statement extraction, compatibility filtering and natural language inference screening, multi-model judge verification, and thematic analysis, with contradictions confirmed by majority vote of a three-model LLM panel. Applied to two corpora of website privacy policies, 123 collected in 2026 (OPPT) and 115 collected in 2015 (OPP-115), the pipeline finds the same category patterns recurring across the 11-year gap, with third-party sharing contradictions the majority of confirmed cases in each primary run, consistent with structural factors in policy composition rather than necessarily intentional deception. At least one panel-confirmed contradiction appears in 12.2% of OPPT companies (15/123; 9.8% excluding legacy pairs) and 36.5% of OPP-115 companies (42/115). A stability re-run seven months later, with a fully separated configuration (new extraction models, judges from three Chinese providers absent from both corpora, matched filters, no judge-submission similarity threshold), reproduces the OPPT prevalence under the original protocol (13.0% vs. 12.2%), finds sub-threshold pairs confirm at rates of the same order as those above (raising prevalence to 20.3% and 40.9%), and shows the third-party majority is panel-sensitive while the recurrence of the same category pairs is not. Two caveats govern all figures: panel verdicts are not validated against human expert judgment, so precision is unknown and prevalence figures are lower bounds; and the two primary runs used different filter configurations, so their prevalence difference is not interpretable as a corpus or era effect (the matched re-run reduces the gap to roughly twofold but does not eliminate it).
[NLP-62] A Tri-Agent Framework for Evaluating and Aligning Question Clarification Capabilities of Large Language Models
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在交互式系统中对用户意图理解不精准的问题，尤其针对用户查询存在模糊或信息不足时的澄清对话能力不足这一关键挑战。其解决方案的核心在于提出一种新颖的三智能体框架（tri-agent framework），通过三个基于LLM的代理协同实现对澄清对话能力的稳健评估：（1）提问澄清代理（Question Clarifying Agent, QCA），作为待评估的系统主体，负责识别查询中的歧义并生成澄清性问题；（2）回应代理（Respondent Agent, RA），模拟真实用户的多样化响应，包括无关或具有挑战性的回复，以增强测试场景的真实性；（3）评估代理（Evaluator Agent, EA），作为“大模型裁判”（LLM-as-a-judge），依据涵盖歧义处理能力、问题质量、对话效率、语言得体性及最终意图对齐度在内的多维度指标对对话质量进行量化评估。该框架以供应链领域为例，构建了合成数据生成方法，并验证了评估代理与人工判断的一致性，从而为对话式大模型的澄清能力提供了一套结构化、可复现的基准测试与优化路径。
链接: https://arxiv.org/abs/2609.02054
作者: Yikai Zhao,Saurabh Pandey,Pradeep Kumar Misra
机构: Amazon(亚马逊)
类目: Computation and Language (cs.CL)
备注:
Abstract:Large Language Models (LLMs) are increasingly deployed in interactive systems where understanding user intent precisely is paramount. A key capability for such systems is effective question clarification, especially when user queries are ambiguous or underspecified. This paper introduces a novel tri-agent framework for the robust evaluation of an LLM’s ability to engage in clarifying dialogue. Our framework comprises three distinct LLM-based agents: (1) a Question Clarifying Agent (QCA), the system under evaluation, tasked with identifying ambiguities and posing clarifying questions; (2) a Respondent Agent (RA), designed to simulate human user responses, potentially including irrelevant or challenging replies; and (3) an Evaluator Agent (EA), an LLM-as-a-judge, which assesses the quality of the dialogue based on a comprehensive set of metrics. We detail a methodology for synthetic data generation in the supply chain domain as an example. We propose metrics evaluating ambiguity handling, question quality, dialogue efficiency, language appropriateness, and final intent alignment. We also briefly discuss the validation of the EA against human judgments. This work provides a structured approach to benchmark, validate, and improve the clarification capabilities of conversational LLM applications.
[NLP-63] he Dynamics of Continuous Mixture Collapse in Language Models
【速读】： 该论文旨在解决生成式人工智能（Generative AI）中大语言模型（LLM）在隐状态推理（latent-state reasoning）过程中无法有效保持连续混合状态的问题。现有方法试图用词元嵌入的加权混合等连续状态替代离散中间词元，以保留多种可能的推理路径，但预训练语言模型往往无法维持这些混合状态的完整性。其核心解决方案的关键在于揭示并分析导致混合状态失真的三大独立机制：首先，Transformer架构本身即对混合状态的几何结构产生扭曲，且训练过程进一步放大此效应；其次，即使模型能完美线性传递混合状态，其最终的Softmax读出与自回归反馈共同构成一个动力学系统，会将微小差异放大至某一成分主导，或使不同混合状态相互收缩而不可区分；第三，当混合成分数量增加时，精确保持混合状态通常需要依赖上下文相关的校正项，其维度需求随成分数量增长。研究通过理论分析与受控实证验证了上述机制，并发现预训练模型的推理轨迹普遍位于放大区，印证了理论预测。
链接: https://arxiv.org/abs/2609.02049
作者: Ali Backour
机构: Massachusetts Institute of Technology (麻省理工学院)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注:
Abstract:LLMs latent-state reasoning methods replace discrete intermediate tokens with continuous states, such as weighted mixtures of token embeddings, to retain multiple possible reasoning directions rather than committing to one. Yet pretrained language models often fail to preserve these mixtures. We study why through a combination of theoretical analysis and controlled empirical investigations on a variety of models. We identify three independent, distinct sources of failure. First, transformer architectures already distort mixture geometry, and training substantially amplifies this effect. Moreover, the failure can occur even if the model transports mixtures perfectly linearly: the softmax readout and autoregressive feedback form a dynamical system that either amplifies small differences until one component of the mixture dominates or contracts different mixtures until they become indistinguishable. We verify this theoretical prediction empirically: the observed transition between contraction and amplification occurs near the theoretical threshold derived by our analysis, and pretrained-model rollouts lie predominantly on the amplifying side. Finally, we generalize to mixtures of many components and show that exact preservation generally requires context-dependent correction, whose required dimensionality can grow with the number of components.
[NLP-64] How Output Format Confounds Data Quality and Capability in Instruction Tuning
【速读】： 该论文旨在解决当前大模型指令微调（instruction-tuning）评估中存在的一种根本性偏差问题：即模型性能的衡量结果严重依赖于输出表面格式（surface format），而这一格式本身并未被充分考虑。具体而言，现有方法通过质量指标评估指令微调数据、通过基准测试评估模型能力，但这些评估均通过同一输出接口进行，导致评估结果被接口形式所扭曲。其解决方案的关键在于揭示了模型内部更新方向（update direction）携带真实的能力信号，而接口变化引起的残差（interface-varying residual）并非噪声，而是精确标识每个神经元对应目标任务的结构化信息。研究发现，谱统计量如有效秩虽对界面旋转具有不变性，但对语义扰动不敏感；相反，梯度更新方向能够准确反映模型实际能力。更重要的是，模型能力本质上是相对于训练时所用接口定义的——同一技能在不同接口下可能表现出超过40个百分点的准确率差异，甚至在生成预算调整后使微调效果从增益转为显著损失。因此，当前评估范式往往报告的是“接口”而非“内容”，该研究通过预注册干预实验界定了此几何结构所能实现控制的边界，强调数据质量与模型能力均为接口相关量，亟需建立去接口化的评估体系。
链接: https://arxiv.org/abs/2609.02015
作者: Chengguang Gan,Hanjun Wei,Yunhao Liang,Qinghao Zhang,Shiwen Ni,Zhixi Cai
机构: Independent Researcher(独立研究员); University of Chinese Academy of Sciences(中国科学院大学); Pusan National University(釜山国立大学); Shenzhen University of Advanced Technology(深圳先进技术研究院); Monash University(蒙纳士大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Instruction-tuning data are judged by quality metrics, and tuned models are judged by benchmarks, but both judgments pass through an output interface: the surface format in which an answer is written. Using gradient signatures across 12 tasks, four semantically equivalent interfaces, three model families, and controlled corruptions, we show that this interface confounds both measurements. Spectral statistics such as effective rank are provably invariant to interface rotation and empirically blind to semantic corruption, while the direction of the update carries the quality signal. The interface-varying residual is not noise: it identifies each unit’s own target task perfectly across all three families. Capability itself is stored relative to the training interface: a skill that raises accuracy by more than 40 points under the training format can be nearly invisible under every other, and correcting a single generation budget flips the measured effect of fine-tuning on GSM8K from a gain into a large loss. Pre-registered interventions delimit where this geometry stops short of control. Data quality and model capability are interface-conditioned quantities, and current practice often reports the interface instead of the content.
[NLP-65] rain What You Deploy: Closing the MLP Reachability Gap in Low-Rank Clone Distillation
【速读】： 该论文旨在解决生成式 AI 模型压缩中“训练可达到的权重空间”与“推理时实际部署的权重”之间不一致的问题，即在知识蒸馏过程中，学生模型（student）虽以全宽（full-width）结构进行训练，但其训练过程被限制于教师模型（teacher）诱导的低秩子空间，导致大量独立线性自由度在推理阶段无法被激活，造成“闲置容量”（stranded capacity）。核心问题在于：训练阶段未能覆盖最终部署权重的实际参数空间，从而浪费了潜在性能。解决方案的关键是提出“训练即部署”（train what you deploy）的原则，通过两个可合并的实现方式（Dense-LRC 和 CORE-LRC），在保持原有部署形状、参数量及推理浮点运算量（FLOPs）不变的前提下，将训练目标扩展至整个部署矩阵本身，而非受限于教师诱导的低秩子空间。这一设计使学生模型能够充分探索并利用全部可达权重空间，显著提升性能——在多个教师模型（如 Llama3.2-3B、Llama3.1-8B、Qwen2.5-3B）上，相较于相同预算的基线模型，平均准确率提升达 +2.36/+2.71/+10.45（Avg9），尤其在最宽教师模型（Qwen）上实现 2 倍的训练效率提升；且通过严格控制实验验证，性能增益主要归因于扩大了可训练参数空间，而非额外参数或训练方案改变。最终，仅需约 100 亿蒸馏样本和少量监督微调（SFT），一个半参数量（1.5B）的学生模型即可逼近其约 9 万亿令牌教师模型的多任务平均表现，而 2.7B 学生模型甚至在仅使用约 900 倍更少的蒸馏样本下超越 Meta 官方对 Llama3.1-8B 的压缩结果，证明了该方法在极低数据成本下的高效性与优越性。
链接: https://arxiv.org/abs/2609.02006
作者: Wenhui Chen,Zhifeng Li,Jie Zhou,Navan Preet Singh,Madalina Ciobanu,Chenghua Wang,Qingqing Mao,Ritankar Das
机构: 未知
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注:
Abstract:A compressed student has two shapes that need not agree: the weight it deploys at inference and the weight family its training can reach. We show that a state-of-the-art weight-inheritance distiller, Low-Rank Clone (LRC), deploys a full-width student MLP but ties training to a teacher-induced slice, leaving 62.5-81.4% of each deployed matrix’s independent linear degrees of freedom unreachable-paid for at inference, never trainable. Our principle is one line: train what you deploy. From the identical LRC warm start, we make the training object the entire deployed matrix, with no change in deployed shape, deployed parameter count, or inference FLOPs, via two mergeable realizations (Dense-LRC and CORE-LRC) that both collapse to one deployed weight. This recovers stranded capacity: taking the stronger realization per teacher, +2.36/+2.71/+10.45 Avg9 over matched-budget plain-LRC baselines across three teachers (Llama3.2-3B, Llama3.1-8B, Qwen2.5-3B), with the largest gain on the widest teacher (Qwen), where it reaches the original recipe’s approx. 20B-token accuracy at 10B tokens (2x token efficiency); there the strictly same-lineage arm still recovers +6.39, the fully controlled figure. Controls strongly support attributing the gain to the enlarged reachable set, rather than to added parameters or the recipe. From approx. 10B distillation tokens plus a short SFT, a half-parameter 1.5B student matches its approx. 9T-token teacher’s 9-task macro-average, within evaluation noise and with a residual MMLU deficit, and a 2.7B student beats Meta’s own official compression of Llama3.1-8B at ~900x fewer compression tokens (a token count under unmatched recipes, not a compute claim). All results are from single-seed runs on the LRC backbone.
[NLP-66] NS-Copilot: An LLM -Driven Agent System for Autonomous Neuroscience Analysis EMNLP2026
【速读】： 该论文旨在解决神经科学领域中生成式人工智能（Generative AI）应用受限的核心问题，即跨学科壁垒导致的模型整合与应用效率低下。具体而言，尽管针对生理数据的预训练神经模型发展迅速，但其异构架构和模态特异性限制了系统性集成、选择与评估；同时，现有基于大语言模型（LLM）的智能代理系统普遍缺乏神经科学领域的专业知识，难以有效协调多样化的预训练模型并处理该领域特有的数据类型。为此，本文提出NS-Copilot——一个由大语言模型驱动的多智能体系统，其关键在于通过自然语言接口统一整合领域专用的预训练模型，并构建具备规划、自适应控制、代码生成与结果合成等专业化角色的协同智能体，实现对脑电图（EEG）和细胞外尖峰数据等关键神经科学模态的端到端自动化分析。该系统在阿尔茨海默病、帕金森病及工作记忆尖峰解码等基准任务上均显著优于强基线方法，验证了其在无需特定数据集启发式规则的前提下，实现高效、可扩展神经科学研究分析的能力。
链接: https://arxiv.org/abs/2609.01971
作者: Wuche Liu,Yiran Qiao,Linlin Hou,Rui Yang,Shusen Pu,Song Wang,Jing Ma
机构: Case Western Reserve University(凯斯西储大学); University of West Florida(西佛罗里达大学); University of Central Florida(中佛罗里达大学)
类目: Computation and Language (cs.CL)
备注: Accepted to Findings of EMNLP 2026. 20 pages, 9 figures, 7 tables
Abstract:AI is rapidly advancing neuroscience, yet many laboratories fail to fully unleash its potential due to significant interdisciplinary barriers. While pre-trained neural models for physiological data are progressing quickly, their heterogeneous architectures and modality-specific constraints hinder systematic integration, selection, and evaluation. Despite recent advances in large language model (LLM)-based agent systems for intelligent scientific applications, existing approaches often still lack the domain expertise required to effectively select and coordinate diverse neuroscience pre-trained models and handle unique data types in this domain. We present NS-Copilot, an LLM-driven multi-agent system for neuroscience analysis that autonomously supports end-to-end workflows for diverse professional tasks. It unifies domain-specific pre-trained models and supports key neuroscience modalities, including EEG and extracellular spike data, through a natural-language interface. Given raw data and a task description, NS-Copilot orchestrates agents with specialized roles for planning, adaptive control, code generation, and result synthesis, enabling analysis without dataset-specific heuristics. We evaluate NS-Copilot on neuroscience benchmarks spanning Alzheimer’s disease, Parkinson’s disease, and working memory spike decoding. Across 8 trials per task, the system consistently outperforms strong baselines on the primary metric, demonstrating the ability of NS-Copilot for effective and scalable neuroscience analysis.
[NLP-67] Sparse Readout Prism: Explaining Logit-Lens Scores in Features Instead of Tokens
【速读】： 该论文旨在解决语言模型中“透镜方法”（lens method）在解读中间隐藏状态时存在的歧义问题，即透镜读取结果不仅受隐藏状态影响，还高度依赖于用于解码的读出矩阵（readout，即未嵌入矩阵）。由于不同透镜可能基于不同语料库训练，即使面对相同的隐藏状态，其输出的词元（token）也可能不同，这种现象被称为“语料条件性”（corpus conditionality）。为克服这一问题，论文提出稀疏读出棱镜（Sparse Readout Prism, SRP），通过仅利用读出矩阵本身的权重进行分解，将任意词元的对数几率或对数几率差表示为若干稀疏读出特征贡献之和。这使得读出特征成为分析透镜读取的新基本单元，揭示了传统词元身份所掩盖的内在结构，并支持跨词元、上下文、层和透镜的可比性分析。实验表明，用SRP的稀疏近似替代原始读出矩阵，能比六种基于读出行间几何关系的基线模型多重建8.9–17.3个百分点的对数几率差异；特征消融实验显示，对数几率差异的变化与其在SRP中的贡献成比例。尽管词元读取结果随训练语料变化，但主导读出特征保持稳定。由于SRP构建不依赖任何特定语料库，因此为透镜分析提供了一个与语料无关的对照基准。
链接: https://arxiv.org/abs/2609.01936
作者: Matteo He,William F. Shen,Xinchi Qiu,Nicholas D. Lane
机构: University of Cambridge(剑桥大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 55 pages, 33 figures, 42 tables. Under review. Code: this https URL Dictionaries: this https URL
Abstract:A language model’s prediction of its next token develops across layers, and lens methods track this process by decoding intermediate hidden states into tokens. But a lens reading reflects both the hidden state and the readout (the unembedding matrix) used to decode it. Many lenses are fit on a corpus, and we show that two lenses differing only in their fitting corpus can report different tokens for the same hidden states. We call this dependence corpus conditionality. To examine readout structure independently of the fitting corpus, we introduce Sparse Readout Prism (SRP), which decomposes the readout using only its weights and expresses any token logit or logit difference as a sum of contributions from sparse readout features. This reveals readout features as a new unit of analysis for lens readings, exposing structure that token identities can obscure and enabling comparisons across tokens, contexts, layers, and lenses. Replacing the original readout with SRP’s sparse approximation reconstructs 8.9-17.3 percentage points more of the tested logit differences than the strongest of six baselines built on geometric relations among readout rows. Ablating features shifts logit differences in proportion to their SRP contributions. Although token readings vary with the fitting corpus, the dominant readout feature remains stable. Because SRP uses no corpus in its construction, it provides a control independent of the fitting corpus for lens analyses.
[NLP-68] CRISP: Cliff-awaRe Input-adaptive Sparse Prefilling with Structural-Mass-Motivated Routing EMNLP2026
【速读】： 该论文旨在解决长上下文大语言模型（LLM）推理中注意力预填充阶段（attention prefilling phase）的计算瓶颈问题，其核心挑战在于自注意力机制随上下文长度呈二次方增长，导致显著的计算开销。现有稀疏注意力方法多依赖固定模式或离线分析，缺乏对输入依赖注意力结构的动态适应能力；而近期动态方法虽通过实时路由头至稀疏模式提升灵活性，但其依赖间接路由代理（proxy），存在额外计算开销及预算分配机制忽视后softmax质量分布层级结构的问题。本文提出CRISP（Cliff-awaRe Input-adaptive Sparse Prefilling），关键创新在于识别并解决动态路由范式中的两大结构性问题：首先，发现路由决策可直接从代理注意力图的结构中读取，因而以C_struct替代传统的Jensen-Shannon散度（JSD）路由，该结构代理通过测量垂直斜杠（Vertical-Slash）兼容位置的质量分布，实现与JSD等效的路由效果，同时消除池化矩阵乘法和后续KL散度计算的开销；其次，形式化定义了后softmax质量悬崖（post-softmax mass cliff）现象，并理论证明严格累积覆盖率阈值在长上下文中会引入O(n)量级背景噪声，为此提出基于噪声底限的“汇点感知”（sink-aware）阈值策略以有效规避。实验表明，在InfiniteBench、RULER和LongBench多个基准上，CRISP在两种模型族中均表现最优，于检索密集型任务上达到相对于基线最高+28.0个百分点的性能提升，且在512k token长度下实现最高5.30倍的注意力加速，主要归因于其在选择阶段实现的O(n)噪声消除，同时保持了原始注意力结构完整性。
链接: https://arxiv.org/abs/2609.01925
作者: Huu Huy Nguyen,Chien Van Nguyen,Franck Dernoncourt,Ryan A. Rossi,Linh Ngo Van,Jieyang Chen,Thien Huu Nguyen
机构: University of Oregon (俄勒冈大学); Adobe Research (Adobe 研究院); Hanoi University of Science and Technology (河内科学技术大学)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026 (Main Conference). 16 pages, 7 figures, 12 tables
Abstract:The attention prefilling phase of long-context LLM inference scales quadratically, making self-attention a severe computational bottleneck. Traditional sparse attention methods mitigate this through fixed patterns or offline profiling, but lack the flexibility to adapt to input-dependent attention structure. Recent dynamic methods address this by routing heads to sparse patterns in real-time, but rely on indirect routing proxies with overhead and budget allocation mechanisms that overlook the post-softmax mass hierarchy. We present CRISP (Cliff-awaRe Input-adaptive Sparse Prefilling), which identifies and addresses two structural challenges in this dynamic routing paradigm. First, we show that the routing decision can be read directly off the structure of the proxy attention map. We replace the Jensen-Shannon Divergence (JSD) routing with C_struct, a structural proxy that measures mass at Vertical-Slash compatible positions and reproduces JSD’s routing decisions while eliminating both the pooled matmul and subsequent KL divergence overhead. Second, we formalize the post-softmax mass cliff and demonstrate theoretically that strictly cumulative coverage thresholds accumulate O(n) background noise at long contexts. CRISP navigates this via a sink-aware threshold grounded in the noise floor. Empirically, across InfiniteBench, RULER and LongBench on two model families, CRISP is the strongest sparse method overall and matches or exceeds exact dense attention on retrieval-heavy benchmarks, recovering up to +28.0 pp on retrieval tasks over baselines and achieving up to a 5.30x attention speedup at 512k tokens, driven primarily by our O(n) noise elimination during selection while preserving structural integrity.
[NLP-69] Grounded Compute-Efficient LLM Policy Agents for Energy-Poverty Equity in Physically-Constrained Peer-to-Peer Energy Markets EMNLP2026
【速读】： 该论文旨在解决生成式人工智能（Generative AI）在社会公益领域中对能源贫困（Energy Poverty）问题关注不足的短板，尤其针对现有研究多依赖静态检索/问答或高碳足迹的云端大模型（LLM），形成“计算悖论”的困境。其核心解决方案在于构建一个闭环仿真系统EqGrid，通过低频、开源权重的大语言模型（LLM）政策代理设定电价、碳排放上限及定向补贴策略，而高频多智能体强化学习（Multi-Agent RL）交易者则在物理电网约束（基于IEEE-33-bus系统与动态运行包络）下执行连续双拍卖机制以实现电力市场出清。关键创新点包括：（1）基于真实区域社会人口特征构建具有负荷曲线真实性的家庭人格化模拟主体，确保行为建模的实证基础；（2）引入正式的能源贫困公平性度量指标（如能源负担、能源负担吉尼系数、低收入住房成本占比），量化显示干预措施在不增加电网总成本的前提下显著降低负担不平等；（3）提出计算效率前沿，评估从2350亿参数教师模型压缩至亚10亿参数可部署于笔记本设备的轻量模型时，公平性性能的保留程度，并以每决策能耗/碳排放为单位衡量可持续性。采用解耦安全设计（由LLM设定边界，由验证-投影电网门控执行），实现了零电网约束违规，相较直接由LLM控制的55次违规大幅优化。实验表明，该框架使能源负担吉尼系数从0.351降至0.305，平均负担下降28%，且30亿参数模型可保留95%效益，推理能耗仅为教师模型的1/9；0.8亿参数本地模型亦保持92%效益，能耗降低约24倍，充分体现了高效、可部署、高公平性的技术路径。
链接: https://arxiv.org/abs/2609.01918
作者: Kunal Jadhav,Siddhesh More
机构: Arizona State University (亚利桑那州立大学)
类目: Computation and Language (cs.CL)
备注: 9 pages, 2 figures, 4 tables. Accepted to the 5th Workshop on NLP for Positive Impact (NLP4PI) at EMNLP 2026
Abstract:Energy poverty is nearly absent from NLP-for-social-good, and the little existing work is either static retrieval/QA or relies on carbon-intensive cloud LLMs, a self-defeating “computational irony” for a humanitarian setting. We present EqGrid, a closed-loop simulation in which a low-frequency, open-weight LLM policy agent sets price and carbon bounds and targeted subsidies over a community of empirically-grounded household personas, while high-frequency multi-agent RL traders clear a continuous double auction constrained by a physical distribution grid (IEEE-33-bus with Dynamic Operating Envelopes). Our contribution is threefold and directly addresses how to measure the social impact of AI: (i) grounded personas (region-matched socio-demographics) whose load curves are checked for shape and level realism against real smart-meter data; (ii) formal energy-poverty equity metrics (Energy Burden, Gini of EB, LIHC) showing the intervention reduces burden inequality without raising net grid cost; and (iii) a compute-efficiency frontier that measures how much equity performance survives compressing the policy agent from a 235B teacher down to a sub-1B model deployable on a laptop, in estimated energy/carbon per decision. A decoupled-safety design (the LLM sets bounds; a validate-and-project grid gate executes) yields zero grid-constraint violations versus 55 under direct LLM control. On energy-poverty equity, the LLM policy lowers the Gini of energy burden to 0.305 (from 0.351) and mean burden by 28% while cutting cost (outperforming a tuned rule baseline), and a 3B-active model retains 95% of the benefit at roughly 9x lower inference energy than the teacher, with even a 0.8B on-device model retaining 92% at roughly 24x lower energy. We will release code and configs.
[NLP-70] Accurate in space unreliable in time: how LLM s represent national cultural change
【速读】： 该论文旨在解决当前大型语言模型（LLM）在文化对齐评估中普遍存在的“静态化”问题，即现有评估方法将文化视为单一时间点的快照，仅关注模型对当前文化状态的表征准确性，而忽视了文化随时间演变的动态特性。研究指出，根据文化心理学理论，文化价值观在不同国家以不同速率和方向持续变迁，因此一个真正具备“文化意识”的模型应不仅能反映文化的当下状态，还需捕捉其历史演进轨迹。为此，作者基于超过二十年的世界价值观调查（World Values Survey）数据，对比40个国家在英格尔哈特-韦尔策尔文化地图（Inglehart-Welzel cultural map）上的真实文化轨迹与四种先进（SOTA）LLM生成的文化轨迹。研究发现，尽管模型在空间位置上大致接近各国最新调查结果，但其表征存在显著的时间滞后（数年），且未能充分捕捉实际变化幅度，甚至在无明显变动的国家中引入虚假移动，并极少复现文化轨迹的反转现象。这些结果揭示了模型在时间维度上的“扁平化”缺陷，表明仅依赖静态快照的评估可能掩盖模型在文化演化理解上的根本不足，进而影响模型评价的全面性、引发代表性偏误，并对具备文化意识的AI系统的治理提出挑战。解决方案的关键在于引入动态时间维度，构建能够追踪文化变迁轨迹的多时相评估框架，从而实现对模型文化感知能力的更全面、更真实的衡量。
链接: https://arxiv.org/abs/2609.01902
作者: Yalda Daryani,Miranda Bogen,Madeleine I. G. Daepp
机构: University of Southern California(南加州大学); Center for Democracy and Technology(民主与技术中心)
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 38 pages, 12 figures, 36 tables; includes Supplementary Materials
Abstract:Assessments of cultural alignment have become an important part of the development and improvement of large language models (LLMs). However, the majority of the evaluations treat culture as a single snapshot, investigating only whether a model represents a society accurately at the current time. Research in cultural psychology shows that cultural values change at different rates and directions over time. Therefore, a “culturally aware” model should capture not only where a culture is today but also how it has changed over time. We examine this missing dimension of cultural awareness using more than two decades of the World Values Survey data. We compare the cultural trajectories of 40 countries with the trajectories produced by four state-of-the-art (SOTA) LLMs on the Inglehart-Welzel cultural map. Our findings show that while models generally place countries close to their most recent surveyed positions, these representations tend to lag several years behind that position. They also capture only part of the magnitude of the observed change, introduce movement where little occurred, and rarely reproduce reversals in countries’ trajectories. These findings point to temporal flattening and suggest that snapshot accuracy can give an incomplete picture of cultural awareness in LLMs and have implications for model evaluation, representational harms, and the governance of culturally aware AI systems.
[NLP-71] GAPS: Dimension-Level Gates for Conditional Activation Steering
【速读】： 该论文旨在解决生成式 AI（Generative AI）在文本生成过程中难以有效抑制不良行为（如毒性内容输出）的同时保持模型能力的问题。现有条件化激活调节方法（如CAST和DSAS）虽通过决定干预时机改善了行为与能力之间的权衡，但其在激活时对所有隐藏维度施加完整的密集向量，缺乏对具体神经元的精细化选择，导致资源浪费且可能影响模型性能。本文提出的关键解决方案是引入维度级条件控制（dimension-level conditioning），作为对原有时间粒度条件控制的补充，实现更精细的选择性干预。所提出的GAPS（Gated Activation Steering via Posterior and Separability）方法不依赖训练，包含两个轻量级门控机制：一是静态可分性门（基于AUROC衡量神经元概念信息的统计可靠性），用于筛选具备可靠概念表征的神经元；二是动态后验门（基于高斯模型判断当前激活是否更符合不良概念解释），仅在必要时对相关神经元进行调节。该方法每标记增加约O(D)计算开销，可无缝集成至已有条件化方法中。实验表明，在Gemma-3（4B）和Qwen-3（1.7B）上，GAPS在毒性缓解（RealToxicityPrompts）与概念移除（OneSeC）任务中持续达到或超越其令牌级对应方法的帕累托前沿；在固定能力预算下，DSAS+GAPS将Gemma-3的毒性率从6.52%降至0.48%，显著优于仅使用DSAS的3.52%。消融分析显示，性能提升主要归因于动态后验门的作用。
链接: https://arxiv.org/abs/2609.01878
作者: Moghis Fereidouni,Muhammad Umair Haider,Hassan Sajjad,A.B. Siddique
机构: University of Kentucky(肯塔基大学); Dalhousie University(达尔豪西大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Activation steering suppresses undesired behaviors in language models by adding a steering vector to the hidden state during generation. Recent conditional methods such as CAST and DSAS improve the behavior-capability trade-off by deciding when to intervene, but once active, they apply the full dense vector to all hidden dimensions, regardless of whether a neuron carries concept information or already lies in the desired regime. We introduce dimension-level conditioning as a complementary axis of selectivity that also decides which neurons to intervene on. Our method, GAPS (Gated Activation steering via Posterior and Separability), combines two training-free gates: a static separability gate that restricts steering to neurons with statistically reliable concept information (via AUROC), and a dynamic posterior gate that steers a neuron only when its current activation is better explained by the undesired concept under a Gaussian model. The gates add O(D) overhead per token, and they plug into existing conditional methods. On toxicity mitigation (RealToxicityPrompts) and concept removal (OneSeC) with Gemma-3 (4B) and Qwen-3 (1.7B), GAPS consistently matches or improves the Pareto front of its token-level counterparts; under a fixed capability budget, DSAS+GAPS reduces Gemma-3’s toxicity rate from 6.52% to 0.48%, versus 3.52% for DSAS alone. Ablations attribute most of the gain to the posterior gate.
[NLP-72] hinking effort aligns between humans and reasoning models in abductive reasoning EMNLP2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）与人类在语言及非语言任务中行为一致性的问题，尤其关注生成式人工智能在推理过程中的“思考成本”是否与人类认知努力存在对齐。现有标准LLM多通过偏好对齐进行优化，而大型推理模型（Large Reasoning Models, LRMs）则采用可验证奖励的强化学习训练方式，以鼓励产生正确推理结果而非表面迎合。为更准确评估两者在认知努力上的相似性，研究转向溯因推理（abductive reasoning），因其难度无法从形式结构推断，且难以通过捷径机制被模型模仿，从而提供了更可靠的实证基础来检验模型与人类是否真正经历相似的认知搜索过程。研究发现，LRM与人类在推理成本上表现出进一步的一致性，并且二者倾向于犯下类似的错误；更重要的是，采用允许多路径探索的解码方法显著提升了三类模型在推理成本上的对齐程度，揭示了推理过程多样性对增强人机认知一致性的关键作用。
链接: https://arxiv.org/abs/2609.01867
作者: Henry Arthur
机构: Center for Mind/Brain Sciences, University of Trento (特伦托大学心智/大脑科学中心), Italy
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 14 pages, 5 figures. To appear in Findings of EMNLP 2026
Abstract:A major question in cognitive modeling concerns the behavioral alignment between large language models and humans across linguistic and non-linguistic tasks. Unlike standard LLMs, large reasoning models (LRMs) are optimized with reinforcement learning from verifiable rewards, encouraging correct solutions to reasoning tasks rather than preference-aligned responses. Recent work (de Varda et al., 2025) investigates the cost of thinking in humans and LRMs by comparing human reaction times with model reasoning traces across a range of reasoning tasks. We isolate this alignment by turning to abductive reasoning: unlike deductive tasks, its difficulty cannot be inferred from formal structure and offers no shortcuts a model could exploit to mimic effort without genuine search, providing firmer ground for empirical claims of shared effort. We find further evidence of alignment between LRM and human reasoning effort, as well as evidence that models and humans tend to make similar errors. Finally, we show that decoding methods that let models explore multiple reasoning paths increase alignment in reasoning cost between humans and LRMs across the three models tested.
[NLP-73] he Memory Trust Gap: Capability-Dependent Failures in Persistent-Memory Agents NEURIPS2026
【速读】： 该论文旨在解决生成式 AI（Generative AI）在持续记忆（persistent memory）机制下因存储事实过时而导致的“记忆信任鸿沟”（Memory Trust Gap）问题，即模型过度依赖已过时的记忆信息，即使存在权威性更强的当前证据也未能正确识别。其核心问题是：随着模型能力的提升，这种对过时记忆的过度信任何时开始并如何演变？解决方案的关键在于揭示了“过度信任”并非源于认知混淆，而是由模型规模与特定特征共同决定的能力门控现象。研究通过构建包含“益处”（Benefit）和“安全”（Safety）两套测试场景的基准，在同一模型系列（Qwen3 0.6/1.7/4/8B）上验证发现：在“益处”场景中，所有规模模型均以0.92–1.00的高频率输出过时事实；而在“安全”场景中，损害程度（Δₘₑₘ）随模型规模增长而加剧，当过时信息被伪装成最新时，大模型会迅速崩溃。多因素实验表明，触发过度信任的机制取决于模型规模与具体特征的交互作用，例如移除标签或引入虚假时效性特征（如旧内容标注为新）会显著放大大模型的信任偏差。此外，源权威性（source authority）表现平庸且不随规模变化，位置效应从正向转为负向。通过跨规模直接对比验证了这些尺度依赖关系。缓解策略同样具备能力依赖性：仅对能力强的模型暴露元数据可提升准确性，而只有提前化解冲突才能恢复小模型的性能。该模式在独立的 Llama-Instruct 模型系列及两个外部数据集（RGB、MisBench）中复现，且控制实验显示记忆标签本身并无稳定优势——在小模型中，模型更信任过时文档而非过时记忆，而在8B模型中二者无显著差异。
链接: https://arxiv.org/abs/2609.01852
作者: Jundong Hu,Shekar Ramachandran
机构: PayPal AI
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Preprint. Under review at a NeurIPS 2026 workshop. 14 pages, 7 figures, 11 tables
Abstract:Persistent memory supports personalized agents, but a stale stored fact can override current authoritative evidence without warning. We study when this harm begins as model capability changes. We evaluate a frozen, closed-set, action-scored benchmark with 2 suites that represent 2 different meanings of “no memory” (a Benefit suite, unsolvable without the stored fact, and a Safety suite, in which an authoritative tool always holds the correct value), on a same-family model-size series (Qwen3 0.6/1.7/4/8B). The Memory Trust Gap reflects over-trust rather than confusion. In the Benefit suite, models answer with the stale value 0.92-1.00 of the time at every scale. In the Safety suite, harm below the no-memory baseline under the trap conditions ( \Delta_\mathrmmem ) is capability-gated, with the larger models collapsing most once a stale note is made to look current. In a 2\times2\times2\times2 factorial, which feature triggers over-trust depends on both the feature and model scale. Removing a label amplifies over-trust at every size, and a recency feature (stale dated newer) fools the larger models harder. Source authority is weak and scale-flat, and position changes from positive to negative across the Qwen3 model-size series. We confirm these scale interactions with direct cross-size contrast tests rather than overlapping per-model intervals. Mitigation is likewise capability-dependent: exposing metadata improves accuracy for the capable models, but only pre-resolving the conflict restores accuracy for the 2 smaller checkpoints. The same pattern appears on the capable models in an independent Llama-Instruct model-size series and on 2 external datasets (RGB, MisBench). A framing control finds no consistent advantage for the memory label: at the 3 smaller scales, models trust a stale document more than a stale memory; at 8B, the difference is not significant.
[NLP-74] Candidate Generation and Definition-Guided Verification for Sentence-Level Depression Symptom Recognition
【速读】： 该论文旨在解决句级抑郁症状识别中的核心挑战：相似语言表达在症状相关性上可能存在显著差异，而现有语言模型的推理过程缺乏对诊断定义的充分依赖。为应对这一问题，论文提出一种两阶段框架，其关键在于将症状候选生成与基于诊断定义的验证相分离。第一阶段采用对比微调的句子编码器，从每句话中生成潜在的症状候选；第二阶段则通过微调的语言模型，结合上下文、句子内容及特定候选的诊断定义，判断该症状是否存在，并在输出前对照定义进行校验。该方法在多个基线（包括编码器、基于推理的模型、医学和通用大模型以及单阶段监督分类器）上均取得最优的准确率与F1分数，且生成的解释理由与专家标注高度一致。初步临床审计显示，模型输出与诊断定义具有中等程度的一致性，且解释质量高度依赖于预测准确性。研究结果支持将症状识别分解为候选生成与定义驱动验证的范式，但对罕见类别仍存在性能瓶颈。
链接: https://arxiv.org/abs/2609.01833
作者: Weiming Li,Catarina Barata,Miguel Constante,Joao Sanches
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:Sentence-level recognition of depression symptoms is challenging because similar expressions can differ in symptom relevance, and language-model inference is insufficiently grounded in diagnostic definitions. This study proposes a two-stage framework separating symptom-candidate generation from definition-grounded verification. A contrastively fine-tuned sentence encoder generates a symptom candidate per sentence, and a fine-tuned language model verifies whether the candidate is present or absent using the sentence, its context, and a candidate-specific diagnostic definition, checking its judgment against that definition before answering. Evaluated against encoder, inference-based, medical, and general LLM baselines and a matched single-stage supervised classifier, the proposed pipeline attains the best accuracy and F1 scores of all methods, with rationales matching expert-authored annotations. A preliminary clinical audit indicates moderate alignment with diagnostic definitions, with explanation quality strongly dependent on prediction correctness. The results support decomposing symptom recognition into candidate generation and definition-grounded verification, though performance remains limited for rare categories.
[NLP-75] Interpretable Symptom Vectors for Depression in a Large Language Model
【速读】： 该论文旨在解决抑郁症患者症状异质性在临床实践中被简化为单一严重程度评分所带来的问题，其核心挑战在于如何实现对患者自然语言中多样化抑郁症状的精准、可解释的量化评估。现有大型语言模型（Large Language Models, LLMs）虽具备从言语中捕捉多维度症状的潜力，但其内部表征机制不透明，限制了临床信任度。本文的关键解决方案是通过机制可解释性技术分析Gemma-3-27B-PT模型的残差流（residual stream），发现症状群在第21层通过多种距离度量实现了最优几何分离；进一步利用语义投影（Semantic Projection）将自然语言文本映射至基于临床量表构建的症状向量，结果表明各症状维度的系数仍保持与临床标注一致的排序关系，且第21层存在一个独立的“抑郁向量”（depression vector），可有效区分抑郁与非抑郁文本（AUC = 0.789），并作为情绪效价门控机制控制症状投影范围。研究揭示了模型内部存在解耦且与临床判断一致的症状信号，为开发可解释的抑郁症评估工具提供了可读取的机制基础。
链接: https://arxiv.org/abs/2609.01832
作者: Fangyi Zhu,Ajay Subramanian,Allison Constant,Camille Wang,Ravish Gupta,Corey J. Keller
机构: Stanford University School of Medicine (斯坦福大学医学院); Wu Tsai Neurosciences Institute, Stanford University (吴泰神经科学研究所，斯坦福大学); Kaiser Permanente (凯撒医疗集团); BigCommerce (大商城); Veterans Affairs Palo Alto Healthcare System (帕洛阿尔托退伍军人事务医疗系统)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Neurons and Cognition (q-bio.NC)
备注: 26 pages, 6 figures
Abstract:Patients with depression present with diverse symptom profiles, yet clinical practice routinely reduces this variation to a single severity score. Large language models (LLMs) can potentially capture various symptoms and their severity from patient speech. However, how depressive symptoms are represented inside LLMs remains poorly understood, limiting clinical trust. To examine whether internal model activations match clinician judgment, we analyzed the residual stream of Gemma-3-27B-PT using mechanistic interpretability techniques. Recording activations across symptom descriptions drawn from validated clinical instruments, we found that symptom groups geometrically separated the most at layer 21 across multiple distance metrics. Using Semantic Projection, we then projected held-out naturalistic text onto Symptom Vectors constructed from these instruments. The resulting per-symptom coefficients preserved clinician-annotated rank ordering across mood, somatic, and suicidality axes. Furthermore, a single depression vector in Layer 21 separates held-out depressive from non-depressive text (AUC = 0.789), which can be used as an emotional valence gate that restricts symptom projection to depressive speech. These results reveal a decorrelated, clinician-aligned symptom signal readable directly from internal activations, offering a mechanistic foundation for interpretable depression-assessment tools.
[NLP-76] AVERT: Audio-Verified Adjudication for Spoken Dialogue State Tracking
【速读】： 该论文旨在解决语音对话状态追踪（Spoken Dialogue State Tracking, DST）中因自动语音识别（ASR）错误导致的槽位值不一致问题，这些问题在多轮对话中持续存在，使得该任务既涉及生成又涉及编辑。现有方法虽通过逐轮文本编辑器修正部分错误，但仅基于转录文本，仍无法处理三类可恢复错误：跨轮次预测不一致的值、遗漏的槽位以及音频不支持的值。为此，作者提出AVERT框架，其核心在于结合跨轮一致性评分与经过音频条件训练的验证器，对候选值进行综合评估，并采用三种受限操作——投票（vote）、添加（add）和替换（swap），分别针对常见于特定槽位的错误类型进行修复。实验表明，在SpokenWOZ数据集上，基础语音大模型（speech-LLM）的联合准确率（JGA）为33.04，文本编辑器提升至38.34，而AVERT达到40.13，无需重新训练即可逼近使用完整语音历史的10亿参数端到端系统（39.32）的表现，且仅依赖两个10亿参数解码器。音频验证器带来统计显著的性能增益，同时限制每种操作仅作用于相关槽位子集至关重要——若取消此限制，无约束投票会覆盖正确的分类值，导致性能下降至编辑器水平以下。
链接: https://arxiv.org/abs/2609.01828
作者: Chunggi Lee,Hanspeter Pfister
机构: Harvard University (哈佛大学); Cambridge, Massachusetts, USA
类目: Computation and Language (cs.CL)
备注:
Abstract:Spoken dialogue state tracking recovers slot-value pairs from speech, where ASR errors concentrate in entity values and persist across turns, making it both a generation and an editing problem. A strong per-turn text editor corrects much of this but, operating on the transcript alone, leaves three recoverable errors: a value predicted inconsistently across turns, an omitted slot, and a value the audio does not support. We present AVERT, which scores each candidate value by combining cross-turn agreement with a trained audio-conditioned verifier and resolves the three error types with three operators, vote, add, and swap, each restricted to the slots where its error is common. On SpokenWOZ, a base speech-LLM reaches 33.04 JGA, a text editor 38.34, and AVERT 40.13, without retraining either. This is in the range of a 1B end-to-end system that consumes the full spoken history (39.32), though AVERT uses two 1B decoders rather than one. The audio verifier contributes a statistically significant gain, and restricting each operator to a selected slot subset matters: removing it lets unrestricted voting overwrite correct categorical values and fall below the editor.
[NLP-77] alkFa: A Unified Benchmark for Farsi Dialogue Generation and Understanding
【速读】： 该论文旨在解决波斯语（Farsi）在对话生成与理解任务中缺乏综合性基准评测体系的问题。当前，尽管波斯语拥有超过1.2亿使用者，但尚无系统性、多维度的高质量数据集支持相关自然语言处理研究。为此，论文提出 TALKFA，一个统一的基准框架，包含三个互补的数据集：(1) WIKI-FADIAL，基于维基百科知识的4.2K对话数据，用于知识增强型对话生成；(2) DAILYDIALOG-FA，6.6K标注了对话行为（Dialogue Act）与情感（Emotion）的日常对话数据；(3) PLAYDIAL-FA，2.1K带有情感标签的戏剧化对话数据。解决方案的关键在于采用大语言模型（LLM）辅助构建数据，同时通过母语者主导的多阶段人工审核与修订流程确保数据质量，仅发布经人工验证的最终版本。实验表明，使用低秩适配器（LoRA）微调可显著提升对话生成性能，且仅需25%-50%的训练数据即可恢复超过90%的性能增益；在分类任务中，FABERT在对话行为识别上表现最佳，LORA-MISTRAL-7B在情绪识别中领先，而MISTRAL-24B在情感分析任务中取得最高得分。人类评估与外部独立验证证实了该基准的可靠性，且与GPT-4.1作为评判模型的对比显示，现有自动评估指标对对话质量存在明显高估。零样本测试进一步表明，TALKFA仍为前沿大模型带来挑战。所有数据集、标注规范、代码及模型检查点将公开发布。
链接: https://arxiv.org/abs/2609.01810
作者: Neda Jamshidi,Kamyar Zeinalipour,Fahimeh Akbari,Monica Bianchini,Marco Maggini,Marco Gori
机构: University of Siena (锡耶纳大学); DIISM (信息与系统科学系)
类目: Computation and Language (cs.CL)
备注:
Abstract:Farsi, spoken by more than 120 million people, lacks a comprehensive benchmark for dialogue generation and understanding. We introduce TALKFA, a unified benchmark comprising three complementary datasets: (1) WIKI-FADIAL, 4.2K Wikipedia-grounded dialogues for knowledge-grounded generation; (2) DAILYDIALOG-FA, 6.6K dialogues annotated for dialogue acts and emotions; and (3) PLAYDIAL-FA, 2.1K theatrical dialogues with sentiment labels. While LLMs assist data construction, every dialogue undergoes multi-stage review and revision by native Farsi speakers, and only the final human-approved dialogues are released. Experiments with six LLAMA and MISTRAL models show that LoRA substantially improves dialogue generation while requiring only 25-50% of the training data to recover over 90% of the final performance gains. Across classification tasks, FABERT achieves the best dialogue-act performance, LORA-MISTRAL-7B performs best on emotion recognition, and MISTRAL-24B achieves the highest sentiment score. Human evaluation and independent external validation demonstrate the reliability of the benchmark, while comparisons with GPT-4.1 as an LLM judge reveal that automatic metrics substantially overestimate dialogue quality. Zero-shot evaluation with frontier LLMs further shows that TalkFa remains a challenging benchmark. We will release all datasets, annotation guidelines, code, and checkpoints.
[NLP-78] How Do Prompt Variations Affect Energy Consumption in On-Device LLM s? EMNLP2026
【速读】： 该论文旨在解决生成式 AI（Generative AI）在移动端部署时面临的能耗问题，特别是提示（prompt）设计对本地化大语言模型（LLM）推理能耗的影响尚未被充分研究这一关键瓶颈。其核心问题是：提示的两个属性——认知负荷（cognitive load）与表述模式（phrasing pattern）——如何影响设备端推理的能耗行为。解决方案的关键在于通过细粒度的阶段级能效分析（区分预填充阶段与解码阶段），系统性地揭示不同提示属性对能耗的差异化影响机制：认知负荷主要影响每令牌的能耗成本，而表述模式则通过改变令牌使用量显著影响总能耗。进一步的能效-质量联合分析表明，不同模型对提示设计的响应存在异质性，强调了在实现高效能设备端推理时必须采用模型感知型提示设计策略。
链接: https://arxiv.org/abs/2609.01798
作者: Wei Hu,Xiaolong Tu,Dawei Chen,Yitao Chen,Kyungtae Han,Haoxin Wang
机构: Georgia State University(佐治亚州立大学); Toyota Motor North America(丰田汽车北美公司)
类目: Computation and Language (cs.CL)
备注: Accepted to the EMNLP 2026 Main Conference; camera-ready version
Abstract:Large language models (LLMs) are increasingly deployed on mobile devices, making energy efficiency a key deployment constraint, yet the energy impact of prompt design remains underexplored. This paper aims to understand how two prompt properties, cognitive load and phrasing pattern, shape the energy behavior of on-device LLM inference. We conduct a broad empirical study covering prompt properties, datasets, models, and devices, with phase-level profiling that separates prefill and decode energy. We find that cognitive load primarily affects the energy cost per token, while phrasing pattern affects energy largely through token usage. Our energy-quality analysis further shows that prompt design reshapes the attainable frontier differently across models, highlighting the need for model-aware prompt design in energy-efficient on-device LLM inference. Code, datasets, and scripts are available at this https URL.
[NLP-79] Disentangling Statistical Preemption from Entrenchment in Language Models Avoidance of Overgeneralization EMNLP2026
【速读】： 该论文旨在解决学习者如何避免诸如“Tom laughed me”这类过度泛化现象，而无需显式的否定性证据。为此，建构主义提出了两种解释间接否定性证据的机制：预占（preemption，即优先接触近义结构如“She made him laugh”）与固化（entrenchment，即对动词所有语法用法的全部暴露，包括“He laughed”）。研究通过在基于儿童-照料者对话训练的语言模型（LMs）上开展受控培育实验，系统性地移除具有预占性的证据与非预占性证据，以区分这两种假说。结果发现，尽管语言模型能够避免过度泛化，但在动词特定层面并未表现出显著的预占效应，仅显示出微弱但非零的抽象预占证据。结合对模型训练动态的分析，研究进一步表明，在动词特定条件下，模型将竞争性结构视为间接正向证据而非负向证据。这一发现暗示，若预占是人类避免过度泛化的更合理路径，则神经网络学习者需具备对间接否定性证据的敏感性，从而为未来验证抽象预占的人类实验提供了新方向。
链接: https://arxiv.org/abs/2609.01794
作者: Yixuan Wang,Freda Shi,Kanishka Misra
机构: University of Waterloo(滑铁卢大学); Vector Institute(向量研究所); Canada CIFAR AI Chair(加拿大CIFAR人工智能主席); Department of Linguistics(语言学系); University of Texas at Austin(德克萨斯大学奥斯汀分校)
类目: Computation and Language (cs.CL)
备注: Accepted to EMNLP 2026 Main
Abstract:How do learners avoid overgeneralizations such as Tom laughed me without explicit negative evidence? Constructionists have posited two proposals that describe indirect negative evidence against overgeneralizations: preemption (which privileges exposure to near-synonymous construction—e.g., she made him laugh) vs. entrenchment (all exposures to a verb’s grammatical usages, including cases like He laughed). We disentangle these hypotheses by running controlled rearing experiments on LMs trained on child-caregiver conversations, where we systematically remove preemptive vs. non-preemptive evidence. We find that while LMs avoid overgeneralizations, they do not show preemption at a verb-specific level, instead showing weak but non-zero evidence of abstract preemption. Combined with results from analyzing the LMs’ training dynamics, we find that LMs treat competing structures as indirect positive—as opposed to negative—evidence in the verb-specific condition. Insofar as preemption is the more plausible route to avoiding overgeneralizations in humans, our results point the need for there to be sensitivities to indirect negative evidence in neural network learners, and suggest new human experiments to test abstract preemption.
[NLP-80] VakyArth: Evaluating Prag matic Competence in LLM s across Indic Languages EMNLP2026
【速读】： 该论文旨在解决现有语用评估体系在印地语系（Indic）语言中长期缺失的问题，这些语言具有丰富的语言与文化多样性，但当前主流的语用理解评测仍集中于英语等高资源语言，难以覆盖其特有的语用现象。为此，研究提出VakyArth——首个针对印地语系语言（包括印地语、旁遮普语、泰米尔语和马拉雅拉姆语）的语用基准测试，作为诊断性评估工具，涵盖指称（deixis）、言语行为（speech acts）、隐含意义（implicature）、社会语用（social pragmatics）和连贯性（coherence）五大核心语用现象。其解决方案的关键在于：构建由母语者原创的多任务评估框架，包含多项选择题（MCQ）、自然语言推理（NLI）和翻译任务，以系统性捕捉模型在特定语言文化背景下的语用理解能力。实验结果表明，尽管大语言模型（LLM）在不同语言间表现不一，但普遍在基于印度语言文化惯例的语用理解上存在系统性失败；同时发现，传统自动翻译评估指标无法有效反映输出在语用层面的忠实度，尤其在隐含意义与指称现象上表现不佳，凸显了构建文化敏感型语用评测的重要性。
链接: https://arxiv.org/abs/2609.01788
作者: Usneek Singh,Poorvaja Veera Balaji Kumar,Parth Nanda,Anand Madhusoodanan,Geyang Guo,Wei Xu,Junyi Jessy L
机构: Georgia Institute of Technology(佐治亚理工学院); University of Texas at Austin(德克萨斯大学奥斯汀分校)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Findings of EMNLP 2026
Abstract:Real-world communication often requires pragmatic reasoning: interpreting meanings implied through context and cultural convention rather than stated literally. Existing pragmatic evaluation remains largely limited to English and high-resource languages, leaving Indic languages unexplored despite their linguistic and cultural diversity. We introduce VakyArth, the first pragmatic benchmark for Indic languages, designed as a diagnostic evaluation covering Hindi, Punjabi, Tamil, and Malayalam. VakyArth evaluates models across five phenomena: deixis, speech acts, implicature, social pragmatics, and coherence; through multiple-choice questions, natural language inference, and translation, with all items authored by native speakers. Across multilingual large language models (LLMs) of varying families and sizes, we find consistent failures on pragmatic meanings rooted in Indic linguistic and cultural conventions. Our analysis shows systematic differences across languages and tasks: MCQ accuracy exceeds NLI accuracy in all model-language combinations, translation performance does not reliably track pragmatic understanding, and Indo-Aryan languages show a translation advantage over Dravidian languages. We further show that automatic translation metrics can miss fluent but pragmatically unfaithful outputs, especially for implicature and deixis.
[NLP-81] MemeCULT-1K: Benchmarking South Asian Cultural Context and Humor Understanding of Multimodal Models EMNLP2026
【速读】： 该论文旨在解决当前视觉语言模型（Vision-Language Models, VLMs）在理解南亚地区梗图（meme）时存在的文化语境缺失问题，尤其针对其在隐含文化知识与语用推理方面的不足。现有模型通常仅依赖图像和文本表面信息，难以捕捉梗图背后深层次的文化背景与社会语境，导致理解偏差。为此，研究提出MemeCULT-1K，一个包含1,000张南亚梗图的多语言基准数据集，涵盖孟加拉语、英语和印地语，并为每张梗图配备文化背景注释及三份人工撰写的解释，辅以54个孟加拉地方方言梗图子集。解决方案的关键在于引入最小文化上下文（minimal cultural context）作为外部增强信号，在“仅梗图”与“上下文感知”两种评估设置下验证其有效性。实验结果表明，提供文化上下文可显著提升所有模型在不同语言中的表现：平均SBERT相似度从44.6提升至56.4（+11.8），BLEURT得分从37.3增至42.3（+5.0），LLM-as-a-Judge评分从2.57升至3.43（+0.86）。细粒度错误分析揭示，闭源模型主要失败于实体识别与指代误判，而开源模型则受限于更广泛的跨文化知识鸿沟，且语言与语音层面的错误具有最强的上下文鲁棒性，凸显了文化知识显式建模在梗图理解中的核心挑战。该工作推动未来研究向显式集成文化知识的方向发展。
链接: https://arxiv.org/abs/2609.01772
作者: Tawsif Tashwar Dipto,Mehedi Ahamed,Radib Bin Kabir,Mueeze Al Mushabbir,Mohammed Saidul Islam,Mir Rayat Imtiaz Hossain,Md Tahmid Rahman Laskar,Sabbir Ahmed
机构: Islamic University of Technology(伊斯兰科技大学); South East University(东南大学); Vector Institute(向量研究所); University of British Columbia(不列颠哥伦比亚大学); York University(约克大学); Queen’s University(皇后大学)
类目: Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by EMNLP 2026 (Main Conference)
Abstract:Meme understanding goes beyond recognizing visual content or literal text; it requires implicit cultural knowledge and pragmatic inference that most vision-language models still lack. We introduce MemeCULT-1K, a multilingual benchmark of 1,000 South Asian memes in Bengali, English, and Hindi, where each meme is paired with a cultural context note and three human-written explanations, along with a supplementary set of 54 Bengali regional dialect memes. We evaluate thirteen popular Vision Language Models (VLMs) under two settings: meme-only and context-aware. Providing minimal cultural context yields consistent gains across all models and languages: mean SBERT similarity improves from 44.6 to 56.4 (+11.8), BLEURT from 37.3 to 42.3 (+5.0), and LLM-as-a-Judge scores from 2.57 to 3.43 out of 5 (+0.86). Fine-grained error analysis reveals that closed-source models fail mainly on entity and reference misidentification, while open-source models are bottlenecked by broader cultural knowledge gaps, with linguistic and phonological failures proving the most context-resistant across both. These results highlight the difficulty of culturally grounded meme understanding and motivate future work on explicit cultural knowledge integration. Our dataset and code are publicly available at TawsifDipto17/MemeCULT-1K.
[NLP-82] When Can a Machine Trust a Statute? A Survival Certificate for Machine-Extracted Legal Logic
【速读】： 该论文旨在解决机器在解析法律条文时因提取器间不一致而导致的逻辑可靠性问题，尤其是在面对高噪声环境下，如何确保从机器提取的法定上下文所生成的蕴含关系（implication）仍具备形式逻辑上的可生存性。其核心解决方案是构建一种“被动存活证明”（passive survival certificate），用于验证杜昆-吉约（Duquenne-Guigues）蕴含基（implication basis）中各蕴含关系在存在属性级提取误差下的稳健性：通过测量不同提取器间每项属性的分歧率，在1,000次蒙特卡洛模拟中重现这些分歧，并仅当蕴含关系的单侧威尔逊95%置信下限达到0.95时才予以认证；所有被认证的蕴含均附带前提片段及最小反例。该方法在密苏里州29,365个条款和印度中央法案502个条款上进行了预注册的留出集检验，结果显示在严格精确（7个标题内10个法典族完全匹配）与宽松容忍（11个标题内16个法典族匹配，允许5%误差）条件下均通过验证。然而，在一个全局部署的错误模型下，93.2%的留出章节低于信息量阈值，且2×2因子实验表明此现象主要归因于校准率迁移而非选择偏差。研究结论指出，该证书虽具可用性但具有脆弱性，建议按章节进行校准或采用容错策略部署。相关代码、数据产品及审计轨迹（含一项撤回主张）均已公开发布。
链接: https://arxiv.org/abs/2609.01741
作者: Surya Saka
机构: JudicialMind
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 18 pages, 9 figures, 13 tables (6 main text, 7 appendix). Code, data products, and preregistration to be released on GitHub
Abstract:Statutes are increasingly parsed by machines before people read them, and the parsers disagree: on Missouri’s statutes, two independently written extractors diverge on numeric-threshold presence at a false-negative rate of 0.43. We ask what formal logic survives such noise. We build a passive survival certificate for the Duquenne-Guigues implication basis of machine-extracted statutory contexts: per-attribute inter-extractor disagreement is measured, replayed against the basis in 1,000 Monte Carlo trials, and an implication is certified only when a one-sided Wilson 95% lower bound on survival reaches 0.95; every certified implication carries premise spans and a minimal counterexample. On 29,365 Missouri sections and 502 Indian central-Act sections, the preregistered held-out gate passes (10 statute families across 7 Titles exact; 16 across 11 with 5% tolerance), yet under one globally deployed error model 93.2% of held-out chapters fall below the informativeness floor, and a 2x2 factorial assigns that to calibration-rate transfer, not selection. The certificate is usable but fragile: deploy it per-chapter-calibrated or error-tolerant. Code, data products, and the audit trail, including one retracted claim, are released.
[NLP-83] SpeakPay: Domain-Adaptive LoRA Fine-Tuning of Whisper for Low-Resource Nepali Financial Speech Recognition
【速读】： 该论文旨在解决尼泊尔移动支付应用因图形化界面设计而导致视障用户难以访问的核心问题，提出以语音为首要交互方式的数字钱包SpeakPay，并聚焦于低资源金融领域语音识别的领域自适应技术难题。其解决方案的关键在于构建并利用首个针对尼泊尔语金融语音指令的专用数据集NepFinSpeech-403（包含403条语音命令，涵盖“转账、充值、余额查询”等操作，涉及237个唯一数值），通过引入低秩适配（LoRA）对Whisper large-v2模型进行微调，实现高效的领域迁移。实验表明，经领域自适应后的模型将词错误率（Word Error Rate, WER）从零样本基线的129.95%显著降低至42.58%，相对减少67.2%；同时，天城文数字识别准确率从0.0%提升至73.9%。更关键的是，尽管词级指标存在局限性，任务级表现显著改善：交易成功率达33.33%，较原始系统提升约20倍，且在个体语音片段和各类指令类型上均具统计显著性（符号检验，p < 10⁻¹⁷）。数据效率分析进一步显示，仅需100条领域特定语音即可使零样本WER减半，性能在约300条样本后趋于饱和。误差分析揭示了系统主要失败模式为数值层面的系统性混淆（如零值插入/删除、前缀幻觉），为后续优化指明方向。最终，该模型已部署为公开可访问的语音优先网络应用，相关代码、数据集、模型权重及论文全文均已开源。
链接: https://arxiv.org/abs/2609.01737
作者: Biraj Subedi
机构: 独立研究者(Independent Researcher); github.com/subedibiraj/speakpay
类目: Computation and Language (cs.CL)
备注: 12 pages, 2 figures. Code, dataset, and model weights: this https URL
Abstract:Mobile payment applications in Nepal are graphically mediated and largely inaccessible to visually impaired users. This paper presents SpeakPay, a voice-first digital wallet, and documents the central technical contribution: a controlled study of domain adaptation for low-resource financial speech recognition. We introduce NepFinSpeech-403, a 403-utterance dataset of Nepali financial voice commands (send, load, and balance operations spanning 237 unique numerals), and fine-tune Whisper large-v2 with LoRA. On the held-out test set, the domain-adapted model reduces Word Error Rate from 129.95% (zero-shot baseline) to 42.58% — a 67.2% relative reduction — and improves Devanagari numeral recognition accuracy from 0.0% to 73.9%. We find that word-level metrics understate the practical task-level impact: domain adaptation improves the Transaction Success Rate from 1.67% to 33.33%, a roughly 20x gain. The improvement is consistent at the individual-utterance level (sign test, p 10^-17 ) and across all command types. A data efficiency analysis shows that as few as 100 domain-specific utterances are sufficient to halve the zero-shot WER, with performance plateauing around 300 examples. Error analysis reveals systematic numeral confusion patterns (zero insertion/deletion, prefix hallucination) that account for the majority of remaining transaction failures. The trained system is deployed as a publicly accessible voice-first web application. All code, dataset, model weights, and this paper are released at this https URL.
[NLP-84] Learning Evidence Sufficiency Boundaries for Selective Answering in Grounded Multi-Hop QA
【速读】： 该论文旨在解决多跳问答（multi-hop QA）中生成式模型在证据不足或部分支持情况下仍强行回答的问题，即如何实现“基于证据的可选择性回答”（grounded selective answering），确保模型仅在证据充分时才作答。其核心挑战在于：部分证据可能使不成立的答案显得合理，导致模型误判。解决方案的关键是提出一种名为“证据充分性边界训练”（Evidence Sufficiency Boundary Training）的生成式原生训练框架，通过构建有序证据链，并直接监督模型从“拒绝回答”到“开始回答”的临界点（即证据充分性边界）。该方法融合了层级监督、边界翻转损失（boundary flip margin）、边界后稳定性约束以及答案召回保护机制，从而精确刻画模型在证据量变化下的响应行为。实验基于HotpotQA、2WikiMultiHopQA和MuSiQue数据集构建证据链，在链级指标、原始问答性能及外部非可回答集上的未支持回答率等维度验证，结果表明该方法显著提升了边界定位精度（0.807对比基线0.781）与降低未支持回答率（0.095对比0.101），同时保持了优异的原始问答F1分数，证明了在训练中显式标记“拒答转为回答”的证据阈值对提升可选择性回答能力具有关键作用。
链接: https://arxiv.org/abs/2609.01687
作者: Haruto Sato,Yuki Tanaka,Ren Nakamura,Aoi Kobayashi,Mei Ito
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:Grounded question answering systems should answer only when the supplied evidence supports the answer. In multi-hop QA, this requirement is difficult because partial evidence can make an unsupported answer appear plausible. We study selective answering through evidence sufficiency boundaries: for the same question, a model should abstain under unsupported or partially supported context, answer when the context first becomes sufficient, and keep the answer stable when redundant evidence is added. We introduce Evidence Sufficiency Boundary Training, a generation-native training framework that constructs ordered evidence chains and supervises the abstain-to-answer transition directly. The method combines level supervision, a boundary flip margin, post-boundary stability, and answer recall protection. We build evidence chains from HotpotQA, 2WikiMultiHopQA, and MuSiQue, then evaluate models with chain metrics, raw QA utility, and unsupported-answer rates on external non-answerable sets. With Qwen2.5-3B-Instruct and LoRA adaptation, Evidence Sufficiency Boundary Training gives the strongest boundary localization among the tested systems, with flip accuracy of 0.807 compared with 0.781 for a token-level abstention baseline. It also achieves the lowest overall unsupported-answer rate on external non-answerable evaluation, 0.095 compared with 0.101 for the same baseline, while retaining competitive raw QA F1. The results show that grounded selective answering improves when training marks the evidence level where refusal should give way to answering.
[NLP-85] Ranked by the Matcher: A Reproducibility Audit of Knowledge Graph Extraction from Threat Reports
【速读】： 该论文旨在解决当前威胁情报（Cyber Threat Intelligence, CTI）知识图谱抽取系统评估中存在的一致性与可比性问题，核心在于现有基于三元组F1分数的评估指标高度依赖于预测三元组与标注金标准之间的匹配规则，而不同系统采用的匹配策略不透明且差异显著，导致评估结果不可靠。其解决方案的关键是构建一个名为CTIForge的可复现评估框架，通过引入确定性验证层（deterministic validation layer）实现提取过程字节级一致的前提下，独立调控验证策略，从而分离出组件效应（如后端处理、解码方式、提示工程等）对性能的影响。实验表明，不同的匹配协议可逆转多达11组系统间的相对排序，且同一预测集在不同协议下F1值跨度达0.16–0.70；同时，基于大语言模型（LLM）的判别器在外部校准集上达到86%的共识率，显著优于传统机械匹配器（最高71%），凸显了人工判断与语义理解在评估中的重要性。研究进一步揭示，验证策略的变化不仅影响精度，还引发约2.8倍的实体类型争议行为增加，反映出规则设计与抽取器开发背景间的隐含耦合关系。该工作通过公开完整流水线、评估协议及逐三元组审计记录，推动了威胁情报抽取评估的透明化与标准化。
链接: https://arxiv.org/abs/2609.01671
作者: Safayat Bin Hakim,Houbing Herbert Song
机构: University of Maryland, Baltimore County (马里兰大学巴尔的摩县分校)
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Code and configurations are available at this https URL
Abstract:Security teams and researchers choose knowledge-graph extraction tooling for threat reports on the strength of published triple-F1 scores, yet those scores depend on how predicted triples are matched to gold annotations. We could reimplement the stated matching rule for only five of twelve inspected systems. Re-scoring ten system outputs on shared documents under eight protocols reverses eleven of forty-five pairwise orderings; one fixed prediction set spans 0.16-0.70 F1. On GRID’s external 378-item calibration set, no mechanical matcher (lexical, embedding, or entailment) agrees with multi-reviewer adjudication above 71%, whereas an LLM judge reaches 86%. To separate component effects from matcher rewards, we build CTIForge, whose deterministic validation layer can vary while extraction is held byte-identical. Across seven tested deployment configurations, validation raises precision for all four hosted backbones and lowers it for all three offline backbones. Because backbone, decoding, and backend-specific prompting covary, this is a descriptive split rather than an isolated serving effect. It coincides with a roughly 2.8-fold increase in actions explicitly disputing entity type, consistent with hand-written rules encoding the conventions of the extractor against which they were developed. We release the pipeline, protocol suite, and per-triple audit records.
[NLP-86] Beyond Textual Chain-of-Thought: A Survey on Action-Grounded Reasoning in Autonomous Driving EMNLP2026
【速读】： 该论文旨在解决生成式人工智能（Generative AI）在自动驾驶场景中从传统文本型思维链（Chain-of-thought, CoT）向动作锚定推理（action-grounded reasoning）转变所面临的根本性挑战。传统CoT依赖于离散的、语言化的中间步骤，而自动驾驶决策需输出连续的动作序列，其推理过程必须具备与物理世界一致的时空结构。因此，核心问题在于如何构建可真实落地、实时响应且可在安全关键系统中验证的中间表示（intermediate representations）。该研究提出以“表示”为中心的分类体系，将130篇方法类论文归纳为语言基、视觉-空间、隐变量-动态和外部化四类，进一步细分为13种子类型，聚焦不同兴趣区域的中间状态建模。其解决方案的关键在于：发展能够与现实世界对齐、与实时控制动作耦合、并支持安全性验证的中间表征机制，从而推动自动驾驶智能体实现可解释、可验证、可部署的推理能力。
链接: https://arxiv.org/abs/2609.01659
作者: Zhengxu Tang,Xiaozhou Zhang,Guofeng Cui,Ziyu Gong,Zi Wang,Yunfei Shi,Ruifeng Deng,Chengzhi Qi,Ke Chen,Sachin Patil,Tianjun Xiao,Langechuan Liu,Pichao Wang
机构: NVIDIA(英伟达)
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL); Robotics (cs.RO)
备注: Accepted by EMNLP 2026
Abstract:Chain-of-thought (CoT) reasoning powers generative models by eliciting intermediate steps before producing an answer. In autonomous driving, the answer is a continuous action. Thus its reasoning must share the same spatiotemporal structure as the physical world. This survey studies the resulting shift from textual CoT to action-grounded reasoning. Surveying 171 papers, including 130 method papers and 41 benchmarks, datasets, surveys, and analysis papers, we propose a representation-centered taxonomy that treats the form of the intermediate state as the organizing axis. We systematize the 130 methods into four categories: language-based, visual-spatial, latent-dynamic, and externalized reasoning, further divided into 13 subtypes tied to distinct regions of interests. Our synthesis shows that the open frontier of reasoning in driving agents lies in intermediate representations that can be grounded in the real world, coupled to real-time action, and verified under safety-critical systems. Project page: this https URL.
[NLP-87] PRO-Step: Step-level Process Reward Optimization for Retrieval-Augmented Generation EMNLP2026
【速读】： 该论文旨在解决检索增强生成（Retrieval-Augmented Generation, RAG）在多跳推理（multi-hop reasoning）任务中因早期检索失败导致错误传播的问题。现有方法通常仅对最终答案进行优化，无法有效识别和纠正中间步骤中的检索与推理错误；尽管已有基于过程的监督方法引入了步骤级信号，但仍以最终答案为标准评估每一步，从而可能奖励那些通过偶然正确检索而得到正确答案的虚假成功。为此，本文提出PRO-STEP框架：通过训练一个生成式偏好模型（Generative PRM）来同时评估每一步逻辑有效性与证据支撑性，利用PRM引导的价值树搜索构建对比有效步骤与错误步骤的偏好对，并采用步骤级直接偏好优化（Direct Preference Optimization, DPO）对策略进行优化。该方法实现了对中间推理过程的双重监督，显著提升了多跳问答任务的准确率。实验结果表明，PRO-STEP在五个基准数据集上的平均精确匹配（EM）与F1分数均达到最优水平。
链接: https://arxiv.org/abs/2609.01658
作者: MinKeon Kim,Namjun Lee,Jaekwang Kim
机构: Sungkyunkwan University (成均馆大学); Department of Applied Artificial Intelligence; Convergence Program for Social Innovation
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 22 pages, 7 figures, 23 tables. Accepted to EMNLP 2026
Abstract:Retrieval-Augmented Generation enhances Large Language Models by grounding responses in external knowledge, but multi-hop reasoning remains vulnerable to error propagation, where early retrieval failures confound subsequent steps. Standard outcome-based optimization only rewards the final answer, leaving intermediate retrieval and reasoning errors undetected. While existing process-based methods introduce step-level signals, they still score each step against the final answer, rewarding spurious successes where flawed retrieval coincidentally produces the correct answer. Step-level supervision in RAG requires evaluating both logical validity and evidential grounding at each step. We introduce PRO-STEP: we train a generative PRM that evaluates both dimensions, employ PRM-guided value tree search to construct preference pairs contrasting valid steps against flawed ones, and optimize the policy via step-level Direct Preference Optimization. Experiments on single and multi-hop QA datasets demonstrate that PRO-STEP achieves the best average EM and F1 across five benchmarks. Code, models, and training data are publicly available at this https URL.
[NLP-88] Whose Judgments Count? Representation Gaps in Crowdsourced Content Moderation Produce Unequal Protection from Perceived Toxicity
【速读】： 该论文旨在解决在线内容治理中一个核心问题：在平台依赖人工判断构建内容删除机制的背景下，不同用户群体在面对被认为具有毒性（toxic）的内容时，其受保护程度是否存在不平等，以及这种不平等如何由审核人员（moderator）群体的种族、性别等人口学特征所结构性地决定。其解决方案的关键在于结合大规模的判断数据与反事实模拟（counterfactual simulations），系统性地追踪审核员群体的人口构成如何影响不同用户群体在感知毒性内容下的受保护分布。研究通过对来自16,221名美国受访者对推特（Twitter）、Reddit和4chan上102,463条评论的删除判断进行分析，发现审核需求存在显著的人口异质性，并揭示出“同群保护”（in-group protection）的普遍模式——即与审核员具有相同人口身份的用户更可能获得对其感知毒性内容的保护。尤为关键的是，即便审核员池与Prolific平台上自我申报的审核员人口结构一致，也会加剧保护不平等；而即使采用全国代表性样本，黑人（Black）和性少数群体（LGB）用户仍处于持续被低估保护的状态，除非其在审核员群体中的比例远超其人口占比。这一发现表明，内容审核中对毒性的判定标准存在阶层化（stratified），其聚合过程本身便导致了结构性的保护不平等，因此审核输入的人口构成是决定谁真正获得网络保护的核心因素。
链接: https://arxiv.org/abs/2609.01625
作者: Zhaodi Chen,Byungkyu Lee
机构: University of South Carolina (南卡罗来纳大学); New York University (纽约大学)
类目: ocial and Information Networks (cs.SI); Computation and Language (cs.CL); Computers and Society (cs.CY)
备注:
Abstract:Content moderation is a central form of digital governance, yet people disagree over what content should be removed from shared online spaces. While platforms aggregate human judgments to build moderation systems, it remains unclear how this process shapes which users are protected from content they perceive as toxic. We address this gap by combining large-scale judgment data with counterfactual simulations that trace how the demographic composition of moderator pools shapes the distribution of protection across users. Applying this framework to removal judgments from 16,221 U.S. respondents evaluating 102,463 comments from Twitter, Reddit, and 4chan, we find demographic heterogeneities in moderation demand. We further reveal a consistent pattern of in-group protection: reductions in perceived toxicity accrue disproportionately to users who share the demographic identities of the moderator pool. Crucially, moderator pools that mirror the demographic composition of self-identified moderators on Prolific widen these disparities relative to a nationally representative baseline, while even fully representative pools fail to ensure equal protection: Black and LGB users remain underprotected unless they are represented well beyond their population share. These findings show that unequal protection from perceived toxicity can arise structurally from the aggregation of stratified removal standards, making the demographic composition of moderation inputs a key determinant of who is protected online.
[NLP-89] EvalDetectBench: A Benchmark for Measuring Evaluation Awareness in Frontier Language Models
【速读】： 该论文旨在解决前沿大型语言模型（Large Language Models, LLMs）在评估过程中表现出的“评估意识”（evaluation awareness）问题，即模型能够识别自身正处在评估情境中，并因此改变行为，导致评估结果与实际部署场景下的表现不一致，从而削弱现有人工智能安全框架中评估结果的有效性。其核心解决方案是提出EvalDetectBench——一个开放式的检测管道与基准测试框架，兼容任意支持Inspect的评估流程，使从业者可对当前及未来基准进行评估意识检测。该框架包含一套新整理的对话转录集，涵盖当前前沿模型的系统卡评估与多样化的实际部署来源。EvalDetectBench具备双重功能：一是量化前沿大模型识别评估情境的可靠性，二是评估各基准测试本身的可检测性。研究发现，现有文献中的两个方法论选择引入了系统性偏差：生成部署转录的模型身份贡献了11.25%的测量方差，并可能改变模型排名；而针对某一模型优化的诱导提示在其他模型上表现接近随机。为纠正上述偏差，EvalDetectBench采用基于模型的探针校准（per-model probe calibration）和分层生成器调和（stratified generator-harmonisation）机制，实现更公平、鲁棒的评估意识测量。
链接: https://arxiv.org/abs/2609.01611
作者: Xinning Li,Kemunto Ochwang’i,Aryasomayajula Ram Bharadwaj,Alexandra Souly,Robert Kirk
机构: University of Pennsylvania (宾夕法尼亚大学); LASR Labs (LASR实验室); UK AI Security Institute (英国人工智能安全研究所)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 24 pages, 12 figures, 10 tables. Code: this https URL Data: this https URL
Abstract:Frontier large language models can often recognize when they are being evaluated, a capability known as evaluation awareness. If models behave differently in evaluations than in deployment, this undermines the validity of evaluation results, which are a crucial component of current AI safety frameworks. We introduce EvalDetectBench, an open pipeline and benchmark for measuring evaluation awareness that works with any Inspect-compatible evaluation, allowing practitioners to test against current and future benchmarks. EvalDetectBench ships with a newly curated transcript suite covering current frontier system-card evaluations and diverse deployment sources. The benchmark serves two purposes: measuring how reliably frontier LLMs recognize that they are being evaluated, and assessing how detectable individual benchmarks are as evaluations. We identify two methodological choices in the existing literature that introduce systematic bias: the identity of the model that generated the deployment transcripts accounts for 11.25% of measurement variance and can reorder model rankings; and elicitation prompts selected for high performance on one model can perform near chance on others. EvalDetectBench corrects for both via per-model probe calibration and a stratified generator-harmonisation procedure.
[NLP-90] wo Centuries of Sexism in British Parliament: A Computational Analysis of Womens Representation in the Hansard Corpus EMNLP2026
【速读】： 该论文旨在揭示英国议会两百年来关于女性权利辩论中隐含的系统性性别偏见，即使在支持女性权利的言论中亦存在此类偏见。其核心问题是：立法机构语言如何在看似支持女性权利的语境下，仍持续编码并传播性别歧视模式。解决方案的关键在于运用大语言模型对6,531份跨200年的英国议会辩论记录（Hansard, 1803–2005）进行分析，结合“矛盾性别观量表”（Ambivalent Sexism Inventory, ASI）框架，识别并区分反对与支持女性参政权言论中的性别歧视类型。研究发现，反对女性政治代表权的言论中有54%包含性别歧视内容，而支持者仅21%；且两类立场使用不同类型的性别歧视：反女权话语融合敌意型与善意型性别歧视，而支持者则主要表现为善意型性别歧视。此外，女性议员支持女性政治权利的比例（93%）显著高于男性议员（70%），该差距仅在女性获得选举权后缩小。研究结果证实了矛盾性别观理论在政治话语中的适用性，表明善意型与敌意型性别歧视在不同修辞语境中被策略性地使用。
链接: https://arxiv.org/abs/2608.30485
作者: Mohammad Omar Khursheed,Mandira Sawkar,Ashiqur R. KhudaBukhsh
机构: Rochester Institute of Technology (罗切斯特理工学院)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Machine Learning (cs.LG)
备注: Accepted at EMNLP 2026 (Main Conference Track)
Abstract:The language a legislature uses to debate women’s rights, even in favour of them, encodes systematic patterns of sexism that persist across two centuries. In this work, we analyse 6,531 speeches over 200 years of UK parliamentary debate (Hansard, 1803-2005) by using large language models to classify a speaker’s perspective towards women’s suffrage and political representation, as well as analyse sexist speech in parliament from the lens of the Ambivalent Sexism Inventory. We also release this parliamentary dataset, an organized and metadata-enriched version of the publicly available Hansard Corpus optimized for computational social science research, with 6.7 million speeches across 1.2 million debates, with 89% gender-matching for speeches by MPs from the House of Commons. We find that 54% of speeches opposing women’s representation contain sexist content, compared to 21% of speeches that are for the cause, and that the two sides use fundamentally different types of sexism: anti-suffrage rhetoric combines hostile and benevolent framing, while pro-suffrage sexism is overwhelmingly benevolent. Female MPs support women’s political rights at 93% compared to 70% for male MPs, a gap that closes only after enfranchisement. Our findings are evidence that benevolent and hostile sexism are used in different rhetorical contexts in a manner consistent with the theory of Ambivalent Sexism.
[NLP-91] Dutch Books for Language Models
【速读】： 该论文旨在解决生成式语言模型在提供概率性预测时存在的概率一致性（coherence）问题。具体而言，用户常依赖语言模型进行涉及不确定性判断的重要决策，如重大人生事件、自然灾害或经济走势的预测，而这些预测隐含地依赖于模型具备一个内在一致的世界观。然而，现有语言模型生成的概率预测可能存在逻辑不一致现象，即其输出的概率分布无法满足概率公理，从而可能被套利者利用（Dutch book）实现无风险获利。为此，论文基于de Finetti定理设计了一种评估机制：通过从股票收益率数据生成一系列事件，向语言模型获取其对这些事件的概率估计，并构建线性规划模型以计算最大可能的套利收益（即“荷兰赌利润”），以此量化模型预测的不一致性程度。该方法无需真实结果标签，可在事件未发生或不可观测的情况下进行评估。研究发现，语言模型的概率预测存在显著不一致性，且当事件间逻辑关系更复杂时，不一致性加剧；此外，无关上下文信息甚至可使不一致性提升一个数量级。因此，论文指出，未来可通过改进训练策略（如引入概率一致性约束或强化逻辑推理能力）来提升语言模型在概率预测方面的内在一致性。
链接: https://arxiv.org/abs/2609.02797
作者: Isaiah Andrews,Suproteem Sarkar
机构: MIT(麻省理工学院); NBER(国家经济研究局); University of Chicago(芝加哥大学)
类目: General Economics (econ.GN); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 14 pages, 6 figures
Abstract:People increasingly use language models to support life decisions. Many such decisions involve a probabilistic forecast: How likely is a major life event, a natural disaster, or an economic outcome? Users of language models may implicitly trust that these forecasts fall out of a coherent world model. In this paper, we evaluate the coherence of language model probabilistic forecasts through a procedure that builds on a theorem due to de Finetti. We elicit forecasts from language models across events generated from stock returns data. We then use linear programs to compute the largest Dutch-book profit - the profit an arbitrageur could guarantee by betting against model-generated probabilities - which we use as a measure of incoherence. Our procedure does not require outcome labels, so we can evaluate coherence even in settings where outcomes are not observed or have not yet resolved. We find substantial evidence of incoherence in language model forecasts. Such incoherence increases when there are richer logical relationships between events, and irrelevant contextual details can increase incoherence by an order of magnitude. We conclude by discussing how alternative training strategies may improve probabilistic coherence.
信息检索
[IR-0] Incremental Pooled LLM Evaluation for Cost-Effective Retrieval Model Selection
链接: https://arxiv.org/abs/2609.02745
作者: Max Nelson,Hanoz Bhathena,Aviral Joshi,Saket Sharma
类目: Information Retrieval (cs.IR); Computation and Language (cs.CL)
备注: 10 pages, 1 figure
Abstract:Selecting a retrieval model for a production RAG system requires reliable comparative evaluation, but obtaining relevance judgments at scale is expensive and difficult to repeat as new candidate systems arrive. We study pooled LLM evaluation, in which an LLM judges the union of documents retrieved by the current set of candidate systems, and the pool is then expanded incrementally as new systems are introduced by judging only the new documents they contribute. These judgments are reused to evaluate all systems on a common basis. We validate this approach on four retrieval benchmarks with 11 systems spanning dense, sparse, and hybrid configurations, and deploy it to compare 62 retrieval configurations for a financial news QA system. Pooled LLM rankings correlate strongly with gold-standard evaluation across datasets, and 97% of pairwise system orderings are preserved once bootstrap uncertainty in the qrels is taken into account. In production, document overlap yields 65-80% judgment reuse and up to 4.9x lower evaluation cost, allowing teams to benchmark new retrieval candidates without re-judging previously assessed documents. These results suggest pooled LLM evaluation is a practical and cost-effective workflow for incremental retrieval model selection in deployed systems.
[IR-1] Recommender System as Slow and Fast Thinkers
链接: https://arxiv.org/abs/2609.02671
作者: Zichen Yuan,Xiaoxuan Dong,Linkun Dai,Jinwei Yang,Jining Luan,Dexu Yu,Chunxiao Li,Joemon M. Jose,Youhua Li,Hanwen Du,Junchen Fu
类目: Information Retrieval (cs.IR)
备注: 12 pages, 4 figures
Abstract:Sequential recommendation models are foundational to modern personalized services, yet their effectiveness varies substantially across heterogeneous user environments. In particular, static one-pass recommenders often perform well on common behavior patterns but degrade on operationally challenging user groups, such as users with longer histories or less mainstream item profiles. To address this limitation, we propose \textscDS-Frame, an adaptive fast–slow inference framework for sequential recommendation. \textscDS-Frame combines a Fast System for efficient routine prediction, a Slow System for iterative latent refinement, and a learned selector that routes each sample under a controllable computation budget. Experiments on five real-world datasets show that \textscDS-Frame consistently improves representative sequential recommendation backbones, with larger gains on challenging groups and effective accuracy–efficiency trade-offs. This highlights the potential of adaptive inference for more efficient and robust recommendation. Code is available at \hrefthis https URLthis link.
[IR-2] raining seeds and model-selection stability in recommender-system evaluation RECSYS2026
链接: https://arxiv.org/abs/2609.02499
作者: Juan Manuel Rodriguez,Oleg Lesota,Antonela Tommasel
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注: Accepted RecSys 2026 (ResearchPractice Notes)
Abstract:Recommender-system experiments often rely on a single random training seed, assuming that run-to-run stochasticity has limited impact on evaluation conclusions. This assumption is risky, as a training seed may influence several algorithm-dependent mechanisms, including parameter initialization, mini-batch ordering, dropout, masking, latent sampling, and training-time negative sampling. We examine this assumption by fixing the data partition and varying the training seed across hyperparameter configurations. We analyze seed effects at three levels: user-level metric sensitivity, validation-based model selection and recommendation-list agreement. Results show that seed variation is often detectable. Its impact depends on whether configurations are clearly separated, whether validation results transfer to test, and whether similar scores lead to similar top- k lists. Findings suggest that reporting single-seed results can overstate the stability of recommender system evaluation, and that training seeds should be treated as part of the evaluation protocol rather than as incidental implementation noise.
[IR-3] ViSAR: Training-Free Adaptive-k Retrieval for Visual Document Question Answering
链接: https://arxiv.org/abs/2609.02486
作者: Adrien Mialland,Marc Plantevit,Julien Gallois,Céline Robardet
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV)
备注: 13 pages, 5 figures, 4 tables
Abstract:Document Visual Question Answering (DocVQA) often leverages Retrieval-Augmented Generation (RAG), where late-interaction encoders are commonly used to identify document pages relevant to a user query, before answer generation by a Large Vision-Language Model (LVLM). Existing approaches typically retrieve a fixed top- k number of pages regardless of query complexity, which increases LVLM latency and may degrade answer accuracy. We introduce ViSAR (Visual Semantic Activation Retrieval), a training-free adaptive- k retrieval method for late-interaction visual document retrieval. ViSAR operates directly in the embedding space to construct a query-conditioned page-level similarity matrix that highlights query-relevant semantics and dynamically determines the number of pages to retrieve. Across multiple encoders and LVLMs, ViSAR retrieves compact, query-adapted page sets that reduce RAG latency by up to 58.7%, while maintaining or improving answer accuracy compared with fixed top- k and adaptive retrieval heuristics. Furthermore, we show that the similarity matrix structure correlates with answer accuracy, suggesting future directions for retrieval quality-aware document understanding.
[IR-4] MultiGhostBench: A Multilingual Benchmark for Long-Form LLM -Generated Text Attribution under Distribution Shifts
链接: https://arxiv.org/abs/2609.02379
作者: Matteo Greco,Anudeex Shetty,Andrea Tagarelli,Jey Han Lau
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Digital Libraries (cs.DL); Information Retrieval (cs.IR)
备注:
Abstract:While existing work on LLM authorship attribution (AA) has made progress, available benchmarks remain limited, often focusing on English, controlled settings, or relatively outdated models, with the few multilingual studies considering only relatively short texts. We introduce MultiGhostBench, a multilingual benchmark comprising 928 books generated by five recent LLMs across six languages and three scripts, with an average length of approximately 59K words per book. The benchmark supports evaluation under domain, author, and language shifts. Evaluation of representative AA methods shows that no single method consistently performs best across settings, and performance generally degrades under distribution shifts. Transformer-based detectors can retain generator-related information across languages, although transfer effectiveness varies by language pair, whereas statistical and fingerprint-based detectors are more language-dependent. We envision MultiGhostBench as a valuable resource for the development and evaluation of robust AA methods. The dataset and code can be found at this https URL.
[IR-5] Adaptive Test-Time Inference for Text2Cypher with Trace Budgeting and Selective Refinement
链接: https://arxiv.org/abs/2609.02324
作者: Makbule Gulcin Ozsoy
类目: Information Retrieval (cs.IR)
备注:
Abstract:Large language models have enabled natural language interfaces for structured databases, but generated queries may still contain syntactic errors, violate database schemas, or fail during execution. Test-time inference strategies improve generation reliability without additional training, but existing approaches often use fixed inference budgets and uniform refinement strategies, leading to unnecessary computation across questions with different complexity levels. In this work, we investigate adaptive test-time inference for Text2Cypher and introduce two strategies: adaptive trace budgeting, which dynamically adjusts the candidate generation budget based on question difficulty, and selective execution-guided refinement, which applies correction only when additional inference is expected to be beneficial. Experiments on Gemma-2-9B and Qwen-2.5-7B show that adaptive trace budgeting reduces the average generation budget by 30.7% and wall-clock inference time by 21-25% while maintaining comparable generation quality. Selective refinement preserves nearly all execution success gains of full refinement, reducing execution success by only 0.2-0.5% while avoiding unnecessary refinement for simpler questions. Experiments show that a single correction model (Gemma-4) effectively refines outputs from a different model family, suggesting refinement transfers across families.
[IR-6] Counter-GEO-Bench: Evaluating Defenses Against Information-Distorting Generative Engine Optimization EMNLP2026
链接: https://arxiv.org/abs/2609.02316
作者: Bing Zheng,Zongyao Zhao,Wenming Yang
类目: Information Retrieval (cs.IR); Computation and Language (cs.CL); Cryptography and Security (cs.CR)
备注: Accepted to EMNLP 2026 (Main Conference). 17 pages, 5 figures
Abstract:Generative engine optimization (GEO) enables content producers to increase the visibility of their web pages in generative search engines, but the same techniques can deliver targeted misinformation when adversaries publish ordinary-looking GEO-optimized documents that victim large language models (LLMs) retrieve and synthesize into distorted answers. No existing benchmark evaluates defenses against this threat under controlled conditions. Therefore, we present Counter-GEO-Bench, a defense benchmark that pairs 247 human-verified, quality-gated queries with information-preserving and information-distorting GEO rewrites, and evaluates defenses on attack success rate (ASR), false positive rate, and answer quality across three victim LLMs. Under Counter-GEO-Bench, three off-the-shelf defenses (Granite Guardian, Llama Guard 3, and NeMo Self-Check Fact-Checking) reduce ASR by at most 5.7% relative, while Granite Guardian’s reduction is not statistically significant. Safety-taxonomy guardrails target policy violations, while GEO misinformation passes through them as fluent informational content. To this end, a lightweight benchmark baseline, C-GEO Guard, is proposed, reducing ASR by 47.6% relative with near-zero utility loss, which proves threat tractable.
[IR-7] Genuine Information Needs of Social Scientists Looking for Data
链接: https://arxiv.org/abs/2609.02303
作者: Andrea Papenmeier,Thomas Krämer,Tanja Friedrich,Daniel Hienert,Dagmar Kern
类目: Information Retrieval (cs.IR)
备注:
Abstract:Publishing research data is widely expected to increase its reuse and to inspire new research. In the social sciences, data from surveys, interviews, polls, and statistics are primary resources for research. There is a long tradition to collect and offer research data in data archives and online repositories. Researchers use these systems to identify data relevant to their research. However, especially in data search, users’ complex information needs seem to collide with the capabilities of data search systems. The search capabilities, in turn, depend to a high degree upon the metadata schemes used to describe the data. In this research, we conducted an online survey with 72 social science researchers who expressed their individual information needs for research data like they would do when asking a colleague for help. We analyzed these information needs and attributed their different components to the categories: topic, metadata, and intention. We compared these categories and their content to existing metadata models of research data and the search and filter opportunities offered in existing data search systems. We found a mismatch between what users have as a requirement for their data and what is offered on metadata level and search system possibilities.
[IR-8] Group-Aware Adaptive Retrieval for Evidence Navigation EMNLP2026
链接: https://arxiv.org/abs/2609.02188
作者: June Park,Jun Kwon,Jonghyo Kim,Jongwuk Lee
类目: Information Retrieval (cs.IR)
备注: Accepted to EMNLP 2026; 20 pages, 11 figures, 13 tables
Abstract:Reasoning-intensive retrieval addresses queries whose relevance cannot be identified by surface-level matching, thereby requiring multi-step reasoning. Because relevant documents rarely appear in the initial candidate set, retrieval systems suffer from the bounded recall problem. Existing methods iteratively expand a candidate pool at the document level over a corpus graph, examining each neighbor in isolation and drifting toward a narrow region of the corpus. To address this problem, we propose Group-Aware Adaptive Retrieval for Evidence Navigation (GAREN), which explores the corpus graph through group-level expansion. GAREN organizes documents into semantically coherent and distinguishable groups based on their connections in the corpus graph. The information in each group indicates what can be accessed by expanding through it, providing guidance beyond individual document-level signals. At each iteration, GAREN uses a group-level navigator to identify promising expansion directions, retrieves documents from the selected groups, and applies a document-level reranker to evaluate the updated candidate set. Extensive experiments show that GAREN achieves up to 8.0% improvement over the strongest baseline on BRIGHT. The source code is available at this https URL
[IR-9] GenCAR: Generative Counterfactual Alignment with Risk-Controlled Selection for Out-of-Distribution Recommendation
链接: https://arxiv.org/abs/2609.02162
作者: Qianqian Wang,Yunshan Li,Jiawen Zeng,Wenwu Gong,Lili Yang
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注: 19 pages, 8 figures, 7 tables
Abstract:Serving useful recommendations under distribution shift is crucial for balancing utility and risk in out-of-distribution (OOD) recommendation. However, most existing OOD methods improve ranking or construct counterfactual candidates without controlling the proxy-label false discovery rate (FDR) of the served set. In this work, we formulate OOD serving as the \alpha -Valid Counterfactual Recommendation ( \alpha -VCR) problem to retain candidate support learned from counterfactual supervision while controlling proxy-label FDR, and propose GenCAR, which couples preference-grounded counterfactual supervision with calibrated set selection. In particular, GenCAR fixes the stable-preference representation while intervening on the environmental factor, grounds offline large language model proposals through preference anchors and trust-radius filtering, and uses conformal p -values for Benjamini–Hochberg selection. We theoretically bound conditional counterfactual approximation error and prove finite-sample, distribution-free control of proxy-label FDR under exchangeability and positive regression dependence, with a Benjamini–Yekutieli guarantee under arbitrary dependence. Extensive experiments audit realized proxy false discovery proportions and demonstrate that GenCAR consistently enhances OOD candidate recovery across diverse benchmarks.
[IR-10] Beyond Modality Harmony: Orthogonal Purification and Topology-Guided MoE for Conflict-Aware Multimodal Recommendation ACM-MM2026
链接: https://arxiv.org/abs/2609.02152
作者: Jialin Liu,Zhaorui Zhang,Ray C. C. Cheung
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注: Accepted to ACM Multimedia 2026 (ACM MM 2026)
Abstract:Multimodal Recommender Systems (MRSs) typically rely on a flawed “modality harmony” assumption, presuming that multimodal features are inherently beneficial and strictly aligned with users’ collaborative interaction patterns. However, modality-topology conflicts are ubiquitous in real-world scenarios due to deceptive visual clickbaits and mismatched semantics. Blindly integrating these noisy modalities inevitably pollutes the pristine collaborative space, causing severe representation distortion. To address this, we propose Orthogonal purification and topology-guided MoE for conflict-aware multimodal Recommendation (OrthoRec). At its core, OrthoRec introduces Collaborative-Guided Orthogonal Purification (CGOP), which geometrically decouples multimodal features into directions parallel and orthogonal to a pure collaborative anchor. By adaptively truncating the orthogonal noise with an energy-preserving normalization, CGOP rectifies deceptive semantic directions while preserving the modality’s intrinsic representation capacity. Furthermore, we design a Topology-Aware Routing Mixture-of-Experts (TAR-MoE). Guided by the collaborative topology, TAR-MoE employs decoupled sigmoid gating to break the zero-sum bottleneck of traditional softmax attention, autonomously determining the injection scale for each purified modality. Finally, a safe-SSL objective is introduced to dynamically penalize the forced contrastive alignment of contradictory pairs. Experiments on three real-world Amazon datasets show that OrthoRec consistently outperforms competitive recent baselines and exhibits improved robustness under modality noise and item sparsity.
[IR-11] A Power Law in Logarithms Clothing: On the Scalability of Graph-Based Vector Search
链接: https://arxiv.org/abs/2609.02143
作者: Sajad Faghfoor Maghrebi,Navid Eslami,Niv Dayan
类目: Databases (cs.DB); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注: 17 pages, 10 figures
Abstract:Most vector databases rely on graph-based indexes, notably HNSW and Vamana, for approximate nearest neighbor search. With embedding models widely adopted, the datasets these databases store grow rapidly. At a fixed accuracy, how does search cost scale with dataset size? The prevailing answer is poly-logarithmic growth. Yet the claim is proven only under special conditions and asserted without proof for the indexes used in practice. It is also largely untested: standard benchmarks measure cost at one dataset size, not across sizes. We put the claim to the test. The answer depends on the scale itself. While the dataset size N is small relative to the data’s intrinsic dimensionality, search cost grows as N^c for a constant 0c1 . We call this scaling the Sublinear Power Law. Once N is large enough, growth slows to subpolynomial, consistent with the poly-logarithmic claim. The Sublinear Power Law appears on every dataset, mostly up to its full size, at every recall target, query hardness level, and index configuration we test. The transition to subpolynomial growth appears on the two datasets that grow large enough relative to their intrinsic dimensionality. One mechanism underlies both behaviors: a dataset’s intrinsic dimensionality grows with its size until the data resolves its underlying distribution. Higher intrinsic dimensionality packs more vectors into the query neighborhood the search must examine. We present a unifying theory of beam-search cost that explains our observations. For exact and bounded-degree constructions, we prove the Sublinear Power Law and the eventual transition to poly-logarithmic scaling, and derive the scale at which it occurs. We also develop models that predict the power-law exponents for any recall target and index configuration. These models give a principled way to navigate trade-offs among search cost, insertion cost, and recall as data grows.
[IR-12] Beyond Context Windows: Persistent Discovery Context for Data-Centric Agents
链接: https://arxiv.org/abs/2609.02129
作者: Jalal Mahmud
类目: Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
备注:
Abstract:Data-centric agents repeatedly perform a discovery step before planning or execution: identifying the data objects relevant to a task. Yet successful discovery outcomes are typically discarded rather than reused. We introduce persistent discovery context, a lightweight memory layer that stores prior intent-to-object mappings and reuses them to augment future retrieval. Across three structured data environments, persistent discovery context consistently improves retrieval quality over metadata-only search, remains effective with automatically generated memories, and exposes a reproducible interference failure mode. In lexically sparse domains, memory-only retrieval can even outperform metadata-based retrieval. These findings suggest that discovery outcomes constitute a useful form of reusable context for data-centric agents.
[IR-13] SPAR: Enhancing Industrial-Scale Generative POI Recommendation via Real-World Spatial Perception
链接: https://arxiv.org/abs/2609.02062
作者: Fangye Wang,Yunjin Gu,Haowen Lin,Yifang Yuan,Song Yang,Xiaojiang Zhou,Pengjie Wang
类目: Information Retrieval (cs.IR)
备注:
Abstract:Generative Point-of-Interest (POI) recommendation, autoregressively generating a target POI’s semantic ID (SID), holds great promise for Location-Based Services, where a recommendation helps only if the user can reach it. Yet, existing methods operate within an interest space defined by behavior sequences and collaborative signals, where geography enters only as a textual attribute of the SID, leaving no explicit mechanism to learn or preserve how urban places are related by distance, direction, and reachability; their predictions are thus behaviorally plausible yet far from the user’s real-time location. We argue that such services require injecting real urban spatial knowledge into the interest space, rather than inferring geography from behavior alone. Hence, we propose SPAR, a unified framework whose three synergistic stages jointly construct, cultivate, and preserve urban spatial knowledge: (1) at the tokenization level, Spatially-Intrinsic SID (SI-SID) explicitly encodes longitude–latitude coordinates into a sinusoidal geospatial embedding and fuses it with the textual semantic embedding, producing identifiers via RQ-Kmeans that are simultaneously semantically and geographically consistent; (2) at the cognition level, Multi-Granular Geospatial CPT (MG-CPT) continually pre-trains the base LLM on 25 curated geospatial datasets organized into three tiers of basic attributes, pairwise relations, and city-scale navigation, so that scattered POIs cohere into a connected urban space; and (3) at the adaptation level, Task-Vector Anchored SFT (TV-SFT) anchors the acquired spatial knowledge as a frozen parameter-space task vector to prevent its catastrophic forgetting during behavioral fine-tuning, thereby fusing the two spaces. Extensive quantitative and visualization experiments on two public and four industrial-scale datasets demonstrate the effectiveness of SPAR.
[IR-14] GeoStore: Finding Small Storefronts in Large Scenes – A Fine-Grained POI Localization Benchmark with Global-to-Local Asymmetric Matching ICASSP2027
链接: https://arxiv.org/abs/2609.02012
作者: Lu Han,Xiting Sun,Hao Wang,Zhiqiang Cao,Ruihuan Du,Ziquan Zeng,Chunlong Lv
类目: Computer Vision and Pattern Recognition (cs.CV); Information Retrieval (cs.IR)
备注: 6 pages, 3 figures. Submitted to ICASSP 2027
Abstract:Point-of-interest (POI) localization – matching a user’s close-up storefront photograph against large-scale geo-tagged street-view imagery – underpins map construction, POI verification, and location-based services. Its closest existing paradigm, visual place recognition (VPR), assumes symmetric, whole-image matching of the same scene at a comparable scale; POI localization instead must match a close-up query, in which the target fills the frame, against wide references in which the same POI occupies only a small, off-center region among visually similar shops, under a substantial capture-domain gap. We introduce GeoStore, to our knowledge the first benchmark dedicated to this asymmetric, fine-grained, open-set formulation, and show that global-descriptor methods tuned for symmetric VPR are systematically limited on it, since a single global vector dilutes the small target. We further propose GLAM (Global-to-Local Asymmetric Matching), which couples a retrieval-anchoring global descriptor with an asymmetric local pathway: each reference is kept as a compact set of pooled region tokens and matched against a single query probe through a learnable soft late interaction; at inference, the same tokens enable a lightweight mutual-nearest-neighbor re-ranking. GLAM surpasses strong global and two-stage baselines on Recall@1/5/10 and mAP, with ~5x smaller re-ranking features and ~two orders of magnitude lower per-pair matching cost than prior local re-ranking. The benchmark and code will be publicly released.
[IR-15] Seed-Anchored Budget-Bounded Graph Rendering for Question Answering on Industry-Standard Power-Grid Information and Exchange Models
链接: https://arxiv.org/abs/2609.02011
作者: Jayakumar Manoharan,Yamini Sehgal
类目: ystems and Control (eess.SY); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
备注: Submitted to Engineering Applications of Artificial Intelligence
Abstract:Large language model question answering over power-grid models must respect a fixed context budget. We introduce seed-anchored graph rendering, a deterministic method that prioritizes query-local graph evidence without adding method-specific tuned or learned parameters beyond the shared hop bound and context budget. The method provides a checkable condition under which predefined seed-local answer-bearing render units are preserved in a greedy bounded-context prefix. We evaluate the approach on Common Information Model (CIM) network models exchanged through the Common Grid Model Exchange Standard (CGMES). On two budget-binding CGMES encodings, naive descriptions-first rendering retains local evidence for every single-hop item but only 0.12 and 0.00 of multi-hop items, whereas seed-anchored rendering retains all such evidence. On a preregistered fresh 100-item bank from the SmallGrid topology family, accuracy rises from 0.450 to 0.970 under a fixed 8,000-character context budget. Under a common retrieval and rendering pipeline, the standards-native seed-anchored graph matches or exceeds extracted graph representations produced by LightRAG, Microsoft GraphRAG, and HippoRAG, while avoiding LLM graph-construction tokens. The results are specific to the evaluated CIM/CGMES models, reader, and context budget; they concern budget-bounded retrieval rather than general question answering.
[IR-16] MERGED: Multimodal Entity Resolution via Generated Expert Reasoning Distillation
链接: https://arxiv.org/abs/2609.01913
作者: You-Lin Chen,Kyoungjun Park,Bin Xu,Prithviraj Sen,Pedro Herrero-Vidal
类目: Information Retrieval (cs.IR)
备注: 8 pages, 3 figures, 4 tables
Abstract:In product entity resolution, relationship definitions constantly evolve with business needs, yet adapting to each change traditionally requires slow, costly human annotation that is often noisy and carries no reasoning. Large vision-language models (VLMs) prompted zero-shot can adapt to a new definition immediately and supply the reasoning that human labels lack, but their cost and latency are prohibitive at production scale. We present MERGED, a distillation framework that transfers not just labels but structured reasoning from large teacher VLMs into a compact 7B-parameter student, requiring no human annotation. Multiple teachers label each product pair and articulate the reasoning behind their decision: agreement pairs supply supervised fine-tuning, while disagreements are resolved by a meta-judge into preference pairs for Direct Preference Optimization. Evaluated against human-labeled ground truth on a multilingual e-commerce dataset, the resulting student improves PR-AUC by 13.79% over the same backbone trained on human labels and surpasses the larger Qwen2.5-32B-VL baseline by 6.32% at 6x lower cost, while also yielding tighter label-reasoning alignment (over 10% above Qwen2.5-32B-VL). Moreover, re-applying MERGED from an existing checkpoint adapts to a new relationship definition with only 10K samples, improving PR-AUC by 6.97% over zero-shot and outperforming from-scratch training. MERGED enables rapid adaptation to evolving relationship definitions, supporting a new one in days rather than months, at a cost and latency suitable for large-scale industrial deployment.
[IR-17] ExecRetrieval: Measuring the Functional-Correctness Gap in Code-Embedding Retrieval EMNLP2026
链接: https://arxiv.org/abs/2609.01865
作者: Aaryan Kapoor,Md Abdullah Al Hafiz Khan
类目: oftware Engineering (cs.SE); Computation and Language (cs.CL); Information Retrieval (cs.IR)
备注: Accepted to EMNLP 2026 (Main Conference). Camera-ready version. 17 pages, 6 figures, 8 tables
Abstract:Embedding-based code retrieval is a core component of coding agents and retrieval-augmented code generation, where retrieving correct code matters more than retrieving lexically similar code. Existing code-retrieval benchmarks do not plant controlled, execution-verified single-edit variants of each query’s canonical implementation in the search pool, leaving the question of whether embeddings can functionally discriminate correct from near-clone-but-incorrect code unanswered in a retrieval setting. Resolving this requires a benchmark whose search pool itself contains the relevant counterfactuals – execution-verified buggy variants near-identical to each canonical – so that a retriever’s rank ordering can be directly tested for functional discrimination rather than topical or identity overlap. We introduce ExecRetrieval, 939 Python tasks each paired with one execution-verified canonical implementation and up to four execution-verified buggy distractors, each generated by a mechanical mutation making a single targeted edit, and evaluate 23 dense embedding configurations plus BM25 under provider-native invocation with paired McNemar tests and query-level bootstrap intervals. With near-clone counterfactuals in the pool, the top hosted system reaches exec@10 = 1.00 but only exec@1 = 0.331; rank-1 misses are paired buggy variants 91.5-99.4% of the time across the four leading systems, and the canonical scores below at least one of its four paired distractors in 67-78% of queries on the leading systems. The full dataset, execution oracle, embedding matrices, environment snapshot, and pairwise statistical tests are released at the URL in Appendix D.
[IR-18] Cite or Decline: A Strict Course-Grounded Chatbot for STEM Lecture Videos EMNLP2026
链接: https://arxiv.org/abs/2609.01846
作者: S M Masrur Ahmed,Jaspal Subhlok
类目: Computation and Language (cs.CL); Computers and Society (cs.CY); Information Retrieval (cs.IR)
备注: To appear in the Proceedings of the 2026 Conference on EMNLP 2026, Industry Track
Abstract:Recorded lecture videos, often enhanced with search and summarization features, are a standard study resource. However, students cannot easily ask course specific questions or verify answers against an instructor’s lecture. We report a semester-long deployment of VideoPoints platform with a retrieval-augmented chatbot that answers from course lecture materials and returns timestamped citations. The chatbot retrieves only from the active course, uses chapter summaries to guide transcript ranking, and returns clickable timestamped citations. Students used it for quick lookups and exam review. Across 833 messages, 70.5% included citations, none crossed a course boundary, and when no lecture evidence matched, the chatbot usually declined rather than answering. Among the users, citations were the most consistently useful feature, while practice-question generation was the strongest unmet request. We also evaluated the design on the real-world test split of EduVidQA, a public multimodal benchmark for lecture-video question answering. Our design improved correct-lecture retrieval by 6.3 percentage points over dense-only retrieval. Together, the results show that effective deployment depends on course isolation, supported citations, and alignment with students’ study practices.
[IR-19] Index-Free Dynamic Edge Retrieval with Energy-Tail-Aware Partial Scans
链接: https://arxiv.org/abs/2609.01820
作者: Mohammad Arif Rasyidi,Omar Alhussein
类目: Information Retrieval (cs.IR)
备注: Accepted for presentation at the 2026 IEEE 12th World Forum on Internet of Things (WF-IoT)
Abstract:Dynamic maximum inner-product search (MIPS) returns the K stored vectors with the largest dot products with a query while allowing the dataset to change through insertions, replacements, and deletions. For edge retrieval, the challenge is to achieve high recall and fast queries without making updates expensive. Full-vector scanning keeps updates simple but compares each query with every stored vector, while indexed methods reduce query cost at the expense of maintaining additional structures during updates. We propose ETAR, an index-free method that reduces query work while preserving simple updates. ETAR keeps the query coordinates with the largest squared values until they cover most of its total squared magnitude and treats the rest as a low-magnitude tail. It estimates similarity from the retained coordinates using a compact lower-precision representation, corrects for skipped coordinates, and reranks a fixed number of candidates using full-precision vectors. Across five runs on nine static datasets, ETAR averages 99.2% Recall@10, the fraction of exact top-10 results recovered, while running over 4 \times faster than exact scanning at a representative setting. This speedup also extends to an ARM-based mobile device, where ETAR is up to 6.9 \times faster across four synthetic distributions. Under five streaming workloads, it maintains 100% Recall@10 at every measured point without index rebuilds. Overall, ETAR offers a practical middle ground for dynamic MIPS by reducing query cost while retaining simple, index-free updates. Code is available at this https URL.
[IR-20] hLLM : Single Pass Decoding for Generative Reranking
链接: https://arxiv.org/abs/2609.01807
作者: Emil Laftchiev,Prachi Agrawal,Moe Kayali,Bixing Yan,Qi Xu,Zijie Lei,Chen Qiu,Zhi Hua,Ke Li,Luke Simon
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
备注: 10 pages
Abstract:Large language models (LLMs) achieve state-of-the-art generative ranking quality, but the ranking they produce must be decoded, and autoregressive decoding spends one sequential forward pass per emitted token. We observe that the only tokens a ranker must emit are the N ordinal values naming the items in ranked order, and that this narrow, permutation-structured output format admits decoding strategies which are much more efficient than left-to-right generation. We introduce hLLM (Hungarian LLM), a format-specialized decoding strategy that decodes all N ordinals in O(1) forward passes. hLLM reads an N \times K item-position score matrix off the LLM’s prefill hidden states with a lightweight self-attention head, then decodes the ordinals as the optimal bipartite assignment of that matrix via the Hungarian algorithm, yielding a valid permutation by construction rather than by repair. Through a systematic study of training signals and backbone adaptation, we show that LoRA-based fine-tuning combined with teacher ranking distillation reaches 28 ms end-to-end inference, a speed-up of 64\times while maintaining ranking quality on par with the teacher. We provide a complete ablation decomposing the contributions of architecture, training signal, and backbone adaptation. Our framework connects generative ranking to combinatorial optimization, opening a path toward other O(1) -decode mechanisms for real-time ranking.
[IR-21] KGVoyager: Knowledge Graph Agnostic Question Answering via Agent ic Navigation
链接: https://arxiv.org/abs/2609.01780
作者: Essam Wisam,Chengkai Li
类目: Information Retrieval (cs.IR)
备注: 9 pages, 8 figures, 4 tables
Abstract:Knowledge Graph Question Answering (KGQA) over RDF graphs remains challenging in domain-specific settings, where formal ontologies and curated text-SPARQL pairs are often unavailable. We present KGVoyager, a KG-agnostic agentic architecture that generates SPARQL queries from natural language questions by dynamically discovering graph structure and semantics, requiring only a query endpoint of the underlying graph. Using a think-act-observe loop with search, exploration, and execution tools, KGVoyager maps terms to graph IRIs, uncovers structure, and refines queries through execution feedback - all without pre-existing ontologies or examples. Unlike the prior state of the art, KGVoyager requires only a lightweight class index which renders it applicable for far more real-world endpoints. Across four benchmarks, KGVoyager improves F1 by ~8 points while cutting cost and runtime by ~22% each.
[IR-22] NeoMME: A Single-Tower Multimodal-Native Multilingual Foundation Encoder for Efficient Fine-Tuning and Inference
链接: https://arxiv.org/abs/2609.01657
作者: Aurélien Lac,Tony Wu
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Multimodal models often build on architectures designed for generative vision-language modeling, typically combining separately pretrained vision encoders with causal language models. Visual document retrievers such as ColPali repurpose these models as encoders, carrying over the parameter and compute overhead of a VLM for a non-generative task. We introduce NeoMME, a family of 260M and 800M-parameter Multimodal and Multilingual bidirectional Encoders that process multilingual text and raw image patches in a single bidirectional Transformer encoder. Both models are pretrained from scratch with a masked discrete-diffusion text objective, conditioned on visible image patches for multimodal examples. Both support a 16,384-token context, enough to encode up to two standard 4K UHD images. To demonstrate its downstream capabilities, we fine-tune NeoMME with jointly trained dense and late-interaction heads. On the ViDoRe v3 benchmark, the resulting NeoMME-Retriever 260M outperforms all evaluated models strictly below 800M parameters with 0.523 nDCG@10, while NeoMME-Retriever 800M reaches 0.556. At a matched 2048x2048 image input size on an NVIDIA L40S, NeoMME-260M encodes pages with about 2x the throughput of ColModernVBERT. Hierarchical token pooling and asymmetric quantization compress late-interaction multimodal document embeddings by 255x while preserving over 95% of baseline nDCG@10. We contribute NeoMME to Hugging Face Transformers and release the pretrained backbone and retrieval-compatible checkpoints under Apache 2.0 at this https URL. Subjects: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2609.01657 [cs.IR] (or arXiv:2609.01657v1 [cs.IR] for this version) https://doi.org/10.48550/arXiv.2609.01657 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[IR-23] From Feature Interaction to Feature Transport - A Unified Block for Scalable Recommendation Models KDD
链接: https://arxiv.org/abs/2609.01655
作者: Zichen Luo,Jiachen Guo,Keming Gu,Jie Zhang
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注: 13 pages, 8 figures, accept to KDDCUP2026 Workshop
Abstract:Unified recommendation models aim to jointly model non-sequential multi-field features and sequential user behaviors, but existing interaction-centric designs mainly focus on mixing heterogeneous tokens within each layer. We argue that scalable unified recommendation also requires controlling how intent information is carried, filtered, and preserved across stacked blocks. Inspired by flow-based representation dynamics, we introduce feature transport, a view that treats deep unified recommendation as a discrete context-conditioned representation evolution process. We propose CRAFT, a Contextual Residual Adaptive Feature Transport block, which summarizes non-sequential features into a reliability-aware contextual field and uses it to generate residual displacement and memory-preserving signals for intent and sequence representations. In this way, non-sequential context acts as an active controller of representation evolution rather than a passive object of interaction. In the TAAC2026 advertising recommendation competition, CRAFT achieves a test AUC of 0.838090, surpassing the previous leaderboard-best score of 0.83798. Scaling experiments further show that CRAFT benefits from both depth and width expansion: stacking CRAFT to six blocks improves test AUC to 0.838148, while increasing the hidden dimension reaches 0.838106. These results demonstrate the effectiveness, scalability, and generalization potential of the feature transport paradigm. Source code: this https URL
[IR-24] MELON: A Large-Scale Dataset for Multi-Event Text-to-Long-Video Retrieval
链接: https://arxiv.org/abs/2609.01654
作者: Chan Hur,SeungWoo Song,Jeong-hun Hong,Won Jun Oh,Hyeyoung Park,KyungTae Lim
类目: Information Retrieval (cs.IR)
备注:
Abstract:Existing text-video retrieval datasets primarily consist of short-form clips containing a single dominant event. While suitable for measuring basic vision-language alignment, they are limited in capturing real-world retrieval scenarios, where long-form videos naturally contain multiple semantically distinct events and a single text query may correspond to several non-contiguous temporal segments. To bridge this gap, we introduce MELON, the first large-scale dataset designed to extend text-video retrieval to long-form videos featuring complex, multi-event structures. MELON explicitly annotates multiple event intervals per video along with their corresponding textual descriptions, enabling both training and evaluation of multi-event understanding in long, untrimmed videos. In addition, we propose a multi-event aware loss that encourages models to differentiate between full-event and partial-event matches, yielding substantial improvements in retrieval accuracy. Together, the MELON dataset and our proposed loss establish a robust foundation for expanding text-to-video retrieval to complex long-form scenarios and provide a more realistic evaluation setting for future research in the field.
[IR-25] Not All Matches Are Equally Valuable: An Online Experiment of Retention-Focused Recommendation in a Job-Matching Platform RECSYS
链接: https://arxiv.org/abs/2609.01652
作者: Tatsuya Ute,Chiaki Ichimura,Yuta Saito
类目: Information Retrieval (cs.IR)
备注: RecSys in HR '26: The 6th Workshop on Recommender Systems for Human Resources, in conjunction with the 20th ACM Conference on Recommender Systems, September 28–October 2, 2026, Minneapolis, MN, United States
Abstract:Recommender systems in two-sided matching platforms are commonly optimized for immediate engagement signals such as click-through rate, reply rate, or the total number of successful matches. However, in real-world marketplaces, maximizing matches alone may be misaligned with business goals such as user churn rate and platform revenue, especially when users with fewer matches are at substantially higher risk of churn. In this paper, we study a real job matching platform and show that users with very few recent matches are indeed much more likely to leave the platform, while additional matches for already successful users provide limited marginal value for retention. Motivated by this empirical finding, we formulate a retention-aware recommendation problem and implement a simple post-processing method that adjusts the baseline match-focused ranking to prevent user churn. Specifically, the implemented method gives a score boost to churn-risk users with the goal of increasing their likelihood of obtaining matches and improving retention. We evaluate this practical approach in an online experiment on a real job-matching platform. The treatment group showed directionally lower user churn than the control group, although the estimated effect was not statistically significant at conventional levels, while company-side churn showed no evidence of deterioration. To our knowledge, this is among the first online experimental studies to investigate retention-focused recommendation in a real reciprocal job-matching platform.
[IR-26] he Vocabulary Gap Is an Equity Gap: Register Mismatch in Retrieval Systems for Public-Benefits Access
链接: https://arxiv.org/abs/2609.01645
作者: Krish Sapru
类目: Information Retrieval (cs.IR)
备注: 6 pages, 3 figures, 2 tables
Abstract:Retrieval-augmented question answering is increasingly used to help people navigate public-benefits eligibility, yet the documents these systems retrieve from are written in agency register while intended users often ask questions in plain, informal, or non-native English. We show that this register mismatch can turn a high-performing retrieval system into an inequitable one. We construct a controlled benchmark of 51 publicly documented federal benefit-eligibility rules and 25 information needs, each phrased in both agency register and plain user register while keeping the gold passage fixed. Across BM25, TF-IDF, and a term-graph retriever, formal-register evaluation is nearly perfect (Recall@5 96-100%), but plain-register retrieval collapses (Recall@5 36-44%). For BM25, Recall@1 falls from 84% to 16% and Recall@5 from 100% to 44%, a 56-point equity gap on identical information needs. We trace the mechanism to a measurable vocabulary gap: formal queries share 0.63 of their content terms with the gold passage, while plain queries share only 0.11, a 5.9x reduction. A deliberately simple, auditable plain-to-formal lexicon bridge recovers much of the failure, lifting plain-query BM25 Recall@5 from 44% to 80%. The contribution is not a new retriever; it is an evaluation protocol, benchmark, mechanistic diagnosis, and transparent mitigation for a high-stakes social-impact failure mode that standard retrieval evaluation hides.
[IR-27] Imagine Before Retrieval: Prospective Skill Retrieval for LLM Agents
链接: https://arxiv.org/abs/2609.01642
作者: Shuo Liu,Yutong Yang,Haohao Xiao,Mouxing Yang,Xi Peng
类目: Information Retrieval (cs.IR)
备注:
Abstract:Skill retrieval has recently emerged as a promising paradigm for identifying the desirable execution guidelines from the skill gallery, thus equipping large language model (LLM) agents with the procedural knowledge to accomplish the specified task. To this end, most existing methods customize the retrieval model or reconfigure the retrieval pipeline to prioritize skills that are most semantically relevant to the task query. However, we empirically reveal that task queries and skills are naturally formulated from different perspectives, namely, objective-oriented and procedural-oriented, leading to an under-explored problem termed Query–Skill Misalignment (QSM). Clearly, it is daunting and even impossible to associate the desirable skills in the context of QSM, thus hindering the agent from correctly executing the task. As a remedy, inspired by human prospective cognition, we propose SkillDreamer, a novel framework to alleviate the negative impact of QSM problem. In brief, SkillDreamer first infers the capabilities necessary for task execution, then imagines how to realize these capabilities by generating pseudo skills, and finally leverages such prospective information to bridge the gap between objective-oriented task queries and execution-oriented skills. Extensive experiments on SkillRet and SkillUsage not only verify the effectiveness of SkillDreamer in both skill retrieval and end-to-end task execution, but also demonstrate its generalizability across diverse retrieval models and pipelines. The code will be released upon acceptance.
[IR-28] GRAND-HC: Graph-Refined Author Name Disambiguation
链接: https://arxiv.org/abs/2609.01636
作者: Yuanhao Sun,Zhouyang Jin,Yi Xu,Luoyi Fu,Jiaxin Ding,Xiaoying Gan,Xinbing Wang,Chenghu Zhou
类目: Information Retrieval (cs.IR)
备注:
Abstract:From-Scratch Name Disambiguation (SND) groups papers sharing an ambiguous name into clusters of distinct real-world authors. Existing methods suffer from two critical limitations: (1) inherent long-tailed author distribution biases representation learning, causing over-merging of tail authors; (2) existing cluster number estimation methods are unreliable for long paper sequences, hindering large-scale deployment. We propose \textbfGRAND-HC, a complete end-to-end SND framework. We construct a heterogeneous paper graph via co-author, co-organization, and co-venue relations, using a graph attention network as the embedding backbone. \textbfHarmony Contrastive Learning (HCL) dynamically reweights training loss to suppress overfitting to prolific authors, learning discriminative embeddings. A \textbfGraph-Refined Distance Matrix (GRDM) leverages graph topology to optimize pairwise distances, further preventing tail author over-merging. Meanwhile, a lightweight \textbfPaper Compression Module (PCM) achieves accurate cluster number estimation across varying scales. Finally, Hierarchical Agglomerative Clustering outputs the final clusters. Extensive experiments demonstrate state-of-the-art macro F1 performance. GRAND-HC has been deployed in a billion-scale academic database. Source code: this https URL.
[IR-29] Graph Neural Team Recommendation: An Integrated Approach
链接: https://arxiv.org/abs/2609.01631
作者: Md Jamil Ahmed,Mahdis Saeedi,Hossein Fani
类目: ocial and Information Networks (cs.SI); Information Retrieval (cs.IR)
备注:
Abstract:Team recommendation aims to select an optimal subset of experts who can form an almost surely successful collaborative team for a given set of required skills. State-of-the-art methods are neural multi-label classifiers that transfer dense vector representations of skills into a sparse occurrence vector representing the optimal subset of experts. Such methods, however, overlook experts’ relational and structural information encoded in the expert collaboration graph and, thus, fall short of capturing complex inter-dependencies among experts and their associated skills within teams. Moreover, the skills’ dense vectors are pretrained disjointly and independently of the underlying neural classifier, hence, preventing end-to-end optimization. In this paper, we propose to reformulate the team recommendation problem into end-to-end link predictions in the expert collaboration graph to consume multi-hop intra-team and cross-team collaborations among experts while eschewing the unnecessary complexities of the disjoint two-phase training procedure. Our experiments on two large-scale datasets from various domains with distinct distributions of skills in teams demonstrate the superiority of the end-to-end approach and establish a new state of the art. Our code is available at this https URL.
[IR-30] Marginal Expected Revenue for Jointly Ranking Auction and Fixed-Price Listings in E-Commerce Sponsored Search SIGIR
链接: https://arxiv.org/abs/2609.01628
作者: Greg Kocher,Sanjana Arun
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注: Accepted at the SIGIR eCom’26 Workshop, July 24, 2026, Melbourne, Australia
Abstract:E-commerce search ranking must balance multiple objectives–relevance, user engagement, and platform revenue–when allocating impression slots to competing listings. Estimating the expected revenue component is well understood for fixed-price items, but becomes challenging when marketplace inventory includes mixed listing formats such as pure auctions and hybrid “Auction with Buy It Now” (ABIN) items, where prices evolve dynamically and the final transaction value is unknown at ranking time. Yet auction and ABIN listings account for a meaningful share of inventory and transaction volume on platforms such as eBay, and are a popular format for individual sellers and for unique items with unclear value. We extend the standard Expected Cost-per-Mille (eCPM) framework to auction and ABIN listings by deriving a marginal eCPM (meCPM) that captures the incremental value of showing one more impression of an item whose price is still evolving. The resulting formulation extends the familiar fixed-price eCPM–which is already inherently marginal–to auction dynamics, allowing unified ranking of fixed-price, auction, and ABIN listings under a single objective. We then describe a practical production implementation that approximates this objective, addressing cold-start challenges by bootstrapping from existing engagement models. Online A/B tests at a large e-commerce platform showed positive revenue gains and statistically significant improvements to user metrics, and the system was deployed to production.
[IR-31] he Utility of LLM s in Recommender Systems Explanation Evaluation RECSYS2026
链接: https://arxiv.org/abs/2609.01627
作者: Kathrin Wardatzky,Oana Inel,Luca Rossetto,Abraham Bernstein
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注: 10 pages, 5 figures, Accepted at ACM RecSys 2026
Abstract:Explanations play a crucial role in creating trustworthy recommender systems (RS), yet choosing a good explanation method presents challenges. Many explanation methods exist, but little guidance exists on which is best for which setting. Existing explanation generation methods often produce abstract outputs that require further formatting to become user-friendly, with a seemingly endless pool of options. Running user-based evaluations of all possible options is usually unfeasible, while automated evaluation metrics often either assess only the explainer’s abstract output or require comparison with a ground truth, which is generally unavailable. Recent studies have shown that large language models (LLMs) can serve as ``judges’’ for explanation evaluation, but their reliability has not yet been thoroughly explored. This paper studies the utility of LLMs in selecting an effective explanation method for a given application. We first explore their ability to generate explanation prototypes given varying information about the RS and the user. Specifically, we generate 18 distinct explanation prototypes, which are subsequently evaluated by 14 LLMs of varying sizes across two temperature settings. We compare these against human ratings derived from a user study. Our results show that while LLMs exhibit human-like rating patterns and achieve moderate rank correlation with human raters, their absolute rating agreement is low and varies substantially by model size and evaluation construct. We derive four practical recommendations: keep explanation-generation prompts concise, prefer larger models for evaluation, pre-test evaluation constructs, and audit explanations for factual accuracy, as neither humans nor LLMs reliably detect non-factual content.
[IR-32] RecEvolve: A Knowledge-Driven Autonomous Agent System for Recommender Systems RECSYS’26
链接: https://arxiv.org/abs/2609.01622
作者: Weidi Pan,He Ma,Shuhao Ye,Palaksh Rungta,David McPeek,Junyi Jiao,Arnab Bhadury,Mingyan Gao,Onkar Dalal
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 8 pages, 4 figures, target conference: RecSys '26
Abstract:The rise of agentic AI has catalyzed a shift toward self-iterating systems, opening new frontiers for the autonomous optimization of production recommender models. This paper presents the empirical validation of a knowledge-driven autonomous agent system, deployed directly on a production large-scale Two-Tower retrieval model. By delegating the entire research lifecycle, spanning idea generation, code implementation, offline training, and metric evaluation, to a continuous closed-loop autonomous framework, the agent system executed over 40 completed autonomous training runs from scratch. Executing these runs under rigorous production-scale evaluations, the system systematically navigated hidden architectural bottlenecks on the latest production model to achieve a breakthrough ~20% relative improvement in NDCG, a gain that translated directly to a +3.77% increase in user satisfaction in live production traffic. Furthermore, the deployment exposed critical vulnerabilities in standard evaluation protocols, as the agent system autonomously discovered reward-hacking shortcuts. These findings prove that an autonomous pipeline can dramatically accelerate the pace of machine learning research and stress-test the rigorousness of underlying experimental infrastructure, while also exposing novel challenges such as reward hacking and redundant exploration of failed hypotheses.
[IR-33] When Literature Data Mislead Artificial Intelligence in Materials Discovery
链接: https://arxiv.org/abs/2609.01621
作者: Qian Wang,Ying Li,Ryuhei Sato,Hidemi Kato,Shin-ichi Orimo,Hao Li,Eric Jianfeng Cheng
类目: Information Retrieval (cs.IR); Materials Science (cond-mat.mtrl-sci); Computational Engineering, Finance, and Science (cs.CE); Machine Learning (cs.LG)
备注: 3 main figures, 2 supplementary tables, research article on data reliability for AI-driven materials discovery
Abstract:Artificial intelligence (AI) increasingly treats scientific literature as a data source for building databases, training predictive models, and guiding discovery. Yet literature-derived datasets often assume that reported experimental values are internally consistent and directly reusable. Here, we analyze this assumption using solid electrolyte (SE) conductivity data as a representative materials-science case. By tracing values from source articles to curated datasets, we identify recurrent text-figure mismatches, ambiguous axis annotations, unit inconsistencies, and missing measurement context. These discrepancies are often numerically plausible and therefore difficult to detect through routine preprocessing, but they can propagate as structured label noise during database construction and machine-learning reuse. A cross-database example shows how ambiguous reporting can create a 100-fold conductivity error. Our analysis reframes data accuracy as an infrastructure requirement for artificial-intelligence-driven discovery and motivates traceable reporting, curation, and validation practices for reusable scientific data. Keywords: AI for science; Data reliability; Scientific databases; Structured label noise; Literature-derived data; Materials informatics; Solid electrolytes Comments: 3 main figures, 2 supplementary tables, research article on data reliability for AI-driven materials discovery Subjects: Information Retrieval (cs.IR); Materials Science (cond-mat.mtrl-sci); Computational Engineering, Finance, and Science (cs.CE); Machine Learning (cs.LG) ACMclasses: I.2.6; H.2.8; J.2 Cite as: arXiv:2609.01621 [cs.IR] (or arXiv:2609.01621v1 [cs.IR] for this version) https://doi.org/10.48550/arXiv.2609.01621 Focus to learn more arXiv-issued DOI via DataCite
[IR-34] MGDiff: Multi-Interest Sequence Recommendation with Masking GNN-Guided Diffusion
链接: https://arxiv.org/abs/2609.01619
作者: Wenjing Xiao,Hao Ding
类目: Information Retrieval (cs.IR)
备注: 8 pages, 4 figures
Abstract:We propose a novel Multi-Interest Sequence Recommendation Framework with \underlineMasking \underlineGNN-Guided \underlineDiffusion Model (MGDiff), designed to generate accurate, bias-free user interest information during the diffusion process. First, we propose a semantics-enhanced Dual-layer Semantic Guidance (DSG) framework, which decomposes guidance into two synergistic stages: extracting latent item semantics and decoupling multidimensional user intent. We design a Weight-adaptive Masking Graph Neural Network reconstructs missing links to uncover deep item relationships beyond superficial co-occurrence, while a Dynamic Multi-Expert Network projects user preferences into distinct semantic subspaces to suppress irrelevant interference. This hierarchical design yields structured guidance that significantly improves the generation accuracy of diffusion models. Second, We propose a Popularity-Aware Guidance (PAG) mechanism that performs spatial geometric adjustments on the outputs of diffusion models: by using item popularity as a differentiable adjustment signal to recalibrate similarity metrics, we enable DMs to generate diverse recommendations free from popularity bias. Finally, we compare MGDiff with multiple baseline models across four widely used datasets, demonstrating its superior performance and validating its effectiveness.
[IR-35] Multi-Agent Retrieval-Augmented Generation for Efficient Cloud Knowledge Base Search in Telecom SNOC Environment
链接: https://arxiv.org/abs/2609.01618
作者: Harish Saragadam,Sudhanshu Sharma,Ipsha Routray
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注:
Abstract:Telecom Service and Network Operations Centers (SNOCs) rely on large collections of cloud documents, including Standard Operating Procedures (SOPs), vendor technical manuals, incident reports, and configuration guides, to maintain uninterrupted network operations. During critical incidents, engineers must quickly retrieve accurate information, yet traditional keyword based and single stage retrieval approaches often struggle to provide precise results. This paper presents Athena for Cloud Knowledge Base, a fully offline, multi agent Retrieval Augmented Generation (RAG) framework designed for enterprise cloud document search in Vodafone Idea’s SNOC environment. The system integrates dense retrieval using E5 Large V2 embeddings, BM25 sparse retrieval, and Knowledge Graph expansion within a LangGraph based orchestration framework. Retrieved candidates are fused using Weighted CombSUM, followed by cross encoder reranking and Maximal Marginal Relevance (MMR) to obtain a diverse and relevant evidence set. To improve answer reliability, the framework performs per chunk LLM evaluation with explicit attribution verification, assessing each MMR selected chunk independently before generating a response. Unsupported or weak evidence is discarded, and if no chunk satisfies the verification criteria, the system automatically evaluates multiple chunks together as a fallback. Experiments on a corpus of 4200 SNOC cloud documents containing 312000 indexed chunks show that the proposed approach achieves an MRR at 10 of 0.910 and an Exact Match (EM) score of 78.4 percent, outperforming single stage dense retrieval by 14.6 percentage points. The entire pipeline operates in a fully offline environment, satisfying enterprise data sovereignty requirements while delivering accurate and grounded responses for cloud document search. Subjects: Information Retrieval (cs.IR); Machine Learning (cs.LG) Cite as: arXiv:2609.01618 [cs.IR] (or arXiv:2609.01618v1 [cs.IR] for this version) https://doi.org/10.48550/arXiv.2609.01618 Focus to learn more arXiv-issued DOI via DataCite
[IR-36] Hybrid Retrieval-Augmented Generation with Knowledge Graph Expansion RRF Fusion and Per-Chunk Grounded Evaluation for Enterprise Document Search
链接: https://arxiv.org/abs/2609.01617
作者: Harish Saragadam,Sudhanshu Sharma,Meghana Pujari
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Getting accurate, grounded answers out of large enterprise document repositories is a difficult problem. Dense vector retrieval alone frequently performs poorly on queries that mix technical terminology, vendor-specific acronyms, or require reasoning across several non-adjacent sections. DocuSearch was built to address exactly this gap - an offline, multi-agent document intelligence system developed and evaluated in a production telecom network operations environment. Rather than relying on a single retrieval signal, DocuSearch pulls together three complementary sources of evidence: semantic search over a Qdrant vector store using BGE-Large embeddings, BM25 full text search over an SQLite FTS5 index, and Knowledge Graph neighbour expansion from a structured edge table. These three ranked lists are merged through Reciprocal Rank Fusion with signal weights of 0.50 for vector search, 0.35 for BM25, and 0.15 for the knowledge graph, using a smoothing constant of 60 to stabilize scores. A cross-encoder then reranks the fused list, and Maximal Marginal Relevance with a balance factor of 0.65 prunes results for relevance and diversity. What makes DocuSearch distinctive is a per-chunk evaluation loop treating each chunk as its own mini-retrieval problem: an LLM decides whether the chunk needs more context, whether it fully answers the query, and whether the answer is grounded in retrieved text. Ungrounded answers are not returned; the system falls back to a multi-chunk merge instead. On a telecom corpus, DocuSearch reaches Precision@10 of 0.69, Recall@10 of 0.79, and a grounding rate of 89.6% - gains of 15, 16, and 18.4 percentage points over a dense-only RAG baseline. Index Terms: retrieval-augmented generation, knowledge graph, reciprocal rank fusion, enterprise document search, agentic evaluation, BM25, cross-encoder reranking, on-premise deployment, LangGraph, telecom AI.
[IR-37] Incident Memory: Training-Free Operational Memory through Sequential Pattern Mining and Velocity-Stratified Retrieval
链接: https://arxiv.org/abs/2609.01616
作者: Adarsh Agrawal,Rahul Suresh Babu
类目: Information Retrieval (cs.IR)
备注: 14 pages, 5 figures, 8 tables (main text); includes appendix
Abstract:Incident response is a memory problem: teams accumulate tickets, traces, postmortems, and wiki pages, but the knowledge needed for the next incident is rarely stored with its order, freshness, and provenance intact. We present Incident Memory, a deterministic system that accumulates operational knowledge without model training. It combines (i) velocity-stratified retrieval, which ages structural, behavioral, contextual, and ephemeral facts at different rates; (ii) fingerprint-conditioned PrefixSpan mining, which extracts ordered playbooks from successful investigations; and (iii) provenance-aware metric definitions, which detect conflicting definitions through executable checks. On the UCI ITSM event log, containing 141,712 events across 24,918 incidents, Incident Memory extracts 23,110 ordered traces, mines 39 playbooks, and covers 84.3% of 6,934 held-out incidents. On controlled benchmarks with known ground truth, it achieves 99.2% ordered playbook precision (controlled), an architectural staleness guarantee versus 36% stale returns for a flat baseline, and conflict-detection F1 of 0.876. A direct Claude Haiku baseline on 19 fingerprint groups reaches 0.661 ordered precision, compared with 0.985 for PrefixSpan. The central result is not that language models are weak at incident response; it is that repeated incident histories are low-entropy once conditioned on fingerprint and previous action. In that regime, exact memory is a stronger primitive than open-ended generation.
[IR-38] Skim and Skip: Hierarchical Adaptive Inference for Efficient Multimodal Retrieval
链接: https://arxiv.org/abs/2609.01613
作者: Meng Gao,Yizhen Zhang,Yang Ding,Ziqi Dai,Shuoshuo Zhang,Junjie Wang,Taiqiang Wu,Chufan Shi,Lei Ji,Jian Jiao,Linfeng Zhang,Yeyun Gong,Yujiu Yang
类目: Information Retrieval (cs.IR)
备注:
Abstract:Universal multimodal retrieval (UMR) increasingly adopts multimodal large language models (MLLMs) as unified embedding backbones, but their strong retrieval performance comes at substantial inference cost. Existing methods typically rely on uniformly dense inference, where all input tokens are processed through the entire model and matched using the final-layer [EOS] representation. However, this paradigm overlooks two key forms of heterogeneity in multimodal retrieval: token contributions to the final retrieval embedding are highly uneven, and different queries require markedly different amounts of inference depth. To address this, we propose Skim and Skip (SAS), a hierarchical adaptive inference framework for efficient multimodal retrieval. SAS first performs token-level evidence selection to preserve only the input information most relevant to the final retrieval embedding, and then performs depth-adaptive inference to determine whether the current representation is already sufficient for reliable matching. Experiments on 12 MMEB retrieval tasks show that SAS retains about 99% of the dense baseline’s average retrieval performance while achieving up to 1.64 times end-to-end speedup and up to 66.3% FLOPs reduction.
[IR-39] MESSY STREETS: A Benchmark for Geocoding Real-World Addresses
链接: https://arxiv.org/abs/2609.01612
作者: Edward Gaere,Florian von Wangenheim
类目: Information Retrieval (cs.IR); Computation and Language (cs.CL)
备注:
Abstract:We introduce MESSY STREETS, a benchmark for evaluating geocoders on verbatim web addresses, with existence verification and controlled measurement of surface-form divergence. Unlike conventional benchmarks based on clean or synthetically perturbed addresses, MESSY STREETS contains addresses whose surface forms diverge from canonical representations and whose components may be missing, repeated, malformed, or incomplete. The benchmark is constructed from the December 2024 Web Data Commons corpus, with reference locations established from OpenAddresses or OpenStreetMap. The strongest commercial geocoders outperform open-source systems by up to 49 percentage points in recall. This gap is driven primarily by differences in candidate return rates on non-canonical addresses; once a candidate is returned, positional accuracy is broadly comparable across systems. Non-canonical surface form alone accounts for up to 25 percentage points of recall loss. Examining Nominatim’s query-processing pipeline, we show that its conjunctive matching lets a single unrecognised token zero an otherwise valid query. The results demonstrate that geocoder choice is a consequential design decision for applications processing noisy address data, and that normalisation and preprocessing could substantially narrow the gap between open-source and commercial geocoders. Subjects: Information Retrieval (cs.IR); Computation and Language (cs.CL) Cite as: arXiv:2609.01612 [cs.IR] (or arXiv:2609.01612v1 [cs.IR] for this version) https://doi.org/10.48550/arXiv.2609.01612 Focus to learn more arXiv-issued DOI via DataCite
[IR-40] Making Revisions Understandable: A Survey of Edit Intentions Methods and Applications ACL2026
链接: https://arxiv.org/abs/2609.01610
作者: Fangping Lan,Qi Zhang,Eduard Dragut
类目: Information Retrieval (cs.IR)
备注: 17 pages, 6 figures, 4 tables, accepted by Findings of ACL 2026
Abstract:Text revision is a core process in document creation, capturing how authors iteratively refine, reorganize, and improve written content. With the increasing availability of large-scale revision histories from platforms such as Wikipedia and arXiv, NLP research has begun to move beyond modeling what changes are made to understanding why they are made, i.e., the underlying edit intentions. To our knowledge, this is the first survey that synthesizes text revision research through the lens of edit intentions, providing a unified view of datasets, taxonomies, identification methods, and applications. We review prior work across the full revision workflow, including revision corpus construction, edit intention taxonomy design, and edit intention identification. We further categorize representative datasets and methods, summarize downstream applications such as writing assistance and document edit summarization, and highlight key open research directions.
人机交互
[HC-0] BuildOcc: A Large Language Model Occupant Agent Platform for Building Energy Research
链接: https://arxiv.org/abs/2609.02729
作者: Wooyoung Jung
类目: Human-Computer Interaction (cs.HC)
备注: 28 pages, 4 figures, 11 tables, 3 appendices. Software: this https URL
Abstract:Occupants are a primary source of uncertainty in building energy consumption and management, yet existing occupant behavior models cannot capture adaptive and reasoning responses considering the occupant’s personal history, current context, and the type of energy signal being delivered. This study presents BuildOcc, an open-source Python platform that grounds large language model agents in the American Time Use Survey (ATUS), a nationally representative diary dataset covering 16,684 respondents. Through BuildOcc, each simulated occupant agent can be instantiated with a demographic persona drawn from ATUS population statistics, a memory stream that accumulates and reflects on timestep-level observations, and an activity scheduler that samples empirically from ATUS time-at-activity distributions. The platform exposes a three-layer interface - Python library, REST API, and Model Context Protocol server - so that any building energy tool (EnergyPlus, Home Assistant) can integrate behavioral intelligence without bespoke coupling code. A plugin registry lets the community add new occupant strata, custom schedulers, and alternative memory backends as separate installable packages. Two validation tiers show that ATUS-grounded sampling reproduces empirically calibrated activity distributions and that demographic priors propagate into persona-consistent agent reasoning across timesteps, establishing internal consistency across strata. BuildOcc provides the building energy community with a reusable, openly available implementation of the occupant behavioral layer. BuildOcc is openly released at this https URL under the Apache License 2.0 and installable via pip install buildocc.
[HC-1] Large Language Model-Driven Context-Aware Eco-Feedback Generation and Evaluation
链接: https://arxiv.org/abs/2609.02719
作者: Wooyoung Jung,Prosper Babon-Ayeng
类目: Human-Computer Interaction (cs.HC)
备注: 37 pages, 12 figures, 12 tables. Author’s original submitted manuscript (pre-peer-review). The peer-reviewed and published version appears in Energy and Buildings 370 (2026) 118038
Abstract:The objective of this study was to demonstrate the potential of generating eco-feedback that accounted for unique household contextual information, named as context-aware eco-feedback, through a large language model-integrated framework. Previous studies have introduced personalized eco-feedback, mostly relying on household energy use patterns; however, they frequently did not reflect distinct household characteristics, including their persona or non-negotiable routines, leaving eco-feedback ineffective and sometimes superficial. To address these limitations, we introduced a contextual engineering framework that generated eco-feedback using a self-consistency with chain-of-thought prompt that leveraged household energy analysis data, utility rate structures, and characteristic information. We conducted a rigorous empirical validation and a combinatorial evaluation analysis to assess this framework systematically. The former aimed to test the framework’s ability to generate accurate and data-driven eco-feedback, customized to given contexts by comparing it with reference interventions. The latter aimed to reveal the framework’s adaptability across diverse household contexts by investigating how context-aware eco-feedback changed. Key findings were the following: our proposed framework generated eco-feedback that aligned with reference solutions at a mean accuracy of 92.0% across different household configurations, accurately leveraging the provided household data for feedback generation (95.7% of data citation accuracy). Also, it was largely adaptive to diverse household contexts, significantly shifting targeted appliances and energy-saving strategies. Ultimately, this study contributes to realizing the next level of context-aware interactions between occupants and buildings which paves the way for higher occupant living quality and sustainability.
[HC-2] he PIONEER Project: A PrIvacy companion for mOtivatioN and knowlEdge transfER
链接: https://arxiv.org/abs/2609.02700
作者: Simon Althaus(1 and 3),Nina Gerber(1 and 3),Sara Hahn(1 and 2 and 3),Andreas Heinemann(2 and 3),Angela Menig(1 and 3),Max Mülhäuser(1 and 3),Christian Reuter(1 and 3),Ephraim Zimmer(1 and 3) ((1) Technical University of Darmstadt, (2) University of Applied Sciences Darmstadt, (3) ATHENE - National Research Center for Applied Cybersecurity)
类目: Human-Computer Interaction (cs.HC); Computers and Society (cs.CY)
备注: Accepted as poster at the Symposium On Usable Privacy and Security (SOUPS) 2026
Abstract:Remaining control over their private data is one of the key challenges in this century for users. We know from prior work that users are often neither in a position to fully grasp the content of the usually complicated texts, nor are they motivated to spend the time necessary to do so. We report on the progress made by the PIONEER project on a privacy support tool that combines knowledge transfer and persuasive elements to increase users’ privacy awareness and motivation; thus empowering them to more privacy sovereignty. Throughout the research and design process, we consider user group specifics that may result in different requirements, e.g., for children, adolescents, parents, or elderly people. We further target sustainable behavior change by addressing different states of change, precisely: spark initial motivation, facilitate the creation of new habits, and encourage habituation of these habits in the long term (volition). Finally, we provide a privacy support tool demonstrator that can be utilized for research and education purposes, e.g., in school contexts.
[HC-3] Beyond Problem Solving: Large Language Models for Emotional and Reflective Support in Mathematics Learning
链接: https://arxiv.org/abs/2609.02611
作者: Vera Rief,Mirella Hladký,Minju Yoo,Stephanie Heel,Shintaro Sato,Tomohiro Nagashima
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:Intelligent Tutoring Systems (ITSs) traditionally focus their adaptive support on cognitive aspects of learning. Although effective, little is known about how such systems can be enhanced by addressing students’ emotional states. In particular, the role of mindful interventions for supporting student learning and experiences in adaptive math learning remains underexplored. We developed “Math with Matt”, an ITS that leverages Large Language Models (LLMs) to provide both cognitive and emotional support in algebra learning. The system offers 1) an LLM-based mindful chat that delivers context-sensitive emotional support through a pedagogical agent Matt, and 2) mindful feedback and hint messages (not just evaluative) to enhance learning experiences and reduce math anxiety. We conducted a classroom study with 7th graders, comparing a Mindful version against a version with cognitive support only. Overall, the ITS reduced executive state-math anxiety and improved students’ math learning, though no significant differences emerged between the conditions. However, students with the mindfulness interventions showed higher learning efficiency and well-balanced problem-solving behavior, since they achieve a similar level of math learning with less learning time and fewer requested hints compared to the Cognitive version. Additionally, they reported that the pedagogical agent felt more supportive and caring than students in the cognitive condition. Our study demonstrates the feasibility and scalability of integrating mindfulness into ITSs through LLM-based interactions and positions LLMs as an adaptive, socio-emotional layer within cognitive math tutoring.
[HC-4] EEG-based Visual Retrieval and Reconstruction: From Neurally Visible Optimal Layer to Hierarchical Diffusion Generation
链接: https://arxiv.org/abs/2609.02582
作者: Minyi Wang,Zhenqin Wu,Rihui Li
类目: Human-Computer Interaction (cs.HC)
备注: 11 pages, 6 figures
Abstract:Decoding visual perception from electroencephalography (EEG) is important for non-invasive brain-computer interfaces (BCIs). However, most existing visual decoding pipelines directly align EEG features with semantic features from pretrained vision models. Those EEG signals carry information at more than one level and this practice disregards the varying neural visibility of different visual components in EEG signals, leading to cross modal mismatches and incomplete information use. In this work, we address this limitation through layer-wise contrastive learning. For each subject, the intermediate CLIP layer that maximizes retrieval performance is selected as the Neural Visibility Optimal Layer (NVOL). Built on NVOL, a hierarchical framework couples retrieval and generation through a shared intermediate representation. The retrieval branch fuses multi-NVOL features, aligns them to image embeddings via contrastive learning, and applies cross-domain similarity local scaling (CSLS) at test time to mitigate hubness. The generation branch reconstructs subject-specific NVOL features from EEG using a conditional diffusion prior, maps them to CLIP space through a lightweight adapter, and drives a pretrained Stable Diffusion XL model. Experimental validation on THINGS-EEG showed that, NVOL-based retrieval achieves 78.1% mean Top-1 accuracy in 200-way retrieval, rising to 86.4% with CSLS. Two-stage NVOL-to-semantic reconstruction also outperforms single-stage final-layer diffusion on semantic and structural metrics. By aligning EEG with layer-wise neural visibility rather than fixed high-level semantics, the proposed framework improves both retrieval accuracy and image reconstruction in EEG-based visual decoding.
[HC-5] Beauty is in the AI of the beholder: MLLM s systematically overrate facial attractiveness
链接: https://arxiv.org/abs/2609.02512
作者: Santiago Grandas,Juan Sebastian Cely-Acosta,Mohit Mendiratta,Shafee Hassan,Macken Murphy
类目: Computer Vision and Pattern Recognition (cs.CV); Human-Computer Interaction (cs.HC)
备注:
Abstract:Beauty assessments from Multimodal Large Language Models (MLLMs) are increasingly popular amongst users, companies, and aestheticians. This raises the question of whether these AI models can accurately reflect human judgments of attractiveness. In a pre- registered exploratory study, we compared the attractiveness ratings of 2,513 human participants to four widely used commercial AI models: Claude, Gemini, GPT, and Grok. Results showed that MLLMs systematically rate faces more favourably and within a narrower range than humans and, at the time of study, do not reproduce human ratings in absolute terms. However, MLLMs exhibit strong correlations with human attractiveness judgments, accurately tracking the rank-ordering of faces. MLLMs may judge faces by different cues than humans; only face age was a predictor of facial attractiveness in both humans and MLLMs, with inconsistent patterns across models for ethnicity and gender. AI models strongly agree with one another, except for Grok, which also showed the lowest agreement with humans. Our findings suggest that while they may be able to approximate rank-orderings of human attractiveness, current off-the-shelf commercial MLLMs systematically overrate the beauty of human faces.
[HC-6] Orthogonal Ensembles and Tested Explanations for Performer-Independent Body-Motion Emotion Recognition
链接: https://arxiv.org/abs/2609.02510
作者: Naoto Nishida,Yoshio Ishiguro
类目: Computer Vision and Pattern Recognition (cs.CV); Human-Computer Interaction (cs.HC); Machine Learning (cs.LG)
备注: 8 pages, 3 figures, 3 tables. Accepted to ACII2026 workshop
Abstract:We study body-only, 12-class acted-emotion classification from skeleton motion under leave-performer-out (LPO) evaluation, a hard, underdetermined setting: chance is 8.3%, and a protocol-matched reproduced STGCN++ baseline reaches only 25.73 +/- 4.03% Macro-F1. We show that reliable gains come not from a new architecture but from combining eleven models with orthogonal error modes: under 10-fold LPO cross-validation on the labeled training performers, an equal-weight logit-mean ensemble reaches 36.80 +/- 4.00% per-fold Macro-F1, a protocol-matched +11.07 pp (+43% relative) over the same-split reproduced baseline. Our central contribution is a tested explanation suite: for a strong ensemble member, part-masking and counterfactual edits show (rather than assert) that its decisions depend on motion-grounded body-region evidence, and this region saliency aligns with rule-based Laban Movement Analysis (LMA) attributes far more than with classical kinematics: region-level saliency-LMA Spearman rho = +0.500 versus +0.033, roughly 15x, and the alignment holds for the submitted 11-way ensemble itself at rho = +0.517; the audit is post hoc and needs no retraining. The same suite faithfully reports a negative: within-window temporal saliency is diffuse rather than localized.
[HC-7] Decoding Decision Correctness from EEG Under High Cognitive Workload in Virtual Reality: Implications for Collaborative Brain-Computer Interface Teams
链接: https://arxiv.org/abs/2609.02436
作者: Christopher Baker,Stephen Hinton,Tom Reed,Stephen Fairclough
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:Collaborative Brain-Computer Interfaces (cBCIs) offer a promising mechanism to augment team decision-making, but existing approaches rely exclusively on evidence available only after a decision has been made and reported, such as reaction time or stated confidence. This limits their use to explaining or discounting a decision after the fact, rather than informing a team’s response before it is finalised. We tested whether spatial-covariance EEG features could instead provide a genuinely pre-emptive signal of an operator’s decision correctness, available within the response window itself, and whether such a signal depends on cognitive workload. Using a continuous virtual reality target-detection task, participants (N = 23) completed a within-subject workload manipulation (High vs. Low). At the team level, weighting votes by this pre-emptive neural signal, available before a response is committed, produced substantial accuracy gains on contested (evenly-split) trials under High Workload (57% to 88% as team size increased from 2 to 16), but was actively detrimental under Low Workload. Critically, this advantage held even against post-hoc behavioural signals: confidence was the strongest single team-level signal overall, but by definition cannot inform a decision still in progress, whereas the neural signal can. These findings indicate that EEG-based decision-reliability signals are not a general-purpose team augmentation tool, but a workload-conditional one, with clear implications for when and how cBCI systems should be deployed in operational teams.
[HC-8] Removing Speech Keeping Activities: A Privacy Firewall for Acoustic Sensing in Assisted Living
链接: https://arxiv.org/abs/2609.02376
作者: Pavlos Nicolaou,Christos Efstratiou
类目: ound (cs.SD); Cryptography and Security (cs.CR); Human-Computer Interaction (cs.HC)
备注: 42 pages, 8 figures, 4 tables. Submitted to Pervasive and Mobile Computing. Preprint also available on SSRN: this https URL
Abstract:Acoustic sensing offers a promising non-intrusive approach for monitoring daily activities of older adults, yet speech privacy concerns remain a critical barrier to real-world deployment. We present a privacy firewall pipeline based on a U-Net encoder-decoder, trained entirely on synthetic data, that removes speech from ambient audio while preserving environmental sounds indicative of daily activities. Activity recognition is performed using VGGish transfer learning with an SVM classifier. Evaluated on the ESC-50 and SINS datasets across multiple speech content levels, the proposed model reduced residual speech to 0% VAD-detectable speech (Silero Voice Activity Detection) under all tested conditions, outperforming Facebook Denoiser (6.55% residual), SepFormer (36.34%) and ConvTasNet (47.21%) on ESC-50 at the 100% speech level. On ESC-50 at 40% speech level, classification performance recovers to 85% precision and 85% recall after speech removal, compared with 81%/75% before removal and an 84%/83% speech-free baseline. Evaluation on real-world participant home recordings collected with the AudioHive app showed 0% VAD-detectable speech after processing while maintaining 76% precision and recall. The pipeline enables privacy-preserving acoustic sensing without sacrificing activity recognition performance, addressing a key obstacle to the adoption of ambient monitoring in elderly care.
[HC-9] owards a Foundational Ontology for Identifying and Resolving Contradictions in Dialogue-based Human-Robot Interactions ICRA2026
链接: https://arxiv.org/abs/2609.02364
作者: Maitreyee Tewari,Michele Persiani
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注: 5 pages, 1 figure, Accepted at the 2nd edition of the Joint Workshop on Ontologies, Semantic Maps and Autonomous Robotics Standardization (J-WOSMARS 2026) collocated with ICRA 2026, Austria
Abstract:Existing Human-Robot Interaction (HRI) literature has focused on identifying and structuring errors, failures, conflicts, and knowledge issues (called in this work as contradictions) in domain-specific dialogue-based interactions. However, there is still lack of a formal computational framework to represent and define these contradictions, interoperable and usable across HRI and human-agent interaction (HAI) domains. Thus, this research project aims to capture, represent, and evaluate the notion of (1) dialogue-based collaborative interaction and (2) related contradictions in a foundational ontology. METHONTOLOGY, a systematic approach to build domain-independent ontologies was applied. In the conceptualisation stage of the presented ontology, concepts and models from Activity Theory were used. Preliminary results presented in this short article are: (i) Natural language definitions of dialogues and related contradictions in HRI, (ii) Set Theoretic definitions of dialogues and contradictions, and (iii) First Order Logic (FoL) formulation of the contradiction concepts and three novel principles guiding dialogue-based interactions between humans and robots. In summary, we report on ongoing work to develop a foundational ontology based on Activity Theory called Activity Theory-based foundational ontology (ATFOt) to capture and represent the notion of contradictions in HRI.
[HC-10] OmegaUse-SOP: SOP Engineering for Professional Computer Use from Human Demonstrations
链接: https://arxiv.org/abs/2609.02149
作者: Yixiong Xiao,Lang An,Hucheng Yang,Pinxue Ma,Yongquan Chen,Jingjia Cao,Yusai Zhao,Ting Wang,Ting Liu,Siqi Bao,Jingbo Zhou,Hua Wu
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models (LLMs) are increasingly evolving from conversational assistants into agents capable of operating external digital environments. Graphical user interface (GUI) agents play an important role in this transition, as many real-world workflows remain accessible only through user-facing software interfaces. However, despite recent progress on general computer-use benchmarks, domain-specific professional standard operating procedures (SOPs) remain challenging for GUI agents because they often involve implicit domain knowledge, software-specific conventions, and task-level verification requirements. We introduce OmegaUse-SOP, a human-in-the-loop SOP Engineering system for transforming human demonstrations of professional computer use into reusable SOP skills for GUI agents. Analogous to prompt engineering, SOP Engineering iteratively refines demonstrations, execution rules, and domain knowledge to convert professional SOPs into reusable GUI-agent skills. OmegaUse-SOP consists of four modules: Observe, Reason, Configure, and Execute. Together, these modules record expert operations as multimodal GUI traces, abstract low-level events into semantic step-level instructions, incorporate domain rules and task-specific parameters, and execute the resulting skills in live GUI environments through step-wise grounding, action generation, and verification. To demonstrate its effectiveness, we collaborate with a power-sector client and test OmegaUse-SOP on photovoltaic simulation workflows in PVsyst 7.2. The results suggest that OmegaUse-SOP can improve GUI-agent reliability on professional SOP tasks, highlighting a practical path toward deploying GUI agents in domain-specific professional software environments.
[HC-11] Reconciling Kinesthetic Mismatches: A Somatic Alignment Mindset for Musical Body Transformation
链接: https://arxiv.org/abs/2609.01981
作者: Ziyue Piao,Isabelle Cossette,Marcelo M. Wanderley
类目: Human-Computer Interaction (cs.HC)
备注: 5 pages, 1 figures. Accepted to the CHI 2026 Workshop on Body Transformation Experiences: A Workshop on How to Elicit, Assess and Support them through Multisensory Technology
Abstract:Mastering musical performance requires precise multisensory coordination, yet learners encounter a kinesthetic mismatch, which is a discrepancy between the internal perception of an action and the actual physiological state of the body. While multisensory Body Transformation Experiences (BTE) provide tools to bridge this gap, existing designs often focus on external correction rather than internal alignment. To address this, we propose the Somatic Alignment Mindset (SAM), a conceptual lens that integrates Taoist philosophy to shift the focus of HCI design from prescriptive feedback toward holistic embodied unity. By positioning technology as a reflective medium, SAM operationalizes the principles of Adaptation, Assessment, and Awareness to reconcile somatic discrepancies and foster deep, self-aligned musical mastery.
[HC-12] Knowing Is Not Enough: Information Retrievability as a Precondition to Effective LLM Oversight
链接: https://arxiv.org/abs/2609.01976
作者: Xinyu Fu,Narayan Ramasubbu,Dennis Galletta
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models (LLMs) are increasingly embedded in organizational work, yet their errors often pass human review. Prior research locates such failures in users’ capability to review LLM output or their engagement in doing so. We develop an alternative, retrieval-based account of human oversight and posit that error detection is more effective when oversight-relevant information is accessible to users at the moment of review. Across two randomized lab-in-the-field experiments with 640 customer-facing employees, we show that self-generated explanations improve error detection and strengthen recall of verification-relevant reasoning, while cues that reactivate such reasoning help sustain detection under repeated LLM use. Theoretically, we identify information retrievability as a distinct precondition for effective oversight and specify generative encoding and cue-supported reactivation as mechanisms that build and sustain it. Practically, lightweight onboarding self-explanations and daily retrieval cues can make human oversight more resilient as LLM use becomes routine.
[HC-13] Exploring Breathing-Music Coupling: Using the Breathing Mirror for Somatic Reflection in Piano Performance
链接: https://arxiv.org/abs/2609.01974
作者: Ziyue Piao,Yohei Wada,Isabelle Cossette,Marcelo M. Wanderley,Akira Maezawa
类目: Human-Computer Interaction (cs.HC)
备注: Proceedings of the International Conference on New Interfaces for Musical Expression (NIME)
Abstract:While breathing is essential to living and for sound production in some instruments, for pianists, it is often a hidden and automatic process, making it difficult to analyze or refine. A critical gap exists between data and awareness: while sensors record precise physical metrics, they fail to capture the performer’s somatic experience. Conversely, the high cognitive load of performance makes it nearly impossible for musicians to recall their internal states with temporal precision. To address this, we present a system, Breathing Mirror, and associated methodology designed to externalize the pianist’s internal somatic experience through three analytical lenses: a Baseline View (synchronized signals), a First-Person View (subjective recall), and an Interpersonal View (collaborative reflection). Through a four-week longitudinal study with a skilled amateur pianist (35 years of experience), we evaluated the system’s effectiveness by recording respiratory data using textile-integrated strain sensor belts. The results show that the Breathing Mirror reveals some patterns of breathing-music coupling and identifies critical blind spots where objective data diverges from subjective perception. Furthermore, we propose four somatic themes regarding the link between breathing and musical elements, offering a foundation for future large-scale validation across a broader range of pianists. This work provides a new way to study body signals, transforming breathing from an internal biological function into an articulate expressive parameter. Comments: Proceedings of the International Conference on New Interfaces for Musical Expression (NIME) Subjects: Human-Computer Interaction (cs.HC) Cite as: arXiv:2609.01974 [cs.HC] (or arXiv:2609.01974v1 [cs.HC] for this version) https://doi.org/10.48550/arXiv.2609.01974 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Journalreference: Piao, Z., Wada, Y., Corssette, I., Wanderley, M., Maezawa, A. (2026). Proceedings of the International Conference on New Interfaces for Musical Expression, 60–69 Related DOI: https://doi.org/10.5281/zenodo.20784052 Focus to learn more DOI(s) linking to related resources
[HC-14] Beyond Instruction-Driven Editing: Source-Grounded Problem Discovery with User-Governed Repair for Scientific Posters
链接: https://arxiv.org/abs/2609.01813
作者: Xingda Lyu,Honglin Lu,Xinye Luo,Shiqi Yang
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:Interactive editors usually assume that users already know what to change. Yet an important interaction state comes earlier: a user may recognize that an artifact is not working without knowing what intervention to request. We call this the articulation gap. We introduce PROS (Proactive Refinement Of Scientific Posters), which separates epistemic initiative from behavioral authority: the system can surface source-grounded candidate problems, while users decide which become repair goals and whether resulting changes are committed. Accepted issues hand off to native-object PPTX editing with validation and reversible preview. We also introduce PROS-Bench, a source-linked collection of 120 papers and 320 editable PPTX posters, including a 120-poster matched primary core and a separate conference representation challenge. On the primary core, PROS achieves a mean VLM-rated stage-balanced diagnosis quality score of 67.2 on a 0-100 scale and 87.6% operator-verified target resolution among accepted diagnoses. Temporally blinded automated scoring yields a +22.7-point paper-macro accepted-target uplift, yet 14.8% of assessable accepted targets decline. This divergence shows why problem discovery, local resolution, and realized outcome should be evaluated separately. More broadly, intelligent editors can support problem discovery before a concrete edit request exists without taking authority over consequential change.
[HC-15] VirSqueezer: Generating Realistic Deformations and Squeezing Dynamics in VR from Fine-Grained Squeezing Controls
链接: https://arxiv.org/abs/2609.01698
作者: Qian Zhang,Xiaoming Chen,Xiaorui Ma,Haisheng Li,Weidong Cai
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:Squeezing is one of the most natural forms of hand manipulation, inherently involving fine-grained, temporally evolving, per-finger flexion. In VR content creation, squeezing plays a unique role in enabling particular visual effects such as localized deformations and dynamic behaviors, e.g., bursting a Coke can or juicing a fruit, thereby expanding the expressive possibilities of VR content. However, existing techniques, such as 3D Gaussian splatting-based methods and diffusion-based video generation models, are limited in their ability to simulate fine-grained virtual squeezing effects. We introduce VirSqueezer, a framework designed to generate both localized deformations (primary effects) and complex squeezing dynamics, such as rupture and overflow (secondary effects). VirSqueezer captures squeezing control signals using a SenseGlove and provides the user with inferred resistance force feedback during the squeezing process. By estimating object contact areas, inferring physical properties, and simulating physical responses, VirSqueezer computes conditions that guide generation models for visual effect generation, ensuring both visual coherence and temporal synchronization with the simulation. Consequently, VirSqueezer enables the generation of physically realistic visual effects directly from continuous, fine-grained squeezing control signals. Our extensive evaluation demonstrates VirSqueezer’s ability to reproduce realistic localized deformations, generate convincing visual dynamics, and maintain consistency in fine-grained squeezing controls.
[HC-16] Slow-Fast Brain-Computer Interfaces: Preventing Neuroadaptive Overfitting in AI-Mediated Neural Interfaces
链接: https://arxiv.org/abs/2609.01767
作者: Aarthy Nagarajan
类目: Neurons and Cognition (q-bio.NC); Human-Computer Interaction (cs.HC); Signal Processing (eess.SP)
备注: Commentary, 1 figure, 1 table, 12 pages including references
Abstract:Artificial intelligence (AI) is transforming brain-computer interfaces (BCIs) from task-specific neural decoders into adaptive systems that complete language, smooth movement, regulate rehabilitation support and adjust stimulation. These capabilities can increase speed, fluency, usability and clinical reach, yet conventional performance metrics may overlook losses in intent fidelity, authorship, agency, therapeutic challenge and durable clinical benefit. I define neuroadaptive overfitting as a closed-loop failure mode in which an AI-mediated BCI becomes over-optimized to short-term proxies of success, including reduced effort, rapid acceptance, lower workload or smooth task completion, while drifting from the user’s durable goals. I then propose Slow-Fast BCI, a framework for pacing AI assistance according to decoder evidence, uncertainty, contextual and clinical stakes, fatigue, and user- or clinician-defined goals. The framework distinguishes fast assistance when intent is clear and stakes are low, guarded assistance under uncertainty and slow assistance when misalignment could compromise safety, agency, authorship, motor learning or therapeutic value. Across communication, motor-control, neurorehabilitation and closed-loop neuromodulation applications, I outline corresponding safeguards and evaluation measures. This Perspective argues that AI-mediated BCIs should be evaluated not only by decoding accuracy and task performance, but also by how AI assistance is deployed: when systems act autonomously, seek confirmation, preserve user effort or return control to the user.
计算机视觉
[CV-0] SolarWM: Open Data and Scalable Training for Long-Horizon Video World Models
链接: https://arxiv.org/abs/2609.02886
作者: Junchao Huang,Guian Fang,Shengju Qian,Xianghao Kong,Zhuoran Zhao,Wei Huang,Yihua Du,Zixin Zhang,Justin Cui,Yuchao Gu,Yukang Chen,Xinting Hu,Tianyu He,Shaoshuai Shi,Zhuotao Tian,Xin Wang,Mike Zheng Shou,Li Jiang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: this https URL
Abstract:We introduce SolarWM, a fully open foundation for building interactive video world models from data preparation through long-horizon inference. Training across heterogeneous data sources and video backbones is challenging: datasets differ in temporal scale, camera geometry, visual quality, motion, and captioning styles, while video generators use distinct representations and architectures. Naive data mixing and model-specific implementations therefore produce inconsistent supervision and make results difficult to reproduce and compare. SolarWM addresses this coupling with a reconfigurable multi-source data engine and a backbone-native adaptation framework. The engine converts 1.43 million canonical clips from 10 datasets into a unified, frame-aligned contract covering visual observations, metric camera geometry, captions, quality metadata, selection decisions, and provenance, while decoupling source processing from mixture construction. Under shared camera-conditioning, training, and inference interfaces, we instantiate four 5B–33B models based on Wan2.2, LTX-2.5, and MiniMax-H3 while preserving their native representations and objectives. A unified three-stage recipe combines bidirectional adaptation, teacher-forced autoregressive initialization, and distribution matching distillation. The resulting causal models enable real-time interaction over rollouts ranging from minutes to hours after being trained on only 5s sequences. By releasing the resulting data, pipeline, recipes, weights, and framework, SolarWM provides a reproducible and extensible foundation for interactive world-model research.
[CV-1] hinking in Pictures: A Systematic Benchmark for Reasoning -driven Image Generation
链接: https://arxiv.org/abs/2609.02864
作者: Yutong Liu,Nan Huang,Xu Cao,James M. Rehg
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Recent advancements in unified generative models (UGMs) and world simulators have achieved unprecedented results in visual perception and synthesis. However, these models primarily rely on surface-level event alignment, leaving the capacity for high-level visual reasoning underexplored. True visual generative intelligence demands “Reasoning-to-Generation”, an ability to infer latent rules from visual inputs and manifest solutions through precise, logically constrained visual outcomes. We introduce RIG-BENCH, a novel comprehensive benchmark that systematically evaluates Reasoning-driven Image Generation (RIG) across four cognitively demanding domains: Concept-based, Transformation-based, Pattern Structure, and Scenario-based. Featuring 2000 curated samples, RIG-BENCH serves as a rigorous stress test for RIG. Our extensive evaluations of state-of-the-art UGMs and image/video generation models reveal a significant reasoning-generation gap, wherein models frequently produce locally plausible but globally illogical outputs. RIG-BENCH provides a vital diagnostic framework to guide the development of next-generation, logically grounded UGMs and world simulators.
[CV-2] PlantC2USeg: Cross-Scale Consistent Pre-Training for Few-Shot Unified Plant Point Cloud Segmentation
链接: https://arxiv.org/abs/2609.02860
作者: Yu Tian,Xintong Jiang,Jan Franklin Adamowski,Shiv O. Prasher,Shangpeng Sun
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 27 pages, 20 figures
Abstract:Modern crop breeding demands precise organ-level analysis for trait quantification, making plant point cloud segmentation (PPCS) increasingly important. However, conventional deep learning approaches rely heavily on densely annotated datasets that are labor-intensive to acquire. Unified PPCS adaptation from distribution-shifted examples with minimal additional training remains challenging. To address this, we propose PlantC2USeg, a deep transfer learning framework featuring cross-scale consistency learning to explicitly align features across spatial scales and an information-restricted decoding strategy that prevents reconstruction shortcuts and promotes robust adaptation. The resulting pre-training enables stable few-shot generalization across species and sensing conditions, while unified fine-tuning with inherited thresholds further reduces adaptation overhead. Under full supervision on Soybean3D, PlantC2USeg achieves the highest semantic IoU and instance mWCov among compared methods, at 91.91% and 94.62%. With 20 labeled samples, it leads both metrics at 89.78% and 90.27%; with only 10 samples, it retains the highest mWCov of 83.23% while achieving 83.19% IoU. Across HR3D, 10-shot transfer to tobacco, tomato, and sorghum averages 78.41% IoU and 79.42% mWCov, while 22-shot transfer to SYAU-Maize achieves the highest IoU and mRec at 92.75% and 93.51%. Furthermore, a leading category-averaged mIoU of 85.0% on ShapeNet Part demonstrates the framework’s capability to handle diverse shape variations beyond agricultural domains. These results demonstrate that PlantC2USeg reduces overall adaptation effort under distribution shifts, enabling scalable plant phenotyping and transferable 3D representation learning beyond agriculture.
[CV-3] MuyBridge: Mobile Human Center-of-Mass Estimation from Monocular Video via Sparse Fusion
链接: https://arxiv.org/abs/2609.02854
作者: Aidan Bradshaw,Marco Giordano,David Rode,Andreas Habersack,Elif Basokur,Annika Kruse,Markus Tilp,Michele Magno,Peter Wolf,Luca Benini,Christoph Leitner
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:The 3D center of mass (CoM) is a primary quantity in the biomechanical analysis of sport, rehabilitation, and clinical movement, yet existing 3D pose tracking, mesh recovery, and multi-view triangulation methods either optimize 3D keypoint accuracy without anatomical constraints or carry compute and capture infrastructure too heavy to deploy where CoM tracking is most useful. As a result, the metric CoM remains difficult for coaches and movement analysts to measure from a single camera where athletes train and compete. In this work, we introduce MuyBridge, an on-device system that estimates the athlete’s segmental center of mass trajectory from a single phone camera video stream. MuyBridge couples a compact 2D pose network and a distilled single-step monocular depth network through an analytic metric fusion that uses anatomical and physical priors to anchor the metric CoM, requiring no 3D or task-specific supervision. Evaluated on the athletic movements of AthletePose3D (running, track and field, and figure skating), MuyBridge achieves 33-41 mm vertical CoM error and 2.3-6.6% absolute-relative range error (AbsRel) under a one-time calibration, and produces CoM estimates at the 63 FPS pose-estimation rate using asynchronous 2.86 Hz depth updates on iPhone 15. Code is available at: this https URL
[CV-4] RoGe: Novel View Synthesis via End-to-End Implicit Reconstruction and Generation
链接: https://arxiv.org/abs/2609.02847
作者: Xiaolei Lang,Ze Kang,Zehao Huang,Naiyan Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Novel view synthesis from sparse inputs requires both geometric grounding from the observed views and generative priors of unobserved regions, motivating recent hybrid methods that combine reconstruction and generation. However, existing methods bridge the two with rendered images or explicit 3D representations such as point maps or 3D Gaussians. Generation is thus conditioned on a lossy and imperfect projection of the scene, inheriting its errors, and reconstruction receives no signal from generation to correct them. We present RoGe, an end-to-end unified reconstruction and generation framework that removes this explicit bridge. It targets roaming within a scene anchored by sparse views: given a few posed images and a camera trajectory, it synthesizes a temporally coherent video along that trajectory. From the sparse input views, RoGe builds an implicit scene representation with a feed-forward reconstruction model, and queries it with target camera rays to obtain per-view geometric features. These features are injected into a video diffusion model as conditioning, without any 3D intermediate. Both modules are trained jointly, so the generation objective directly shapes its own geometric conditioning. We conduct experiments on DL3DV, where RoGe outperforms reconstruction-based, generation-based, and hybrid baselines on image-level metrics and video-level temporal consistency. Ablations confirm that ray-queried implicit features outperform both raw reconstruction tokens and rendered RGB as conditioning, and that joint training brings further gains.
[CV-5] Efficient All-in-One Weather Restoration using Spectral Harmonization
链接: https://arxiv.org/abs/2609.02839
作者: Paula Garrido-Mellado,Daniel Feijoo,Yuning Cui,Alvaro Garcia,Marcos V. Conde
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Technical Report
Abstract:Adverse weather conditions such as rain, haze, and snow significantly degrade image quality, posing challenges for both human perception and physical AI. Existing restoration methods require large computational budgets, struggling to process high-resolution images and handle different degradations. In this paper, we present Frequency Reconstruction via Spectral Harmonization, a novel lightweight all-in-one restoration method that explicitly decomposes feature representations into high- and low-frequency components at each scale of a hierarchical encoder-decoder architecture. By combining spectral decomposition with spatial processing through Fourier-based skip connections, FReSH-IR captures complementary frequency information without sacrificing spatial detail. Our approach achieves similar restoration quality with 80% fewer parameters and operations than transformer-based models. Extensive experiments demonstrate that our method offers a great efficiency-performance trade-off, highlighting its practical applications in constrained-resource systems.
[CV-6] Benchmarking RAW and RGB Restoration in Image Signal Processors BMVC2026
链接: https://arxiv.org/abs/2609.02831
作者: Zihao Lu,Radu Timofte,Marcos V. Conde
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted BMVC 2026: The 37th British Machine Vision Conference
Abstract:Modern cameras transform RAW sensor measurements into sRGB images through an image signal processor (ISP). We benchmark two placements for blind restoration around a fixed ISP: (A) pre-ISP restoration in the RAW domain and (B) post-ISP restoration in the sRGB domain. The benchmark covers four smartphone device groups, two learned ISPs, three degradation regimes–noise, blur, and joint noise and blur–, and several representative RAW and RGB restoration models. Our results show that placement alone does not determine performance. The RAW restoration strategy outperforms the best generic RGB restoration models. However, RGB restoration models trained considering the ISP transformations, achieve the best overall performance. Our novel benchmark demonstrates that the image reconstruction performance strongly depends on the alignment between the restoration model and the target imaging pipeline. We consequently recommend reporting restoration placement and ISP-aware supervision as key experimental factors. Our code is available at this https URL
[CV-7] GDB-Reward: From Evaluation Metrics to Training Rewards for Graphic Design
链接: https://arxiv.org/abs/2609.02813
作者: Adrienne Deganutti,Purvanshi Mehta,Simon Hadfield,Andrew Gilbert
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Text-to-image models excel at natural image synthesis but struggle with graphic design, where success depends on satisfying precise constraints on typography, layout, color, and visual communication. While prompt optimization offers an attractive alternative to expensive diffusion model fine-tuning, learning prompts for frozen image generators requires informative reward functions despite the entirely non-differentiable generation process. Reinforcement learning does not require differentiable objectives; it requires only scalar rewards capable of ranking candidate outputs. This raises a simple question: can design evaluation metrics themselves become reinforcement learning rewards? Our central contribution is GDB-Reward, a framework that systematically transforms heterogeneous graphic design evaluation metrics into a unified reinforcement learning reward. Experiments demonstrate that GDB-Reward provides an effective optimization objective, substantially improving adherence to the design specification in perceptual quality, rendering fidelity, and spatial accuracy while keeping the image generator entirely frozen. More broadly, our results demonstrate that heterogeneous, non-differentiable evaluation metrics can move beyond passive benchmarking to become effective optimization objectives for reinforcement learning in domains where differentiable supervision is unavailable.
[CV-8] AutoCompass: Accurate Visual Localization on Public Maps by Learning from Weak Labels ECCV2026
链接: https://arxiv.org/abs/2609.02798
作者: Javier Tirado-Garín,Alan Savio Paul,Shuai Chen,Axel Barroso-Laguna,Tommaso Cavallari,Daniyar Turmukhambetov,Victor Adrian Prisacariu,Eric Brachmann
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: ECCV 2026
Abstract:Neural map matchers estimate an image’s 3-DoF pose relative to a 2D map. These models are trained on large-scale datasets of geo-referenced images, whose position and heading labels often contain noise that affects the trained models. To address this, we present AutoCompass, a supervision approach for training neural map matchers from inaccurate absolute pose labels. First, we show that heading labels are unnecessary: trained from raw GPS labels, models learn to predict accurate headings, automatically. Second, defining a tolerance region around raw GPS improves positional accuracy. Third, if available, our supervision uses relative poses between training images, obtained via SLAM or SfM, which provide a more accurate training signal. Across driving and egocentric benchmarks, AutoCompass consistently outperforms counterparts trained with the usual strong reliance on absolute pose labels.
[CV-9] Video-Based Palm-Vein Authentication under Challenging Conditions
链接: https://arxiv.org/abs/2609.02776
作者: Xiaofeng Yan,Kechen Liu,Abhilash Venkatesh,Cathy Zhang,Xia Zhou,Salvatore Stolfo
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Palm-vein biometrics are increasingly used for secure, contactless authentication. Yet real-world deployment exposes them to surface noise (sweat, dirt), illumination and motion variation, and temperature-driven changes in vascular visibility, which remain underexplored for lack of data captured under such conditions. To study these effects, we introduce the Columbia University Palm-vein (CUP) dataset, to our knowledge the first public video-based palm-vein dataset. CUP records every palm under four surface conditions (a clean baseline, warm, wet, and dirty) and pairs each subject with physiological and demographic metadata. On it we benchmark twenty-one recognizers spanning static, video, and multi-frame aggregation architectures. Models that verify reliably on clean palms lose most of their accuracy on dirty ones, and the mean equal error rate (EER) roughly quadruples. We recover much of that robustness along both axes of the capture. Temporally, a consensus over the few frames the sensor already returns cancels transient corruption; spatially, a test-time matcher that adds no learned parameters fuses the global cosine with a saliency-steered region-level optimal transport that routes the comparison around corrupted regions. The full design leads on every surface of CUP in EER, TAR@FAR=0.01, and Rank-1, at 4.3M parameters and 3.1 GFLOPs, a fraction of the video models’ cost. Attached to four frozen state-of-the-art backbones it cuts their mean EER by 29-37% without retraining, and on four public single-image datasets the regional matching alone still helps. A preliminary audit across ten demographic and physiological traits finds two warm-condition gaps, along body water and gender, that survive multiple-comparison correction. CUP will be released for non-commercial research use at this https URL upon publication.
[CV-10] Multi-Tool Image Editing Attribution in Facial Forgery
链接: https://arxiv.org/abs/2609.02751
作者: Sheng Liu,Qiang Sheng,Danding Wang,Yu Li,Chenming Zhou,Juan Cao
类目: Computer Vision and Pattern Recognition (cs.CV); Multimedia (cs.MM)
备注: Accepted to ACM Multimedia 2026 (MM 2026)
Abstract:As generative AI tools become increasingly powerful and easy to use, people can easily edit portrait images with a prompt, necessitating the task of image editing attribution, which predicts the involved editing tools from the given image. Existing attribution methods hold the single-tool assumption and can only attribute a specific editing tool, but struggle to handle the more complex and increasingly common multi-tool editing scenarios, where artifacts left by different editing tools are composite and overlapped. To address this gap, we explore Multi-Tool Image Editing Attribution (MIEA), which aims to identify multiple editing tools involved in a multi-tool edited facial image. To simulate the real-life editing operations on facial images, we then construct a new dataset, MultiEdit, which contains 500k+ edited facial images and covers six types of editing tools that support face swapping (Deepfake) and various facial enhancements. Inspired by the findings from data analysis, we design DPEC, a multi-tool attribution method that can capture distinguishable, locality-aware editing tool traces from both spatial and frequency domains with the support of an error-based curriculum learning strategy. Experiments show \Method\ outperforms nine methods for facial images edited in at most five steps.
[CV-11] Balancing Frequencies and Pixels in Flow Matching
链接: https://arxiv.org/abs/2609.02748
作者: Lucas Degeorge,Paul Couairon,Arijit Ghosh,Alexei A. Efros,David Picard,Vicky Kalogeiton
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Natural images follow a 1/f^2 spectral distribution: most signal energy lies in the low spatial frequencies, while the perceptually important structures such as textures and edges occupy sparse high-frequency bands. Pixel-space reconstruction objectives, however, treat all spatial errors uniformly, causing low frequencies to dominate the optimization signal and delaying the learning of fine-scale details. In this work, we identify this objective-level spectral imbalance as a key inefficiency in training pixel-space flow models. To address it, we propose a Focal Log-Frequency Loss (f-loss), a spectrally balanced objective that equalizes the learning signal across frequencies, emphasizing high-frequency components that are otherwise underrepresented in pixel-space objectives. Building on this, we introduce a simple training strategy that combines frequency and pixel supervision: we first emphasize frequency-domain learning early to capture all frequencies, and then transition to standard pixel-space v-loss for spatial refinement. This balancing mitigates the low-frequency bias of pixel losses and aligns the training signal with the evolving needs of the model. Our approach is conceptually simple, requires no architectural changes, and acts as a drop-in replacement for flow matching losses. Across multiple model scales, it accelerates convergence by up to 40% while consistently improving FID and perceptual fidelity. We will release code and models.
[CV-12] InceptionGS: Generative Bootstrapping for Large-Scale Gaussian Splatting under Unstructured View Sampling
链接: https://arxiv.org/abs/2609.02747
作者: Tianheng Lu,Guangyu Wang,Ruqi Huang,Lu Fang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Achieving truly immersive large-scale scene digitization necessitates consistent and visually pleasing rendering across all possible viewing perspectives. However, collecting multi-view images covering every fine detail of a large-scale scene is prohibitive due to scene complexity, capture cost, negligence, or accessibility constraints. As a result, the sampled views tend to be highly unstructured – the majority of the scene is well covered yet certain regions inevitably lack sufficient observations. Existing reconstruction based methods are vulnerable to view scarcity while generation based approaches suffer from generalization, controllability, and 3D consistency issues. To address this challenge, we propose InceptionGS, which bootstraps Gaussian splatting by subtly balancing reconstruction and generation. Starting from an initial Gaussian splatting, InceptionGS reasonably rethinks and repairs problematic regions caused by view scarcity while preserving the quality elsewhere, by softly incorporating scene- and view-adaptive generative priors. Extensive experiments on real-world large-scale scenes demonstrate the superiority and broad applicability of our approach in handling unstructured imagery and boosting high-fidelity Gaussian splatting. Please refer to the supplementary video for better visual demonstrations.
[CV-13] RVSD: Retrieval Vision Sparse Decoding for Mitigating Visual Hallucinations in Large Vision-Language Models
链接: https://arxiv.org/abs/2609.02731
作者: Canjie Liu,Jiawen Kang,Jinbo Wen,Zishao Zhong
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Large vision-language models have achieved remarkable success in vision-language tasks. However, they remain prone to Visual Hallucinations (VHs), undermining their reliability in real-world applications. Existing solutions typically require curated datasets, additional training, or multi-round decoding, resulting in considerable computational overhead. In this paper, we propose \textbfRVSD (\underlineRetrieval \underlineVision \underlineSparse \underlineDecoding), a training-free and plug-and-play decoding framework that, for the first time, unifies token sparsification and \textbfSemantic-Space Visual Retrieval (SSVR) within a single decoding pass. Within RVSD, we introduce a \textbfsemantics-directed token selection strategy that selectively sparsifies redundant tokens while preserving critical visual information. We further propose the SSVR mechanism, which reformulates visual compensation as an on-demand cross-modal retrieval process within a shared semantic space. Extensive experiments demonstrate that RVSD achieves state-of-the-art performance in mitigating VHs while maintaining robust suppression capabilities under long-context generation settings. Our code is available here.\footnotethis https URL
[CV-14] MV-dVRK: A Multi-Viewpoint Benchmark for Spatial Surgical Perception
链接: https://arxiv.org/abs/2609.02717
作者: Guido Caccianiga,Sergey Prokudin,Yutong Chen,Bernard Javot,Rachael L’Orsa,Omer Burak Aladağ,Yarden Sharon,Jens Rolinger,Ivan Capobianco,Anton Deguet,Siyu Tang,Katherine J. Kuchenbecker
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注:
Abstract:Large-scale training and refined optimization techniques have greatly improved sparse multi-view 3D reconstruction. Despite their relevance to surgery, such methods have never before been rigorously evaluated on real endoscopic images. Current clinical telerobots deploy a single stereo camera inside the patient, making multi-viewpoint data extremely rare. This paper presents MV-dVRK, the first ex-vivo surgical dataset to combine multiple exposure-synchronized stereo viewpoints with accurate surface geometry and camera poses. The static subset of the benchmark provides dense SfM reference geometry, validated against an industrial 3D scanner, together with ground-truth camera poses and sparse-view test sets. We use MV-dVRK to systematically compare zero-shot monocular, stereo, multi-stereo, and multi-view 3D reconstruction methods as the number of viewpoints increases. With two endoscopes, multi-stereo reconstruction achieves the highest coverage. With a third viewpoint, optimization-based multi-view methods perform best, covering 67% of ground-truth surface points within a 1 mm tolerance and recovering highly accurate relative camera poses. By contrast, feed-forward foundation models cover only 43% of the ground-truth surface in the same setting. MV-dVRK also includes ten dynamic sequences spanning multiple surgical tasks, with increasing kinematic complexity and tissue deformation, providing a basis for future research in multi-viewpoint surgical perception. The project is available at: this https URL.
[CV-15] A Top-Down Framework for Metric-Scale Athlete Localization from Single Broadcast Frames
链接: https://arxiv.org/abs/2609.02705
作者: Thanh-Khoi Nguyen,Hoang-Phuc Nguyen,Linh-Huynh,Minh-Triet Tran
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Accurate world-coordinate localization of athletes from single-frame broadcast footage is inherently challenging due to extreme scale disparities in ultra-high-resolution imagery. In this paper, we propose a top-down framework for metric-scale athlete localization from a single calibrated frame. Our approach centers on three key contributions. First, we propose Boundary-Aware Adaptive Tiling, a semantics-guided extension of standard sliced inference. By iteratively expanding tile boundaries based on coarse bounding-box predictions, it systematically ensures full object containment, effectively mitigating boundary-splitting artifacts through a lightweight pipeline adaptation without architectural modifications. By substantially mitigating recall degradation under extreme scale variance, Boundary-Aware Adaptive Tiling enables us to isolate perspective distortion as the primary source of residual localization error. Second, we adapt the RTMPose-X architecture into a specialized two-keypoint estimator (pelvis and ground projection), employing a reformulated Gated Attention Unit optimized for this geometrically coupled point pair, and then deterministically lift the 2D ground projections into world coordinates via camera-calibrated ray casting. On the public test set, our method achieves a LocSim score of 97.44 and an mAP of 0.9128, outperforming the baseline by over 21 % and establishing a robust solution for high-resolution scale variance.
[CV-16] Generating Medical Image Counterfactuals using Causal Explanations
链接: https://arxiv.org/abs/2609.02697
作者: David A. Kelly,Tom Yaacov,Nathan Blake,Sander Beckers,Hana Chockler
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Deep learning models have achieved impressive performance in medical image diagnosis, yet their deployment in clinical settings remains constrained by limited explainability. Counterfactual images provide one means of auditing model behavior by showing how an image would need to change for a classifier to produce a different prediction. Existing approaches typically generate such explanations using auxiliary models, including generative adversarial networks and diffusion models. While often capable of producing visually realistic images, these methods explain one black-box model using another, making it difficult to separate the classifier’s decision-making process from the inductive biases of the generator. We propose a novel counterfactual-generation framework that requires no generative model. Instead, counterfactuals are constructed directly from causal evidence extracted from the classifier. The resulting approach is deterministic, requires no additional model training, and enables controllable edits within user-specified regions of interest. Experiments on real-world medical imaging datasets demonstrate that the proposed method successfully changes classifier predictions while remaining closer to the original image than generative baselines, providing a more direct and transparent view of the classifier’s decision boundary. Subjects: Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2609.02697 [cs.CV] (or arXiv:2609.02697v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2609.02697 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-17] GaLe: memory-efficient Global Approximate and Local Exact features
链接: https://arxiv.org/abs/2609.02689
作者: Alberto Ancilotto,Elisabetta Farella
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Embedded devices typically lack the resources of GPU-equipped machines, and existing inference methods suffer from either high computational overhead (patch-based) or accuracy loss (approximation-based). We propose GaLe, a memory-efficient technique that enables the deployment of pretrained networks on constrained devices without retraining. GaLe partitions feature maps into two components: a local exact (Le) representation that preserves fine details and a global approximate (Ga) representation that retains long-range dependencies. Unlike standard tiling, GaLe supports global operations and attention mechanisms found in hybrid CNN-transformer models. Validated on ImageNet, our method matches exact-inference performance while achieving up to 65% speedup and 90% RAM reduction on a Cortex-M33 compared to patch-based inference. We further demonstrate GaLe’s versatility across classification, detection, and generation tasks, highlighting its potential as a foundation for resource-efficient architecture design.
[CV-18] Genesis: A Generative Engine for Hierarchical Satellite Image Synthesis
链接: https://arxiv.org/abs/2609.02683
作者: Subash Khanal,Yangzhi Cui,Daniel Cher,Eric Xing,Brian Wei,Srikumar Sastry,Nathan Jacobs
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to SIGSPATIAL 2026: Application Track (Oral)
Abstract:Earth observation is fundamentally multi-scale; geospatial tasks span varied resolutions, and satellite imagery is organized into cascading tile pyramids that nest fine detail within wide coverage. Current generative models of satellite imagery, however, operate along a single axis: they either zoom to enhance a single tile’s resolution or pan to extend imagery at a fixed scale. As a result, no existing method produces a complete pyramid that stays consistent across both scale and space, where a high-zoom tile must agree with the coarse context it refines and with the neighbors it meets. Motivated by this gap, we introduce a new task, multi-scale tile completion: given a sparse set of seed tiles at arbitrary zoom levels and positions, synthesize a complete, uniform quadtree that is globally consistent across both scale and space. We approach this task with Genesis, a generative engine that brings both axes together by composing two specialized operators over the quadtree: a vertical super-resolution model and a horizontal mask-based outpainting model, producing pyramids that are consistent across zoom levels and seamless across neighboring tiles. Each operator achieves state-of-the-art results on its subtask, and the engine propagates sparse seeds into seamless, multi-resolution maps from any initial configuration. To evaluate the task and benchmark Genesis, we introduce dense500, a fully observed multi-scale pyramid dataset spanning diverse geographic regions, together with a suite of pyramid-level metrics. Code, models, and our dataset are available at this https URL.
[CV-19] Query Rewriting for Complex Object Segmentation in 4D Gaussian Representations
链接: https://arxiv.org/abs/2609.02664
作者: Thanh-Khoi Nguyen,Thien-Phuc Tran,Minh-Triet Tran
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Recent 4D Gaussian representation frameworks have demonstrated strong performance in language-guided dynamic scene understanding. However, these methods remain highly sensitive to verbose and narrative-style queries that contain noisy contextual information. In this paper, we investigate the impact of query rewriting for complex object segmentation in 4D Gaussian representations. Inspired by recent findings in retrieval-augmented language models and keyword-guided query reformulation, we propose a training-free reinterpretation strategy that transforms long descriptive queries into concise keyword-grounded forms. Our approach progressively reduces linguistic noise while preserving semantic anchors relevant to object-centric representations. Experiments on HyperNeRF and Neu3D demonstrate that concise rewritten queries significantly improve both temporal localization and spatial segmentation performance. In particular, our method improves average temporal accuracy from 60.92% to 92.21% and average vIoU from 20.08% to 76.94% without any additional fine-tuning. Extensive ablation studies further reveal that shorter, keyword-focused queries consistently yield stable video-feature similarity distributions and better alignment with object-centric Gaussian representations
[CV-20] Characterizing Text Branch Sensitivity in Medical Vision-Language Segmentation via Evidence Decoupling
链接: https://arxiv.org/abs/2609.02663
作者: Ziquan Liu,Zhewei Zhu,Xuyang Shi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 16 pages, 3 figures
Abstract:Pretrained vision-language models (VLMs) have shown promising performance in medical image segmentation by incorporating clinical text. However, it remains unclear how much textual information actually contributes to pixel-level predictions. In this work, we systematically investigate the role of text in multimodal medical image segmentation. We first analyze several commonly used fusion strategies and find that segmentation performance is largely insensitive to the choice of fusion module. To further understand modality interactions, we propose an Evidence Decoupling Decoder (EDD) based on evidential deep learning and deep supervision. EDD serves as an internal representation analysis tool that decomposes image evidence and text-modulated evidence throughout the decoding process while maintaining competitive segmentation performance. Experimental results show that the sensitivity to text perturbation varies substantially across datasets. On BUSI and BTMRI, removing text causes catastrophic performance drops, indicating strong model reliance on textual input. On ISIC and Kvasir-SEG, text exerts relatively marginal influence. We further find that text affects predictions mainly through global semantic modulation rather than independent spatial localization, and that the specific semantic components driving text sensitivity differ across datasets. These findings provide a deeper understanding of modality interaction in multimodal medical image segmentation and offer practical insights for future model design.
[CV-21] Physics-Driven Independent Pair Generation for Iterative Self-Supervised Low-Dose CT Denoising
链接: https://arxiv.org/abs/2609.02654
作者: Xianlei Han,Shaoyu Wang,Jiancheng Fang,Weiwen Wu,Qiegen Liu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 10 pages, 11 figures, 3 tables. Includes a 1-page appendix. Submitted to IEEE Transactions on Circuits and Systems for Video Technology
Abstract:Low-dose computed tomography (LDCT) measurements contain mixed Poisson-Gaussian noise. However, most self-supervised methods rely on generic image statistics and do not explicitly model this noise, which may limit their ability to effectively suppress realistic LDCT noise. To address this issue, we propose a physics-driven framework with cross-domain iteration for self-supervised LDCT denoising. The proposed framework proceeds in three main steps. First, a learned sinogram prior and the LDCT noise model guide posterior inference of photon counts, enabling separation of the Poisson and Gaussian components. Second, the separated Poisson and Gaussian components are respectively processed by binomial thinning and Gaussian data thinning to construct two branches, and residual scaling matches each branch’s noise level to that of the observation, yielding a training pair with approximately independent noise realizations from one low-dose measurement. Finally, the pair is used to train an image-domain network whose forward-projected outputs update the prior. Through cross-domain iteration, the prior and the training pair are progressively refined while maintaining consistency with CT acquisition physics. Experiments on simulated data from AAPM, LIDC-IDRI, and LoDoPaB-CT and on real LDCT data show consistent gains over the evaluated self-supervised baselines across dose levels, with performance comparable to the evaluated supervised baseline.
[CV-22] Learning to Attract and Repel: Dual Quality Margin Learning for Face Recognition (DQM-Face) ECCV2026
链接: https://arxiv.org/abs/2609.02644
作者: El Ouanas Belabbaci,Bhavesh Wani,Philipp Terhörst
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: ECCV 2026. Code: this https URL
Abstract:Face recognition in unconstrained environments remains highly challenging due to diverse and extreme variations encountered in real-world scenarios. To mitigate these effects, existing margin-based approaches model sample quality through feature magnitude. However, magnitude-based modeling alone is susceptible to identity-agnostic noise, which can degrade the reliability and discriminative power of learned representations. In this paper, we propose Dual Quality Margin Learning for Face Recognition (DQM-Face), a novel framework that enables refined attraction and repulsion dynamics during representation learning. Our approach unifies conventional magnitude-based quality estimation with a newly introduced semantic quality learning mechanism, realized via squeeze-and-excitation semantic attention. By jointly leveraging magnitude and semantic cues, we construct enhanced quality-aware margins that adaptively strengthen intra-class compactness through improved attraction during learning. To further enhance inter-class discrimination, we introduce a repulsion margin formulation that explicitly enlarges inter-class separation. The unified integration of semantic quality modeling with dual attraction-repulsion margin optimization results in a more structured and discriminative feature geometry. Extensive experiments on multiple challenging benchmarks demonstrate that DQM-Face consistently outperforms state-of-the-art face recognition methods. Moreover, we show that the quality learned for margin optimization is highly effective for face image quality assessment within the proposed framework, demonstrating that the learned quality signal is intrinsically aligned with the recognition objective. The code is publicly available: this https URL
[CV-23] From Detection to Localization: A Unified Forensics Framework for Fully Synthetic and Tampered Images
链接: https://arxiv.org/abs/2609.02640
作者: Annalisa Gallina,Marco Fiorucci,Marco Brigo,Federica Battisti,Lamberto Ballan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at the DFF Workshop, ACM Multimedia 2026
Abstract:The rapid advancement of generative models has significantly worsened the problem of manipulated image detection, as these methods are capable of producing highly realistic forgeries, reinforcing the importance of multimedia forensics. Conventional approaches typically frame image manipulation detection as a binary classification task (real vs. generated), which limits the capability to distinguish and localize different forms of manipulation. To address these constraints, this work extends an existing detector by introducing a unified multiclass framework (real vs. fully generated vs. tampered). In addition to classifying image authenticity, the framework incorporates a segmentation branch to enable pixel-level localization of tampered regions. The proposed approach outperforms selected recent benchmarks, offering an efficient solution with improved classification accuracy and higher IoU scores for the localization task. Find the code at this https URL.
[CV-24] AffectDelta: Beyond Emotion Labels for Image Editing
链接: https://arxiv.org/abs/2609.02616
作者: Xingzu Zhan,Lin Gu,Ruogu Fang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 12pages, 6 figures
Abstract:Emotion-driven image editing aims to evoke a specified target emotion by modifying emotion-relevant visual cues in a source image, while preserving the overall composition and semantic-structural coherence of the original scene. Existing scene-level editors typically specify the target with a single emotion category and often learn visual transformations from operation-level text instructions. A category collapses a mixed affective endpoint into one dominant label, while language cannot precisely quantify how coexisting emotions should increase, decrease, or remain stable. We introduce AffectDelta, a source-aware editor that treats editing as a transition between eight-dimensional emotion distributions. A frozen Emotion Distribution Predictor estimates the source state, and the signed source-to-target difference encodes the direction and magnitude of the requested transition. Within AffectDelta, an internal transition encoder and a source-aware diffusion backbone jointly translate this signal into context-dependent semantic and appearance changes. To train this formulation, we construct AffectPair-249K, comprising 248,841 source-target pairs with predicted eight-dimensional distributions and spanning both cross-category and within-category transitions. Experiments against six baselines, combining quantitative evaluation with qualitative comparisons, demonstrate improved affective alignment and content preservation, while ablations validate our design choices. Code and dataset will be made publicly available upon acceptance.
[CV-25] Generalizable Brain Tumor Segmentation with Self-Training and Tumor-Aware Deformations MICCAI
链接: https://arxiv.org/abs/2609.02600
作者: Henrique Zan Grande,Jeovane Honorio Alves,Rayson Laroca,Andre Gustavo Hochuli
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted for presentation at the 2026 International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) - BraTS Cluster of Challenges
Abstract:This work presents an approach to the Generalizability Across Tumors (BraTS-GoAT) task of the BraTS 2026 Challenge, which focuses on robust segmentation of brain tumor sub-regions across a heterogeneous patient population. The proposed method employs the nnU-Net framework with a large residual encoder architecture, integrating a semi-supervised learning technique with pseudo-labels generated from the unlabeled training data and a tumor-aware deformable augmentation that locally deforms the lesion while preserving the surrounding anatomy. We evaluate the individual contributions of each component, as well as their combination, using varying proportions of the most confident pseudo-labeled cases. The submitted configuration for the generalization task achieves Dice and NSD scores of 0.881 and 0.473 for Whole Tumor, 0.817 and 0.490 for Tumor Core, and 0.775 and 0.533 for Enhancing Tumor on the BraTS-GoAT validation set, improving over the labeled-only baselines across all tumor regions and confirming that self-training and the proposed augmentation are complementary. Our source code is publicly available at this https URL.
[CV-26] Deeply Interleaved Text-Image Contexts for Multimodal LLM s Assessment
链接: https://arxiv.org/abs/2609.02573
作者: Zihao Wang,Xi Xiang,Yuwen Sun,Yingyu Li,Yabo Zhang,Yihan Zeng,Fan Li,Wangmeng Zuo
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Current evaluations and training of multimodal models predominantly focus on multi-image tasks, largely overlooking interleaved text-image scenarios. In such multi-image tasks, text typically serves merely as task instructions, lacking deep semantic interaction with the visual content. In contrast, realworld applications like text-image co-creation, character tracking, and spatial reconstruction require constant interaction between text and images. Consequently, models must possess a deep understanding of these interleaved contexts. To bridge this gap, we introduce a novel benchmark, TIC-Bench (deeply interleaved Text-Image Contexts), designed to evaluate the capability of models to integrate text-image clues and recover the ground truth facts within deeply interleaved contexts. This benchmark encompasses three core domains: Logical, Temporal, and Spatial Association, which are further categorized into eight specific types, comprising a total of 2,280 questions. We evaluated 10 state-of-the-art MLLMs and observed a substantial performance gap compared to human experts, together with persistent difficulties in integrating evidence distributed across interleaved visual and textual inputs. Ultimately, this benchmark provides a valuable analytical tool for assessing and advancing the ability of multimodal models to effectively integrate text and image information in deeply interleaved contexts. TIC-Bench is publicly available at this https URL
[CV-27] MARS: What Retrieval Signals Are Hidden in Multimodal Large Language Models for Text-Video Retrieval? EMNLP2026
链接: https://arxiv.org/abs/2609.02565
作者: Uicheol Jung,Juyoung Hong,Geuntaek Lim,Yukyung Choi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 16 pages, 6 figures. Accepted to the Main Conference of EMNLP 2026
Abstract:Text-video retrieval requires representations that can distinguish videos with similar scenes, actions, and temporal patterns. Recent multimodal large language models have been adapted as embedding models, but they often represent each input using a single token from the final layer. This can compress diverse video-text cues into a single vector and limit fine-grained retrieval. To address this limitation, we propose MARS, a multi-layer and multi-slot embedding framework for text-video retrieval. MARS constructs multiple adaptive representation slots by combining hidden states from different decoder layers, compares corresponding text and video slots, and aggregates their similarities for retrieval. To better handle confusing candidates, we further introduce a hard-negative-aware slot specialization objective that encourages the slots to capture discriminative matching cues. Experiments on four text-video retrieval benchmarks show that MARS achieves state-of-the-art results in both direct similarity-based retrieval and reranking settings. Ablation studies and analyses demonstrate that multi-layer fusion, multiple slots, and hard-negative-aware slot specialization provide complementary gains. Code is available at this https URL.
[CV-28] Stereo 4D Radar for 3D Object Detection: Integrating Geometric Alignment and Absolute Velocity Estimation
链接: https://arxiv.org/abs/2609.02560
作者: Seung-Hyun Song,Dong-Hee Paek,Woong-Chan Byun,Seung-Hyun Kong
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Four-dimensional (4D) Radar is a powerful sensing modality capable of detecting surrounding three-dimensional (3D) objects under diverse weather conditions and providing Doppler-based motion information. However, raw 4D Radar signals contain significant clutter from road surfaces, guardrails, and surrounding vehicles, along with multipath-induced ghost reflections and the receiver’s inherent noise floor. Consequently, preprocessing algorithms designed to remove such invalid measurements often make the Radar data excessively sparse. Moreover, the Doppler measurements provided by 4D Radar describe only the radial component of an object’s velocity, limiting their ability to recover the full motion state. In this paper, we introduce a stereo 4D Radar-based 3D object detection framework that exploits the geometric disparity between left and right Radars to estimate the absolute velocity of objects and achieve more robust perception through the fusion of their complementary features. The effectiveness of the proposed framework is validated on our in-house stereo 4D Radar dataset, demonstrating performance gains of 8.82 points in AP 3D and 9.0 points in AP BEV over state-of-the-art mono 4D Radar baselines. These results demonstrate that absolute velocity estimation combined with stereo geometry-aware feature fusion leads to substantial improvements in 3D object detection.
[CV-29] RGB-to-IR image translation for infrared vehicle detection in unseen UAV domains
链接: https://arxiv.org/abs/2609.02556
作者: Thijs A. Eker,Ella P. Fokkinga,Jan Erik van Woerden,Elfi I.S. Hofmeijer,Sebastiaan P. Snel,Klamer Schutte,Friso G. Heslinga
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Submitted to SPIE Sensors + Imaging 2026
Abstract:Synthetic training data is crucial for developing vision AI when real-world data is scarce, as in thermal infrared (IR) aerial vehicle detection. While abundant UAV RGB imagery motivates RGB-to-IR translation for data augmentation, unobservable thermal traits (e.g., engine heat) make learning transferable mappings challenging. This work investigates whether modern generative translators can overcome this cross-modal gap to improve infrared vehicle detection on unseen UAV target domains. Translators are trained on paired RGB-IR source datasets and applied to RGB training images from held-out target datasets to generate synthetic IR data. Evaluated methods include supervised GANs, ControlNet-based diffusion models, and foundation-model editing via LoRA. The resulting synthetic IR imagery is used to train RF-DETR vehicle detectors, which are evaluated on unseen IR target test splits across five aerial datasets, with Kust4K and VTUAV serving as target domains. Synthetic IR consistently outperforms RGB and grayscale baselines. Stable Diffusion 3.5 with ControlNet yields the best results, improving mAP from 50.8 to 60.1 on Kust4K and from 25.6 to 38.4 on VTUAV compared to models trained only on source-domain IR data. Increasing output diversity via multiple seeds (+1.1 mAP) and prompt variations (+3.3 mAP) provides additional gains on VTUAV. Although a performance gap to real target IR data remains, generative RGB-to-IR translation effectively mitigates IR data scarcity and improves cross-domain aerial vehicle detection.
[CV-30] Spatially Aware World Action Model via Geometric Latent Diffusion
链接: https://arxiv.org/abs/2609.02531
作者: Javier Alejandro Lopetegui Gonzalez,Paul Pacaud,Cordelia Schmid
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注:
Abstract:World Action Models (WAMs) leverage the capabilities of large-scale pretrained video diffusion models to jointly predict future observations and actions, inheriting rich visual and physical priors from internet-scale video. This has made them a promising paradigm for robot policy learning, yet the prevailing models operate exclusively on RGB observations and do not leverage 3D information. To bridge this gap, we introduce a Spatially Aware World Action Model (SA-WAM), which repurposes a pretrained video model for joint action, RGB, and depth prediction, enabling 3D-aware world modeling and action prediction within a single diffusion backbone. We use a nonlinear encoding that maps the unbounded depth signal into the bounded input domain expected by the frozen VAE tokenizer. This allows us to reuse the tokenizer without 3D-specific fine-tuning, incorporating geometric information without sacrificing the pretrained priors. SA-WAM achieves state-of-the-art results on the RoboCasa and LIBERO-Plus benchmarks, while simultaneously improving future-state predictions. Furthermore, SA-WAM outperforms strong baselines in real-world evaluation using a UR5 robotic arm, with strong gains in randomized environments. We analyze the correlation between world model prediction quality and rollout success, providing insights into WAM performance and avenues for its improvement.
[CV-31] Fine-Grained Anomaly Perception in Wild UGC-Enhanced Images: A Comprehensive Dataset and Difference-Fusion Framework
链接: https://arxiv.org/abs/2609.02529
作者: Yan Zhong,Gefei Chen,Qiufang Ma,Zhen Wang,Zhiwei Fan,Lei Shi,Tingting Jiang
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Image enhancement and restoration have become standard back-end operations on short-video and social media platforms to boost UGC visual experience. Yet these processes inevitably introduce visual anomalies–especially in faces, texts, and textures–that directly undermine perceptual fidelity and viewer trust. While existing IQA methods perform well on classic distortions, they target holistic quality assessment and fail to capture the specific, localized anomalies caused by enhancement algorithms in real-world UGC. To bridge this gap, we formally define a new task-quality Anomaly Perception for UGC image Enhancement (UEAP), and contribute the first UEAP benchmark dataset, named UEAP-4k, curated from the real business scenarios. It provides fine-grained annotations for anomaly categories, localization and severity levels. Furthermore, we propose a Difference-Fusion Anomaly Perception Method (DFAP-UGC) for wild UGC-enhanced images, which leverages explicit problem-reference difference fusion with dense spatial querying, regional verification, and quality-aware ranking, enabling robust anomaly identification in challenging scenarios. To handle the inherent coupling of subtasks in this new task, we propose a Locality-Aware Dynamic Task Prioritization (LADTP) training strategy that enables effective end-to-end learning and eliminates multi-stage overhead. Extensive experiments show that our method outperforms baselines adapted from classical approaches for this task, validating the value of this dataset and the superior of DFAP-UGC for robust UGC-enhanced image anomaly perception. Code and data will be public.
[CV-32] Doppio: A Dataset for Contactless Weight Estimation of Falling Particles
链接: https://arxiv.org/abs/2609.02528
作者: Simon Kiefhaber,Jan-Martin O. Steitz,Julia Grabinski,Christoph Reich,Paul Wagner,Max Zimmermann,Simone Schaub-Meyer,Stefan Roth
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Measuring the mass of powder, including falling particles, is a common task in industrial applications. While scales are effective for static measurements, many applications require contactless sensing, where existing solutions are often costly, application-specific, and technically complex. In this work, we investigate computer vision as a practical alternative for contactless mass estimation. As an accessible real-world case study, we focus on coffee grinding and introduce \emphDoppio, a novel video dataset capturing videos of falling ground coffee, paired with precise, per-frame ground-truth weight measurements. To demonstrate contactless measuring, we evaluate deep learning-based approaches ranging from purely spatial feed-forward networks to recurrent spatio-temporal models. These models are analyzed with respect to their predictive accuracy and computational trade-offs. We demonstrate that deep learning-based computer vision models accurately estimate the cumulative weight of falling particles, establishing a solid foundation for future vision-based contactless measurement solutions.
[CV-33] SR-Edit: Region-Aware Image Editing via Self-Refinement
链接: https://arxiv.org/abs/2609.02504
作者: Andong Wang,Zehua Chen,Yuxuan Jiang,Jun Zhu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:With the recent rapid progress in generative models, image editing has made remarkable advances, yet achieving faithful edits that precisely modify only the target regions while strictly preserving all other regions remains challenging. Since externally provided region annotations are often difficult to obtain in practice, a growing body of work seeks to improve preservation by automatically inferring edit and non-edit regions, and then enforcing consistency on the latter. However, these approaches still suffer from inaccurate region estimation and heuristic correction strategies that distort the native inference process, making methods designed for fidelity themselves a new source of artifacts. We propose SR-Edit, an image editing framework that overcomes these issues via iterative self-refinement. Specifically, at each iteration, SR-Edit first (i) extracts progressively precise and self-consistent region separation from the model’s own predictions by lightweight post-processing, and then (ii) enforces preservation in non-edit areas through correction updates that remain aligned with the original sampling dynamics. Extensive experiments demonstrate that SR-Edit achieves superior preservation and overall image quality compared to existing editing techniques.
[CV-34] Blending Concepts: Benchmarking Visual Metaphor Generation in Text-to-Image Models
链接: https://arxiv.org/abs/2609.02502
作者: Chuer Chen,Zichen Wang,Yi He,Zhengxi Yu,Nan Cao
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Text-to-image (T2I) models have achieved remarkable success at faithfully rendering specified objects and attributes, yet their ability to produce visual metaphors, images that convey abstract ideas by combining elements from two distinct domains, remains largely unexamined. To bridge this gap, we introduce VMetaphor-Bench, the first benchmark for evaluating visual metaphor generation in T2I models. It comprises 1,500 visual metaphors curated from real-world creative imagery, organized into three levels and ten categories, with each sample paired with two prompts of differing specificity. For evaluation, we develop a hybrid framework within an MLLM-as-judge paradigm, combining a multiple-choice question (MCQ) based protocol of 9,594 questions across four levels of metaphorical fidelity with a dimension-based scoring protocol along three perceptual dimensions. Extensive evaluation of 11 representative T2I models reveals that even the strongest proprietary models struggle with compositional structuring and cross-domain mapping, key aspects of metaphorical expression, highlighting visual metaphor generation as an important frontier for future T2I research.
[CV-35] UnCapsTSR: An Unsupervised Transformer-based Image Super-Resolution Approach for Capsule Endoscopy Images
链接: https://arxiv.org/abs/2609.02476
作者: Anjali Sarvaiya,Shubh Kawa,Lalit Agrawal,Jagrit Joshi,Kishor Upla,Kiran Raja
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted manuscript of the article published in Neurocomputing, Volume 665, Article 132161, 2026
Abstract:Wireless Capsule Endoscopy (WCE) captures and streams video while passing through a patient’s Gastrointestinal (GI) tract and is used to examine its irregularities. Although advantageous over conventional endoscopy, WCE suffers from limitations related to capsule size and wireless transmission, resulting in images with coarser resolution. This work presents UnCapsTSR, an unsupervised transformer-based Generative Adversarial Network (GAN) framework for improving the spatial resolution of Low-Resolution (LR) WCE images. The proposed method accomplishes SR without explicit degradation estimation of real-world LR data and eliminates the need for true LR-HR pairs. UnCapsTSR employs a Bilateral Total Variation (BTV) loss to ensure spatial continuity in SR images. A newly curated dataset from the Kvasir Capsule dataset is also presented for training WCE SR models. Generalizability is validated on KID and GIANA datasets that are not used during training. A new non-reference metric, Endoscopy Quality Metric (EndoQM), is introduced for quantitative evaluation of domain-specific WCE data. Experiments demonstrate consistent improvement over state-of-the-art unsupervised SR approaches using NIQE, BRISQUE, PIQE, and EndoQM. Statistical evaluation shows 40 to 80 percent improvement in EndoQM from LR to SR across the evaluated datasets.
[CV-36] Learning to Track from Privileged Target Appearances
链接: https://arxiv.org/abs/2609.02471
作者: Xin Chen,Jiao Xu,Dong Wang,Huchuan Lu,Kede Ma
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 13 pages, 2 figures
Abstract:Target templates define what a visual tracker searches for, yet the templates available at inference trade off localization certainty with appearance freshness: the initial ground-truth template is exact but becomes stale, whereas recent templates better reflect the current appearance but are cropped from uncertain predictions. We quantify this bottleneck with a non-deployable oracle that supplies an exact current-frame target crop, improving AUC on LaSOT by 15.2 percentage points. This gap reveals a training-only opportunity: frame-level ground truths provide exact current- and future-frame target crops, although such crops are unavailable at deployment. We introduce Privileged Appearance Transfer for Tracking (PATT), a teacher-student training framework that transfers these privileged appearances to a deployable tracker through multi-level representation prediction. The privileged teacher observes exact target crops from past, current, and future frames, whereas the student receives only past-frame templates and learns to predict the teacher’s search representations. To avoid transferring unreliable teacher signals, PATT weights this transfer by the teacher’s relative localization advantage over the student and its absolute localization accuracy. After training, the teacher, latent predictor, reliability weights, and privileged crops are removed, leaving standard student-only inference. Across seven benchmarks at two model scales, PATT achieves consistent gains under both long- and short-term tracking protocols.
[CV-37] VIPS: Vehicle-Infrastructure Cooperative Planning Benchmark via Pseudo-Simulation ECCV2026
链接: https://arxiv.org/abs/2609.02462
作者: Hoonhee Cho,Jae-Young Kang,Giwon Lee,Hyemin Yang,Heejun Park,Kuk-Jin Yoon
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ECCV 2026 Spotlight
Abstract:End-to-end autonomous driving in urban environments requires robust decision-making under partial observability and complex multi-agent interactions. Severe occlusions and dense traffic at intersections limit the perception capability of single-agent systems, motivating recent efforts on Vehicle-to-Infrastructure (V2I) cooperation for perception and planning. However, existing evaluation protocols face a fundamental trade-off: open-loop evaluation fails to capture error accumulation and recovery from deviations, while closed-loop evaluation is costly, difficult to scale, and often relies on simulated environments that may suffer from domain gaps. To bridge this gap, we propose VIPS, a benchmark for cooperative autonomous driving in V2I settings based on pseudo-simulation. VIPS extends pseudo-simulation by integrating vehicle and infrastructure observations. This enables scalable yet realistic evaluation of robustness and error propagation without full simulation. We further present CoS-V2X, a cooperative planning framework based on sparse representations. CoS-V2X models vehicle-infrastructure interactions using compact features for efficient communication and robust decision-making under heterogeneous observations. Code and dataset are available at this https URL.
[CV-38] WiFlow: Estimating Optical Flow using WiFi Channel State Information
链接: https://arxiv.org/abs/2609.02452
作者: Thomas Weigel,Simon Kiefhaber,Fabian Portner,Matthias Hollick,Simone Schaub-Meyer
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Knowing where and how fast objects are moving within a scene is important across various domains. Usually, cameras are used to capture the data necessary for this task, but adding cameras often raises privacy concerns, and the quality of captured frames is heavily influenced by lighting conditions. In this work, we explore using WiFi channel state information (CSI) instead of camera frames for optical flow estimation. We propose WiFlow, a CSI based flow estimator, a preprocessor evaluation for CSI, and three model architectures that offer different trade-offs between accuracy and complexity. Further, we create the first dataset for training and evaluating CSI-based optical flow estimators, and our experiments provide insights into key design elements for this task. Code and data are available at this https URL.
[CV-39] Adapting a Foundation Model for Lunar Surface Height Estimation
链接: https://arxiv.org/abs/2609.02448
作者: Patrick Bauer,Marius Schwinning,Melanie Siegel,Andreas Weinmann,Hichem Snoussi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Digital elevation models (DEMs) can provide accurate height information, making it invaluable for analyzing the lunar surface. As the European Space Agency (ESA) prepares for future lunar missions that aim to land on the Moon, a precise method for height estimation will be essential for hazardous terrain that could endanger the landing approach. Traditional approaches to generate DEMs from imagery, such as shape from shading (SfS) and stereophotogrammetry (SPG) have been proven highly valuable for this task. However, due to advancements in machine learning, especially computer vision, the focus has shifted towards monocular depth estimation via deep learning. The lunar surface is covered by rocks and craters, and classic hazard detection methods rely solely on 2D image data. Our goal is to address this issue by developing a relative lunar surface height estimator that can provide additional information for hazard localization. In this letter, we present a methodology that builds on the well-known zero-shot relative depth estimation model Depth Anything V2 (DAV2). Other works have been using it as a state-of-the-art comparison for their proposed lunar DEM estimation method, but without adaptations to the target domain. Thus, it may underperform. Therefore, we propose a fine-tuning strategy with publicly available SPG-derived DEM data of the lunar surface. Our results demonstrate a significant improvement in performance compared to the zero-shot model, effectively transforming DAV2 into a reliable relative depth estimator of the lunar surface.
[CV-40] Uncertainty-Guided Adverse Weather Restoration via Gated Transformer Network
链接: https://arxiv.org/abs/2609.02434
作者: Zheke Jin,Yuning Cui,Tianle Jin,Alois Knoll,Hu Cao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Restoring images degraded by adverse weather remains challenging due to spatially heterogeneous degradations. Many existing weather-specific restoration models rely on weather-agnostic global aggregation, naive cross-scale fusion, and deterministic objectives, which struggle to handle heterogeneous degradations in all-in-one adverse-weather settings. To address these limitations, we propose an Uncertainty-guided Adverse-weather Restoration Network (UAR-Net), a weather-specific AiO framework that integrates a gated transformer with balanced multi-scale skip connections. Specifically, we employ Gated Dual-scale Transformer Blocks (GDTB) to jointly model selective global interactions and multi-scale local structures, a progressive Balanced Multi-scale Skip Connection (BMSC) for balanced multi-scale feature integration, and an Uncertainty-Aware Refinement Head (URH) that performs artifact removal, detail enhancement, and predictive uncertainty estimation. The model is supervised by a Brightness-Aware Energy Loss (BAE-Loss) to encourage accurate reconstruction with well-calibrated uncertainty. Extensive experiments demonstrate that our method achieves state-of-the-art performance across multiple adverse-weather benchmarks. The codes will open source upon acceptance.
[CV-41] he Diagnosis a Reporter Leaves Unspoken: Surfacing Frozen Tumor Features for Brain-Tumor MRI Reporting MICCAI2026
链接: https://arxiv.org/abs/2609.02411
作者: Khawaja Murad ul Hassan,Ruqiyya Adil,Adil Qayyum,Rida Hassan,Asad Mansoor Khan,Muhammad Usman Akram,Mehran Ebrahimi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 10 pages, 2 figures, 3 tables. Accepted at MLCN 2026, a workshop held in conjunction with MICCAI 2026; to appear in Springer LNCS
Abstract:A capable brain-MRI report generator can still be, in effect, diagnostically silent. When a multi-chain chain-of-thought (CoT) reporter built on a medical Mistral-7B backbone is evaluated on held-out cohorts, it names most meningiomas and almost all metastases “glioma” (diagnosis recall 0.44/0.07). Yet the answer is not absent from the model: a supervised linear probe applied to its frozen segmentation features recovers the three tumour cohorts at 0.82 macro-F _1 (5-fold cross-validation; chance \approx 0.33). We introduce NeuroFusion, an assistive reporter that surfaces this latent signal rather than overriding it: discriminative field-classifier heads over per-lesion features condition a fast, single-pass draft-then-review decoder on their committed outputs. Built on the identical Mistral backbone, this restores the diagnosis (meningioma 0.92, metastasis 0.75) and wins 8 of 9 prose-content comparisons across three held-out cohorts (RaTEScore, RadGraph-F _1 , GREEN; Holm-corrected paired BCa), with no significant loss on the ninth, at 5-6x lower latency ( \approx 80 vs. 457 s/case). A controlled negative result sharpens the mechanism: a learned diagnosis pin that overrides the decoder instead of merely informing it collapses out-of-distribution metastasis recall to 0.03. Grammar-constrained decoding keeps 92.3% of records schema-valid, making every sentence entailment-checkable (7.5% contradicted vs. 36.8% for the direct baseline). In a blinded nine-case pilot, two board-certified neurologists independently rated NeuroFusion highest in every tumour type, the only system with zero critical errors, and gave it the top-rated sign-off in eight of nine cases (six outright, two ties).
[CV-42] CA-OPD: Confidence-Aware On-Policy Distillation for Structured Visual Prediction
链接: https://arxiv.org/abs/2609.02401
作者: Menghao Li,Linjie Mu,Yin Wang,Haotian Hu,Yannian Gu,Lujiayi Xue,Fanyi Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Autoregressive vision language models unify heterogeneous perception tasks but are highly susceptible to compounding errors. On-policy distillation (OPD) bridges the training-inference mismatch by training students on their own rollouts. However, unreliable student predictions, especially early in training, can derail the trajectory and degrade the quality of teacher supervision. While recent interleaved distillation methods allow the teacher to verify and replace student tokens, they primarily rely on rigid ranking metrics rather than exact teacher confidence, and they overlook how intervention decisions can inform token-level supervision. To address this, we introduce Confidence-Aware On-Policy Distillation (CA-OPD), a framework that couples reliable rollout construction with adaptive supervision. CA-OPD utilizes teacher confidence to selectively correct unreliable student transitions, gradually transferring rollout control to the student via a strict-to-relaxed schedule. Crucially, CA-OPD aligns knowledge transfer with these intervention decisions: corrected positions receive direct cross-entropy supervision from the teacher’s prediction, while retained positions benefit from the teacher’s full predictive distribution. Evaluated in a multi-teacher setting for GUI grounding and optical character recognition, CA-OPD substantially improves the Qwen3.5-0.8B baseline across all six target benchmarks, including gains of 9.50 points on ScreenSpot-Pro and 6.72 points on OCRBench-v2 English. Controlled studies further show that the gains depend on intervention placement, progressive rollout control, and intervention-aligned supervision, rather than intervention frequency alone.
[CV-43] ProSR: Semantic-Prototype-Guided Discrete Modeling for Physically Consistent SAR Super-Resolution ECCV2026
链接: https://arxiv.org/abs/2609.02377
作者: Byoungwoo Kim,Munchurl Kim
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to ECCV 2026
Abstract:High-resolution Synthetic Aperture Radar (SAR) imagery is critical for precision analysis such as automatic target recognition, yet its acquisition is costly. Although generative image super-resolution (ISR) models offer a promising alternative, current smooth-approximation based diffusion frameworks often struggle to preserve the coherent scattering statistics, causing stochastic structural distortions that are less consistent with real SAR physics. To address this, we propose Semantic Prototype-Guided Super-Resolution (ProSR), reformulating SAR ISR as a semantically-guided discrete token prediction task within a quantized latent space. By mapping signal features to discrete scattering primitives, ProSR preserves the impulsive nature of SAR without over-smoothing. Furthermore, we integrate a Self-Supervised Learning backbone into SAR ISR to extract label-free semantic priors, overcoming label scarcity. Guided by these priors, we introduce Semantic-Aligned Detail Encoding to decouple high-frequency signals into discrete scattering primitives. In parallel, the Semantic Prototype Map Generator explicitly constructs semantic prototype maps, allowing Prototype-Map-Guided Attention to route the information flows within identical categories and mitigate inter-class interference. To validate our approach, we present a large-scale 0.25m resolution benchmark from the Umbra Open Dataset. Experimental results show ProSR achieves superior visual quality while preserving essential scattering characteristics required for practical SAR applications.
[CV-44] Information Density Imbalance in Visual Object Detection
链接: https://arxiv.org/abs/2609.02369
作者: Ziwei Zhao,Yanxi Lu,Yuwei Hu,Shiyang Su,Mingxuan Wang,Chenyue Zhou,Jiayi Chen,Hehan Li,Xiaoshuai Hao,Andi Zhang,Yanbiao Ma
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 12 pages, 6 figures
Abstract:In object detection, the number of instances is typically used to determine whether a dataset exhibits a long-tailed distribution, implicitly assuming that the model will perform poorly on categories with fewer instances. This assumption has led to extensive research on category bias in datasets with imbalanced instance numbers. However, even in datasets where instance numbers are relatively balanced, models still exhibit category bias, indicating that instance count alone cannot explain this phenomenon. In this work, we first introduce the concept and measurement of information density. We then observe a significant negative correlation between a category’s information density and its accuracy, and we investigate how the training process impacts this relationship. Empirical studies suggest that information density imbalance may be a potential source of category bias. To preliminarily validate the potential of information density, we made simple improvements to three advanced object detection loss functions using this concept. Experiments on the Pascal VOC, COCO-LT, and LVIS datasets demonstrate that information density can significantly reduce model bias while effectively enhancing the overall performance of existing loss functions. This study provides a new perspective for understanding the generalized bias phenomenon in object detection models and offers new tools for designing fairer loss functions and training strategies.
[CV-45] he Missing Temporal Link: Temporal Context Routing for Script-Driven Audio-Video Generation
链接: https://arxiv.org/abs/2609.02367
作者: Yichen Liu,Quanwei Zhang,Haozhe Wang,Donghao Zhou,Xiaojie Li,Yang Shi,Jiaming Liu,Ruihua Huang,Yingtian Zou,Daquan Zhou
类目: Multimedia (cs.MM); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Joint audio-video generation models have made substantial progress in visual quality and audio-visual synchronization. However, they still provide limited control over when shot transitions occur and dialogue is spoken. This limitation constrains their application in script-driven content creation, where timing errors can undermine narrative coherence and the viewing experience. Current joint generators align video and audio representations on a shared temporal axis, yet the precise timing of shots and dialogue specified in a structured prompt is encoded only in the prompt’s text representation and remains unaligned with the temporal coordinates of either modality. Consequently, video and audio may remain synchronized with each other while both fail to follow the script timeline. This mismatch motivates us to extend temporal alignment beyond video and audio to include the structured script. We therefore introduce Temporal Context Routing (TCR), which maps the script timing onto the shared temporal axis of video and audio generation and routes each prompt’s guidance to the corresponding positions in both modalities. Compared with the baseline on 200 test scripts, TCR reduces Shot Boundary MAE by 96%, from 1.11 s to 0.042 s, and raises Dialogue Acc@0.5 s from 28.3% to 84.1%. TCR achieves these improvements while maintaining visual quality and audio-visual synchronization comparable to those of the baselines. A user study further shows that participants prefer TCR on all five evaluated dimensions.
[CV-46] mpoGround: State-Aware Streaming Visual Grounding with Vision-Language Models
链接: https://arxiv.org/abs/2609.02359
作者: Leqian Ding,Junning Qiu,Manwen Yang,Yu Guo,Fei Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Visual grounding maps language referents to spatial targets and is central to open-vocabulary perception with vision-language models. Existing methods have made substantial progress on single-frame and video-based visual grounding, yet under streaming inputs they still suffer from identity drift, cross-frame inconsistency, and fragile localization under partial occlusion. To address these issues, we present TempoGround, a VLM-native framework that detects cross-frame object correspondence and explicitly models object presence states, thereby enabling accurate and consistent visual grounding under streaming inputs. The key is a curriculum prediction mechanism guided by state-aware cross-frame correspondence: TempoGround resolves 2D instance association, predicts whether each object newly enters, continues in, or leaves the view, decodes the 2D box, and then lifts it to a camera-frame 3D box. As token-level supervision alone cannot capture the geometric objectives of streaming grounding, we further introduce Streaming Grounding Reinforcement (SGR), which optimizes TempoGround with verifiable Grounding, Identity, and Consistency rewards, jointly reinforcing persistent localization and temporally consistent predictions. We carefully design a three-stage training strategy and train TempoGround on large-scale data. We evaluate visual grounding under causally streaming inputs on multiple challenging benchmarks: TempoGround improves F1_2D@0.5 and F1_2D@0.95 by 4.4 and 0.5 on average, and F1_3D@0.25 and AP_3D by 6.2 and 7.5, respectively. These results demonstrate that TempoGround provides a practical foundation for visual grounding under streaming inputs.
[CV-47] LookStep: Efficient Vision-Language Navigation with Linguistic Foresight and Event Driven Memory EMNLP2026
链接: https://arxiv.org/abs/2609.02350
作者: Kun-Yang Yu,Yingzhe Li,Hongyu Xu,Shi-Yu Tian,Zhi Zhou,Yang Chen,Ming Yang,Sheng Wang,Qing Yu,Lan-Zhe Guo,Yu-Feng Li
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注: 19 Pages, 7 Figures. Accepted in EMNLP 2026 Main
Abstract:Vision-Language Navigation (VLN) requires an embodied agent to follow natural-language instructions in unseen environments. Recent progress has been largely driven by Multimodal Large Language Models (MLLMs). Existing methods follow a next-step action prediction paradigm, supervising only the expert action, which requires a high quantity of data for training. They also rely on cognitive maps, accumulated historical frames, or external 3D tools to maintain states, leading to high computational and memory overhead. To realize resource efficiency VLN, we propose LookStep, a unified end-to-end framework that combines Language Centric Future State Modeling and Event Driven Rolling Memory that uses language labels to generate coarse-grained navigation progress and future states for each candidate action, while autonomously deciding whether to write each observation into a bounded rolling memory with a semantic role. We validate LookStep empirically. On VLN-CE tasks, LookStep outperforms existing methods under the same training settings, achieving a 49.7% success rate on R2R-CE Val-Unseen with better memory efficiency and less data usage. Code and model is available at this https URL.
[CV-48] GlyphAnchor: Enhancing Visual Text Rendering via Position-Anchored Glyph Priors
链接: https://arxiv.org/abs/2609.02349
作者: Qiang Xiang,Shuang Sun,Binglei Li,Yibo Chen,Xu Tang,Yao Hu,Junping Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Rendering accurate text remains difficult for image generation and editing models, especially when the target contains long, complex, and densely arranged text or rare characters. Existing approaches either improve native text rendering through stronger backbones and data-centric training without explicit glyph priors, or incorporate glyph priors through specialized designs that remain insufficiently accurate and robust under challenging scenarios. We introduce GlyphAnchor, a novel text-rendering enhancement method for both text-to-image and image-editing diffusion transformer models. GlyphAnchor enhances the backbone with lightweight glyph patch conditions whose positions are anchored to the target image through the model’s native positional encoding. We train this capability with staged supervised finetuning and further refine it with text-aware post-training to improve robustness. We also introduce InfoTextBench, a benchmark for evaluating text-rich visual text rendering in both generation and editing settings. Experiments across multiple backbones and benchmarks, including long, complex, and densely arranged text and rare character scenarios, show that GlyphAnchor consistently improves text fidelity while preserving overall image quality.
[CV-49] Structured-Prior-Guided Diffusion Inpainting with Physical Consistency for Traffic Sign Augmentation
链接: https://arxiv.org/abs/2609.02348
作者: Luo Li,Chongchong Huang,Jun Jia,Qiang Gao,Xinlong Liu,Gui Yang,Liang Cao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Traffic sign detection faces a long-tailed data distribution. Many rare signs matter as much as common ones from a regulatory standpoint, yet they have very few samples. Generative data augmentation is one way out. General-purpose inpainting models, however, distort digits, deform geometry and perspective, and shift colours when applied directly to sign regions. We trace this to a single gap: the conditioning signal is too abstract for the physical composition of a sign. We propose a structured-prior-guided diffusion inpainting framework with physical consistency. It injects the semantic, appearance and geometric priors of a sign through three orthogonal pathways: a JSON-formatted text prompt, a front-view vector template rendered with measured dominant colours (via IP-Adapter), and an affine-aligned vector template (via ControlNet). Two physical consistency losses constrain colour with a CIELAB chromaticity L_1 term and edge structure with a Sobel gradient term. We train by self-supervised reconstruction on a large set of images collected in-house at AMAP, then evaluate zero-shot on the public TT100K-2021 dataset, a different source. Our method uses a Stable Diffusion 1.5 backbone of about 1.4B parameters. It beats seven representative competitors on every metric of reconstruction fidelity, physical consistency and semantic controllability. Its OCR exact-match rate reaches 91.1%, against 44.2% for the 12B industrial model FLUX.1 Fill [dev], and it needs only 1/14 of that model’s inference time. Leave-one-out ablations confirm that each of the three prior pathways and both loss terms contribute on their own. In downstream detection, the synthetic data raises the group-pooled AP50 of rare classes by 1.23\times to 7.40\times over a real-data-only baseline. Code and pre-trained models are available at this https URL.
[CV-50] owards Zero-Shot Transfer Across Embodiments For Driving VLAs
链接: https://arxiv.org/abs/2609.02341
作者: Caio Azevedo,Stefano Sabatini,Sascha Hornauer,Fabien Moutarde
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Vision-Language-Action models (VLAs) have shown strong potential in autonomous driving by leveraging multimodal pretraining for instruction following, visual reasoning, and scene-level generalization. In robotic manipulation, scaling VLA fine-tuning across multiple robot setups–especially when unifying representations across embodiments–has been shown to improve in-dataset performance and cross-embodiment generalization; in autonomous driving, however, VLAs remain largely trained on individual datasets and are rarely evaluated for zero-shot transfer to unseen datasets and camera rigs; furthermore naively adding more datasets to the training data does not necessarily lead to better performance within seen embodiments. To address these problems, we study multi-dataset training for the driving task and BEV-Forcing, an auxiliary objective that transfers ground-plane object-layout information from a specialized Bird’s-Eye-View model into the VLA backbone. By encouraging the model to represent object position through a shared BEV spatial interface, we show that an auxiliary task such as BEV-Forcing can improve both in-distribution and out-of-distribution performance when training on a small number of camera rigs. As the number of training embodiments increases, however, the benefits of the auxiliary task are reduced; we present this as evidence that new techniques in the literature may see their benefits diminish when simply scaling up training diversity, which motivates presenting results taking into account data scaling.
[CV-51] ORB-SVM : An Innovative Hybrid Framework for Efficient Brain Tumor Detection from MRI Scans
链接: https://arxiv.org/abs/2609.02333
作者: Amirhosein Azarpour
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 6 pages , 7 figures
Abstract:Brain cancer remains one of the most significant challenges in modern medicine, where the accuracy of early stage diagnosis is a decisive factor in patient survival and treatment efficacy. Although Magnetic Resonance Imaging (MRI) is the established gold standard for visualizing neurological structures, the interpretation of these high dimensional scans is often complicated by subjective variability among practitioners and the inherent noise present in complex medical images. While contemporary approaches frequently rely on high parameter deep learning architectures, such models often involve significant computational costs and require extensive data for effective training. This study introduces a hybrid framework that utilizes the Oriented FAST and Rotated BRIEF (ORB) algorithm for precise feature extraction and a Support Vector Machine (SVM) for classification [1], [2]. The proposed approach achieves a sub- stantial data reduction of approximately 99.5%, which effectively minimizes the influence of non informative background data while preserving critical diagnostic patterns essential for tumor identification. By balancing feature sparsity with a robust kernel based classifier, this methodology addresses the limitations of over parameterized systems while maintaining high diagnostic integrity. Experimental evaluations conducted on the Br35H dataset demonstrate that the framework attains a classification accuracy of 97.5%. The findings suggest that the integration of localized feature representation and optimized classification provides a reliable and resource efficient alternative for medical image analysis, offering a structured solution that maintains per- formance without the need for extensive computational overhead.
[CV-52] YesTrack: Referring Multi-Object Tracking via MLLM -based Yes/No Verification ECCV2026
链接: https://arxiv.org/abs/2609.02318
作者: Quansheng Hu,Qin Sun,Qiansen Dai,Jin Ding,Wan Zhang,Xue Zhou,Jianxiao Zou
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to ECCV 2026
Abstract:Referring multi-object tracking (RMOT) aims to track every instance in a video that matches a given language expression. Despite the recent integration of multimodal large language models (MLLMs) to enhance generalization, existing methods predominantly relegate them to the role of caption generators, necessitating external modules for final decision-making. This paradigm not only introduces extra latency but also severely underutilizes the inherent vision-language alignment capabilities of MLLMs. To address these limitations, we propose YesTrack, a novel two-stage RMOT method that reformulates referring as a discriminative task, directly leveraging MLLMs for Yes/No verification without explicit text generation. To further enhance the reliability and efficiency of this MLLM-based verification, we introduce two lightweight temporal consistency constraints: Temporal Confidence Prior (TCP) and Temporal Reference Propagation (TRP). We further validate the generality of this discriminative paradigm by proposing YesTrack-MOT, a straightforward yet highly effective instantiation for generic multi-object tracking (MOT). Experiments on Refer-KITTI and Refer-KITTI-V2 show that YesTrack significantly outperforms existing state-of-the-art methods while maintaining high efficiency, even when implemented with the smallest variant of Qwen3-VL. Code is released at this https URL.
[CV-53] Domain shift-robust object detection with GenAI image editing
链接: https://arxiv.org/abs/2609.02299
作者: Isabel D. Stein,Thijs A. Eker,Sebastiaan P. Snel,Ella P. Fokkinga,Klamer Schutte,Luca Ambrogioni,Friso G. Heslinga
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Submitted to SPIE Sensors + Imaging 2026
Abstract:Object detectors often degrade under domain shifts such as changes in lighting, weather, or occlusion. These shifts alter object appearance and expose a reliance on visual shortcuts learned from the training distribution that do not generalize across domains. Acquiring sufficient real-world samples to capture such domain variation is particularly difficult in specialized, low-data settings. Recent advances in diffusion-based generative image editing have shown promise for improving the in-domain performance of object detectors through synthetic data augmentation. However, their potential to improve out-of-domain robustness remains largely unexplored. We hypothesize that generative image editing can simulate a controlled domain shift in training data, effectively bridging the gap between source and target domains. To test this, we studied camouflaged military vehicle detection as a challenging domain shift scenario. Detectors trained on uncamouflaged data demonstrate substantial degradation on real test imagery containing foliage, netting, and multi-spectral camouflage across 15 vehicle classes in close-up, ground-level imagery. We used two diffusion-based editing models, Qwen Image Edit 2509 and Flux.2 Dev, to synthetically add camouflage to the training data, alongside a LoRA fine-tuned version of Qwen. A non-generative black-bar occlusion baseline served as a lower bound on augmentation quality. Using a GroundingDINO detector trained on real and synthetic data, generative camouflage augmentation yielded substantial mAP improvements for foliage (+20.1) and netting (+14.4) camouflage. Generating multi-spectral camouflage proved more challenging, but LoRA fine-tuning improved performance by 4.4 mAP over the uncamouflaged baseline.
[CV-54] VoRTeC: Taming Foundation Flow for One-step Real time Video Compression
链接: https://arxiv.org/abs/2609.02291
作者: Yichong Xia,Qinhong Wu,Qinhong Wu,Jinpeng Wang,Zeyuan Chen,Haoqian Wang
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Ultra-low bitrate video compression still faces critical challenges: traditional neural video compression inevitably introduces blurring artifacts, while diffusion-based generative video compression suffers from excessive decoding latency and poor temporal consistency. To address these issues, we propose \mathttVoRTeC , a Video Compression framework built upon a foundational flow model (Wan2.1). By compactly encoding latent video representations, predicting the positions of compressed representations along flow trajectories, and integrating multi-scale priors, \mathttVoRTeC enables the compressor to harness generative video flow priors effectively. Without accessing the parameters or gradients of flow matching networks, our framework achieves one-step decoding and reconstructions with high perceptual fidelity. Meanwhile, we maintain consistency across frame groups via tail-frame reuse and prior caching. Extensive experiments demonstrate that our method reduces bit consumption by 58% compared to prior diffusion-based approaches, with decoding speed boosted by 3 to 197 times: \mathttVoRTeC achieves a decoding speed of 13 FPS at 720p and 32 FPS at 480p.
[CV-55] If It Moves Radar Knows: A Physics-Aware Radar Transformer for Class-Agnostic Moving-Object Detection ICRA
链接: https://arxiv.org/abs/2609.02289
作者: Yinghao Sun,Shuguang Li,Jinliang Shao,Tieshan Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 8 pages, 5 figures, 4 tables, submitted to 2027 ICRA
Abstract:Detectors trained on closed-set annotations can miss rare moving objects outside the training taxonomy. Automotive radar provides category-independent Doppler motion cues and is less affected by adverse illumination and weather, but sparse, noisy returns hinder class-aware 3D box detection. Surface location and velocity remain useful for motion reasoning and collision avoidance when full box geometry is difficult to recover. We present the Physics-Aware Radar Transformer (PART), a fully sparse radar-only detector that predicts existence confidence, a representative surface point, and 2D ground-plane velocity for each moving-object hypothesis. Doppler-Aware Query Initialization (DAQI) replaces scene-independent learned queries with input-dependent proposals by clustering radar returns in position and velocity, easing query-object assignment in sparse scenes. Physics-Guided Cross-Attention (PGCA) incorporates radial-Doppler consistency and radar cross section (RCS) into query-point association. Uncertainty-aware supervision randomly masks ground-truth objects and assigns soft existence targets to ambiguous radar-supported queries, reducing reliance on exhaustive annotations. With only 1.1 million parameters, PART achieves a class-agnostic average precision (CA-AP) of 0.8827, a mean average surface translation error (mASTE) of 0.3188 m, and a mean average velocity error (mAVE) of 0.8084 m/s on nuScenes. It attains 0.9203 recall on rare and safety-relevant categories excluded from the standard evaluation and remains effective at night, in rain, and under severe occlusion. Inspection of apparent false positives shows that some predictions correspond to moving objects absent from the nuScenes annotations. Code and pretrained model weights will be publicly available at this https URL.
[CV-56] Diffusion-Encoding Gaussian Field for Joint k-q dMRI Reconstruction ALT
链接: https://arxiv.org/abs/2609.02288
作者: Zhibo Chen,Yajuan Huang,Yu Guan,Qiuyun Fan,Dong Liang,Qiegen Liu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 11 pages, 8 figures. Preprint submitted to IEEE Journal of Biomedical and Health Informatics
Abstract:Diffusion MRI requires repeated k-space acquisitions over multiple diffusion-encoding directions, making acquisition time dependent on both spatial and angular sampling. Existing joint k-q methods either associate directional parameters with fixed voxels or separate spatial reconstruction from angular completion. However, diffusion-weighted images acquired under different directions share the same anatomical organization, while their local signal intensities vary with diffusion encoding. Existing formulations do not fully exploit the complementarity between shared anatomy and direction-dependent signal variation. Consequently, residual spatial errors may be misinterpreted as genuine angular variation and propagated to unobserved directions. We propose a subject-specific spatial-angular Gaussian field for self-supervised joint k-q dMRI reconstruction. Shared 3D Gaussian primitives provide local spatial support, with each primitive carrying a continuous q-conditioned tensor-residual response. The signal at each location is synthesized from multiple overlapping primitive responses, coupling neighboring spatial regions and diffusion directions. The field is progressively optimized from undersampled k-space measurements of observed directions, without fully sampled targets or held-out-direction supervision. Experiments on three HCP diffusion shells under multiple acceleration settings demonstrated consistent improvements in missing-direction DWI reconstruction, tensor-derived metrics, and principal diffusion orientation estimation. Comments: 11 pages, 8 figures. Preprint submitted to IEEE Journal of Biomedical and Health Informatics Subjects: Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2609.02288 [cs.CV] (or arXiv:2609.02288v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2609.02288 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-57] RouteGraph-Mona: Confusion-Aware Routing Fine-Tuning for Mineral Image Classification
链接: https://arxiv.org/abs/2609.02282
作者: Jierui Li,Zhiyuan Qi,Hao Zhu,Yufan Liu,Jixian Liu,Shaojie Jiang,Jianda Wang,Yaqi Liu,Xiaotong Li,Wei Wang
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Mineral image classification is important for geological exploration and resource development, but it remains challenging due to substantial intra-class variations in appearance and high inter-class visual similarity. Multi-cognitive Visual Adapter (Mona) is a vision-oriented parameter-efficient adapter that adapts pre-trained visual models by tuning only a few parameters. However, Mona statically aggregates responses from multiple scales, limiting its ability to accommodate sample-specific scale preferences and model confusion among visually similar mineral categories. To address this issue, we propose \textbfRouteGraph-Mona, a lightweight route-space regularization method built on Mona. Specifically, we replace Mona’s static multi-scale aggregation with sample-adaptive routing. The resulting branch-selection behavior defines a compact routing space that captures each image’s scale preferences. We then regularize the resulting routing signatures with class-wise route anchors and confusion-weighted margins. The route anchors encourage class-consistent routing patterns, while the margins promote greater separation between visually similar categories in the routing space. Experiments on three public mineral image datasets with two visual backbones show that RouteGraph-Mona consistently outperforms Mona in mean accuracy and remains competitive with representative fine-tuning methods and mineral image classification baselines.
[CV-58] Retrosynthesis of Synthetic Media for Explainable AI Provenance Forensics
链接: https://arxiv.org/abs/2609.02268
作者: Yijie Lin,Ching-Chun Chang,Isao Echizen,Hui Li,Chin-Chen Chang
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Multimedia (cs.MM)
备注: 12 pages, 10 figures. This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible
Abstract:With the rapid proliferation of generative models on Machine Learning as a Service (MLaaS) platforms, reliably tracing the provenance of synthetic media without modifying generator architectures or parameters remains a major challenge. In this work, we propose a self-referential retrosynthesis framework for explainable AI provenance forensics under a fixed-generator setting. The framework leverages a jointly optimized encoder-decoder pair to implement a self-embedding mechanism that enables round-trip consistency verification. During inference, client inputs are first encoded and then processed by the generator to produce outputs with high visual fidelity. For forensic verification, the consistency between the resynthesized image and the query image is analyzed to determine whether the image originates from the target generative model. Our approach eliminates the need for watermark embedding or modifications to the generation process. Experimental results show that images generated from encoded inputs maintain visual quality comparable to original generator outputs, while decoded images reliably trace back to their corresponding source inputs. Furthermore, the framework provides interpretable evidence for generative content provenance, establishing a practical tool for explainable generative AI forensics.
[CV-59] MAOL: Morphology-Aware Ordinal Learning for Fine-Grained Industrial Defect Severity Grading ICME2026
链接: https://arxiv.org/abs/2609.02266
作者: Zhaoyang Wang,Haiyong Chen,Binyi Su,Kun Liu,Kun Wang,Xianen Zhou,Atik Shahariar
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at IEEE ICME 2026
Abstract:Fine-grained defect severity grading is essential for industrial inspection, yet remains challenging due to the ordinal nature of severity labels, the strong dependence on morphology-related cues, and the train-test discrepancy between clean annotated instances and noisy predicted instances in two-stage pipelines. We propose MAOL, a Morphology-Aware Ordinal Learning framework for fine-grained industrial defect severity grading. MAOL formulates severity grading as an instance-level ordinal learning task, incorporates explicit morphological features to enhance representation learning, introduces class-conditional adaptive ordinal thresholds to model defect-specific grading boundaries, and employs prediction-aware training via localization perturbation to improve robustness to imperfect predicted instances. Extensive experiments under both clean-ROI and predicted-instance settings demonstrate that MAOL consistently outperforms rule-based methods, nominal classification models, and existing ordinal baselines, especially in the predicted-instance setting. The proposed approach ranked third in the IDA 2026 Challenge on Fine-Grained Severity Grading for High-Precision Manufacturing.
[CV-60] 2LSC-Bench: Benchmarking Localized Semantic Control in Text-to-Image Generation
链接: https://arxiv.org/abs/2609.02255
作者: Yan Wang,Xinyi Hou,Weiguo Lin,Junjun Si,Siwei Ma
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Recent text-to-image models have become increasingly capable of rendering explicit text, but reliable localized text control requires more than generating the correct string. In applications such as product labeling, signage, and interface design, target text should be rendered within a designated text-bearing region without altering the predefined subject identity or surrounding scene semantics. We refer to violations of this requirement as target-text-associated semantic leakage, in which target-text semantics are expressed through non-textual visual content beyond the designated anchor. Existing visual-text benchmarks primarily evaluate readability, spelling accuracy, and layout, leaving this form of semantic leakage largely unexamined. We introduce T2LSC-Bench, a controlled diagnostic benchmark comprising 50 seed subjects and 1,200 prompt cases per model, yielding 7,160 evaluated images across six models. Its factorized design varies semantic relation, scene openness, prompt mode, and language. A dual-branch protocol combines OCR-VLM text verification with structured VLM semantic judgments to measure Text-at-Anchor Accuracy (TAA), Semantic Subject Preservation (SSP), Semantic Leakage Rate (SLR), and Conditional Semantic Leakage Rate (cSLR). Under stress-test conditions, SLR increases from 1.2% to 18.1% and cSLR from 1.3% to 18.2%, whereas TAA decreases only from 91.4% to 90.9%. Anti-leakage prompting reduces SLR from 16.6% to 8.4% without degrading rendering accuracy. Human validation on 420 images shows strong agreement between automatic and adjudicated annotations. These results show that accurate text rendering does not guarantee local containment of target-text semantics.
[CV-61] Handwriting Trajectory Recovery via Autoregressive Ordered Stroke Instance Prediction
链接: https://arxiv.org/abs/2609.02251
作者: En-Guang Wang,Yan-Ming Zhang,Fei Yin,Cheng-Lin Liu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 24 pages, 6 figures, 7 tables
Abstract:Handwriting trajectory recovery aims to infer the dynamic writing process hidden behind a static handwritten image. Since offline handwriting preserves only the final spatial ink pattern, temporal information such as stroke order, writing direction, and pen-tip motion is lost, making recovery inherently ambiguous. Existing learning-based methods often directly predict the complete character trajectory without explicitly exploiting the stroke-level organization of handwriting. We argue that recovering the writing process should follow the writing process itself. Accordingly, we propose a two-stage framework that first recovers ordered stroke instances and then reconstructs continuous within-stroke motion. The first stage integrates stroke extraction and stroke-order recovery through autoregressive ordered stroke prediction, while direction-related structural cues further support within-stroke trajectory generation. Experiments on Chinese handwriting show that the proposed ordered prediction is more effective than post-hoc stroke ordering. Even without trajectory simplification, our full-point model achieves numerically better results than those reported by all compared baselines, while a controlled analysis shows that trajectory sampling density substantially affects measured recovery performance. Additional experiments demonstrate generalization to unseen Chinese character categories and cross-language extensibility to English and Tamil handwriting.
[CV-62] SAUF-Net: Structure–Appearance Representation Learning with Uncertainty Feedback for Semi-Supervised Medical Image Segmentation
链接: https://arxiv.org/abs/2609.02247
作者: Qin Lu,Zheyang Jing,Yujie Yang,Jianwang Li,Chen Yi,Shaofeng Jiang
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Semi-supervised learning has shown great potential for reducing annotation costs in medical image segmentation. However, most existing methods mainly exploit unlabeled data through prediction-level consistency, while the reliability of internal feature representations is often overlooked. In medical images, target-related structural cues are easily entangled with unstable appearance variations, which may lead to unreliable pseudo labels and error accumulation during training. To address these issues, we propose SAUF-Net, a Structure–Appearance Representation Learning with Uncertainty Feedback Network for semi-supervised medical image segmentation. SAUF-Net uses the Structure–Appearance Decomposition Module (SADM) to separate bottleneck features into structural and appearance representations. The Disentangled Guidance Module (DGM) injects these representations into the decoding process to enhance structure-aware segmentation. Meanwhile, the Auxiliary Decoder produces branch-specific predictions for reliability estimation and a fused prediction for appearance-swapped consistency. Furthermore, we introduce an Appearance-Swapped Consistency branch to encourage structural representations to remain stable under appearance variations. We also introduce a reliability-map-guided dual-head discriminator with a Validity Head and an Uncertainty Head to provide feature-level uncertainty feedback. Extensive experiments on ISIC-2016 and Kvasir-SEG demonstrate that SAUF-Net outperforms state-of-the-art semi-supervised methods, especially under low-label settings.
[CV-63] InfraPatch: Cross-Task Targeted Grayscale Patch Attacks on Infrared-Adapted Vision-Language Models
链接: https://arxiv.org/abs/2609.02233
作者: Chengyin Hu,Dingyi Lu,Jiaju Han,Xiang Chen,Weiwen Shi,Jiahuan Long,Yiwei Wei,Jiujiang Guo
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Infrared vision-language models (IR-VLMs) have emerged as a promising paradigm for multimodal perception under low-visibility conditions, yet their robustness to targeted adversarial attacks remains poorly understood. Existing adversarial patch methods mainly study RGB-based models or a single downstream task and do not characterize whether localized perturbations can induce an intended semantic target in IR-VLMs. We propose InfraPatch, a white-box, per-instance framework for targeted digital grayscale patch attacks against IR-VLMs. InfraPatch optimizes a compact single-channel patch within an approximately 5% local-area budget, combines proxy-guided placement with task-adaptive semantic objectives, and induces target behaviors in image classification, image captioning, and binary visual question answering. We evaluate ten infrared-adapted model variants on 300 synthetic infrared-style images generated by applying DiffV2IR to a fixed 30-category COCO subset, using clean-conditioned targeted success criteria. InfraPatch achieves targeted attack success rates from 86.00% to 100% across the ten variants. On CLIP and BLIP-2, proxy location search improves success by 6.67 and 10.33 percentage points over optimized random placement, respectively; LLaVA-1.5 remains saturated near 100% under both settings. Patch-area and objective ablations further expose substantial differences in vulnerability across architectures and task formats. These results show that small grayscale patches can inject chosen target semantics across IR-VLM families under a controlled digital threat model, motivating stronger robustness evaluation for infrared multimodal systems.
[CV-64] Signal or Noise? Auditing Rotation-Induced Saliency Drift in Medical and Aerial Imaging
链接: https://arxiv.org/abs/2609.02224
作者: Khawaja Murad ul Hassan,Mehran Ebrahimi
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 11 pages, 3 figures, 6 tables. Code at this https URL
Abstract:Post-hoc saliency maps such as Grad-CAM are increasingly used to audit why a deployed vision model made a decision, yet the heatmap drifts when the input is rotated, even when the prediction is unchanged. In domains with no canonical orientation, such as histopathology and aerial imagery, this undermines using saliency as evidence. We ask whether that drift is faithful signal or noise introduced by the CAM operator, and answer it by measuring equivariance at every stage of the operator rather than inferring it from the network’s output. The instability is not where one would guess: the channel weights are the most rotation-stable stage, and on ResNet-50 exactly stable, because a GAP+linear head makes the class gradient field spatially constant. What moves is the spatial activation tensor, and the classifier’s own pooling discards that movement. A causal test confirms the consequence: occluding the pixels whose saliency drifts costs the model less than occluding random pixels, at either orientation. The drift is carried by degrees of freedom the classifier throws away, which is what makes removing it faithful rather than destructive. EquiGrad-CAM is a training-free wrapper that takes T rotated views, inverse-rotates each view’s saliency into a common canonical frame, and averages. On the full ImageNet-1K validation set it raises equivariance over single-view Grad-CAM by +36.0% (ResNet-50), +87.5% (VGG-16) and +247% (ViT-B/16); a scale-matched ablation isolates alignment before averaging, not the locus of aggregation, as the driver. It beats rotation-augmented training without retraining, lifts zero-shot CLIP by +145%, and yields rotation-consistent explanations on PatchCamelyon and RESISC45. Its by-product PEUM ranks explanations by how reproducible they are, at no cost beyond the views already taken. Code: this https URL
[CV-65] Hardware-Accelerated Instance Segmentation for Resource-Constrained Space Robotics with Criticality Analysis
链接: https://arxiv.org/abs/2609.02219
作者: Siddhant Shete,Hilmi Dogu Kücüker,Udo Frese,Frank Kirchner
类目: Robotics (cs.RO); Hardware Architecture (cs.AR); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Autonomous lunar missions require real-time per- ception under three coupled constraints: extreme low-light conditions, limited onboard compute, and radiation-induced hardware faults that can silently corrupt inference. We present a deployment-oriented instance segmentation framework for resource-constrained lunar robotics that jointly addresses quan- tization calibration and system-level fault exposure under strict compute constraints. First, we introduce Activation Variance Informative Sampling (AVIS), a label-free calibration strategy that deterministically selects calibration samples based on activation variance statistics. Second, we deploy a YOLO-based segmentation model on a Deep Learning Processor Unit (DPU) with architectural modifications that reduce CPU fallback paths and enable statically compiled execution with bounded latency in low-lighting conditions. We further introduce a software-level criticality analysis to estimate fault exposure and guide mitigation under radiation-constrained operation. On a lunar micro-rover platform, AVIS with bias correction recovers 69.8% of quantization-induced accuracy loss while achieving 309 ms inference latency and 5.7 W power consumption. Targeted mitigation reduces global criticality by 31.7%. The results demonstrate an integrated approach and a blueprint for a reliable and safe AI perception framework under space deployment constraints.
[CV-66] FuDU: A Fuzzy Dual-dimensional Uncertainty Framework for Streaming Active Learning in Industrial Defect Detection ECCV2026
链接: https://arxiv.org/abs/2609.02212
作者: Zhaoyang Wang,Haiyong Chen,Binyi Su,Xinwei Lyu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at ECCV 2026
Abstract:Ensuring the reliability of deep learning models in real-time industrial defect detection is critical for high-stakes quality inspection. To mine uncertain samples within continuous industrial media streams, thereby enhancing the reliability of the detection system, this paper proposes a streaming active learning method based on the Fuzzy Dual-dimensional Uncertainty (FuDU) framework. Specifically, we first design a Prototype-based Global Uncertainty Quantification (PGUQ) module on the backbone to evaluate image-level uncertainty via normal/defective feature prototypes. A Dual-entropy defect Uncertainty Evaluator (DeUE) is then integrated into the detection head to quantify box-level uncertainty. Finally, by modeling uncertainty as systematic error, we propose a fuzzy dual-dimensional uncertainty-aware strategy that leverages fuzzy inference to fuse dual-dimensional uncertainties, enabling expert knowledge-driven adaptive sampling decisions. Comprehensive experiments demonstrate that FuDU is efficient and flexible, making it well-suited for challenging industrial inspection tasks such as the detection of nuclear fuel rod defects. Our code is publicly available at: this https URL.
[CV-67] Asymmetric Paired-Annotation Learning for Multi-Structure ULF Pediatric Brain MRI Segmentation MICCAI2026
链接: https://arxiv.org/abs/2609.02210
作者: Ha-Hieu Pham,Dang P.M. Cao,Minh Hoang Pham,Khanh Nguyen Vo Ngoc,Thanh-Huy Nguyen,Ulas Bagci,Huy-Hieu Pham
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at LISA Challenge, MICCAI 2026
Abstract:Portable ultra-low-field (ULF) MRI can expand access to pediatric neuroimaging, but segmentation at 0.064 T remains challenging because anatomical boundaries are weakly delineated, small structures may be only partially visible, and high-field references can be locally misregistered. The LISA 2026 Challenge provides two non-equivalent annotations reflecting different sources of anatomical evidence: a highfield-derived (HF) mask defining the scored target and a low-field-edited (LF) mask aligned with visible ULF anatomy. In this challenge report, we describe AURA, an nnU-Net-based asymmetric supervision strategy that treats these annotations as distinct observations rather than interchangeable ground truths. AURA anchors training to the HF mask and incorporates the LF mask through a bounded reliability gate based on label disagreement, boundaries, predictive uncertainty, class reliability, and training stage. On a 16-case development split, the HF-supervised baseline, AURA, and their ensemble achieved Dice scores of 0.7984, 0.7950, and 0.7988, respectively, while the ensemble achieved an HD95 of 1.8892 and an ASSD of 0.7855. These results provide a preliminary evaluation of AURA within the LISA 2026 Challenge and motivate further assessment on the hidden test set and external ULF cohorts. Our code and pretrained models are available at this https URL A-nnU-Net-based-asymmetric-supervision-strategy.
[CV-68] AME: Temporal-Aware Mixture-of-Experts for Text-Video Retrieval
链接: https://arxiv.org/abs/2609.02204
作者: Uicheol Jung,Juyoung Hong,Hojung Kwon,Yukyung Choi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 17 pages, 6 figures
Abstract:Text-Video Retrieval (TVR) retrieves videos that match a natural-language query, but extending image-text models such as CLIP to videos is fundamentally limited by the lack of temporal modeling. Videos exhibit frame-wise heterogeneity in appearance and motion, and compressing all frames into a single representation often obscures temporal structure and semantic transitions. To address this, we propose Temporal-Aware Mixture-of-Experts for Text-Video Retrieval (TAME), a CLIP-based framework that jointly models frame-level structure and temporal relations. First, we integrate sparse Mixture-of-Experts (MoE) layers into both CLIP encoders and apply frame-consistent routing on the vision branch so that experts specialize according to frame-level visual patterns while preserving the original vision-language alignment. Second, we introduce Frame-Temporal (FT) tokens that aggregate global cross-frame information and feed it back to each frame, enabling the visual encoder to capture long-range temporal dependencies without harming local details. Third, we design a Cross-Temporal Interaction and Aggregation (CTIA) module that refines frame-wise sentence-video similarities through staged temporal filtering and fusion. Experiments on standard TVR benchmarks show that TAME consistently improves over CLIP-based baselines. On MSR-VTT, it improves R@1 by 4.0 over CLIP4Clip, and also achieves consistent gains on DiDeMo, MSVD, LSMDC, and ActivityNet. The code is available at this https URL.
[CV-69] Lightweight Adaptation of General-Purpose VLMs for Multispectral and SAR Image Understanding
链接: https://arxiv.org/abs/2609.02187
作者: Shanji Liu,Kelu Yao,Junxiao Xue,Chenghui Lv,Xiangyang Miao,Yekai Huang,Yaying Chen,Chao Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 18 pages, 7 figures, 17 tables
Abstract:General-purpose vision-language models (VLMs) now support strong visual recognition, instruction following, and generation. However, most pretrained visual encoders are built around three-channel natural images and do not directly accommodate observations such as native multispectral measurements or synthetic aperture radar (SAR). Adapting VLMs to these sensors typically requires dedicated encoders and domain pretraining, slowing the reuse of stronger general-purpose checkpoints. We show that the multi-image interface of general-purpose VLMs offers a lightweight alternative. Our protocol renders each observation as five optical views and one SAR view, names them in the prompt, and adapts the language network and selected visual transformer blocks with LoRA. This exposes band composites, spectral indices, and radar backscatter through an existing visual interface. For land-cover recognition, structured supervision couples predicted classes with sensor evidence. We further construct preference pairs in which a true label is omitted while its supporting evidence is retained, encouraging complete predictions that remain consistent with the observations. On a balanced six-class land-cover benchmark derived from BigEarthNet-v2, the adapted Qwen3-VL reaches 0.8275 micro F1. The same input and adaptation protocol improves all four tested VLM architectures and transfers to Sen1Floods11 flood verification and this http URL captioning. Image removal and mismatch controls show that the adapted models use the supplied sensor observations. Together, these results demonstrate that VLMs can be repurposed for multispectral and SAR tasks through rendered inputs and compact LoRA adaptation, without training a new foundation model.
[CV-70] CC-4DGS: Computational Deformation and Point-Cloud Compression for Storag e-Efficient Dynamic Gaussian Splatting
链接: https://arxiv.org/abs/2609.02184
作者: Kyungdae Park,Chae Eun Rhee
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 16 pages, 8 figures, and 9 tables. Published in IEEE Transactions on Visualization and Computer Graphics. Code is available at this https URL
Abstract:Dynamic four-dimensional (4D) Gaussian Splatting has emerged as a powerful explicit representation for high-quality view synthesis, yet existing methods still require tens to hundreds of megabytes per scene due to their heavy reliance on large multi-resolution hash tables and high-dimensional Gaussian attributes. This paper presents CC-4DGS, a storage-efficient and scalable framework that rethinks both deformation modeling and canonical attribute storage. First, we introduce a computational deformation field (CDF) that replaces large multi-resolution learnable hash tables with deterministic dense hash encoding and compact neural decoders, enabling on-the-fly synthesis of deformation features while reducing deformation storage to only 1–3 MB per scene. Second, we propose a compression of canonical point-cloud attributes (CCA) pipeline that compresses high-dimensional spherical harmonic appearance terms and auxiliary Gaussian attributes via conditional autoencoding, selective quantization, and residual codebooks, achieving 3–5 \times point-cloud reduction with negligible quality loss. Together, these components yield a unified representation that preserves real-time rendering performance while reducing total storage to 20–30 MB. Extensive experiments across the N3DV and Technicolor Light Field datasets demonstrate that CC-4DGS achieves reconstruction accuracy comparable to state-of-the-art methods such as Swift4D, while offering significantly improved storage efficiency and favorable runtime-memory trade-offs.
[CV-71] Progressive Pseudo-Label Optimization for Point-Supervised Change Detection
链接: https://arxiv.org/abs/2609.02171
作者: Hailong Ning,Hao Wang,Yimeng Wang,Tao Lei,Renwei Dian,Asoke K. Nandi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 9 pages,4 figures
Abstract:Point-supervised change detection (PS-CD) aims to identify pixel-level changes between bi-temporal images using only sparsely annotated points. Although point annotations substantially reduce labeling costs, their limited spatial coverage often results in incomplete and noisy pseudo-labels. To address this issue, we propose a two-stage framework that introduces SAM2 priors into PS-CD and progressively adapts them to the target task. In Stage I, SAM2 generates object-aware candidate masks from point annotations on the bi-temporal images, and a bi-temporal mask selection strategy is designed to convert generic segmentation responses into more reliable change pseudo-labels. Subsequently, a lightweight CNN refinement module with an uncertainty-aware loss is employed to improve boundary quality and local structural consistency. In Stage II, we construct a teacher-student self-training framework in which the teacher is updated by exponential moving average and periodically refreshes the pseudo-labels. This design establishes a closed-loop optimization process that alternates between pseudo-label refinement and model re-optimization. Experiments on three benchmark datasets, including WHU-CD, LEVIR-CD, and SYSU-CD, demonstrate that the proposed method outperforms previous weakly supervised approaches on most benchmarks and remains competitive with several fully supervised methods.
[CV-72] World-Coherent Decoding: Self-Verifying Test-Time Planning for World Action Models
链接: https://arxiv.org/abs/2609.02159
作者: Chuhan Zhang,Seiji Ito,Kenta Hoshino,Satoshi Ikehata,Ikuro Sato
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:World Action Models (WAMs) aim to control robots by stochastically generating visual futures and then decoding actions, but empirical observations indicate that the results can strongly depend on which future is selected. We propose World-Coherent-Decoding (WCD), a self-verifying test-time planning framework that treats WAM rollouts as falsifiable future–action hypotheses. At each decision step, WCD samples multiple candidates from a frozen WAM and ranks them using internal generative signals: flow-based video surprisal for visual plausibility and action path effort for action-generation stability. After execution, the realized observation audits the selected imagination, yielding an imagination–reality mismatch that trains a lightweight online predictor for future candidate selection. Thus, WCD converts delayed self-verification into pre-execution reliability estimation without updating the backbone model. On RoboTwin 2.0, WCD improves Hard success under limited randomized-scene supervision from 55.80% to 60.90% , with a +16.43 gains on Horizon-3 tasks, and shows qualitative robustness on real Franka visual-shift tests. These results highlight a simple principle: test-time scaling for WAMs depends less on sampling more futures than on selecting reliable ones.
[CV-73] Synergistic Information Disentanglement for Omni-modal Slide Representation Learning in Computational Pathology MICCAI2026
链接: https://arxiv.org/abs/2609.02118
作者: Mingxin Liu,Chengfei Cai,Anwen Lu,Pengbo Xu,Jun Li,Jinze Li,Depin Chen,Jun Xu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 11 pages, 3 figures. Early accepted by MICCAI 2026 (Oral Presentation)
Abstract:In computational pathology (CPath), developing omni-modal self-supervised learning (SSL) models that integrate histology, genomics, and clinical reports enables transferable representation learning for whole slide images (WSIs). Existing approaches implicitly force heterogeneous modalities into a uniform latent space by contrastive alignment, causing modality collapse where unique, synergistic diagnostic signals (termed as \mathrm\Phi ) are discarded in favor of trivial redundancy. We hypothesize that the strongest task-agnostic SSL training signal stems from distilling the synergistic interactions over merely aligning shared redundancy. To this end, we introduce \textsc \mathrm\Phi -Omni, a synergistic information disentanglement framework grounded in Partial Information Decomposition (PID) theory for slide representation learning. Unlike standard contrastive approaches, \textsc \mathrm\Phi -Omni employs a Synergistic Information Bottleneck (SIB) regulated by the proposed \mathrm\Phi\textID objective, which explicitly suppresses marginal redundancy while maximizing irreducible synergy, thereby distilling high-order cross-modal interactions. Following pretraining on breast ( n =1031) and lung ( n =919) cohorts, \textsc \mathrm\Phi -Omni demonstrates superior few-shot performance across five independent external datasets spanning eight tasks compared to supervised and SSL baselines. Source code is available here.
[CV-74] Disease Burden over Skin Tone: Decomposing the Dermatology-AI Generalization Gap
链接: https://arxiv.org/abs/2609.02111
作者: Nirajan Kunwor,Sanjaya Poudel,Quoc-Huy Trinh,Jahidul Arafat,Sunil Kumar Gaire
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Dermatology artificial intelligence (AI) models are predominantly trained on light-skinned, cancer-focused image collections, yet they are increasingly proposed for deployment in resource-constrained settings where patients differ from training populations along two confounded axes: skin tone and disease distribution. We investigate whether poor generalization is primarily caused by skin-tone underrepresentation or disease-distribution shift. We evaluate a cancer-trained baseline (ResNet-50 fine-tuned on HAM10000 and ISIC 2019), two dermatology foundation models (DermLIP and MONET), and a general-purpose vision model (DINOv3) as frozen feature extractors. Models are evaluated on a tone-stratified disease-matched dataset (Diverse Dermatology Images, DDI) and a disease-shifted tone-diverse dataset (Skin Condition Image Network, SCIN). Our results show that disease-distribution shift contributes more than skin tone in the evaluated settings. The cancer baseline decreases from 0.62 to 0.21 balanced accuracy when transferred to unfamiliar clinical conditions, while the within-disease skin-tone gap is smaller (0.10-0.18) and inconsistent. Label-free representation analysis shows that this failure reflects a representational limitation rather than only missing output labels: cancer-specialized features poorly cluster unfamiliar conditions (kNN purity lift +0.06 over chance), whereas dermatology-pretrained features retain stronger transferable structure (+0.23). Finally, we show that representation quality predicts recoverable performance under lightweight adaptation. Starting from dermatology foundation models, approximately ten labeled examples per clinical category recover most attainable performance. We release the evaluation protocol and code to support reproducible auditing of dermatology AI generalization.
[CV-75] A Unified Rate-Distortion Perspective on Vector Product and Scalar Quantization
链接: https://arxiv.org/abs/2609.02107
作者: Xianghong Fang,Wenlong Mou,Yuan Yuan,Dehan Kong,Tim G. J. Rudner
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注: 26 pages, 2 figure, 8 tables
Abstract:Discrete visual tokenization, predominantly driven by vector, scalar, and product quantization, lacks a unified conceptual framework for understanding quantization tradeoffs. In this paper, we propose a unified rate–distortion perspective on modern discrete visual tokenization. By viewing quantization as lossy compression, we characterize the nominal fixed-length coding rate through token count and codebook size, and quantization error as the distortion. Within this framework, we resolve three central questions. First, we theoretically and empirically show that minimizing distortion, rather than maximizing codebook utilization, is the primary intrinsic objective for reconstruction fidelity, with a direct connection to the STE-induced gradient discrepancy. Second, we establish two critical fairness conditions for intrinsic quantization comparison: controlling latent feature statistics and enforcing identical coding rates. Third, under these conditions, we recover the VQ–PQ–SQ distortion hierarchy in modern visual tokenization and show empirically that modern VQ methods achieve the lowest distortion. This work provides a foundational rate–distortion reframing of modern discrete visual tokenization, resolves ambiguities in quantizer evaluation, and provides a controlled framework for isolating intrinsic quantization effectiveness under fixed-rate constraints.
[CV-76] Federated LoRA Adaptation of BiomedCLIP Across Four International Chest X-Ray Cohorts
链接: https://arxiv.org/abs/2609.02101
作者: Sanjaya Poudel,Nirajan Kunwor,Manish Dhakal,Debesh Jha,Sunil Kumar Gaire
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Federated learning (FL) lets institutions train a shared model without exchanging data, and Low-Rank Adaptation (LoRA) makes this practical at scale by communicating only compact low-rank updates. Biomedical imaging is a compelling setting for this combination: patient data are archived behind privacy regulations, and institutions differ widely in scanners, protocols, and compute. Such heterogeneity raises the question of how federated LoRA updates should be aggregated, increasingly pressing as multimodal vision-language models become central to medical image analysis. We benchmark federated Parameter-efficient fine-tuning (PEFT) of BiomedCLIP for chest radiograph classification across four public cohorts on three continents (USA, Vietnam, Spain). Federated LoRA adaptation improves shared-class AUC on all four cohorts over the unadapted BiomedCLIP backbone (mean 0.687 to 0.802), showing that the gains come from federated adaptation rather than from the pretrained model’s zero-shot ability. Relative to isolated single-cohort training, federation improves the weaker cohorts while largely preserving the strongest and approaches a centralized reference (0.812) that pools all data. The singular value decomposition (SVD)-based product-space aggregation introduced by FlexLoRA is essential to this gain (naive factor averaging drops mean AUC by 0.097), whereas a drift-correcting optimizer (FedProx) shows no benefit over FedAvg in our single-seed runs, consistent with LoRA’s low-rank updates already limiting client drift. Biomedical vision-language models can thus be adapted collaboratively across heterogeneous, geographically distributed institutions without centralizing data.
[CV-77] Evidence-Guided Detection Localization and Explanation for Text-Centric Image Forensics
链接: https://arxiv.org/abs/2609.02097
作者: Peifeng Liu,Bin Li,Qingsong Zhang,Yangxin Yu,Leqing Chen,Xiaoye Qiu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:The rapid progress of AIGC has made text-centric image manipulation increasingly accessible, creating new forensic challenges that require not only authenticity detection but also spatial grounding and evidence-based explanation. This paper presents our solution to the GenText-Forensics Challenge at ACM Multimedia 2026. We propose an evidence-guided detector-localizer-reasoner system, where an image-level detector provides a global authenticity prior, a dedicated localizer extracts tampered regions as spatial grounding evidence, and an MLLM-based reasoner generates structured forensic reports grounded in this expert forensic evidence. These modules are connected through a cascaded evidence flow: the detector gates the subsequent localization and prompting process, the localizer converts tamper responses into grounding boxes, and the reasoner is trained to synthesize the detector decision and localized evidence into the final report. As a key part of our method, we introduce iterative difficulty-aware mining to improve localization quality and apply report-mask consistency post-processing to align report grounding with predicted masks. On the official hidden test set, our system achieves a final score of 0.638 and ranks second in the challenge, validating the effectiveness of the proposed evidence-guided system. The code is available at this https URL.
[CV-78] Rendering-in-the-Loop: An Execution-Driven Agent for Interactive Web Development
链接: https://arxiv.org/abs/2609.02088
作者: Yilong Guo,Hanqi Chen,Zixiao Ye,Guanzhong Wang,Chen Yu,Zeyu Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Multimodal large language models have achieved remarkable progress in front-end web development, generating interactive webpages from multimodal references such as screenshots and interaction videos. However, existing work largely emphasizes visual metrics such as aesthetics and layout similarity, while overlooking the more critical validation of interactive functionality. We present RILA, an execution-driven agent that puts browser rendering in the loop, iteratively editing generated code from runtime interaction feedback. RILA introduces an Action Interaction Verification (AIV) module that replays the reference interaction trajectory on the generated webpage to collect grounded execution-aware observations, and an Execution-aware Rendering Score (ERS) that jointly measures interaction correctness and visual fidelity to guide iterative optimization. We further build an execution-verified data synthesis pipeline that produces diverse, high-quality training data, offering gains complementary to inference-time optimization. On IWR-Bench, RILA consistently improves both interaction and visual fidelity across foundation models. Notably, with our training pipeline, RILA lifts the compact Qwen3.5-9B backbone from 40.40% to 57.52%, surpassing far larger one-shot generators, including the 1T-parameter Kimi-K2.6 (55.61%) and the proprietary GPT-5.5 (55.74%).
[CV-79] C-Next: Zero-Shot Multimodal Cyclone Forecasting
链接: https://arxiv.org/abs/2609.02085
作者: Zhe Wang,Sijie Chen,Yiming Luo,Daehyun Kim,Chien-Yi Chang
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注: 11 pages
Abstract:We present TropicalCycloneNext (TC-Next), a multimodal deep learning model that forecasts tropical cyclone track and intensity at 6 - 24 h leads by leveraging a foundation model’s forecast fields of atmospheric kinematic and thermodynamic fields and GridSat infrared satellite imagery. Trained only on GraphCast forecasts over the Western Pacific (WP), yet reliant only on generic atmospheric variables, TC-Next on GraphCast lowers track error by 15 - 44% and intensity error by a factor of 3 - 6 relative to a conventional, rule-based tracker, TempestExtremes; applied without retraining to the forecast fields of Pangu-Weather and IFS HRES, it stays ahead of TempestExtremes on both. Applied zero-shot to the generic weather fields of WeatherNext Cyclones on the 2025 WP season, TC-Next attains lower intensity error at every lead time, and lower or comparable track error, compared to that model’s specialized direct tracker in a deterministic comparison. Our ablation studies show that our multimodal model is able to utilize the additional modality to improve performance in tracking errors at every lead time and in intensity prediction at longer lead times.
[CV-80] KSG-Net: Key-Sparse and Global-Context Learning for Maritime 3D Ship Detection PRICAI2026
链接: https://arxiv.org/abs/2609.02077
作者: Zhouyuan Huai,Meiqi Wan,Yan Yang,Minshi Chen,Xin Yuan,Wei Wang,Xiao Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by PRICAI 2026
Abstract:Accurate 3D ship detection in maritime environments is critical for autonomous navigation, yet remains challenging due to large-scale vessel variations, sparse point clouds of small vessels, and severe sea-clutter interference. Existing methods, primarily based on 2D features or dense representations, struggle to balance detection accuracy and computational efficiency, while sparse 3D detectors designed for road scenes generalize poorly to maritime scenarios. This paper focuses on two key challenges in maritime LiDAR perception: weak feature representation for small and sparse vessels, and insufficient global structural modeling for large vessels due to the limited receptive field of local sparse convolutions. To address these issues, we propose KSG-Net, a Key-Sparse and Global-Context learning network for maritime 3D ship detection. The core idea is to jointly enhance local discriminative features and global structural awareness within a unified fully sparse detection framework. Specifically, a Key Sparse Multi-scale Aggregation (KSMA) module is designed to enhance the representation of small and sparse vessels by selecting informative key voxels and aggregating cross-scale neighborhood features. Furthermore, a Global Context Aggregation (GCA) module is introduced to capture long-range geometric dependencies through scene-level context modeling with gated residual interactions, thereby improving the representation of large vessels. Extensive experiments on the Thames River vessel dataset and simulated datasets demonstrate that KSG-Net consistently outperforms existing methods in multi-scale vessel detection and exhibits strong robustness in complex maritime environments.
[CV-81] DPA: Decoupling Product-Agnostic Anomaly Representations for Zero-shot Anomaly Generation
链接: https://arxiv.org/abs/2609.02075
作者: Hang Yao,Yansheng Fu,Ming Liu,Zifei Yan,Yanli Ji,Hongzhi Zhang,Wangmeng Zuo
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Industrial anomaly detection benefits from anomaly samples, yet newly deployed products typically provide only normal images, making anomaly samples difficult to collect. Zero-shot anomaly generation offers a promising solution which avoids collection of target-product anomalies. However, existing methods mainly rely on texture images or text descriptions as anomaly sources, which often produce unrealistic anomalies. Observing that similar anomalies can recur across different products, we propose anomaly transfer-based zero-shot generation, which reuses real anomalies from existing source products, making target-product anomalies no longer necessary to generate realistic anomalious samples for unseen target products. Since not every anomaly type suits the target product, an anomaly type filtering mechanism first selects plausible source types. To transfer selected anomaly, we propose DPA, a diffusion-based framework that decouples product-agnostic anomaly representations. Instead of directly extracting anomaly representations, DPA learns product-irrelevant anomaly embeddings through training with the mismatched data pair, enabling transferable anomaly concept learning across products. Furthermore, we design an adaptive mask-guided pipeline that leverages adaptive masks to control the positional and geometric plausibility of generated anomalies during generation. A training-free anomaly labeling module is further introduced to produce pixel-level annotations aligned with generated anomalies. Extensive experiments on MVTec-AD, VisA, and a dedicated anomaly-transfer benchmark demonstrate that the proposed setting and DPA generate more realistic anomalies and significantly improve downstream anomaly detection performance under both zero-shot and few-shot settings. Source code and models will be released.
[CV-82] LaST-SR: Laplace-Inspired Steady-Transient Complex-Frequency Decomposition for Single Image Super-Resolution
链接: https://arxiv.org/abs/2609.02063
作者: Linhao Li,Zhaojie Pan,Langkun Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 16 pages, 3 figures, 3 tables
Abstract:Single-image super-resolution (SISR) requires global context modeling for structurally consistent reconstruction. Fourier operators are increasingly adopted for global feature modeling. However, their periodic spectral bases constrain the representation of localized aperiodic variations, limiting the recovery of irregular structures and fine details. In dynamical systems, the Laplace neural operator extends Fourier modes to complex frequencies and decomposes the output signal into complementary steady-state and transient responses to jointly model periodic and aperiodic information. We derive, for the first time, an approximate steady-transient decomposition for two-dimensional feature maps, providing an analytical basis for the proposed complex-frequency decomposition. Accordingly, we propose LaST-SR, centered on a Complex-Frequency Decomposition module that couples a global full-spectrum Fourier branch for image-wide dependencies and long-range structural consistency with a window-conditioned local complex-frequency branch for localized, content-dependent aperiodic variations. To fuse the resulting features, we further design a Steady-Transient Collaborative Aggregation module for cross-branch interaction and joint aggregation. Experiments on five benchmarks show that LaST-SR achieves the best PSNR/SSIM among the compared methods for \times2 and \times4 SISR. Ablation studies further validate the effectiveness of the proposed architecture and its key modeling mechanisms.
[CV-83] DocHop: Benchmarking Out-of-domain Multi-hop Reasoning in Information-Dense Documents ICML2026
链接: https://arxiv.org/abs/2609.02059
作者: Zhuoran Yu,Le Thien Phuc Nguyen,Jaden Park,Xinyi Gu,Zexue He,Soochahn Lee,Rogerio Feris,Yong Jae Lee
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: Accepted by ICML 2026
Abstract:Multimodal Large Language Models (MLLMs) have achieved strong performance on structured visual understanding tasks such as chart and document question answering. However, existing benchmarks typically evaluate these domains in isolation, leaving underexplored a key capability: whether models can use textual context to determine how chart evidence should be selected, interpreted, and aggregated. We introduce DocHop, a benchmark for integrated chart–context reasoning in document-style images. In DocHop, the document narrative specifies multi-step compositional constraints, while charts provide the corresponding data values. Questions are grounded on a semantic reference label defined in the narrative, requiring models to resolve target entities from context before aggregating evidence across multiple charts. To enable systematic evaluation, we construct DocHop via a stochastic logic-first generation pipeline with controllable reasoning depth and visual density, covering 2,074 examples across six task categories. Experiments on a wide range of proprietary and open-source MLLMs show a substantial gap to human performance: annotators achieve over 90% accuracy, while the best model reaches only 62.83%. Reasoning-enhanced models consistently show improved results, but performance degrades as reasoning complexity increases. Overall, DocHop provides a controlled testbed for challenging multi-hop document reasoning.
[CV-84] st-Time Logit Prompting for Source-Free Missing Modality Adaptation
链接: https://arxiv.org/abs/2609.02039
作者: Taixi Chen,Nancy Guo
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 9 pages
Abstract:Vision-language models (VLMs) have achieved remarkable performance by leveraging complementary information from large-scale image-text pairs. However, missing-modality inputs are commonly encountered during real-world deployment, often leading to significant performance degradation. Existing methods primarily enhance model robustness by learning modality compensation strategies from source training data. However, their reliance on source training data makes them difficult to apply when original data are unavailable due to privacy, storage, or accessibility constraints, such as clinical applications and personalized AI services. This raises an important yet underexplored question: can VLMs be efficiently adapted at test time for visual recognition with missing modalities without accessing source training data? To this end, we propose Test-Time Logit Prompting (TLP), a lightweight source-free test-time adaptation framework for visual recognition with missing modalities. To address missing-induced prediction shifts, TLP optimizes logit prompts with uncertainty-aware adjustment and modality-complete consistency regularization, adaptively adjusting prediction confidence while preserving semantic consistency. Extensive experiments across diverse vision-language benchmarks demonstrate that TLP consistently enhances recognition performance under missing-modality scenarios, achieving up to 8% improvements while requiring only hundreds of tunable parameters and a few test-time optimization steps.
[CV-85] SelfLift: Accelerating Few-Step Diffusion via Self-Recovering Resolution Transition
链接: https://arxiv.org/abs/2609.02036
作者: Tingyan Wen,Chenqian Yan,Xurui Peng,Xiazhang Fang,Shuai Wang,Xueqian Wang,Songwei Liu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Project page: this https URL
Abstract:Few-step diffusion models substantially compress temporal computation, making the spatial cost of each model evaluation an increasingly dominant source of inference latency. Progressive-resolution inference reduces this cost by performing early denoising at low resolution and reserving high-resolution computation for refinement. However, existing methods typically lift intermediate latents directly and rely on subsequent steps to absorb the induced distribution mismatch. In the few-step regime, the limited recovery budget leaves these errors as visible artifacts, constraining how late the transition can occur and, consequently, how efficiently it can be performed. We introduce SelfLift, a self-recovering progressive-resolution framework that derives both transition-repair signals and trajectory-aligned supervision from the generative model itself. SelfLift-zero proposes a training-free Artifact-Aware Consistency Lift, using disagreement between direct latent lifting and pixel-VAE re-encoding as both a localized artifact-risk signal and a model-native correction direction. It enables reliable late transitions without external super-resolution, extra denoiser evaluations, or sampling-schedule modifications. Building on this robust transition, SelfLift-rich performs On-Policy Self Recovery on student-visited states, transferring dense high-resolution guidance from an internal self-teacher while remaining aligned with the altered progressive-resolution dynamics. Across FLUX.2-Klein and Z-Image-Turbo, SelfLift reduces end-to-end latency by 41.5% and 44.1%, respectively. Combined with timestep distillation, it delivers overall speedups of 29.61x and 19.21x over the corresponding 50-step models while preserving competitive generation quality, establishing a stronger speed-quality frontier for few-step diffusion.
[CV-86] Detecting Object Hallucinations in Large Vision-Language Models via Cross-Modal Attention Drifts and Mask-Based Verification
链接: https://arxiv.org/abs/2609.02028
作者: Xuanbing Wen,Boxu Chen,Le Yang,Jiakai Wang,Zhengyu Zhao,Chenhao Lin,Chao Shen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Despite recent advances in large vision-language models (LVLMs), object hallucination remains a major barrier to their reliable deployment. Existing detection methods often characterize visual grounding using attention from individual layers, leaving its evolution across layers underexplored. We propose CADMP, a lightweight object hallucination detection framework that combines adjacent-layer cross-modal attention drift with prediction sensitivity to targeted visual masking. During decoding, CADMP quantifies distributional changes between consecutive cross-modal attention maps to capture abrupt transitions in visual grounding. It then selects the transition with the largest drift, locates the corresponding visually relevant regions, and measures the change in prediction probability after masking these regions. These two signals provide complementary evidence: attention drift characterizes the stability of internal visual grounding, while probability variation verifies whether a prediction truly depends on the identified visual evidence. A lightweight detector integrates both signals to identify hallucinated predictions. Experiments on multiple benchmarks and representative open-source LVLMs demonstrate that CADMP achieves consistently competitive detection performance. Ablation studies further confirm the complementary contributions of adjacent-layer drift modeling and mask-based grounding verification.
[CV-87] Source-Free Class Relearning: Diagnosing Forgetting in Class Unlearning
链接: https://arxiv.org/abs/2609.02018
作者: Zahra Dehghani,Pablo Piantanida,Mohammadhadi Shateri
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Class unlearning aims to remove a model’s ability to recognize designated forget classes while preserving performance on retain classes. However, low forget accuracy after unlearning does not necessarily mean the class structure has been erased. Approximate unlearning methods can alter classifier decision boundaries while leaving recoverable structure in the representation. Prior work has shown that forget classes can be recovered, but existing approaches require real forget or retain samples, auxiliary data, or reference checkpoints. We study class relearning in a strictly source-free setting, asking whether a forget class can be recovered through a classifier-head update using only the unlearned model. Our approach rests on a theoretical analysis establishing a sufficient alignment condition under which a single gradient step on a synthetic probe set increases the expected logit margin of the forget class. Building on this, we propose a white-box Source-Free Relearning Audit (SFRA), which generates candidate embeddings in representation space and uses model-guided confidence filtering to construct high-confidence retain probes and low-confidence boundary-adjacent probes that are relabelled as the forget class. Gaussian sampling and Softmax confidence are used by default, while ablations with alternative proposal distributions and uncertainty criteria show that recoverability is not specific to these choices. To quantify recoverability, we introduce the Relearning Score (RS), which jointly measures forget-class recovery and retain-accuracy preservation, and report class-matched \Delta RS relative to a retrained reference. Experiments on CIFAR-10, CIFAR-100, and TinyImageNet with ResNet-18, ViT-B/16, and Swin-T show that several unlearning methods exhibit substantial source-free recoverability, and that for a subset of methods this recoverability exceeds the matched retrained reference.
[CV-88] InstEditSeg: Instruction-Driven Image Editing for Polyp and Skin Lesion Segmentation
链接: https://arxiv.org/abs/2609.02004
作者: Ziquan Liu,Zhewei Zhu,Xuyang Shi
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 21 pages, 11 figures
Abstract:Accurate segmentation of polyps and skin lesions is pivotal for clinical diagnosis, yet existing methods struggle with low contrast, ambiguous boundaries, and cross-domain distribution discrepancies. Discriminative networks and most diffusion-based segmentation approaches predict standalone binary masks, leaving the visual priors of large-scale pretrained generative models largely unexploited. We propose InstEditSeg, a unified generative framework that reformulates medical segmentation as an instruction-driven image editing problem. Instead of emitting a mask, the model renders a color-coded overlay on the original image, conditioned on a textual instruction, so that the edited output aligns with the natural image distribution learned by latent diffusion models and mitigates the domain gap between natural and medical imagery. To recover fine anatomical structures, we introduce DINOv3 as an auxiliary visual encoder and a DINO Feature Guidance Block that builds a multi-scale feature pyramid. The pyramid is fused into the diffusion U-Net by channel concatenation and zero-initialized convolution so that hierarchical discriminative priors can be injected without perturbing the pretrained weights. A dual-branch classifier-free guidance strategy requiring only two forward passes per denoising step reduces inference cost. On polyp and skin lesion benchmarks the framework achieves accuracy competitive with strong discriminative baselines, and it further demonstrates concrete advantages of the generative formulation: notably better cross-domain generalization on unseen data, more complete multi-lesion segmentation, instruction-conditioned task control, and sampling flexibility. We also analyze the strengths and limitations of the paradigm, including its color sensitivity and unsupported attribute-conditioned selection. Code is available at: this https URL.
[CV-89] InsightSeg: Reusing Correction Insights for Guideline-Consistent Segmentation
链接: https://arxiv.org/abs/2609.02002
作者: Vanshika Vats,Ashwani Rathee,James Davis
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Guideline-consistent semantic segmentation requires more than category recognition, as real-world labeling policies demand fine-grained, task-specific decisions. Recent multi-agent refinement systems improve compliance with such textual guidelines by detecting and correcting errors. However, they are stateless: feedback from the critiquing agent is discarded, causing the same guideline-specific mistakes to be repeatedly rediscovered and corrected across the dataset at the cost of additional refinement. We introduce InsightSeg, an episodic memory mechanism that converts successful correction episodes into reusable, visually grounded insights. A meta-analyzer distills each qualifying episode into directive natural-language insights and anchors them to the local image regions that caused the error using patch-level visual concept vectors. On subsequent images, these concepts are matched against dense patch embeddings to retrieve relevant insights, which condition the segmenting agent before making its first prediction. This shifts the system from correcting recurring errors to preventing them, improving segmentation quality before any refinement occurs. Across Waymo and Cityscapes, InsightSeg improves both first-pass and final guideline-consistent segmentation performance while requiring fewer refinement steps, demonstrating that multi-agent refinement can become more accurate and efficient by drawing on past correction experience.
[CV-90] Who Drives the Probability Game of VLMs? A Temporal Causal Drive Evaluation Framework
链接: https://arxiv.org/abs/2609.02000
作者: Shuyao Xiao,Shengling Wang,Haoyu Niu,Ke Chao,Changwei Xu,Xinran Duan,Chaoyong Jiang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Vision-language models (VLMs) are increasingly evaluated on complex image and video understanding tasks, yet conventional metrics primarily assess final-answer quality and reveal little about how different information sources shape the generation process. We propose a causal and temporal evaluation framework that traces the evolving roles of visual input, question text, and generated prefixes during autoregressive decoding. Grounded in a Structural Causal Model, we use interventions and backdoor adjustment to derive three step-indexed causal-drive metrics—Visual Causal Drive (VCD), Question Causal Drive (QCD), and Prefix Causal Drive (PCD)—for characterizing source-specific generation patterns without requiring reference answers. Experiments on Qwen3-VL-8B-Instruct across MAVIS, LLaVA-Video-178K, and MiraData, together with cross-model validation on InternVL2-8B, reveal a consistent transition from stronger early question and visual guidance toward increasing reliance on generated prefixes. Randomized-intervention validation shows that QCD and PCD reduce recovery error over observational PMI baselines by 34.8% and 47.1%, respectively. On VLMBias, the prefix–visual imbalance score achieves 0.767 AUROC and 0.873 AUPRC for distinguishing prior-driven from visually grounded generations. These results show that causal-drive trajectories provide complementary source-level diagnostics for multimodal generation.
[CV-91] Linear Fusion MultiDiffusion for Fast Training-Free Spherical Panorama Generation ECCV2026
链接: https://arxiv.org/abs/2609.01997
作者: Akio Hayakawa,Yusuke Mukuta,Tatsuya Harada
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: Accepted to ECCV 2026
Abstract:We propose LF-MultiDiffusion, a training-free panorama generation method that extends MultiDiffusion to support linear projections between target and reference image spaces. Our key idea is to reformulate latent aggregation as a regularized least-squares problem and solve it efficiently with a Krylov-based iterative solver inside the denoising loop. This formulation enables denser and more natural mappings than prior training-free methods, yielding more stable generation with far fewer perspective views. As a result, LF-MultiDiffusion reduces the number of image generator evaluations during denoising and significantly improves inference efficiency. Experiments show that LF-MultiDiffusion achieves better visual quality, text alignment, and panoramic consistency than the strongest training-free baseline, while providing a 15.36 \times speedup. Our project page is available at: this https URL.
[CV-92] Morphology signal in whole slide image foundation models can automatically triage slides
链接: https://arxiv.org/abs/2609.01987
作者: Ayushi Sinha,Shashank Yadav,Benjamin Holmes,Pravat Das,Aaron W. Bogan,James S. Lewis Jr.,Santiago Romero-Brufau,Andrew Y. K. Foong,Scott H. Kaufmann,Kathryn M. Van Abel,David M. Routman,Michael R. Lucas
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 12 pages, 3 figures, 4 tables
Abstract:Patient exams in the cancer diagnosis and staging process typically generate several whole slide images (WSIs). One of the initial steps in training models on WSI data is identifying one or a few slides containing tumor or other diagnostic biomarkers necessary for downstream prediction tasks such as estimating recurrence risk or progression-free survival. This step requires tedious manual curation by experienced pathologists. Many published datasets make the artificial assumption of 1 slide per patient. Alternatively, all slides per patient may be used for model training, which may dilute the signal from the few slides containing tumor or other relevant information. In this paper, we present a pipeline to overcome these challenges using publicly available WSI foundation models (FMs). Our evaluations show that ranking WSIs based on predictions from zero-shot classification using WSI FMs accurately identifies slides with the most tumor, indicating that WSI FMs contain sufficient morphology signal to automatically triage slides. We also present a formulation for ranked evaluation to benchmark FM performance in slide triage. We show, on multiple datasets, that tumor slides are identified in the top-2 ranked slides for patients with up to 43 slides.
[CV-93] Aggregating Neighbor Embedding Projection and Rank-Based Manifold Learning for Image Retrieval
链接: https://arxiv.org/abs/2609.01963
作者: Vinicius Atsushi Sato Kawai,Gustavo Rosseto Leticio,Lucas Pascotti Valem,Daniel Carlos Guimarães Pedronette
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Content-based image retrieval (CBIR) has advanced significantly with deep learning, yet effectively ranking similar images remains challenging, particularly in high-dimensional feature spaces, where pairwise distances often fail to capture contextual relationships and the semantic gap between visual features and high-level concepts persists. Manifold learning and rank-based refinement methods have emerged as complementary strategies, respectively improving feature representations and exploiting contextual information embedded in ranked lists, such as neighborhood relationships among images. However, combining these projection-based and rank-based strategies to exploit their complementary properties remains a challenging research problem. To address this, we propose a framework that combines neighbor embedding projections with rank-based manifold learning through rank aggregation. Uniform Manifold Approximation and Projection (UMAP) generates alternative low-dimensional feature representations, and ranked lists obtained from UMAP projections and rank-based re-ranking methods are combined using the Borda Count aggregation strategy. Experiments were conducted on several public datasets using deep learning features extracted from ResNet152, Swin Transformer, and DINOv2 models. Results show that the proposed approach improves retrieval effectiveness in several scenarios, particularly when the baseline representation struggles to achieve high precision. The aggregation strategy also often improves the quality of top-ranked positions, leading to competitive Mean Average Precision (MAP) and Precision values across different datasets and feature extractors. These findings suggest that combining projection-based and rank-based manifold learning strategies through rank aggregation can provide complementary contextual information for image retrieval tasks.
[CV-94] Learning with Volterra Neural Networks: A System Theoretic Perspective
链接: https://arxiv.org/abs/2609.01928
作者: Haoyu Yun,Hamid Krim,Yufang Bao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Higher-order interaction components are important for signal, image, and video modeling, but explicit high-order operators often suffer from rapidly increasing parameter and computational costs. This paper presents kVNN, a learnable kernelized Volterra Neural operator for compact higher-order filtering. The motivation is to use kernelization to improve the efficiency of Volterra-type neural operators while providing a structured interpretation of their higher-order components. The proposed formulation combines the order-wise structure of Volterra filtering with learnable polynomial-kernel atoms, allowing different interaction orders to be represented by separate learnable centers and coefficients. This order-decoupled representation avoids explicit high-order tensor parameterization and can be implemented as a CNN-compatible layer. Experiments on representative vision tasks show that kVNN achieves a favorable accuracy–efficiency trade-off.
[CV-95] Automated Maize Ear Phenotyping Using 3D Reconstructions
链接: https://arxiv.org/abs/2609.01921
作者: Ritwesh A. Kumar,Som Tripathi,Peja Matthews,Srikar Reddy,Talukder Zaki Jubery,Patrick Schnable,Adarsh Krishnamurthy,Baskar Ganapathysubramanian
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Maize kernel traits such as row number, kernels per row, and kernel size vary largely for genetic reasons and are consistently associated with regions of the genome that influence yield. Manual measurement of these traits, however, cannot keep pace with the volume of maize generated in a breeding program. To address this, we developed and validated a fully automated pipeline for extracting these traits from 3D point clouds of corn ears, built on a recently developed video-to-point-cloud platform. Raw video frames are processed through COLMAP and NeRF, the ear is isolated via density-based separation, and the point cloud is distance-calibrated to physical units. The calibrated ear point cloud was Z-axis aligned via PCA and cylindrically unwrapped to a 2D image. We enhanced contrast and performed zero-fine-tuning instance segmentation using Cellpose-SAM. A triple-juxtaposed unwrap strategy was used to prevent double-counting at the seam. The pipeline achieved kernel count R^2 = 0.921 (MAPE = 10.33%) and kernel row number within ±2 rows for 95.2% of ears (MAE = 0.75 rows) on a 168-ear held-out set from the 268-ear labeled dataset. The resulting multi-trait dataset has known genotype identity for each ear, positioning it for phenotype-to-genotype association analyses.
[CV-96] APVid-MV: A Benchmark for Tracking Any Point in 3D Across Multiple Views
链接: https://arxiv.org/abs/2609.01899
作者: Skanda Koppula,Frano Rajic,Abdullah Faiz Ur Rahman,Yi Yang,Ignacio Rocco,Jeet Thakwani,Rishabh Kabra,Andrew Zisserman,Joao Carreira,Siyu Tang,Carl Doersch,Gabriel Brostow
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Multi-camera systems are increasingly practical for robotics, AR/VR, and autonomous driving because complementary views reduce depth ambiguity and preserve visibility under occlusion. Existing point-tracking benchmarks, however, focus on a single video or static multi-camera rigs. None test long-term 3D point tracking across several synchronized views under camera motion. We introduce TAPVid-MV (Tracking Any Point in Video across Multiple Views), the first benchmark for this setting. It contains a curated set of 284 sequences, 1,142 calibrated camera streams, and 109,769 point tracks across seven subsets spanning indoor and outdoor domains, from robotics and human activity to driving and synthetic procedural scenes. We obtain these trajectories using dataset-specific auxiliary modalities: sensor depth, LiDAR, SLAM and SfM points, human meshes, posed object meshes, and simulation. Every sequence and trajectory is visually verified by human annotators. Across more than 30 baselines, no method comes close to solving the task. Surprisingly, existing multi-view point trackers do not consistently outperform monocular point trackers. By evaluating reconstruction and point tracking on the same datasets, TAPVid-MV helps distinguish errors in recovered geometry from errors in point correspondence. Through this joint analysis, we identify geometry recovery as a major bottleneck for accurate 3D point tracking. Beyond multi-view 3D point tracking, our released annotations support monocular 2D and 3D point tracking, future-trajectory prediction, and 4D reconstruction. Subjects: Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2609.01899 [cs.CV] (or arXiv:2609.01899v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2609.01899 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-97] Does Playing it Safe Count as Faithfulness? Reassessing LVLM Hallucination Mitigation Methods
链接: https://arxiv.org/abs/2609.01888
作者: Mehrdad Fazli,Sina Mansouri,Mohit Marvania,Ziwei Zhu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Recent inference-time hallucination mitigation methods for large vision-language models (LVLMs) report strong gains on hallucination benchmarks. However, it remains unclear whether lower hallucination scores reflect improved multimodal grounding or more conservative generation. We evaluate six mitigation methods across three LVLMs and four benchmarks, including hallucination-focused evaluation and the diverse capability benchmark MMStar. Our analysis reveals two consistent patterns. First, hallucination reduction is often coupled with reduced informativeness: methods that lower hallucination rates also reduce object recall, visual coverage, or response detailedness. Second, improvements on hallucination benchmarks do not reliably transfer to broader multimodal capabilities, with methods showing inconsistent or degraded performance on fine-grained perception and reasoning tasks. Our findings suggest that current evaluation protocols may overestimate progress by rewarding conservative generation. We argue that hallucination mitigation should be evaluated as a faithfulness–informativeness–capability trade-off rather than through hallucination scores alone.
[CV-98] SignMatch: Matching Dictionary Signs to Continuous Sign Language Video WWW ATC
链接: https://arxiv.org/abs/2609.01886
作者: Ryan Wong,Youngjoon Jang,Liliane Momeni,Gül Varol,Andrew Zisserman
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 24 pages, 8 figures, Project page: this https URL
Abstract:The objective of this paper is to match dictionary sign videos to corresponding signs in continuous signing videos, where a match is defined by the visual similarity alone - the handshape and motion relative to the body. To achieve this, we learn a prototype-structured sign embedding space from continuous video annotated with signs, where each learnable prototype corresponds to a sign class. Isolated dictionary videos are then mapped into this sign space, enabling the matching between dictionary exemplars and continuous sign instances. This design supports direct dictionary-guided sign matching through embedding similarity and naturally extends to unseen signs using only dictionary exemplars. Experiments on ASL-Citizen dictionary retrieval, ChaLearn OSLWL dictionary-to-continuous sign matching, and using BOBSL’s CSLR2 evaluation for automatic sign annotation demonstrate strong generalisation across datasets, tasks and sign languages. Without benchmark-specific supervision, the learned representation transfers effectively across American, British, and Spanish Sign Languages, outperforming prior methods on all three benchmarks. Project page: this https URL
[CV-99] RAFT-DVC: Resolution-Aware Machine Learning-Based Digital Volume Correlation
链接: https://arxiv.org/abs/2609.01876
作者: Zixiang Tong,Lehu Bu,Jin Yang
类目: Computer Vision and Pattern Recognition (cs.CV); Materials Science (cond-mat.mtrl-sci)
备注:
Abstract:Digital volume correlation (DVC) provides three-dimensional full-field displacement measurements from volumetric images, but how the internal resolution of a machine-learning-based DVC model affects accuracy and operating range remains poorly understood. Here, we present RAFT-DVC, a resolution-aware family of recurrent all-pairs field transforms (RAFT)-based DVC solvers with encoder downsampling factors s = 2, 4, and 8. Using a matched design, we find that the three solvers localize displacement to approximately 0.017 feature-grid voxel, giving an empirical raw-volume error scaling of approximately 0.017s voxel. The solvers exhibit complementary operating regimes governed jointly by displacement reach and volumetric-texture compatibility. Synthetic benchmarks show that RAFT-DVC achieves errors of the same order as tuned classical DVC under fine-texture, small-to-moderate-displacement conditions and becomes competitive or advantageous under coarse-texture, large-displacement conditions. Frequency-swept tests quantify deformation spatial resolution, while tiled inference enables dense estimation on large volumes. Evaluation on confocal volumetric images acquired during indentation illustrates the importance of matching solver operating regime to deformation magnitude and image texture. Tests on micro-CT images of elastomeric foam, despite training only on particle-labeled synthetic data, provide evidence of cross-texture transfer. We also identify coordinate-order inconsistencies in three-dimensional RAFT correlation sampling and introduce a non-cubic impulse test to verify sampler geometry independently of network training. Correcting the sampler improves native-input accuracy and generalization to unseen volume dimensions. Together, these results establish RAFT-DVC as a fast, resolution-aware framework for dense DVC with characterized accuracy and operating regimes.
[CV-100] Cross-Model Distillation of a Human-Pose Foundation Model from Unannotated Infant Video for Markerless 3D Pose Estimation ECCV2026
链接: https://arxiv.org/abs/2609.01840
作者: R. James Cotton,Divya Joshi,Colleen Peyton
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to the ECCV 2026 Workshop MoCha
Abstract:Spontaneous movement is one of the earliest windows onto an infant’s neuromotor health, and structured clinical instruments that score it are validated early predictors of cerebral-palsy risk. However, they require specially trained raters, are time-consuming, and carry inter-rater variability. This motivates automated, video-based markerless assessment, especially as marker-based motion capture is impractical in infants. Yet the foundation models that make markerless capture possible are trained almost entirely on adults: our recent multi-view infant study found that no single model is jointly best, with strong 2D keypoint accuracy and direct 3D body recovery split across different models. While that study identifies this trade-off, it does not resolve it. Here, we perform cross-model distillation from the Sapiens 2 pose model into the SAM 3D Body model, using unannotated infant video alone. A frozen teacher supplies dense pseudo-labels, and a differentiable renderer aligns the predicted mesh to them in the training loop. On eleven held-out infants (18 sessions, 173 recordings) under our prior study’s multi-view protocol, fine-tuning improves same-view 2D keypoint agreement with the Sapiens reference (median body percentage of correct keypoints @ 10px 0.22 - 0.42, face 0.22 - 0.42) and Procrustes-aligned mean per joint 3D position error (25.5 - 22.2 mm). This demonstrates how cross-model distillation improves SAM 3D Body model performance on infants.
[CV-101] SliceBridge: context-consistent repair of corrupted slice intervals in T1-weighted MRI
链接: https://arxiv.org/abs/2609.01827
作者: Jiheng Li,Michael E. Kim,Trent Schwartz,Gaurav Rudravaram,Derek B. Archer,Timothy J. Hohman, theAlzheimer’s Disease Neuroimaging Initiative,Lianrui Zuo,Bennett A. Landman
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Structural magnetic resonance imaging (MRI) images are sometimes corrupted over a contiguous set of slices, where acquisition, motion, hardware, or reconstruction effects leave a single slice or short interval inconsistent with its neighbors while the rest of the image remains usable. Such localized corruption can bias downstream morphometric analysis, yet discarding or reacquiring an otherwise usable image is costly. We formulate this as an image restoration problem: given the location of the affected interval, reconstruct those slices from the surrounding anatomical and imaging context. We propose SliceBridge, a framework for restoring corrupted slice intervals in T1-weighted MRI using rectified flow matching conditioned on the surrounding intact slices and their relative slice positions. Through-plane consistency is encouraged by coupling the slices within the interval through interval-correlated initial noise, a shared flow time, and synchronized sampling. The restored interval is then inserted back, leaving all other slices unchanged. We trained and validated the model on 9,877 T1-weighted brain MRI volumes from four datasets and evaluated it on 581 external subjects using clean interval withholding and controlled corruptions. Compared with a matched model that reconstructed target slices independently, SliceBridge reduced error in slice-to-slice changes within repaired intervals by 32.9%-41.3% across interval lengths and achieved higher SSIM at every interval length. In controlled-corruption cases, SliceBridge reduced the median error in regional brain volume estimates produced by a downstream segmentation model from 1.95% in corrupted volumes to 1.05%.
[CV-102] Kirin: Animal Motion Generation from In-the-Wild Video ECCV2026
链接: https://arxiv.org/abs/2609.01823
作者: Brian Nlong Zhao,Zhuoyang Pan,James M. Rehg,Jiajun Wu,Shangzhe Wu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: ECCV 2026
Abstract:Understanding animal motion is fundamental to modeling animal behavior and biomechanics, yet progress in this area lags far behind human motion research due to the scarcity of high-quality motion data. While human motion can be captured in controlled environments, it is impractical for most animal species, resulting in small, domain-limited datasets that restrict downstream applications such as animation. To address this challenge, we introduce Kirin, a framework that reconstructs motion from video, learns motion priors at scale, and generates realistic motion that can be directly applied to animated assets. Using large collections of in-the-wild animal videos, we reconstruct 3D motion sequences and pair them with captions to create AiM3D, the first large-scale dataset offering aligned video-text-motion tuples for quadruped animals. Building on this dataset, we develop a visual-guided motion generation model that conditions on both text and image to guide the generation of realistic motion across diverse animal species. Finally, by leveraging an off-the-shelf image-to-3D model, we automatically rig and animate 3D meshes using generated motion, producing ready-to-render animated animals. Together, our dataset and framework establish a new foundation for large-scale, text and image conditioned animal motion generation and animation. Project page: this https URL.
[CV-103] Video2Reaction: Training Foundation Video Models to Predict Audience Reaction CVPR2026
链接: https://arxiv.org/abs/2609.01816
作者: Sidong Zhang,Trang Nguyen,Shiv Shankar,Gauri Jagatap,Deepak Chandran,Andrea Fanelli,Madalina Fiterau
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Presented in the Workshop on Emerging Directions in Data for Multimodal Foundation Models at CVPR 2026
Abstract:We introduce Video2Reaction, a multimodal dataset that maps short movie segments to the induced emotional reactions of viewers in the wild, as expressed through social media comments. Video2Reaction captures the natural diversity of emotional responses by aggregating reactions from online comments at scale, modeling labels as distributions over categorical emotions to better reflect the subjective and ambiguous nature of emotional perception. We benchmark two vision-language models (VLMs) finetuned with LoRA, showing that VLMs learn effectively from Video2Reaction and outperform specialized baselines on dominant reaction prediction. We further demonstrate that VLMs pre-finetuned on Video2Reaction transfer effectively to VCE, another induced emotion dataset with a different taxonomy and video domain. Notably, LLaVA-NeXT-Video-7B pre-finetuned on Video2Reaction and adapted on only 1% of VCE training data achieves a top-3 accuracy of 0.682, on par with the best reported VCE performance trained on the full dataset. The dataset is available at this https URL
[CV-104] Integrated Laser Scanning and Image-Based Topology Optimization Techniques for Detection and Quantification of Visible and Subsurface Structural Defects
链接: https://arxiv.org/abs/2609.01808
作者: Mehrdad Shafiei Dizaji,Devin Harris
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 9, 5
Abstract:Reliable characterization of structural defects requires methods capable of resolving both directly observable surface damage and damage that is not visible from the inspected surface. This study presents two complementary non-contact, vision-based approaches for the detection and quantitative characterization of defects in structural components. The first approach employs high-resolution laser scanning to generate three-dimensional (3D) point clouds of damaged steel specimens. Comparative processing of measured and reference point clouds is used to localize damaged regions, quantify geometric loss, and transfer the measured defect geometry to a finite element representation. The second approach combines full-field surface deformation measurements obtained using three-dimensional digital image correlation (3D-DIC) with finite element model updating and topology optimization. In this inverse framework, measured surface response is used to infer subsurface abnormalities through their influence on the spatial distribution of structural response. Experimental steel-beam specimens containing controlled smooth defects and randomly distributed defects are used to evaluate the approaches. Comparisons with milling-based ground-truth measurements demonstrate that both methods can identify and quantify defect geometry, while providing complementary information for visible and subsurface damage assessment. The combined framework establishes a pathway toward high-fidelity, non-contact structural condition assessment and model updating for components with complex and irregular damage.
[CV-105] Consistency as Regularization for Unsupervised Shadow Removal
链接: https://arxiv.org/abs/2609.01806
作者: Anh-Kiet Duong,Petra Gomez-Krämer,Jean-Michel Carozza
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 22 pages, 20 figures
Abstract:Shadow removal is an important preprocessing step for many vision tasks, yet existing supervised methods require paired shadow and shadow-free images, while unsupervised approaches often still rely on shadow masks or shadow-free references. We propose ShadowCLR, an unsupervised framework that learns shadow removal directly from shadow images. Our key observation is that shadows vary across observations while the underlying scene content remains largely consistent. We therefore use consistency across shadow observations as regularization, encouraging the model to recover scene-consistent appearance while suppressing shadow-specific variations. Global and local consistency further enable us to explore visually related images, learn from imperfectly aligned observations, and focus the representation on shared scene information. Experiments on multiple benchmarks show that ShadowCLR achieves competitive and often superior performance over state-of-the-art unsupervised methods, demonstrating that consistency can provide regularization for shadow removal without shadow masks or shadow-free images.
[CV-106] Improved Automatic Target Recognition in Synthetic Aperture Sonar Imagery Using Large Deep Neural Networks
链接: https://arxiv.org/abs/2609.01800
作者: C.J. Moore,Alex Hurt,Jordan Malof
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Automatic Target Recognition (ATR) in Synthetic Aperture Sonar (SAS) is a task largely dominated by deep neural networks (DNNs). Most SAS-ATR models use convolutional neural network (CNN) architectures whereas transformer-based architectures have had much less representation in the literature despite being state of the art in general computer vision (CV) research. Additionally, researchers have had mixed results in attempting to overcome challenges presented by a scarcity of labeled training data by using methods such as data augmentation and the use of pretrained weights from a variety of imaging modalities. In this work, we compare the performance of modern CNN and transformer-based DNNs to determine which architecture and training configurations elicit the highest performance in SAS-ATR. We investigate how network size, architecture, pretraining method, data augmentation and other forms of regularization affect SAS-ATR performance with a focus on producing the highest-performing model and providing a roadmap for training state-of-the-art SAS-ATR models.
[CV-107] Designing Versatile Samples for Learned Trajectory Scoring
链接: https://arxiv.org/abs/2609.01799
作者: Yaguang Li,Jiaru Zhang,Chuheng Wei,Can Cui,Ziran Wang
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Many current end-to-end driving policies emit a pool of candidate trajectories and select one, which makes selection a separable component: a scorer can be retrained while the planner, its backbone, and its trajectory generator all stay frozen. However, many strong planners concentrate their proposals around safe mode, providing limited supervision near decision boundaries. In this work, we design a training dataset that provides more informative supervision for the scorer. In particular, we construct two generators that perturb the logged human trajectory along the two axes a vehicle can be displaced: laterally toward the drivable boundary and longitudinally toward a leading vehicle. The designed dataset produces more informative positive and negative samples than the base planner’s proposal pool. We attach a transformer-based scorer to two frozen generative planners, DiffusionDrive and MeanFuser, and train it on the NAVSIM navtrain dataset. The results of the experiments show that we achieve 90.1 EPDMS on DiffusionDrive and 90.4 EPDMS on MeanFuser when using ResNet-34, with 0.4 and 0.3 EPDMS respectively, from the designed training dataset.
[CV-108] DESA-TTA: Dynamic EMA and Source Anchoring for Test-Time Adaptation
链接: https://arxiv.org/abs/2609.01795
作者: Atif Belal,Lilian Hollard,Marco Pedersoli,Eric Granger
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Vision-language object detectors (VLODs) achieve strong zero-shot performance but remain vulnerable to distribution shifts during deployment. Mean-teacher methods for test-time adaptation (TTA) can improve robustness by updating a student model using teacher-generated pseudo-labels. However, mean-teacher TTA is highly sensitive to the choice of a fixed exponential moving average (EMA) coefficient for teacher updates, and repeated optimization with noisy pseudo-labels can cause cumulative student drift. We propose Dynamic EMA and Source Anchoring for TTA (DESA-TTA), a low-overhead method that jointly regulates teacher updates and student drift through dynamic temporal averaging and source anchoring. Dynamic temporal averaging estimates teacher uncertainty from pseudo-label confidence and box density and uses it to select a sample-wise EMA coefficient within bounds determined by teacher parameter drift. Source anchoring partially restores the updated student parameters toward their pretrained values, with the anchoring strength increasing according to student drift. Experiments across diverse distribution shifts and two VLOD architectures show consistent improvements over existing TTA methods. On VOC-C, DESA-TTA improves AP _50 by 14.5 points over zero-shot inference while achieving 55% higher inference throughput than the previous state-of-the-art TTA method for YOLO-World. Our code: this https URL
[CV-109] Efficient Passive Acoustic Monitoring of Killer Whales Using a Two-Stage Detection and Ecotype Classification Cascade
链接: https://arxiv.org/abs/2609.01792
作者: Daniela Ruiz,Manuel Castellote,Zhongqi Miao,Carl Chalmers,Bruno Demuro,Rahul Dodhia,Pablo Arbelaez,Juan M. Lavista
类目: ound (cs.SD); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Passive acoustic monitoring of killer whales is particularly important for conservation of the endangered Southern Resident killer whale population, but requires accurate models that can operate in real time under severe class imbalance and deployment shift. We propose a lightweight ResNet-based two-stage cascade that first detects killer whale vocalizations and then classifies confident detections into five eastern North Pacific ecotypes, abstaining on ambiguous calls. We train and evaluate the pipeline on the DCLDE 2027 dataset, where the detector achieves 0.960 macro-F1 and the classifier 0.958, outperforming frozen Perch 2.0 embeddings on the five-ecotype benchmark. By separating detection from ecotype recognition, the end-to-end cascade improves seven-class macro-F1 from 0.919 for a single-stage model to 0.933, with the largest gain on the rare OKW ecotype. To assess transfer beyond the benchmark, we use active learning to adapt the Stage 1 to the acoustic environment of Puget Sound, WA, increasing killer whale detection F1 from 0.405 to 0.755 on manually verified detection windows. Finally, each stage processes a 3 s window in approximately 1.4 ms on an NVIDIA H100, enabling faster than real time inference. These results demonstrate that the proposed two-stage cascade pipeline enables reliable killer whale detection and classification, adaptation to new acoustic domains, and real-time monitoring for conservation applications.
[CV-110] CoViT: Instance-Correspondence Contrastive Learning for Vision Transformer
链接: https://arxiv.org/abs/2609.01787
作者: Yisen Wang,Zhirong Wu,Limin Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Vision Transformers (ViT) excel in semantic understanding but fail to discriminate between object instances (e.g., identical embeddings for two dogs), limiting their use in instance-level tasks such as object detection and instance segmentation. We propose Contrastive Vision Transformer (CoViT), a self-supervised learning framework that injects instance-awareness into ViT through geometry-guided contrastive learning. CoViT uniquely coordinates ViT’s attention maps and embeddings by constructing triplets: (1) Attention-guided masking: Refine multi-head attention via adaptive thresholding and morphological operations to generate instance masks, identifying foreground anchors; (2) Hardest contrastive mining: For each anchor, computing pairwise embedding similarities to select the intra-instance hardest positive (least similar patch within its mask) and inter-instance hardest negative (most similar patch from other instances), with intra-instance regions masked during negative search. These triplets drive a contrastive loss that simultaneously compresses intra-instance variance and expands inter-instance margins, forcing ViT to discern subtle geometric and appearance differences between instances. CoViT consistently achieves stable performance gains of over 2 AP points across multiple instance-level perception tasks by using ViT as backbone architecture. Notably, CoViT requires no extra decoders or labels, demonstrating that a pure ViT can learn instance-aware representations via inherent attention priors and targeted contrastive constraints. Code and models will be released.
[CV-111] n Architectures One Error: Shared Failure Modes in Hyperspectral Classification under Spatially Disjoint Evaluation
链接: https://arxiv.org/abs/2609.01786
作者: Ehsan Faghih,Fatemeh Ashrafi,Marguerite Moore,Zahra Saki
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Hyperspectral image classification still relies heavily on random pixel splits within a single scene. The Salinas dataset, randomly split, is among the most widely used datasets for comparing different architectures. However, under a random split method, a large fraction of test pixels fall immediately adjacent to a training pixel, which inflates reported accuracy. This work introduces a leakage-free evaluation protocol linking spatial separation to the model’s receptive field. Applying this protocol across ten different architectures, including classical, spectral, spectral-spatial, transformer, vision-backbone, and state-space families, shows that Macro-F1 drops by 0.147 on average and model rankings change by as many as five places. Furthermore, leakage-free evaluation limits which architectures can be tested on a given benchmark. Since each partition supports patches only within a finite radius, reporting this radius alongside the receptive field is essential for fair comparison. In addition, this study reveals that all ten architectures misclassify largely the same pixels, pointing to a spectral ambiguity in the data that none of them resolves.
[CV-112] Allocate Before You Embed: Adaptive Visual Input Allocation for Video Embeddings
链接: https://arxiv.org/abs/2609.01778
作者: Song Jin,Zhongtao Jiang,Chenglei Shen,Huanxuan Liao,Haozhe Chi,Zhiwei Wang,Kun Xu,Yong Liu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Large-scale video retrieval requires embedding models to encode long and diverse videos under tight visual-input and inference budgets. Existing methods typically sample a small, fixed set of frames at their original resolution, limiting temporal coverage and ignoring frame importance. Our empirical analysis shows that expanding temporal coverage improves retrieval even under a fixed visual-input budget. Gains are larger when the original per-frame resolution is preserved, highlighting the complementary roles of temporal coverage and spatial fidelity. Motivated by this finding, we propose AllocEmbed, an allocate-then-embed framework that reallocates a fixed visual-input budget across more frames. A lightweight allocator uses low-cost previews to assign frame-wise resolutions before the embedding backbone, preserving more detail where it most benefits retrieval while reducing visual cost elsewhere. We further introduce Retrieval-Driven Policy Optimization (RDPO), which learns the allocator directly from retrieval feedback using a rank-validated similarity gap and a confidence-guided efficiency incentive. Operating entirely before the backbone, AllocEmbed integrates with existing retrieval systems without modifying the embedding model or downstream pipeline. Experiments on the MMEB-V2 V-QA and V-RET tasks and our LongRet benchmark show that AllocEmbed achieves the best overall retrieval performance among the evaluated budget-matched methods and transfers across embedding backbones. Our code is publicly available at this https URL.
[CV-113] AlphaRAD: Grounded Zero-Shot Classification in Chest Radiology via α-Corrected Binary Cross Entropy and Factorized Latent Supervision ECCV2026
链接: https://arxiv.org/abs/2609.01757
作者: Jianzhong You,Yuan Gao,Chris McIntosh
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: ECCV 2026
Abstract:Vision-Language Pretrained Models (VLPMs) offer a scalable path to open-vocabulary chest radiology understanding, yet two aspects remain underexplored: how structured clinical semantics extracted from medical reports can reduce in-batch noise during contrastive learning, and how cross-modal fusion can be designed to produce more faithful spatial grounding without added complexity. We introduce AlphaRAD, addressing these opportunities through two contributions. First, we construct a large-scale structured medical concept space from medical reports parsed by a Large Language Model for training, thereby mitigating in-batch learning noise and removing heuristic pair matching in contrastive learning, and thus naturally positioning AlphaRAD as a medical concept discriminator trained via \alpha -Corrected Binary Cross-Entropy. Second, we propose FLaS (Factorized Latent Supervision), an extremely simple yet effective cross-modal feature fusion module that factorizes VLPM representations into independent subspaces, using dedicated alignment supervision to enhance the expressiveness of spatial grounding without introducing additional model parameters. Through extensive empirical validation, AlphaRAD shows strong zero-shot generalization across diverse chest radiology tasks. Notably, it establishes state-of-the-art average performance across 16 classification benchmarks, while achieving individual state-of-the-art results via distinct gains on 7 grounding/phrase grounding and 3 segmentation datasets.
[CV-114] Swin Meets EfficientNet: Lightweight Architectures for GAN-Based Face Forensics
链接: https://arxiv.org/abs/2609.01749
作者: Sejuti Basu,Ashima Sood,Vijay Kumar,Sahil Sharma
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 12 pages, 2 figures, 1 table. Presented at the International Conference on Computational Techniques in Data Science (IICTDS 2025), online. Proceedings forthcoming
Abstract:Modern generative models, such as GANs, diffusion architectures, and autoregressive systems, now produce facial images that are nearly indistinguishable from authentic photographs. This capability makes detecting forged images increasingly difficult, raising serious concerns about identity theft, fraud, and misinformation campaigns. Our research focuses specifically on GAN-generated synthetic faces, which underpin many face-centric deepfakes, and investigates efficient detection approaches using image analysis alone. Existing detection systems rely heavily on either convolutional neural networks (CNNs) or global vision transformers. While CNNs excel at identifying texture-based local features, they struggle with broader contextual understanding. Traditional Vision Transformer (ViT) models can capture long-range structures effectively, but demand substantial computational resources. Our work explores Swin-Transformer-based architectures across three implementations: a compact Swin Transformer trained from the ground up, ImageNet-1K pre-trained Swin-Tiny and Swin-Small models adapted for binary classification, and a novel hybrid combining EfficientNet-B0’s convolutional processing with a Swin Transformer backend. We evaluated all models using the 140K Real and Fake Faces dataset, which includes StyleGAN-generated fake faces alongside authentic images from Flickr and DFDC, with balanced splits for training, validation, and testing. The EfficientNetB0+Swin hybrid achieved 99% accuracy and a 99.44% recall on 5,000 test images, outperforming both pure Swin variants and a previous CNN-only baseline on this dataset. Our results suggest that combining hierarchical CNN features with shifted-window self-attention provides an efficient and computationally lightweight method for detecting GAN-generated synthetic faces.
[CV-115] SCULPT: Training Edge Vision Models for Post-Training Quantization Readiness
链接: https://arxiv.org/abs/2609.01743
作者: Bharadwaj Kavuri,Sourav Babu-PK,Varadhraj Ellapan,Pullarao Maddu,Prasad Deshpande
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Published at Irish Machine Vision and Image Processing Conference (IMVIP) 2026. Paper 043. 8 pages, 3 figures
Abstract:Edge vision models are difficult to deploy on resource-constrained hardware, making low-bit post-training quantization (PTQ) attractive. In practice, standard FP32 training often produces heavy-tailed activation distributions whose outliers destabilize activation quantization: preserving the full range wastes quantization bins on rare extremes, while aggressive clipping causes information loss. Existing solutions typically rely on quantization-aware training (QAT), which adds training complexity and bit-width coupling, or advanced PTQ procedures that repair the model after training. We present SCULPT (Statistical Clipping and Uniform Loss for Post-Training), a training-time method that improves PTQ readiness during ordinary FP32 fine-tuning. SCULPT combines a topology-aware activation regularizer that suppresses quantization-hostile skewness and kurtosis with a stable percentile-based clipping mechanism that learns deployment-ready activation bounds. Unlike QAT, SCULPT does not simulate quantization during optimization; unlike post hoc outlier-repair PTQ methods, it does not require runtime activation transformations. The learned clipping bounds can be exported directly into a standard PTQ workflow for low-bit deployment, including INT8 and lower-bit settings such as W4A8. Comments: Published at Irish Machine Vision and Image Processing Conference (IMVIP) 2026. Paper 043. 8 pages, 3 figures Subjects: Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2609.01743 [cs.CV] (or arXiv:2609.01743v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2609.01743 Focus to learn more arXiv-issued DOI via DataCite Journalreference: Irish Machine Vision and Image Processing Conference (IMVIP), 2026
[CV-116] Evidential Deep Learning for Multi-Modal Anti-UAV Detection
链接: https://arxiv.org/abs/2609.01742
作者: Dmitry Golovchits,Seyed Sahand Mohammadi Ziabari,Ali Mohammed Mansoor Alsahag
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Anti-UAV systems increasingly fuse multiple sensors, yet their detection heads provide no per-modality reliability signal. This study evaluates whether evidential deep learning (EDL) heads, Dempster-Shafer (DS) evidence fusion, and uncertainty-driven temporal sensor gating improve anti-UAV detection through a controlled ablation on three benchmarks: thermal tracking (AntiUAV600), RGB-audio-RF classification (TRIDENT), and RGB-IR tracking (MM-UAV). The EDL training objective improves accuracy over retrained sigmoid baselines (+5.9 percentage points in accuracy and a tripled tracker-on-absent rate in E1; +4.8 percentage points in classification accuracy in E2, surviving a clip-clustered bootstrap, p = 0.011) and ranks classification errors substantially better (entropy UAUC approximately 0.94 vs. 0.51). The remaining components do not support their respective hypotheses. DS fusion does not outperform simple probability averaging. Dirichlet vacuity adds no ranking power beyond predictive entropy and inverts at the detection level, where extreme background imbalance causes it to encode class membership rather than error likelihood, a failure also observed for entropy and sigmoid confidence. Temporal gating preserves accuracy only when nearly inactive and yields no realised latency saving on shared-backbone hardware. The benefit of evidential learning therefore arises primarily from its training objective rather than its uncertainty estimate; a crop-level control further localises the detection-level breakdown to anchor-level evaluation rather than the learned representation.
[CV-117] ZipTok3D: High-Fidelity 3D Tokenization with Compact Token Prefixes
链接: https://arxiv.org/abs/2609.01740
作者: Mingda Lin,Weijie Wang,Zeyu Zhang,Bowen Cui,Yefei He,Haoyu Zhao,Yuanyu He,Donny Y. Chen,Feng Chen,Bohan Zhuang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 25 pages, 9 figures, 6 tables, including appendix
Abstract:Compact token sequences are essential for efficient 3D generation. However, existing 3D tokenizers typically organize latent representations either over spatial regions or as fixed-size sets of global tokens, both suffering sharp reconstruction degradation when compressed to extremely low token budgets. In this paper, we present ZipTok3D, a 3D tokenizer designed for high-fidelity reconstruction from extremely short token sequences. Its key idea is to organize object geometry into progressively informative global-token prefixes and unfold these compact representations through iterative decoding. Specifically, nested dropout randomly truncates the latent sequence after encoding during training and requires each retained prefix to reconstruct the complete object, thereby prioritizing essential geometric information in the leading tokens. The decoder then repeatedly applies a parameter-shared Transformer block to recover fine-grained geometry from each prefix without a separate generative sampling stage. With the same token dimension, ZipTok3D achieves reconstruction quality comparable to the 32-token COD-VAE baseline using only one token on ShapeNet and four on TRELLIS, yielding 32\times and 8\times shorter token sequences, respectively.
[CV-118] UAV Thermal Imagery for Inert Ordnance Screening: Multi Campaign Dataset DevelopmentObject Detection and Practical Recommendations
链接: https://arxiv.org/abs/2609.01738
作者: Chad Melton,PhD.,Annabelle Kelton
类目: Computer Vision and Pattern Recognition (cs.CV); Databases (cs.DB)
备注: Abstract edited to fit format
Abstract:Unexploded ordnance (UXO) continues to restrict civilian access, agricultural activity, infrastructure recovery, and environmental remediation in contaminated areas around the world. This study created a multi campaign UAV thermal image data set of inert ordnance, developed a labeled image set from collected imagery, tested object detection models, and identified practical considerations for humanitarian mine action and demining applications. Data were collected during four field campaigns in Tennessee under summer and winter conditions using inert mines, munitions, and other ordnance placed in short grass, tall vegetation, gravel, mulch, rock, compost, and compacted surfaces. Thermal imagery was collected under flight altitutes of 33 m and 15 m. The final source inventory contained 5,855 thermal image label pairs, including 918 positive images and 4,937 background images. After retaining all positive images and downsampling background images, the 33 m dataset contained 420 training and 106 validation images, while the 15 m dataset contained 629 training and 157 validation images. YOLOV11l and RT-DETR-R50 algorithms were trained and evaluated to develop an automated candidate detection model. Practical recommendations include collecting thermal and RGB imagery together, incorporating varied surfaces and background only imagery, considering periods following changes in solar exposure, balancing survey coverage against target pixel representation, calibrating models with representative local data, and retaining qualified human review. The intended use is screening and prioritization for follow on technical survey or EOD assessment, and not a standalone clearance.
[CV-119] From Visual Cues to Spoken Narration: Rethinking Audio Description EMNLP
链接: https://arxiv.org/abs/2609.01725
作者: Akshita Gupta,Aditya Arora,Federico Tombari,Marcus Rohrbach,Anna Rohrbach
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to EMNLP main conference
Abstract:Audio Description (AD) provides spoken narration of visual events during dialogue gaps, making movies accessible to visually impaired audiences. The problem requires determining both what (which visual event) and when (position for inserting the AD) to narrate, to achieve the best user experience. Prior work has largely reduced the problem to video captioning of pre-segmented video clips, i.e., what is largely predefined and when is ignored entirely. We propose Cue2Narrate, a two-stage pipeline that jointly predicts what and when to narrate in longer untrimmed movie clips. A dual-head audio-visual localizer predicts two temporally distinct windows per AD utterance: a visual cue window and a spoken narration window. A LoRA-adapted VLM then generates concise ADs from the predicted visual evidence, trained with a Description Ranking Loss that ranks captions (negative samples) of the same frames lower than the GT AD. To benchmark this new problem statement, we introduce the LongLSMDC benchmark with up to 8-min movie clips (~6.5min on average). On LongLSMDC, Cue2Narrate outperforms video-only and audio-only localization baselines by 5–12 points in avg. mAP. Under both predicted- and GT-window evaluation, Cue2Narrate improves AD generation over the corresponding fine-tuned base VLM. These results establish the first benchmark for multi-segment AD generation on long-form clips. Data Code: this https URL
[CV-120] FairLens: Benchmarking Fairness in Vision-Language Models for High-Stakes Decision-Making
链接: https://arxiv.org/abs/2609.01691
作者: Vahid Reza Khazaie,Ahmed Y. Radwan,Shaina Raza
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: Code and benchmark resources are available at this https URL
Abstract:Vision-language models (VLMs) are increasingly used to make decisions from visual inputs. We introduce FAIRLENS, a benchmark and evaluation framework for measuring both the fairness and the validity of VLM responses in three high-stakes domains: hiring, legal, and healthcare. FAIRLENS pairs real face images spanning gender, race, and age groups with closed- and open-ended questions, giving more than 100K image-question pairs per model, and evaluates responses from four complementary views: demographic parity over adverse outcome rates, soundness, demographic association over unsupported roles and statuses, and bias in free-text generation. Soundness is the central validity criterion: a response is sound when it follows the evidence stated in the question and abstains when the image cannot support an answer. Evaluating eight VLMs, we find that the primary failure is unwarranted inference rather than unequal treatment. Models routinely infer qualifications, threat, illness, or professional role from a face instead of abstaining, and the weakest model does so on 99% of the questions its input cannot answer. These failures are most severe in legal and healthcare, where recognizing insufficient evidence matters most, and disparity metrics alone would miss them: parity gaps are small in absolute terms, yet when baseline adverse rates are low the same gap means one demographic group receives adverse labels several times as often as another, and a small gap can equally reflect a model that treats every group unsafely. Bias in free-text responses is only loosely coupled to multiple-choice accuracy, so correct structured answers do not imply safe generation. FAIRLENS shows that fair high-stakes VLM behavior requires similar treatment across groups and refusal to infer high-stakes attributes from appearance, and its question suite transfers to any face corpus with demographic annotations.
[CV-121] FORGE: Forward-Only Test-Time Adaptation for Integer-Only Vision Models on Microcontrollers
链接: https://arxiv.org/abs/2609.01683
作者: Muhammad Rehan,Haider Ali,Muhammad Ali Munir,Moaz Amjad
类目: Computer Vision and Pattern Recognition (cs.CV); Hardware Architecture (cs.AR); Machine Learning (cs.LG)
备注: 16 pages, 5 figures, 9 tables. Published in Transactions on Machine Learning Research (2026). OpenReview: this https URL . Code and checkpoints: this https URL
Abstract:Vision models deployed on microcontrollers (MCUs) are quantized to integer-only arithmetic and run in inference-only runtimes that do not carry the machinery backpropagation needs: the standard tool for adapting a model to the distribution shift (sensor noise, blur, lighting) it meets in the field. Existing forward-only test-time adaptation (TTA) methods either run only on server- or edge-GPU-class models (not true microcontroller integer execution), or require the batch-normalization (BN) layers that integer deployment fuses away. We present a forward-only TTA method that operates on deployed, BN-folded, integer-only convolutional networks. The key observation is that fusing BN into the preceding convolution, a mandatory step for integer inference, destroys the statistics that normalization-based adaptation relies on. We restore adaptation by re-normalizing each folded convolution’s per-channel output to its clean training statistics, using only forward-pass estimates. The method (i) recovers most of gradient-based TENT’s accuracy gain (+20.9 vs. +24.9 points) and matches forward-only BN adaptation, while being the only method that runs on a folded integer-only model; (ii) needs to adapt only 3 of 21 layers (selected without seeing the test corruptions) to recover 93% of the benefit; (iii) survives single-sample streaming with a batch-size-scaled momentum; and (iv) generalizes across three datasets (up to 200 classes) and two architectures. We validate bit-exact int8 convolution execution and deploy on an ESP32-S3, where, measured with a Nordic PPK2 power profiler, the forward-only adaptation (a lightweight fp32 recalibration around the int8 convolutions) costs only 8.3 mJ (6.8% of inference energy) and 21.9 ms on the deployed SIMD-optimized model: forward-only adaptation is cheap on a real microcontroller.
[CV-122] LoFi RADIO: A Distilled In-Domain Backbone Applied for Artifact-Severity Grading of Ultra-Low-Field Neonatal Brain MR MICCAI2026
链接: https://arxiv.org/abs/2609.02676
作者: Jonathan B. Martin,Yashwant Kurmi,Charlotte R. Sappo
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
备注: 11 pages, 2 figures, MICCAI 2026 satellite
Abstract:Ultra-low-field MRI makes neonatal brain imaging deploy- able in low-resource settings, but its low SNR, lack of shielding, and long scan duration make it especially prone to acquisition artifacts, motivating automated quality control. We address the LISA 2026 Task 1a challenge: multi-label severity grading (0/1/2) of seven common image artifacts on ULF T2 weighted volumes. We identify that a number of backbones may be successfully paired with a classification MLP, but that no single backbone is uniformly best across artifacts. To improve performance, we evaluate routing complementary foundation model teachers through a per-artifact gate, as well as distilling the teachers into a single in-domain ViT-S student (LoFi RADIO) over an unlabeled low-field MRI corpus. Both of these strategies improve the weighted composite. The distilled backbone matches or exceeds the gate and has the added advantage of not requiring deployment of multiple large foundation models at infer- ence.
[CV-123] Seeing Beyond the Lesion: Disease Recognition from Reactive CNS Tissue
链接: https://arxiv.org/abs/2609.02390
作者: Jan Schnorrenberg,Jan Ernsting,Enrico Küllenberg,Tim Hahn,Benjamin Risse,Christian Thomas
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV); Tissues and Organs (q-bio.TO)
备注:
Abstract:Sampling error yields exclusively reactive, non-lesional brain parenchyma in a significant proportion of intracranial biopsies, leaving the underlying disease undiagnosed. We benchmark four pathology foundation models (UNI2-h, Virchow2, Prov-GigaPath, H-optimus-0) as frozen patch encoders within a shared attention-based multiple-instance learning framework using 245 whole-slide images from 186 patients with confirmed downstream diagnoses. We first show that coarse disease-category prediction can be reproduced largely from slide size alone. After restricting classification to three finer diagnostic distinctions within common tissue categories, this confound no longer explains performance, yet disease labels remain predictable above chance under permutation testing (p \le 10^-4 throughout). Surprisingly, performance is statistically indistinguishable across all foundation-model encoders, suggesting that recovering these weak morphological signatures is not limited by current patch representations. Signed instance-contribution maps and expert review further test whether predictive evidence localizes to reactive parenchyma rather than sampling-induced bias like blood introduced during tissue sampling. These results position acquisition-shortcut auditing via a provenance-only baseline as a necessary control in computational-pathology benchmarks, and show, once that confound is removed, that weakly supervised models still recover disease signal from tissue conventionally regarded as non-diagnostic.
[CV-124] Perceptually Regularized Diffusion Model for Image Super-Resolution
链接: https://arxiv.org/abs/2609.02016
作者: Chuxiangbo Wang,Pavithra Venkatachalapathy,Ying Liang,Min Wang,Jing Qin,Yifei Lou,Weihong Guo
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Image super-resolution, which aims to reconstruct high-resolution images from their low-resolution observations, is fundamental to medical imaging, remote sensing, surveillance, microscopy, and scientific visualization. Traditional model-based methods formulate super-resolution as an inverse problem with hand-crafted regularization priors. While interpretable and theoretically grounded, they rely on fixed assumptions and require computationally intensive iterative solvers. Deep learning methods offer data-driven flexibility by learning nonlinear mappings from low- to high-resolution images, among which diffusion models have achieved particularly impressive perceptual quality. However, the standard diffusion training objective is a pixel-domain noise-prediction loss that does not explicitly enforce perceptual fidelity, which can lead to oversmoothing and loss of fine image structure. To address these limitations, we propose a perceptually regularized diffusion framework that incorporates prior knowledge through perceptual-loss-based regularization, improving training convergence and encouraging the recovery of meaningful image features. Experiments on benchmark datasets demonstrate improved perceptual quality and competitive distortion metrics, highlighting the effectiveness of regularization for diffusion-based super resolution.
[CV-125] Data-Efficient Networks for Multi-Contrast MRI Reconstruction based on a Generalized Content/Style Prior
链接: https://arxiv.org/abs/2609.01959
作者: Chinmay Rao,Efe Ilıcak,Matthias J.P. van Osch,Mariya Doneva,Laurens Beljaards,Navid Jabarimani,Nicola Pezzotti,Marius Staring
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Multi-contrast MR scans contain redundant structural information that can be leveraged during reconstruction and potentially accelerate acquisition times. This idea has inspired end-to-end guided reconstruction models, leveraging one or more contrasts to guide the reconstruction of a different contrast. However, these models require large paired multi-contrast raw datasets for training, limiting their application in low-data regimes. In this work, we propose a modular framework, namely CoSMo-RecNet, for learning guided reconstruction models in the low-data regime. At its core is a reusable multi-contrast representation based on a content/style model, which can be learned from large-scale, publicly accessible, unpaired multi-contrast image datasets, without available k-space data. Using this frozen model as a multi-contrast prior and using a set of reference contrasts, the reconstruction problem reduces to a much simpler refinement problem that can be solved by a lightweight unrolled network and thus learned from small, task-specific reconstruction datasets. We demonstrate the efficacy of CoSMo-RecNet by evaluating it on the low-field 0.3 T M4Raw dataset, showing stable reconstruction quality on decreasing the raw training data budget. CoSMo-RecNet achieved higher reconstruction quality with 5 training subjects or lower compared to a parameter-count-matched MoDL trained on 100 subjects. On a data-limited and severely out-of-distribution ultra-low-field 47 mT Halbach scanner dataset, CoSMo-RecNet was superior to other viable strategies, including classical reconstruction, transfer learning, and zero-shot reconstruction.
人工智能
[AI-0] Discriminative World Models for Web Agents
链接: https://arxiv.org/abs/2609.02885
作者: Kelvin Li,Dhruv Pendharkar,Anish Pahilajani,Chuyi Shang,Leon Oks,Leonid Karlinsky,Rogerio Feris,Trevor Darrell,Roei Herzig
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Recent web agents use world models for test-time action selection by sampling candidate actions, predicting the resulting web states, and ranking them with a ranker model or a Process Reward Model (PRM). These world models are typically trained via supervised next-state prediction to generate fixed representations like HTML or AXTree snapshots. However, this objective is misaligned with the downstream ranker, which relies on predicted states being discriminative across candidates to accurately score them. To address this, we introduce predicted-state matching, a training objective where the predicted representation must distinguish the true resulting state from those reached by alternative actions. We train these models using a branching web-agent dataset derived from WebArena Go-Browse trajectories, where every decision point contains multiple alternative actions and their resulting states. Experiments on our held-out predicted-state matching benchmark show that our approach outperforms world models trained with supervised next-state prediction. We further show that our approach improves PRM-style action ranking on WebPRMBench compared with action-only PRMs and PRMs augmented with supervised-next-state world models. Finally, on WebArena-Lite, using our world model for test-time action selection improves end-to-end task success. Our project page is available at: this https URL.
[AI-1] owards Trustworthy Autonomous Robots: An Explainable AI-Based Decision Framework
链接: https://arxiv.org/abs/2609.02861
作者: Cagri Temel
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: 7 pages. Accepted version. Published in SoutheastCon 2026, IEEE, pp. 1-6
Abstract:Autonomous robots powered by deep learning face a fundamental auditability challenge: when incidents occur, investigators cannot reconstruct why the system made specific decisions. This paper presents TRACE (Transparent Reasoning Architecture for Credible Execution), a decision framework that ensures every autonomous action can be traced back to sensor evidence through documented causal chains. The framework organizes decision-making into four auditable layers: Semantic Perception for evidence-grounded entity recognition, Belief Reasoning for probabilistic state estimation with causal graphs, Action Synthesis for constraint-aware planning with counterfactual documentation, and Execution Verification for compliance monitoring. TRACE is model-agnostic yet designed to integrate learning-based perception modules (CNNs, transformers) while preserving decision-level auditability. We evaluate the framework using three objective metrics: Evidence Traceability (sensor-to-decision linkage), Decision Reconstructability (post-hoc analysis capability), and Temporal Continuity (audit trail completeness). Experimental evaluation on warehouse robot navigation demonstrates that TRACE achieves 98.6% evidence traceability, 99.0% temporal continuity, and 98.1% decision reconstructability across 500 simulated decision cycles. Post-hoc methods like LIME provide feature attributions but lack the artifact structure needed for decision-level reconstruction. The framework addresses EU AI Act requirements for high-risk system transparency and contributes to Explainable AI for safety-critical autonomous systems.
[AI-2] AI Contextual Measurement for Recovering Individual and Group-Level Effects: Validation Against Survey Measures and an Occupational Application
链接: https://arxiv.org/abs/2609.02821
作者: Wenxin Jiang,Xuyang Wang,Yuxiao Wu
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Researchers increasingly use artificial intelligence to construct measures of social, organizational, and occupational characteristics that are absent from conventional surveys. We propose AICOME, AI COntextual MEasurement, a framework for evaluating whether AI-derived respondent-level measures can recover individual and group-level effects in contextual models. The key idea is that an AI measure constructed at the respondent level can be used to derive its group-level aggregate and its individual deviation, allowing researchers to estimate both between-group and within-group associations rather than treating AI measurement as response prediction alone. We validate the framework using the 2022 China Family Panel Studies (CFPS), where occupations provide the empirical grouping structure and several job-related survey variables provide validation benchmarks. For computer use, foreign-language use, weekly hours, and management responsibilities, we compare survey measures with AI-derived measures in response-level, model-level, contextual, and boundary-condition validations. The results show that AI contextual measurement can recover much of the contextual-model information contained in observed survey variables when rich respondent and job characteristics are available. Weekly hours provides the strongest validation case, with AI-derived measures reproducing the large negative between- and within-occupation associations with satisfaction observed in CFPS. The framework also identifies clear boundary conditions: performance deteriorates when information is restricted to occupation and basic demographics, and recovery is weaker when several related concepts are treated as simultaneously unobserved. The findings suggest that AICOME is most useful for recovering a limited number of theoretically important constructs from rich existing datasets. Subjects: Artificial Intelligence (cs.AI); Machine Learning (cs.LG) Cite as: arXiv:2609.02821 [cs.AI] (or arXiv:2609.02821v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2609.02821 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-3] Large Language Models (LLM s) for Telecom Root Cause Analysis (RCA): A Structured Reasoning Framework for Evidence-Grounded Diagnosis
链接: https://arxiv.org/abs/2609.02805
作者: Hao Zhou,Mandar Kulkarni,Hao Chen,Yan Xin,Charlie(Jianzhong)Zhang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Root cause analysis (RCA) is a critical task in telecom network operations, but diagnosing performance degradations in modern 5G and emerging 6G networks remains challenging due to complex cross-layer dependencies. While large language models (LLMs) offer promising capabilities for reasoning and knowledge integration, directly applying vanilla LLMs to telecom RCA often leads to hallucination, unstable reasoning, and poor alignment with structured network evidence. This work first reviews the evolution of telecom RCA from rule-based and machine learning (ML) approaches to emerging LLM-enabled techniques, and provides an overview of recent paradigms, including structured reasoning, retrieval-augmented knowledge grounding, agentic orchestration, and verifiable reasoning. Building upon these insights, we propose a structured reasoning framework for LLM-enabled telecom RCA that aligns diagnostic reasoning with telecom-specific evidence and domain knowledge. The proposed approach first organizes heterogeneous network telemetry into canonical contexts, and then enforces decision-path reasoning during diagnosis, and finally generates evidence-grounded explanations for reliable fault identification. Experimental results on two 5G RCA datasets, TeleLogs and TelecomTS, demonstrate that the proposed framework consistently improves diagnostic accuracy and decision consistency compared with baseline techniques. These cross-dataset results highlight the importance of structured reasoning design for practical LLM-based RCA systems in next-generation telecom networks.
[AI-4] frb100-40 After Two Decades: An Optimality Certificate and a Preregistered Search Study
链接: https://arxiv.org/abs/2609.02804
作者: Onur Uğurlu(İzmir Bakırçay University)
类目: Discrete Mathematics (cs.DM); Artificial Intelligence (cs.AI)
备注: 7 pages, 4 figures, 4 tables. Reproducibility artifact: this https URL
Abstract:For more than 20 years, the Model-RB benchmark frb100-40 remained an open challenge; since 2014, its public record had stood at 99 of 100 variables. We give a directly checkable 100-vertex independent set for its 4,000-vertex graph. Together with a verified partition into 100 cliques of size 40, the witness proves that the maximum independent-set size is 100 and the minimum vertex-cover size is 3,900. The stochastic run that found the witness is kept separate from this proof. We evaluated its added pair and triple repair operators in a preregistered campaign comprising 8,668 valid runs. The primary comparison found no detectable acceleration over base ULSA (hazard ratio 0.967, 95% confidence interval 0.915-1.023; p=0.248), and the factorial ablation reached the same conclusion. On a smaller FRB suite, the group-aware CSP pipeline solved 2,500/2,500 runs, compared with 2,391/2,500 for LibMVC-NuMVC. On frb100-40, full ULSA, base ULSA, and NuMVC each produced 0/56 new certificates. With no events, the planned cross-solver hazard ratios remain unidentified. NuMVC ended with cover size 3,902 in 40 runs and 3,903 in 16. Exhaustive enumeration showed that none of the 108 unique recorded conflict-two states had a strictly improving group-aware CSP neighbor within Hamming radius three. The certificate settles the instance. The experiments characterize the search barrier, and the preregistered comparisons show no heuristic advantage.
[AI-5] SafeEvolve: Harness-Policy Co-Evolution from Agent Experience for Safety Alignment
链接: https://arxiv.org/abs/2609.02786
作者: Qinghua Mao,Wanying Qu,Dadi Guo,Leitao Yuan,Qingyu Liu,Yu Li,Guanxu Chen,Yanwei Fu,Xi Lin,Xia Hu,Dongrui Liu
类目: Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
备注: Project: this https URL
Abstract:The performance of LLM-based agents is jointly shaped by the base model and the harness used when interacting with the environment. This exposes them to safety risks in both harmful final responses and multi-step execution trajectories. Existing safety alignment mechanisms often rely on either external harness updates or policy optimization, yet applying either paradigm in isolation fails to bridge runtime control with intrinsic safety. We propose SafeEvolve, an experience-driven self-evolving framework for agent safety alignment. SafeEvolve leverages safety experience from completed on-policy trajectories to drive a continual loop of harness-policy co-evolution. On the harness side, SafeEvolve converts trajectory-level safety evidence into bounded, component-level updates across safety prompt and hierarchical skills, yielding auditable and reversible harness artifacts. On the policy side, SafeEvolve follows a two-stage SFT-RL paradigm, where harness-use SFT bootstraps the policy to actively leverage evolved harness artifacts, and harness-augmented RL further shapes autonomous safety behaviors during multi-step exploration via verifier-decomposed rewards. Through harness-policy co-evolution, SafeEvolve converts safety experience into an evolved runtime harness and improved policy behavior. Experiments on agentic safety benchmarks show that SafeEvolve achieves a stronger safety-utility tradeoff than existing baselines. For Qwen3.5-4B, SafeEvolve achieves a 3\times ASR reduction on AgentDojo while improving benign utility from 59.79% to 61.86%.
[AI-6] Measurement-Driven Sub-Network Selection for On-Premise Retrieval-Augmented Factory Agents
链接: https://arxiv.org/abs/2609.02760
作者: Vasileios Rizeakos,Georgios Paisios,Alexandros Machairas,Michael Birbas,Athanasios Bachoumis
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:On-premise assistants can give factory workers conversational access to machine documentation, but models capable of the task rarely fit shop-floor hardware. We show that after structural compression and retrieval-grounded adaptation, model size is no longer a reliable predictor of adapted answer quality: general capability falls almost linearly with parameter count, while judged retrieval-augmented answer quality does not. We therefore treat deployment as a post-adaptation selection problem, committing one sub-network per device on judged answer quality and measured on-device throughput under a configurable general-capability floor and memory budget; rules that optimize size, speed, or quality alone each give up capability or throughput. A weight-shared supernetwork trained with sandwich-style in-place distillation keeps this selection inexpensive. In a manufacturing-manual case study, extraction costs 13.7 percent of the unpruned model’s judged quality and retrieval-grounded distillation returns it to within 4.6 percent, recovering two thirds of the loss, and the same assistant runs across three heterogeneous edge tiers at 1.3 to 5 watts standby.
[AI-7] Bilevel Coordinated Reflection: A Game-Theoretic Approach to Multi-Agent LLM Systems
链接: https://arxiv.org/abs/2609.02750
作者: Yihang Chen,Yuxiang Chen,Yuxuan Huang,Meng Fang,Weilin Luo,Jun Wang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Multi-agent LLM systems commonly use an orchestrator to decompose a task for a team of workers and then improve through textual reflection. Despite strong empirical results, these systems lack a unified account of coordination, memory improvement, and the role of external verification. We model orchestrator-worker interaction as a bilevel coordination game: under bounded coupling, the workers’ local-update game is an approximate potential game whose equilibrium slack is controlled by decomposition quality. We then analyse reflection as stochastic movement over semantic memory states. For free-form reflection, we derive a finite-time upper bound, prove worst-case tightness, and give a positive lower bound under a falsifiable persistent-harm condition. We further prove an information-theoretic impossibility result: no gate that observes only the generated transcript can improve uniformly over text-indistinguishable environments, whereas an environment-grounded gate can. Motivated by this separation, we introduce Stochastic Reflective Memory Ascent (SRMA), which accepts a candidate memory only after a grounded evaluation risk strictly decreases. Under calibration and non-degenerate corrective mass, SRMA converges exactly, geometrically or polynomially; matching constructions show that both rate regimes are order-tight. We also provide confidence gating for stochastic evaluation and re-anchoring guarantees for piecewise-stationary environments. Experiments instantiate these objects with environment-grounded metrics and test the predicted coordination and drift laws. On 500 SWE-bench instances, the complete Kimi-based system resolves 72.2% versus a 70.8% public mini-SWE-agent reference. Code: this https URL
[AI-8] Automated Vulnerability Injection in Smart Contracts Using Large Language Models
链接: https://arxiv.org/abs/2609.02624
作者: Luca Migliaccio,Roberto Natella,Naghmeh Ivaki,Nuno Laranjeiro,Marco Vieira
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
备注:
Abstract:Assessing vulnerability detection tools for smart contracts requires datasets with known ground truth, yet such datasets are scarce and difficult to build by hand. We propose an approach that uses Large Language Models (LLMs) to automatically inject vulnerabilities into Solidity smart contracts, and demonstrate it in a case study targeting 49 vulnerability types from OpenSCV. Injected contracts are validated through a multi-step pipeline checking compilation, execution, business logic, and the presence of the intended vulnerability. Applied to real-world contracts from SmartBugs, LLMs generate nearly 1,000 candidate variants; after deduplication and validation, 32 confirmed vulnerable contracts spanning 25 vulnerability types survive (a 16.58% survival rate). Surviving contracts concentrate in structurally simpler targets and vulnerability types with localized syntactic patterns. We report practical challenges including LLMs’ non-determinism and the difficulty of preserving contract semantics. We then use the validated contracts to assess three static analyzers, revealing complementary and incomplete coverage profiles. Results show that LLM-based vulnerability injection is feasible, while exposing key limitations in scalability and diversity.
[AI-9] ProbeMatchDTI: Probe-Driven Multi-Scale Biochemical Pattern Matching for Drug-Target Interaction Prediction
链接: https://arxiv.org/abs/2609.02549
作者: Quan Hao,Mengyue Fan,Zifan Dong,Youru Li,Jianduo Zhao,Lechuan Xu,Hao Zhang,Fei Xia,Jigang Wang,Chong Qiu,Liguo Zhang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Drug-target interaction (DTI) prediction is an important task in AI-driven drug discovery. Although recent biochemical representation learning methods have improved DTI prediction, their passive feature aggregation tends to favor dominant molecular patterns while suppressing weak yet binding-relevant signals, such as functional groups and residue-context patterns, limiting the modeling of multi-scale biochemical correspondences. To address this issue, we propose ProbeMatchDTI, a pattern-probe-driven framework comprising IterProbe and BindingProbe. IterProbe explicitly retains contextual states across refinement depths and uses learnable probes to select them at each position before cross-entity matching, thereby preserving weak biochemical patterns and strengthening associations among functional groups, local motifs, and molecular scaffolds. BindingProbe then characterizes cross-entity drug-protein complementarity at local biochemical-unit and whole-pair levels, jointly modeling fine-grained interactions and multi-scale correspondences while preserving weaker binding-relevant associations. Extensive experiments demonstrate the superiority of ProbeMatchDTI, achieving 2.0% and 0.5% higher AUC-ROC on BindingDB and DrugBank, respectively. Feature-level pattern analyses further characterize its probe-driven behavior in cross-scale biochemical pattern matching. We further connect ProbeMatchDTI predictions with an evidence-guided downstream drug-discovery workflow, demonstrating their utility for candidate refinement and validation planning. Our code is available at this https URL
[AI-10] Learn from Whoever Is Right: Answer-Verified Multi-Teacher Distillation for Multi-Domain LLM s
链接: https://arxiv.org/abs/2609.02548
作者: Xixiang He,Xingming Li,Baiqi Wu,Qiyao Sun,Xuanyu Ji,Ao Cheng,Qingyong Hu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Modern large language models (LLMs) rely on reinforcement learning to build strong capabilities in individual domains, but integrating those capabilities into a single deployable model remains challenging. By routing each sample to the teacher whose domain matches it, existing approaches let a domain label decide which teacher provides supervision. However, domain expertise holds only on average: the matched teacher is not always correct on a given sample, while a teacher from another domain sometimes is. The reliable teacher therefore has to be identified per sample, not per domain. In this paper, we introduce Multi-Teacher Self-Distillation Policy Optimization (MT-SDPO), an on-policy distillation method that unifies several frozen teachers into one student model. MT-SDPO consists of three components: (1) self-anchors, where a rollout is supervised by a correct rollout from its own group; (2) answer-verified eligibility, where a teacher may supervise a sample only if its own answer passes a verifier; and (3) privileged distillation, which merges the anchor and all verified feedback into one context that an exponential moving average self-teacher reads and the student does not, thereby keeping one policy at deployment. Across five students from three model families, MT-SDPO lifts the weakest domain of Qwen3-8B by 14.79 points and narrows its domain gap by 74.7%, a better balance than serving one matched teacher per domain. Verified reliability, not domain membership, should decide who teaches. Code is available at this https URL.
[AI-11] Spectral Initialization and Scheduled Graph Smoothness for Uncertain Knowledge Graph Completion
链接: https://arxiv.org/abs/2609.02519
作者: Md Abrar Jahin,Taufikur Rahman Fuad,Jay Pujara,Craig A. Knoblock
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Uncertain knowledge graphs (UKGs) extend knowledge graphs by assigning each triple a continuous confidence score. Since most possible triples lack observed confidences, recent methods rely on semi-supervised learning to generate pseudo-labels. These methods initialize entity embeddings without using the confidence-weighted graph, discarding its global community and hub structure. We introduce QUEST, which adds no trainable parameters to the standard confidence-distribution learning pipeline. First, QUEST initializes entity embeddings using the smallest non-trivial eigenvectors of the confidence-weighted graph Laplacian, incorporating community and hub structure before training. Second, QUEST applies an unbiased mini-batch Dirichlet energy regularizer to enforce early-stage structural consistency. On two UKG datasets, QUEST improves confidence prediction and link prediction on six of eight metric-dataset pairs over prior methods and matches the previous best on the remaining two, while removing the instability spike observed on dense graphs. These results indicate that spectral structural priors combined with a graph Dirichlet energy regularizer improve accuracy, training stability, and checkpoint reliability in UKG completion.
[AI-12] RINSE: Robust Target-Time Normality Estimation for Zero-Shot Graph Anomaly Detection
链接: https://arxiv.org/abs/2609.02497
作者: Taufikur Rahman Fuad,Md Abrar Jahin,Amir Hussain
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Zero-shot graph anomaly detection seeks to deploy a detector trained on source graphs to unseen, unlabeled targets, yet domain shift can make source-derived notions of normality unreliable. We introduce RINSE (Robust Iterative Normality Self-Estimation), a gradient-free target-time framework that keeps the source-trained detector fixed while sequentially estimating target normality, representation calibration, and evidence reliability from the target graph. Its core idea is to identify a reliable subset of low-residual target nodes, use them to construct a trimmed target-aware normality model, and combine complementary anomaly evidence through reliability-gated rank fusion and encoder ensembling. Across eight unseen target graphs, RINSE achieves the highest average AUPRC among the evaluated methods under two separate preprocessing protocols, while block ablations and sensitivity analyses support the combined design. These results support robust target-time estimation as a practical approach to generalist graph anomaly detection without target labels, gradients, or per-target tuning.
[AI-13] DeepAffinity: Long-Term Aspect Preference Prediction in eCommerce using Small Language Models
链接: https://arxiv.org/abs/2609.02468
作者: Yotam Eshel,Guy Hadad,Guy Feigenblat,Yuri M. Brovman,Matt Gearhart,Bracha Shapira
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:We explore predicting eCommerce user preferences for product aspects such as brand, size, and color - a task we define as Aspect Affinity. Solving this task improves customer understanding and enables fine-grained personalization in recommendation, search, and marketing. We frame Aspect Affinity as a temporal prediction task: forecasting a users future aspect choices from their time-ordered interaction history, capturing long-term preferences that evolve beyond the current session. To this end, we propose DeepAffinity, which leverages Small Language Models (SLMs) with structured prompts and specialized prediction heads fine-tuned for this task. We show DeepAffinity outperforms standard generative fine-tuning methods, while general-purpose open-source LLMs perform poorly without task-specific tuning, highlighting their limits in modeling nuanced behavior. Finally, DeepAffinity enhances recommendation quality on a large-scale multinational eCommerce platform.
[AI-14] CivBench: A Long-Horizon Benchmark for Tool-Mediated Agents in Civilization VI
链接: https://arxiv.org/abs/2609.02459
作者: Austin Tudor David Andrews,Liam Wilkinson,Jamie Heagerty,Harry Coppock,Jakob Nicolaus Foerster,Rui Ponte Costa
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:We present CivBench, an open-source benchmark for evaluating language model agents in long-horizon, tool-mediated environments through the Model Context Protocol (MCP). A single episode spans 300+ turns and produces thousands of tool calls over a large action space, requiring sustained planning, state monitoring, and execution under partial observability. The environment exposes 76 MCP tools and a narration layer that converts visual game state into structured text. We use CivBench to characterise agent behaviour across four model families in 23 admissible runs. The sample is a pilot, not a model ranking: aggregate outcomes do not reliably discriminate models at this scale. Instead, we introduce two interface-level metrics that the environment makes measurable: Proactive Monitoring Rate (PMR), capturing whether agents actively query latent strategic state, and RAG@10, capturing whether commitments stated in structured planning reflections are executed within ten subsequent turns. Across runs we observe two consistent patterns under a shared playbook protocol. Agents under-monitor strategically relevant state that is available but requires explicit querying: despite playbook guidance to query victory progress every 20 turns, agents do so only every 30 to 75 turns, and in 7 of 20 detectable defeats they failed to query within the 20 turn warning window before game end. Agents also frequently fail to execute near-term commitments stated in their own planning reflections (RAG@10 between 48.2% and 65.8% across models). Both patterns arise despite tool access and explicit guidance, and we interpret them as deviations under instruction rather than absences of capability. We release the environment, scenarios, logs, metrics, and analysis pipeline at this https URL Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2609.02459 [cs.AI] (or arXiv:2609.02459v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2609.02459 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Harry Coppock Dr [view email] [v1] Wed, 2 Sep 2026 11:31:39 UTC (5,132 KB)
[AI-15] Addressing Trust in AI Systems through Education: A Didactic Perspective
链接: https://arxiv.org/abs/2609.02453
作者: Pierre Haritz,Hendrik Krone,Thomas Liebig
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI)
备注:
Abstract:Machine learning (ML) education faces two persistent and connected obstacles: many educational tools present ML as an opaque black box, which leaves learners with a superficial understanding, and this same opacity prevents users from forming the calibrated trust that appropriate reliance on AI systems requires. We present ICE-T, a didactic framework that integrates three mutually reinforcing facets: intermodal transfer grounded in Bruner’s enactive, iconic, and symbolic modes of representation, computational thinking operationalized through the Use-Modify-Create progression, and explanatory thinking supported by a process model. Connecting the framework to the empirical literature on algorithm aversion, AI literacy, and mental model formation, and to systematic reviews of the K-12 ML activity landscape, we argue that the three facets supply the cognitive mechanisms that the trust calibration literature identifies as drivers of appropriate reliance: representational richness, graduated process control, and the capacity to contextualize errors. On this basis, we propose that trust calibration be treated as an explicit educational objective, with ICE-T as a principled and scalable means of achieving it.
[AI-16] owards One-for-All Robustness Across a Continuum of Threat Levels
链接: https://arxiv.org/abs/2609.02440
作者: Zhichao Hou,Xiaorui Liu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Adversarially robust models often overfit to a specific attack budget, necessitating multiple specialized models for diverse and dynamic adversarial environments, a strategy that becomes fundamentally intractable as the threat space grows. This raises an open challenge: can we achieve strong robustness across a continuum of threat levels within a single model? We propose the Threat Conditional Network (TCN), grounded in a representation factorization framework that decomposes representation learning into a threat-invariant shared backbone and a lightweight threat-conditional adaptor. TCN conditions a single model on the perturbation level via Fourier-based embeddings and channel-wise affine modulation, and is trained against a distribution over perturbation budgets, enabling flexible and seamless adaptation across an infinite continuum of threat levels during inference. Extensive experiments on CIFAR-10, CIFAR-100, and Tiny-ImageNet show that TCN matches or surpasses a full ensemble of budget-specialized models with a single set of parameters, generalizes to unseen perturbation budgets, and transfers robustly under mismatched threat conditions, with only 4.6% parameter overhead. These contributions chart a promising path toward adaptive and generalizable robustness in dynamic and diverse threat environments.
[AI-17] Coverag e Not Targeting: A Structural Regime in Multi-Turn Agent Credit Assignment
链接: https://arxiv.org/abs/2609.02417
作者: Chenyu Zhou,Qiliang Jiang,Shuning Wu,Xu Zhou
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 22 pages, 7 figures, 8 tables
Abstract:Multi-turn agentic RL increasingly treats credit assignment as a targeting problem: given a terminal verifiable reward, per-turn methods localize credit onto the turns that mattered. We identify the structural quantity that predicts when this is the right move, the verifier information density V_d = k/C (the fraction of an agent’s C-step causal chain whose per-turn correctness the verifier exposes), and show that terminal-state verifiers sit deep in a low-V_d regime where targeting is the wrong axis. In controlled shared-rollout comparisons on tau^2-bench that separate reward density from credit geometry, a continuous dense reward spread uniformly beats the sparse binary outcome reward (net-harmful on 4/5 seeds), while concentrating the same advantage on progress turns or on random turns is equally harmful: targeting is second-order. The mechanism is coverage: terminal-state verification collapses the observable signal to a single final-write turn (k=1 in 98% of rollouts) while success requires a 5-8 step chain of prerequisite tool calls. A synthetic phase boundary places the crossover at V_d* ~ 0.8, whereas measured V_d is ~0.15 on tau^2-bench and ~0.4 on BFCL V3; uniform also wins on BFCL, where a matched-concentration shuffled control is negative on 8/8 seeds. The effect reproduces across model families on ToolACE-2-8B (Delta = -0.048 over 32 pre-registered seeds; an independent 20-seed replication is itself significant), and a pre-registered matched-budget breadth sweep traces a monotone dose-response whose deficit vanishes only at full chain coverage, with a reward-to-go arm reaching full-coverage parity. Uniform redistribution is the zero-information coverage default that per-turn schemes must beat; we contribute the matched-concentration shuffled control that any targeting claim should clear.
[AI-18] Evidence for Shared Routing Geometry and Dynamics in Sparse Mixture-of-Experts
链接: https://arxiv.org/abs/2609.02404
作者: Kirill Labzin,Stepan Kulibaba,Artem Dzhalilov,Artem Gorokhov
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Sparse mixture-of-experts (MoE) models use an independently parameterized router at each sparse layer to select experts for every token. Prior work has shown that routing decisions across depth can often be predicted from earlier routing signals, suggesting that routing is not fully independent across layers. However, the structure behind this predictability remains unclear. In this work, we provide evidence that routing-relevant states across layers share a common geometric structure that is obscured by layer-specific coordinate systems. We isolate the control subspace of each router and align these spaces into a shared canonical representation using generalized orthogonal Procrustes analysis. After alignment, a single linear transition reaches R^2=0.39 – 0.71 and retains 79–90% of the predictive power of separately fitted layer-specific dynamics, indicating that much of routing-state evolution follows a reusable process across depth. We then ask whether this shared dynamics is specific to routing or simply reflects the smooth evolution of hidden representations. A matched-rank comparison shows that residual representations are often easier to predict across layers, while router-control states preserve the model’s expert choices much more faithfully. This separates generic cross-layer predictability from routing-specific information. Finally, we test whether the predicted canonical states remain meaningful when used in place of native routing states. The transported states preserve local routing behavior, while learned state evolution reduces \Delta\mathrmNLL relative to simple persistence by 15.7% on OLMoE and 6.2% over a 10-router horizon on Phi.
[AI-19] Contrastive Explanations in Quantitative Bipolar Argumentation Frameworks
链接: https://arxiv.org/abs/2609.02399
作者: Xiang Yin,Nico Potyka,Antonio Rago,Francesca Toni
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Argumentation frameworks are useful tools for representing and reasoning with information in a variety of settings, e.g. in supplementing AI models as they perform classification tasks, with a notable benefit of providing additional explainability. In this paper, we introduce contrastive explanations for Quantitative Bipolar Argumentation Frameworks (QBAFs), one such formalism. Unlike most existing explanations for QBAFs, which explain the reasoning outcome of a single argument of interest (i.e. a topic argument), contrastive explanations explain the difference between two topic arguments. We introduce a general form of contrastive attribution functions (CAFs) and establish a set of general properties they should satisfy. We introduce CAFs based on removal, gradients and Shapley-values, and study their properties. Finally, to illustrate contrastive explanations, we demonstrate their usefulness in healthcare and bias identification settings.
[AI-20] Percolation Dynamics in Optimization : Variance Cascades and Discrete Scale Invariance
链接: https://arxiv.org/abs/2609.02373
作者: Sai Niranjan Ramachandran,Suvrit Sra
类目: Machine Learning (cs.LG); Disordered Systems and Neural Networks (cond-mat.dis-nn); Statistical Mechanics (cond-mat.stat-mech); Artificial Intelligence (cs.AI)
备注: 43 pages, 9 in the main text
Abstract:We study the dynamics of Stochastic Gradient Descent (SGD), which is known to steer deep neural networks toward invariant sets that correspond to simpler subnetworks. How this steering unfolds over time remains poorly understood. We answer this by modeling the stochastic gradient flow (SGF) as a percolation process, in which architectural symmetries force subnetworks to merge in discrete simultaneous blocks rather than one at a time. These structural transitions register as variance spikes in a macroscopic order parameter, echoing physical phase transitions. We further show this trapping mechanism and its associated scaling cascade extend to Adam and AdamW under an explicit heavy-tailed noise model.
[AI-21] Diagnosing with Insights: Structured Analysis of Agent Failures via Behavioral Abstractions
链接: https://arxiv.org/abs/2609.02371
作者: Jiayi Bi,Yanjie Gao,Yuanmin Xie,Liqun Li,Tianyin Xu,Fan Yang,Mao Yang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:With the proliferation of LLM agents, the ability to understand and diagnose failures in agents is essential to achieving superior effectiveness and trustworthiness. As agent failures often manifest via long and complex trajectories, manually finding the needles in the haystack is untenable. However, traditional diagnosis techniques for software bugs can hardly address LLM agent failures, while completely relying on LLMs as the judge yields unreliable diagnosis results. To overcome these challenges, this paper presents AGENTSCOPE, a new neuro-symbolic approach for agent failure mode diagnosis. The key principle of AGENTSCOPE is to abstract agent behavior, based on its trajectories, into structured representations. Furthermore, AGENTSCOPE introduces the concept of neural invariants to specify agent behavior properties. AGENTSCOPE leverages LLM-guided reasoning atop the structured representation against neural invariants to pinpoint both the failure step and its type in the trajectory. We show the effectiveness of AGENTSCOPE on publicly available agent failure datasets (WhoWhen) and a more comprehensive dataset created by us (AgentErrata), where AGENTSCOPE significantly outperforms the current state of the art in fault localization and attribution accuracy. Our work shows that integrating structured abstractions with LLM-guided reasoning enables effective, reliable, and interpretable diagnosis for agent failures.
[AI-22] Fair Stable Matching: A Nash Social Welfare Approach
链接: https://arxiv.org/abs/2609.02354
作者: Parth Desai,Rasheed M,Ganesh Ghalme,Sujit Gujar
类目: Computer Science and Game Theory (cs.GT); Artificial Intelligence (cs.AI)
备注:
Abstract:While traditional stable matching algorithms, such as the Gale-Shapley algorithm, prioritize stability, they may fall short of achieving equitable outcomes among participants. We study the role of \emphNash social welfare (NSW) as a fairness objective in the classic \emphstable marriage problem. We develop \textttSNSW-Alg that finds a stable matching that maximizes Nash social welfare under rank-induced utilities in \tilde\mathcalO(n^4) time, where n is the number of men or women. We demonstrate that \textttSNSW-Alg balances equity while preserving stability. We empirically evaluate our methods across diverse preference distributions, demonstrating significant gains in fairness without substantial losses in other key measures such as regret, egalitarian criterion, and sex equality. Our findings suggest that the stable matching produced by \textttSNSW-Alg is statistically Pareto-undominated by stable matchings based on other fairness measures - regret, egalitarian, and sex equality. This study offers compelling insights for designing fair-stable matching.
[AI-23] AGI Maze Prediction Datasets: A Compact Benchmark for Learning World Dynamics with Transformers
链接: https://arxiv.org/abs/2609.02339
作者: Alexey Potapov
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:World modeling requires a predictive model to maintain and update an internal state adequate for reasoning about the consequences of actions. We introduce the AGI Maze Prediction Datasets and Benchmark, a lightweight controlled testbed for studying this capability in Transformers and other predictive models. Derived from procedurally generated, stateful grid worlds, the benchmark comprises per-step transition prediction, fixed-horizon state prediction, and sequential textual-observation prediction. Source-maze-disjoint training and validation splits, together with greedy exact-match evaluation, distinguish learning transferable action-conditioned dynamics from memorizing transitions in familiar layouts. We establish from-scratch byte-level Transformer baselines and compare them with two working-memory-augmented architectures. A generic auxiliary latent-memory Transformer can fit some training sets perfectly but does not consistently improve held-out performance. In contrast, a pseudo-video spatial-memory Transformer initializes a two-dimensional latent workspace from the input map and updates it from action history without receiving intermediate maps, positions, or state labels. Under the same data, objectives, and evaluation protocol, this model reaches perfect validation accuracy on selected fixed-horizon tasks where the byte and unstructured-memory baselines do not, and substantially improves sequential text-trace prediction. These results suggest that structured, task-aligned working memory can be more useful than additional latent capacity alone. More broadly, we argue that language grounding is mediated by persistent data structures and computations over them; the benchmark offers a compact setting for testing architectures that couple textual interfaces to learned structured state.
[AI-24] What Is Worth Representing? Representational Empowerment for Continual Model Construction
链接: https://arxiv.org/abs/2609.02322
作者: Fei Dai,Hanqi Zhou,Alison Gopnik,Charley Wu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:The first problem of modeling the world is not just estimating the right parameters or causal structure, but deciding what should be represented at all. We frame this problem as continual model construction: an agent maintains an environment-specific model M of an inaccessible world W and curates a persistent library L of reusable representational elements across environments. We propose Representational Empowerment (RepEmp) to score candidate elements by how much they expand the agent’s future capacity to model and plan, complementing the classic definition of empowerment, but redefined as control over internal representations instead of external states. We realize the framework as a hierarchical Curator-Actor architecture and test it across three experiments. In a closed-vocabulary causal-learning task, human participants construct causal models at varying abstraction granularities to maximize goal reachability rather than fidelity to the world, a signature better predicted by RepEmp than by information-gain alternatives. Matched simulations reveal that RepEmp-guided construction contributes more than exploration to sufficient structure recovery and cross-task transfer. Finally, in an open-vocabulary planning domain, an LLM-augmented Curator builds more compact symbolic libraries, which also generalize better than baselines. Ablating RepEmp eliminates these benefits. Together, these results identify RepEmp as a key principle for continual model construction: deciding what to build, retain, and reuse under bounded resources.
[AI-25] SEAL: Reinforcing Global Safety in Mixture-of-Experts through Shared Expert ALignment CCS2026
链接: https://arxiv.org/abs/2609.02293
作者: Qingyu Meng,Yiwei Zha,Jiahuan Pei,Koen Hindriks,Herbert Bos,Min Chen
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
备注: Accepted at ACM CCS 2026
Abstract:Mixture-of-Experts (MoE) is a scaling architecture for large language models that activates only a small subset of expert modules per token, enabling massive parameter growth with nearly constant computation. Recent Hybrid MoE architecture adds \textitshared experts to capture consistently useful representations, further improving stability and generalization. MoE now powers many flagship open-source and commercial models, yet remains vulnerable to adversarial attacks. Specifically, sparse routing introduces a structural vulnerability: MoE safety hinges on which experts are activated, and adversaries can subvert this selection through jailbreak prompts, malicious fine-tuning, and weight-level pruning of safety-critical neurons. Existing defenses primarily focus on hardening the router, but an adversary may still manipulate or bypass the routing trajectory due to the routing process’s nondeterministic nature, thereby collapsing the defense. To cope with this problem, we first identify theoretically and empirically that shared expert, an always-activated component containing a small proportion of safety-critical neurons, can overcome the uncertainty of sparsely activated routing path and serve as a router-independent anchor to enhance global safety alignment. Based on this insight, we propose SEAL, a training-time parameter-efficient defense that produces a plug-and-play adapter attached to shared expert, and SEAL++, a variant that adds an orthogonal constraint preserving pre-existing safety subspaces during training. We evaluate SEAL and SEAL++ across six attack scenarios that combine three adversarial inputs (harmful prompting, jailbreak, malicious fine-tuning) with and without neuron pruning. SEAL reduces attack success rate (ASR) by up to 60%, at a capability cost of at most 1.4% on a five-benchmark average. Additionally, SEAL can seamlessly integrate with router-level …
[AI-26] Auditory Illusion Benchmark for Large Audio Language Models ICASSP2026
链接: https://arxiv.org/abs/2609.02277
作者: Hayoon Kim,Eunice Hong,Kyogu Lee
类目: ound (cs.SD); Artificial Intelligence (cs.AI)
备注: Accepted to the 2026 IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2026)
Abstract:Perceptual illusions have long served as crucial probes into human cognition, revealing biases and limitations of perception. In the auditory domain, such illusions provide a unique lens for testing whether Large Audio Language Models (LALMs) replicate human perceptual tendencies. Despite their importance, most benchmarks focus on visual illusions or general audio tasks, leaving auditory illusions underexplored. To this end, we present AIB, the first auditory illusion benchmark for LALMs, covering ten representative illusions across music, sound, and speech, each annotated for the presence of knowledge-based priors. Our methodology pairs model evaluation with controlled human listening studies, enabling direct comparison of responses. Results show systematic differences: while most LALMs remain signal-faithful on low-level acoustic illusions, several exhibit more human-like responses when linguistic or musical priors are involved, although no model matches the human perceptual profile. These findings highlight the current limitations of LALMs as cognitive models. By establishing auditory illusions as a rigorous testbed, our work offers a new perspective for probing neural black-box models and advancing understanding of auditory cognition. AIB is publicly available at this https URL.
[AI-27] CrashDiffuser: VLM-Guided Collision Intent Reasoning for Fine-Grained Safety-Critical Traffic Scenario Generation
链接: https://arxiv.org/abs/2609.02270
作者: Shucheng Zhang,Yuang Zhang,Bingzhang Wang,Muhammad Monjurul Karim,Kehua Chen,Yinhai Wang
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:
Abstract:Generating safety-critical scenarios is essential for evaluating autonomous driving systems. However, existing generators primarily focus on inducing collisions and offer limited control over where contact occurs on the target vehicle. In this paper, we study fine-grained safety-critical scenario generation, where success requires both a target collision and a specified head, rear, or side contact region. We propose CrashDiffuser, a closed-loop VLM-guided diffusion framework that decouples semantic collision reasoning from continuous trajectory synthesis through a hierarchical collision-intent interface derived from the requested target contact region. At initialization, the VLM extracts reusable scene-level context; at each replanning step, it predicts a structured action tuple describing speed change, turning behavior, and collision stage. This intent conditions a diffusion model to generate executable adversarial trajectories, while collision-guided sampling, candidate selection, and short-horizon replanning adapt generation to the target vehicle’s evolving behavior. On WOMD-derived closed-loop scenarios, CrashDiffuser achieves a target-collision rate of 50.33% in a single attempt and 67.98% after three attempts, together with a contact-region control success rate of 40.05% and competitive trajectory naturalness. Component ablations further support the proposed design.
[AI-28] DiffuSearch: How Hybrid Trajectory Planning Benefits from Aligned Objectives in Diffusion and Action Space ECCV2026
链接: https://arxiv.org/abs/2609.02252
作者: Steffen Hagedorn,Aron Distelzweig,Alexandru P. Condurache
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: ECCV 2026 Workshop on Emerging Behaviors for Achieving Robust Autonomy
Abstract:In trajectory planning for autonomous driving, hybrid planning architectures are often realized as a collection of disparate modules, each with its own objectives. This lack of a unifying principle can lead to inconsistencies between the initial and refined trajectory, resulting in suboptimal behavior. We address this by introducing DiffuSearch, a novel hybrid planner that uses a unified set of objectives across generation and refinement. Our model encourages all components to follow the same shared driving goals: collision avoidance, drivable area compliance, comfort, and progress. DiffuSearch employs a two-stage architecture. First, a guided diffusion model generates a scene-consistent, joint trajectory prediction, using our driving objectives as differentiable guidance functions to implicitly steer the denoising process. Second, a Monte Carlo Tree Search (MCTS) in a discretized action space performs an explicit, local refinement of this proposal, leveraging the same driving objectives as its reward function. This synergistic design leverages the diffusion model’s strength in finding scene-consistent solutions combined with the explainable, constraint-aware refinement of MCTS. Experiments on nuPlan and interPlan reactive closed-loop benchmarks demonstrate that DiffuSearch achieves strong and often state-of-the-art performance, substantially reducing collisions and improving comfort, particularly in complex, interactive scenarios. Our ablation studies indicate that MCTS refinement is the main mechanism behind the gains, while sharing objectives between implicit guidance and explicit search provides further consistent improvements.
[AI-29] LLM -as-a-Judge Is Not an Oracle: Why Self-Improving Agents Need Deterministic Guardrails
链接: https://arxiv.org/abs/2609.02246
作者: Vansh Wahi
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 20 pages, 4 figures, 5 tables
Abstract:Self-improving agent pipelines have a problem at their center. An optimizer rewrites prompts to score higher, and the score comes from a judge that is itself an LLM. That judge has the last word on whether the system is getting better, and our position is that it has not earned it. The judge should be demoted from oracle to advisor: its verdict becomes one input among several, and every change is gated instead by a deterministic verification layer the judge cannot override. We reached this position by building the alternative and running it. Over months of running autonomous prompt-optimization loops in production across contract analysis, compliance review, and code quality, we cataloged eleven ways the evaluation signal failed, in four classes: judge bias, harness and metric failures, ground-truth errors, and reward hacking. Agents achieved perfect scores by reading cached answer keys from their environment, a 100% pass rate concealing 68% true capability. A corrupted ground-truth label caused the optimizer to delete correct compliance rules to agree with it. A syntactically broken prompt was promoted as the winner because a silent parser fallback improved the metric. Attempts to fix the judge by rewriting its rubric plateaued; the only reliable gain came from a structural constraint on its output order. In response we describe PROCTOR, a Teacher-Student loop in which a stateful orchestrator holds all tool access, stateless subagents diagnose failures and draft mutations they cannot apply, and a Teacher grades those mutations under five deterministic guardrails: hermetic sandboxes, capability-disjoint roles, acceptance checks that outrank the Teacher, frozen holdouts, and canary cases engineered so that a perfect score is itself evidence of cheating. We report the failures this prevented, and, because the Teacher is itself an LLM judge, the failures it did not.
[AI-30] ask-Level Natural Language Priors as Learning Signals for Low-Resource LLM Training
链接: https://arxiv.org/abs/2609.02244
作者: Jian Gao,Xiao Zhang,Xun Zhu,Miao Li,Ji Wu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models (LLMs) often struggle when low-resource training data are ambiguous or incomplete. Task-level natural-language priors can provide useful guidance in such settings, but existing approaches usually treat these priors as input context rather than as learning signals during training. We propose Prior-Guided Tuning (PGT), a training perspective that incorporates natural-language priors as auxiliary learning signals for low-resource LLM training. Under this perspective, we introduce Contrastive Prior Steering (CPS), which keeps the original supervised objective intact while adding positive and negative prior-conditioned auxiliary losses to encourage task-consistent learning and discourage plausible but misleading alternatives. Experiments on AmbiMath, Jigsaw, and MNLI/HANS show that CPS consistently improves over plain and prompt fine-tuning. On AmbiMath, CPS achieves 97.6% average exact-match accuracy. On Jigsaw, CPS improves average Macro F1 by 9.5 percentage points over standard fine-tuning, and with 1/10 of the experimental training data slightly exceeds full-data plain fine-tuning. On HANS, CPS improves non-entailment accuracy by 8.3 and 5.2 percentage points for LLaMA 3.1 8B and Qwen 2.5 7B, respectively, while maintaining comparable in-domain MNLI accuracy. These results support our central claim: task-level natural-language priors can provide useful guidance as auxiliary learning signals for low-resource LLM training. Our code and data will be publicly available.
[AI-31] PGPO: Potential-Guided Policy Optimization for Multi-Turn Agent ic Tasks
链接: https://arxiv.org/abs/2609.02236
作者: Yuyao Zheng,Haipeng Sun,Junwei Bao,Lemao Liu,Hongfei Jiang,Yang Song,Dejing Dou
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Group-based reinforcement learning (RL) has become an effective paradigm for LLM post-training, but in multi-turn agentic tasks with sparse terminal rewards, it often provides coarse credit for intermediate actions. To obtain more fine-grained credit assignment, recent work such as GiGPO introduces step-level advantages for intermediate actions. However, these step-level signals still rely on the final outcome of each individual trajectory. As a result, actions within failed trajectories can remain poorly differentiated, so effective actions can receive the same unfavorable credit as erroneous ones. In this work, we propose Potential-Guided Policy Optimization (PGPO) for multi-turn agentic tasks. PGPO estimates empirical state potentials from anchor-state-group return statistics within each rollout group. It then derives action advantages from potential differences between adjacent states, enabling cross-trajectory credit propagation. This provides finer-grained step-level credit assignment, especially within failed trajectories. Experiments on ALFWorld and WebShop show strong overall performance relative to recent group-based RL methods. Further analysis provides evidence that PGPO yields more informative failure-side credit signals with negligible training overhead.
[AI-32] PhoenixNest-Video: Evidence-Grounded Multimodal Agent Framework for Automated Video Interview Assessment EMNLP2026
链接: https://arxiv.org/abs/2609.02231
作者: Fan Yuxuan,Huang Miaojun,Zhang Haimei,Wu Jingshen,Liu Hao
类目: Artificial Intelligence (cs.AI)
备注: Accepted by EMNLP 2026
Abstract:Interview assessment requires per-criterion judgments grounded in behavioral evidence, yet surging applicant volumes have made human-only evaluation costly and inconsistent, while existing AI approaches yield opaque scores without traceable rationale. We introduce PhoenixNest-Video, an evidence-grounded multimodal agent framework for automated video interview assessment. It builds a semantic video graph as structured working memory, performs rubric-conditioned retrieval with cross-modal verification across visual, audio, and textual streams, and produces per-criterion scores anchored to the candidate’s materials. A Scorer trained via Rubrics-based Reinforcement Learning with dual rewards for rubric alignment and score-level differentiation internalizes the discriminative structure of multi-level rubrics. PhoenixNest-Video attains 91.50% grade-level accuracy on VInterview-2025, outperforming substantially larger proprietary models. A compact, rubric-grounded agent therefore scores candidates in closer agreement with an expert panel than direct prompting of much larger models, and exposes the evidence behind each score for human review.
[AI-33] SkillGLoW: Procedural-Family Skill Consolidation for Self-Improving Agents on Long-Horizon Task Streams
链接: https://arxiv.org/abs/2609.02217
作者: Ao Yan,Xin Zhang,Jiawei Du,Joey Tianyi Zhou
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:LLM agents increasingly self-improve by writing and reusing textual skills, kept either as one global document or as a flat pool of per-task entries, though most of the evidence comes from domains with structurally similar tasks. On long-horizon workloads where each task demands a different solution, the two forms fail in opposite ways: the document collapses into generic discipline, while the pool inflates and its entries stay bound to the instance that wrote them. We argue the missing unit of reuse is the solving procedure shared by a cluster of related tasks, and build SkillGLoW (Global-Local Weave) around it: the local skills a task writes from its own execution are aggregated into procedural families and compressed into de-instantiated global priors, while the instance detail they hold is regenerated per task rather than stored; a commit gate admits a prior only when real execution shows it does not degrade the deployed library. Across four benchmarks (mathematical reasoning, terminal automation, software repair, and embodied control) and three models, the priors gain 17.2 points (hard) over the no-skill baseline on average, with positive gains in all 12 continual-improvement runs, and 18.0 with local regeneration, while the library holds one prior per procedural family, 3.6x more compact than the per-task pool. Under the same protocol GLoW leads a published single-document optimizer on 15 of 21 cells. Unmodified, the library lifts success on unseen ALFWorld tasks from 73.9% to 83.9%, evidence that what transfers is procedure rather than task memory.
[AI-34] PEARL: Path-Entity Aligned Relational Learning with Contextual Subgraphs for Inductive Knowledge Graph Completion
链接: https://arxiv.org/abs/2609.02216
作者: Yunchi Yang,Longlong Li,Cunquan Qu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Inductive knowledge graph completion (IKGC) aims to predict missing links involving entities unseen during training, requiring models to learn transferable relational and structural patterns. Existing subgraph- and path-based approaches often encode relational paths independently of their surrounding query subgraphs, although the predictive relevance of a path may vary across structural contexts. We propose PEARL, a Path-Entity Aligned Relational Learning framework that models paths as context-conditioned reasoning signals. PEARL constructs a query-specific contextual subgraph from the union of the query entities’ neighborhoods and uses a large language model (LLM)-guided retriever to distill semantically relevant paths. It then builds a bipartite interaction graph over paths, contextual entities, and a global subgraph representation, allowing path embeddings to adapt to local and global structural evidence. To suppress noise introduced by the enlarged context, PEARL employs a dual-view contrastive objective that promotes representation consistency under stochastic contextual perturbations. Experiments on WN18RR, FB15k-237, and NELL-995 show that PEARL obtains the best average Hits@10 among the compared IKGC methods on all three benchmarks. Ablation studies, efficiency analyses, and case studies further validate the contributions of contextual subgraph modeling, semantic path retrieval, path-entity interaction, and contrastive regularization.
[AI-35] ASCII Attack: Recontextualising Harmful Requests as Artistic Critique in Large Language Models
链接: https://arxiv.org/abs/2609.02215
作者: Da Cheng Gu,Yifei Dong,Xinghao Yang,Yongshun Gong,Wei Liu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Safety alignment trains large language models to refuse harmful requests stated plainly, but that training is applied mostly to surface form. Requests that only recontextualise the same operational content, changing how the model reads it, are therefore only weakly covered. The ASCII Attack is one such recontextualisation. It is single-turn and black-box: one message, with no access to model internals. It embeds a fully legible harmful request in ASCIl-art characters, presents it as artwork, and asks for feedback. Unlike ArtPrompt, it hides nothing: the request stays readable. The reply is written as artistic critique and can contain operational detail that a plain request would have been refused for. Every framed prompt is paired with a direct-question control, so the contrast is isolated from topic, model and decoding variation. The contrast identifies a bundled surface, not one isolated channel. Across eleven models and eight harm topics, a harm-aware classifier judges 62% of framed prompts harmful against 42% of controls. On the most susceptible model the framed prompt succeeds 93% of the time. A single query matches or exceeds published single-query attacks under four of five harm judges. The effect tracks the model more than the topic and does not diminish with scale. At least one judge dissents from the panel majority on nearly two-thirds of framed rows, which is itself a measurement-validity finding. That pattern is consistent with mismatched generalisation.
[AI-36] SMart: A Multi-source Multi-phase Time Series Representation Transfer Framework
链接: https://arxiv.org/abs/2609.02203
作者: Fang He,Wang-chien Lee
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 11 pages
Abstract:Time series representation learning (TSRL) has attracted growing research interests in recent years. Two recent explorations in TSRL are: i) exploiting a transformer-based framework to learn time series; ii) instead of using only the targeted dataset, borrowing time series from other datasets to to facilitate representation transfer. While these two explorations are shown effective, the self-supervised time series recovery task in (i) and the single-source dataset used in (ii) are technically simple and thus can be enhanced with new ideas. In this work, we propose a new TSRL framework, namely multi-source multi-phase time series representation transfer (SMart), which has two novel mechanisms to address the aforementioned deficiencies: 1) a multi-phase recurrence plots recovery task, in three alternative modes, for guiding the encoder to embed time series dynamics into the time series representation; and 2) a source dataset selector to select multiple suitable source datasets to supplement the original target dataset for pre-training the TSRL encoder. Experimental results show that SMart outperforms several state-of-the-art models for time series representation learning, classification and regression on both uni-variate and multi-variate time series datasets, reducing mean absolute error up to 19.5% for time series regression, and increasing average accuracy up to 1.34% for time series classification.
[AI-37] Examining the Vulnerability of Multi-Agent Medical Systems to Human Interventions for Clinical Reasoning
链接: https://arxiv.org/abs/2609.02191
作者: Benjamin C Liu,Dillon Mehta,Rishi Malhotra,Adam Zobian,Yong Ying Tan,Samir Chopra,Daniella Rand,Natalie Pang,Abhiram Gudimella,Kevin Zhu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Human interventions at fault points can alter the diagnostic accuracy of multi-agent medical systems. We defined fault points as moments in AI agent conversations, in which an agent’s reasoning became most vulnerable to external influence. Using the MedQA dataset, this study analyzed simulated doctor-patient conversations to measure how interventions shifted reasoning and accuracy. Correct intervention methods showed an improvement in baseline diagnostic accuracy of up to 40%, while incorrect or bias-related interventions degraded performance by up to 6% and increased diagnostic drift and uncertainty. Beyond performance changes, our analysis revealed behavioral similarities between cognitive biases in simulated agent environments and real-world clinical practice. Examples included premature closure and susceptibility to misleading cues. Overall, these findings demonstrate that identifying and guiding fault points with human interventions may provide a mechanism for improving diagnostic robustness in multi-agent medical systems.
[AI-38] FUSE: An Evaluating Framework for Dangerous Capabilities of LLM s
链接: https://arxiv.org/abs/2609.02168
作者: Zhengyi Jin,Ru Zhang,Xiao Chen,Xinbo Liu,Jiaxuan Lin,Jia Huang,Jianyi Liu,Zhen Yang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Fragmented safety evaluation undermines the governance of dangerous AI capabilities. We present a modular framework that evaluates each model through three orthogonal pipelines—Knowledge ( K ), Defense ( D ), and Harm ( H )—under a unified protocol, aggregating results into a standardized dangerous-capability profile \phi . Pluggable modules supply scenario seeds, knowledge banks, hazard queries, and judge rubrics, while the core evaluation engine remains unchanged across domains; the CB evaluation is complemented by a cyber pilot demonstrating protocol transfer. Instantiating the framework with a chemical-biological (CB) module, we evaluate 12 commercial LLMs from four families. Our first contribution is a horizontal comparison of dangerous capability across models and model families: the three dimensions expose sharply divergent profiles—models with comparable knowledge differ in refusal resilience, and strong defenders do not generate less harmful content when they do comply—while family-level patterns further separate Claude, DeepSeek, and GPT models. The second is a temporal analysis of capability evolution: tracking K , D , and H against model release dates reveals that dangerous capability has not monotonically declined; newer models deepen knowledge while only partially improving defense, showing that scaling and alignment progress do not uniformly translate into safety. Reliability is established via cross-judge consistency (bootstrap \rho 0.79 , 4 of 5 judges) and pipeline orthogonality ( K – D – H inter-correlations \rho \in [0.32, 0.52] ). Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2609.02168 [cs.AI] (or arXiv:2609.02168v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2609.02168 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-39] GeoSPRINT: Geometric Redundancy-Aware Step Pruning for Inference in Diffusion Trajectories
链接: https://arxiv.org/abs/2609.02160
作者: Arpita Joshi
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Diffusion models achieve high sample quality but remain expensive at inference time because sampling requires many sequential neural function evaluations (NFEs). Existing acceleration methods either use fixed step-skipping schedules, adapt step sizes based on local numerical error, or require additional training. We introduce GeoSPRINT (Geometric Step Pruning for Inference in Trajectories), a training-free framework for constructing non-uniform sampling schedules from the geometry of denoising trajectories. GeoSPRINT detects geometrically redundant steps using a hyperplanarity test in latent space, implemented efficiently via QR factorization, and converts the resulting redundancy profile into a sampling schedule that allocates more steps to high-curvature regions of the trajectory. In addition, we introduce the trajectory projection score \alpha_\mathrmtraj , a residual-variance metric that quantifies trajectory straightness and serves as a model-free diagnostic for rectified flow quality. Across CIFAR-10 ( 32\times32 ), LSUN Church ( 256\times256 ), and Stable Diffusion v1.5 ( 512\times512 latent), GeoSPRINT consistently improves over uniform DDIM (Denoising Diffusion Implicit Models) schedules at matched NFE budgets. On CIFAR-10, GeoSPRINT improves FID (Fréchet Inception Distance) by 0.7-1.1 over DDIM across 49-89 NFEs and surpasses DPM-Solver++ at NFE \geq30 despite using a first-order DDIM solver. On LSUN Church, it reduces FID from 1.48 to 1.26 at 52 steps, and on Stable Diffusion v1.5 it achieves up to 1.93 FID improvement over DDIM. These results show that trajectory geometry provides a useful global signal for allocating inference steps and that schedule quality can substantially improve diffusion sampling efficiency without retraining.
[AI-40] Online Non-Monotone DR-Submodular Maximization Matching the Offline 0.401 Factor
链接: https://arxiv.org/abs/2609.02145
作者: Vaneet Aggarwal,Yiyang Lu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computational Complexity (cs.CC); Machine Learning (stat.ML)
备注:
Abstract:We study online maximization of nonnegative, non-monotone DR-submodular functions over compact convex down-closed subsets of the d -dimensional unit cube. The best known constructive offline approximation factor is 0.401 under the corresponding meta-solvability assumptions, whereas comparable adversarial online guarantees had remained at 1/e . We show that this factor is also achievable online. In the post-decision full-information value-oracle model, our algorithm attains factor 0.401 with sublinear approximate regret when oracle feedback is conditionally unbiased and bounded. The online algorithm does not run the offline construction on a changing objective. Instead, it replaces the offline objective-dependent box step by a weighted online learner that controls the required residual terms cumulatively. An exact asymmetric balance theorem preserves the offline coefficients despite adversarial variation. The direct implementation has O(T^3/4) regret and uses O(dT^1/4) oracle calls per round. More generally, for every \delta\in[0,1/4] , batching gives O(T^\delta) calls per round and O(T^4/5-\delta/5) regret, including a one-call O(T^4/5) endpoint. Under a positive-anchor condition, randomized blocking retains factor 0.401 with O(T^5/6) one-point bandit regret. Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computational Complexity (cs.CC); Machine Learning (stat.ML) Cite as: arXiv:2609.02145 [cs.LG] (or arXiv:2609.02145v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2609.02145 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-41] Semantic Signal-Assisted Inspection and Recovery Allocation in Reverse Logistics
链接: https://arxiv.org/abs/2609.02116
作者: Jiani He,Dingyan Shang,Yihua Xu,Shiqi Huang,Yan Lyu,Jize Li,Shangjing Tang
类目: Artificial Intelligence (cs.AI)
备注: Accepted at the IEEE 4th International Conference on Artificial Intelligence, Blockchain, and Internet of Things (AIBThings 2026). 7 pages, 1 figure, 3 tables. Code and benchmark: this https URL
Abstract:Reverse-logistics operators often decide how to inspect and route returned assets before their condition is fully observed, while full inspection consumes scarce labor. Semantic Signal-Assisted Decision Support converts return notes into a condition factor and a signal-quality score that guide inspection depth and recovery allocation under shared labor capacity. We evaluate the framework in three synthetic benchmark scenarios spanning information technology decommissioning, aircraft maintenance, and consumer-electronics returns. Across 30 paired simulation seeds, the keyword implementation improves net recovery value relative to a structured-feature comparator with noisy full inspection while reducing inspection cost in all three scenarios. A risk-blind comparator that skips inspection altogether still records higher value under the benchmark’s purely economic objective. At matched inspection cost, score-guided targeting adds 53.9 thousand United States dollars per batch in the aircraft scenario but has little economic effect in the other two configurations; phrase and large language model extractors provide further gains in the aircraft scenario. These results show how narrative evidence can support inspection allocation before recovery decisions are made.
[AI-42] MeanField Surrogate Modeling for Scalable Runtime Scheduling of Concurrent Heterogeneous AI Inference on Shared GPUs
链接: https://arxiv.org/abs/2609.02109
作者: Youssef Ennouri,Soonhoi Ha
类目: Distributed, Parallel, and Cluster Computing (cs.DC); Artificial Intelligence (cs.AI)
备注: 4 pages, 3 figures, 2 tables. Accepted for publication in IEEE Embedded Systems Letters
Abstract:Deploying heterogeneous AI models concurrently on a shared GPU introduces resource contention that complicates runtime scheduling. While surrogate models avoid costly online benchmarking, their profiling requirements typically grow combinatorially with the number of co-running models, limiting scalability. We propose a MeanField surrogate that predicts per-model performance from local configuration and aggregate GPU state rather than explicitly modeling all joint interactions. Experiments on concurrent LLM and vision workloads across N \in \2,3,4,5,6\ show high predictive accuracy ( R^2 \approx 0.96 ) with an empirical sample budget that grows approximately linearly in N , in contrast to the combinatorial cost of fully joint profiling. Integrated into a genetic algorithm scheduler, the surrogate scales to an N=5 problem with 78,732 feasible joint configurations, remaining within 0.10% of the exhaustive search with zero SLA violations across eight dynamic workload scenarios, while complete online GA decisions take 26 ms median, about 5\times faster than exhaustive surrogate search.
[AI-43] Git4Data: Database-Native Version Control for AI Agents
链接: https://arxiv.org/abs/2609.02106
作者: Hongshen Gou,Zuyu Zhang,Yuze Sun,Peng Xu,Feng Tian,Long Wang,Jianguo Wang
类目: Databases (cs.DB); Artificial Intelligence (cs.AI)
备注:
Abstract:Large Language Model (LLM) agents increasingly explore many candidate states of relational data in parallel, each of which should remain isolated, reproducible, and auditable, preferably through the same SQL interface used for ordinary data work. Existing tools support this requirement only partially: source-code version control does not scale to large datasets, whereas relational databases manage large data efficiently but rarely expose native branching, comparison, and merging. We present Git4Data, a database-native version-control layer for agentic workflows. Git4Data treats a database as a repository and a table as a versioned object, exposing Git-style operations (snapshot/tag, branch, diff, and merge with explicit conflict-resolution policies) through SQL extensions. Implemented in MatrixOne, a cloud-native relational database, Git4Data leverages immutable object storage and MVCC to make the cost of these operations proportional to the size of the change rather than the size of the data. On the BranchBench agentic branching workloads, Git4Data outperforms DoltDB by up to an order of magnitude. Overall, we believe this work sheds light on how relational databases can better support AI agents through efficient versioning.
[AI-44] READY or Not: Reliable Enterprise Agent Deployment
链接: https://arxiv.org/abs/2609.02095
作者: Veronica Chatrath,Bryan Zhu,Jingxuan Fan,George Pu,Soham Dinesh Tiwari,Soham Dan,Ryan Young,Yuan(Christy)Li,Yuang Yao,Apaar Shanker,Minglai Yang,Daniel Yue Zhang,Yunzhong He,Ying Liu,Chenguang Wang,Zhijun Yin,Yuan Xue
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:An AI agent can perform well on benchmarks and still be unsuitable for deployment. Existing AI-agent benchmarks measure whether an agent can complete realistic professional work, whereas enterprise deployment asks a different question: whether an agent can meet a required reliability level, under acceptable human oversight, and at tolerable cost. We introduce Reliable Enterprise Agent Deployment (READY), a framework for qualifying AI agents for deployment on enterprise workflows. READY preserves each workflow’s own definition of successful execution while applying a common qualification procedure. Given an agent, a workflow, and a class of candidate oversight policies, READY measures the reliability and operating cost of the human-AI system, selects the minimum-cost policy that satisfies a specified reliability target, and statistically qualifies it on held-out cases. The resulting deployment profile characterizes the supported operating point: reliability, human-oversight burden, and cost. READY is implemented as an open testbed that decouples workflow specification, execution, evaluation, and qualification, and runs on existing agent-evaluation infrastructure. In an end-to-end clinical-audit case study spanning 16 agent systems and 750 cases, READY reveals differences hidden by autonomous performance: two systems separated by only 0.3 points in autonomous accuracy (72.8% vs. 72.5%) require 39.2% versus 29.6% human review, respectively, to qualify at the same 76% reliability target under the evaluated oversight policy. READY thus shifts enterprise agent evaluation from how well can the agent perform the work? to under what conditions, and at what cost, can it be reliably deployed? By making those conditions explicit and statistically testable, READY provides a basis for comparing agent systems, setting oversight requirements, and making evidence-based deployment decisions.
[AI-45] Beyond Outcome Gaps: Process-Aware Fairness Diagnosis for LLM -based Multi-Agent Decision Systems EMNLP2026
链接: https://arxiv.org/abs/2609.02092
作者: Yiran Zhao,Lu Zhou,Liming Fang,Yufei Chen,Jiafei Wu,Zhe Liu,Xiaogang Xu
类目: Artificial Intelligence (cs.AI)
备注: Accepted to EMNLP 2026
Abstract:LLM-based multi-agent systems (MAS) are increasingly considered for high-stakes decision-making, yet outcome-based fairness audits can miss where risks arise within the decision trajectory. We present SCOPED-Hiring, a process-aware fairness diagnosis pipeline for LLM-based hiring MAS. SCOPED-Hiring constructs controlled resume variants, runs role-based hiring committees, logs over 311K structured decision trajectories, and converts trajectory fields into quantitative fairness signals organized by six diagnostic lenses: final outcome, counterfactual, process, pathway, dynamic, and design effects. SCOPED-Hiring reveals that balanced final hire rates can mask hidden trajectory unfairness in multi-agent decision trajectories: career gaps trigger suspicion, proxy cues shape qualification judgments, and identity cues lead to unequal investigation. Targeted repair guided by these diagnoses reduces total layered burden by 72.3% while shifting the hire rate by only 1.86 pp, showing that process diagnosis can guide effective repair. Project Page: this https URL
[AI-46] CHIME: Credit-Aware Hierarchical Memory Evolution for Long-Horizon Agent ic Planning
链接: https://arxiv.org/abs/2609.02074
作者: Yongshi Ye,Tian Lan,Feihu Jiang,Muyang Ye,Bin Zhu,Qianghuai Jia,Longyue Wang,Zhao Xu,Weihua Luo,Xiaodong Shi
类目: Artificial Intelligence (cs.AI)
备注: 7 figures, 7 tables
Abstract:Planning is a central capability that enables agents to decompose complex long-horizon tasks into manageable steps. Test-time search and training-based methods improve planning but incur high inference costs or require expensive training data. Self-evolving memory instead accumulates reusable experience from agent interaction outcomes into an external memory bank, so planning capability keeps improving at inference time without parameter updates. However, existing self-evolving memory methods share an inherent credit assignment problem: they rely on final task outcomes as feedback, but such outcomes conflate plan quality with execution errors and environmental factors, so the accumulated planning experience is often biased and noisy. To address this problem, we propose Credit-Aware Hierarchical Memory Evolution (CHIME), a self-evolving memory framework that maintains a separate planning bank and execution bank and follows an attribute-before-memorize principle: CHIME first attributes each task outcome to the plan, the execution, both, or neither, and then updates only the corresponding memory bank. Extensive experiments on four long-horizon agent benchmarks show that CHIME consistently outperforms state-of-the-art training-based and self-evolving memory baselines. Further analyses reveal several interesting findings. For example, CHIME accumulates effective memory with far fewer items. In addition, the learned memory values faithfully reflect downstream utility: high-quality planning memories are more valuable than execution memories. Finally, the accumulated memory effectively transfers across backbone models. Code will be released at this https URL.
[AI-47] oolGate: An Executable Acceptance Pipeline for Tool-Dependent Scientific Benchmark Construction
链接: https://arxiv.org/abs/2609.02067
作者: Ke Zhang,Yankang Liu,Roya Zandi,Maziar Raissi
类目: Artificial Intelligence (cs.AI); Mathematical Software (cs.MS); Software Engineering (cs.SE)
备注: 7 pages, 2 figures
Abstract:Scientific benchmarks are commonly built by domain experts who write tasks and cross-check one another’s work, or who adapt existing material from textbooks, published papers, and online resources. These routes can produce strong evaluations, but they require substantial per-item labor. Language models can reduce this repeated work by proposing candidates quickly. The remaining problem is acceptance. We target scientific questions whose answers require computations with specialist software rather than unaided reasoning alone. A candidate is invalid if its script fails or returns a different answer, or trivial if a model answers it without the software. We present ToolGate, which treats every generated item as a proposal and keeps it only if three gates pass. First, an executable solution script must reproduce the proposed answer when run with the scientific software. Second, randomized no-tool screening rejects candidates that models can already solve from the prompt alone. Third, a tool-using agent must solve each survivor within a fixed time limit. We instantiate ToolGate in FEniCSx with 500 generation attempts. The local-verification gate retains 478 candidates. For final reporting, we rescreen this pool after generation: two randomized no-tool screens exclude 222 from the reported pool, and direct GPT-5.5 API calls at medium reasoning (the API default) exclude another 121. Of the remaining 135, a GPT-5.5 Codex CLI agent with access to FEniCSx solves 130; exact deduplication leaves 128 unique protocol survivors. ToolGate turns repeated answer checking and difficulty screening into an auditable process while leaving domain design and final review to experts.
[AI-48] MineTRACE: An Evidence-Grounded Interactive Reasoning System for Mineral Prospectivity EMNLP2026
链接: https://arxiv.org/abs/2609.02060
作者: Yiran Zhang,Jinwen Liu,Daniel Su,Yisu Chen,Qiang Sun,Chris Gonzalez,Eun-Jung Holden,Marco Fiorentini,Wei Liu,Yihao Ding
类目: Artificial Intelligence (cs.AI)
备注: EMNLP 2026 Demo
Abstract:Mineral exploration requires integrating heterogeneous geochemical, geophysical, and geological evidence, yet existing prospectivity systems often provide only opaque scores or heatmaps. We present MineTRACE, a web-based system for evidence-grounded exploration of eight commodities: Cu, Au, Ni, W, Sn, Co, Ta, and Mn. Users can explore prospectivity maps, query locations or regions, inspect supporting evidence, and interact through natural language. A transparent expert tree, informed by geological knowledge and known deposits, combines multi-source evidence into interpretable prospectivity scores. For a new location, the conversational assistant retrieves the score and supporting evidence from the analysis pipeline and presents them in natural language. The scorer achieves spatial AUC values of up to 0.917 across different test scenarios, while end-to-end evaluation assesses query accuracy and response grounding. MineTRACE makes public geoscience data easier to access, interpret, and verify, supporting more efficient and transparent mineral exploration.
[AI-49] Monitoring Web Agents Without Internal Signals: Observable Trajectories and Key-Step Supervision
链接: https://arxiv.org/abs/2609.02057
作者: Sitong Pan,Yipeng Shen,Yilin Lu,Caiwen Ding,Lu Cheng,Qianwen Wang
类目: Artificial Intelligence (cs.AI)
备注: preprint
Abstract:Reliable web-agent monitoring is difficult when model-internal uncertainty signals such as token logits are unavailable. In this work, we study prefix-level risk prediction for web agents using observable trajectory signals: given an evolving prefix, estimate whether the current execution remains on track or is tending toward failure. We derive two observable trajectory representations: Macro features summarize cross-step agent–environment behavior and feedback, while Micro features measure the consistency of intention, action, and anticipated state change through repeated black-box queries. Instead of inheriting the final result label, we label the first critical error that remains uncorrected in the observed continuation and is associated with final failure as a key-step boundary, preserving valid early prefixes of failed trajectories as on track. Across WebArena-Lite and Online Mind2Web web agent benchmarks with five open- and closed-source backbones, observable trajectory signals are competitive with internal-signal baselines. The resulting predictors also support early intervention under fixed false-cut budgets and transfer across held-out website categories. These findings show that observable trajectory signals support valuable risk prediction abilities.
[AI-50] Modeling What Changes: Sparse Residual World Models for Object-Centric Manipulation
链接: https://arxiv.org/abs/2609.02046
作者: Param Thakkar,Parsika Paresh Shah,Manisha Sushant Gote
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:
Abstract:Monolithic world models predict the entire next state at every step, spending capacity re-predicting the static majority of a scene and injecting error into it. We ask whether explicitly modeling change (a per-object change gate plus a residual delta head that perturbs only the objects the gate flags) is a more effective and interpretable bias for physical prediction and control. On a MuJoCo tabletop pushing benchmark scaling from 3 to 8 objects, the sparse/residual model predicts next-state poses 2.5 to 4.6 times more accurately than a dense multilayer perceptron at 8.6 to 11.1 times fewer parameters, sustains change-detection F1 of 0.80 to 0.87 where the dense baseline is degenerate, transfers across object counts with zero retraining (99.4 percent F1 retention), and reaches about 90 percent of its full-data accuracy with a quarter of the data. In autoregressive rollout it compounds far less error, hugging the no-motion floor while the dense model drifts. Finally, inside a sampling-based planner, prediction-only models fail (though a true-simulator oracle solves the task with the identical planner, confirming the planner is sound), but once featurized and trained for the states a planner visits, the sparse model begins to plan (0.23 plus or minus 0.06 success over three seeds) while the dense monolith stays at zero at every seed. Modeling what changes, rather than re-predicting the whole world, is a simple, effective bias for object-centric physical AI; code, data generators, and all checkpoints will be released upon publication.
[AI-51] HeadWiseKV: Budgeted Per-Head Cache Residency for Hybrid Long-Context Language Models
链接: https://arxiv.org/abs/2609.02029
作者: Renjie Xie,Juncheng Yang,Aoting Hu,Mingxi Zhang,Liyao Wu,Zheheng Hong,Wei Xu
类目: Artificial Intelligence (cs.AI)
备注: 14 pages including appendices, 4 figures, 5 tables
Abstract:Long-context inference retains a growing key–value (KV) cache during decoding, which consumes substantial GPU memory and can reduce generation throughput. This bottleneck remains in hybrid language models because their residual global-attention layers can dominate context-dependent cache demand. We study how to allocate this state under an aggregate KV-residency budget. We introduce HeadWiseKV, a training-free framework that compresses the residual global KV caches of hybrid language models while preserving their native local, recurrent, and linear paths. It assigns each physical KV head a static, multilevel history window, making cache demand predictable before serving. We formulate this allocation as a restricted operational rate–distortion problem and propose SeqCalib as the core policy-generation algorithm in HeadWiseKV. SeqCalib processes layers in execution order and conditions each decision on the lower-layer policy used at deployment, thereby accounting for interactions across depth. A grouped-cache runtime materializes the selected policy as actual per-head KV residency rather than a mask over a full cache. We evaluate downstream quality across four hybrid long-context models and study physical residency and serving behavior on Qwen3.6-27B. HeadWiseKV retains near-Full-KV RULER and LoCoMo quality across the evaluated models. In the fixed-model systems study, it reduces sampled peak device memory by 8.59% at a 112K context length and extends the largest verified successful context from 114K to 161K.
[AI-52] Benchmarking Language Models for Statistical Problem Formulation
链接: https://arxiv.org/abs/2609.01982
作者: Chen Wang,Junzhe Zhao,Xin Cong,Wanlu Deng,Ke Deng
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models (LLMs) are increasingly used as assistants for statistical and data science work, yet existing evaluations largely assume the analysis target is already specified. In practice, users arrive with informal goals and heterogeneous data, leaving the model to decide what statistical task is implied and which data are relevant. We first formalize this upstream step as Statistical Problem Formulation and decompose it into two subtasks: (1) Statistical Problem Classification and (2) Variable Identification Role Assignment. We then introduce StatFormBench, a benchmark built from five cross-domain statistics textbooks and a data science case library, covering diverse problem types, data representations, and scenario styles. It contains 1,013 samples spanning 20 coarse-grained and 85 fine-grained statistical problem categories. Across 14 open- and closed-source LLMs, the best zero-shot models reach only 72.0 fine-grained classification accuracy and 63.2 variable set overlap. No model performs consistently best across the two subtasks, while enhanced prompting strategies yield only limited or inconsistent gains. We release the benchmark data on Hugging Face at this https URL and the evaluation code on GitHub at this https URL.
[AI-53] Post-Training Ternarization of Qwen 3-4B Capability Effective Bit Budget Storag e Compression and Deployment
链接: https://arxiv.org/abs/2609.01962
作者: Anirudh Malik,M Sparsh Mehra,Poojith Devan
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Weight-only post-training ternarization of a 4B-parameter instruction-tuned language model. Activation quantization and end-to-end generation throughput are outside the scope of the primary evaluation
Abstract:Ultra-low-bit language models can reduce storage and memory bandwidth, but a nominal “1.58-bit” label does not fully describe the stored representation, retained capability, or runtime behavior. We study an end-to-end post-training conversion of Qwen, an instruction-tuned 4B-parameter model, using KOTMS rotation, E2M-ATQ ternarization, and GPTQ-style error compensation from TWLA. The experiment is weight-only: activations remain at 16-bit precision, so ILA-AMP is omitted. We evaluate effective bit accounting, task capability retention, perplexity, calibration sensitivity, checkpoint composition, and deployment behavior. The final conversion uses 1.641 effective bits per weight for quantized linear weights, with 81.62% of model parameters targeted. Across ten scored capability comparisons, accuracy falls from 64.5% to 54.7%. Degradation is uneven: BoolQ retains 84.6% chance-corrected teacher performance, while ARC-Challenge retains 43.8%. Perplexity rises from 13.639 to 18.748 on WikiText-2, 24.700 to 31.992 on PTB, and 19.831 to 28.966 on C4. A subsequent packing run preserves the ternary planes and scales, reducing reported model size from 8.29 GiB to 3.96 GiB with essentially unchanged perplexity. A separate third-party packing attempt was lossy and is excluded from the primary artifact claim. The packed artifact has not been benchmarked end-to-end for task accuracy or generation throughput. A preliminary Triton GEMV microbenchmark is 4.6x slower than FP16 cuBLAS on one tested shape. We therefore do not claim that compression alone yields faster inference. Comments: Weight-only post-training ternarization of a 4B-parameter instruction-tuned language model. Activation quantization and end-to-end generation throughput are outside the scope of the primary evaluation Subjects: Artificial Intelligence (cs.AI); Machine Learning (cs.LG) Cite as: arXiv:2609.01962 [cs.AI] (or arXiv:2609.01962v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2609.01962 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Anirudh Malik [view email] [v1] Wed, 2 Sep 2026 00:42:54 UTC (613 KB)
[AI-54] Convergence Theory of Knowledge Distillation in Asynchronous P2P Gossip Learning Network AAAI2027
链接: https://arxiv.org/abs/2609.01952
作者: Lucas Qingyang Fang,Tiyao Liu,Jinhao Jing,Zeji Li,Kaijie Chen,Harikrishna Kuttivelil,Katia Obraczka
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 35 pages, 21 graphes, currently submitting to AAAI 2027 main track (Federated Learning and Decentralized Learning)
Abstract:Decentralized, serverless learning increasingly connects devices running different architectures, where the standard tool, decentralized SGD, is undefined as models with different parameter counts cannot be averaged. Knowledge distillation (KD) exchanges soft predictions rather than weights and sidesteps this obstacle, yet convergence theory for fully decentralized, asynchronous peer-to-peer (P2P) KD is lacking. We provide one, relocating consensus from parameter space to function (output) space: a KD event is a geometric contraction operator in logit space on the peers’ predictive distributions, which we analyse in the Hilbert space of predictions on a reference measure. Under standard smoothness/variance assumptions and two realizability assumptions, one bridging parameter SGD to the functional step and one controlling restricted task/KD alignment, the time-averaged functional stationarity and function-space disagreement converge at rate O(1/(\eta T)) to an O(\eta)+O(B_f^2)+O(\zeta_f^2) neighbourhood. Here B_f is the distance from the task optimum to the peers’ reachable classes and \zeta_f measures persistent local-task heterogeneity. Across homogeneous, width-heterogeneous, and mixed-family networks of the experiments, KD contracts function disagreement by 40-61\times , while isolated training does not. The sampled stationarity diagnostic has late transient exponents 0.99-1.90 on the shared-skeleton main runs, and the four-point step-size sweep exhibits the predicted transient: neighbourhood tradeoff.
[AI-55] On-Policy Distillation Meets Off-Policy GRPO: Training Compact Instruction-Following Rerankers EMNLP2026
链接: https://arxiv.org/abs/2609.01947
作者: Vignesh Prabhakar,Jialing Pan,Anil Babu Ankisettipalli
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: EMNLP 2026 Findings
Abstract:Compact instruction-following rerankers are attractive for deployment, but conventional distillation pipelines typically train students by offline imitation of teacher outputs on a fixed set of examples, constraining supervision to the teacher’s observed ranking space. We revisit reranker distillation through the lens of reinforcement learning. We propose a two-stage framework combining off-policy teacher optimization with on-policy student distillation. In Stage 1, a 4B teacher reranker is strengthened with off-policy GRPO using LLM-judge feedback on 88K instruction-following examples. In Stage 2, a compact 1B student samples rankings from its own policy and receives soft teacher-derived rewards on those rankings, coupling student exploration with knowledge transfer. Our strongest gains appear under distribution shift. On MAIR-11, the original 11-subset, 869-query evaluation, the proposed student reaches 0.7670 nDCG@6, outperforming offline listwise KD by +4.6 points. Controlled comparisons against offline pairwise RankNet KD and on-policy GKD show that neither changing the offline distillation objective nor moving teacher-distribution matching on-policy reproduces the performance of reward-based on-policy distillation over student-sampled rankings. The advantage persists on MAIR-Full: across all 126 tasks and 9,356 queries, the proposed method obtains the highest task-macro point estimates among the evaluated distillation variants, reaching 0.6808 nDCG@6 and 0.7865 MRR@6. It also exceeds two released 7B RL-trained rerankers on the comparable MAIR-11 evaluation, while the same Stage 2 training procedure consistently improves three architecturally distinct alternative student backbones. On the 9,861-query validation benchmark, the resulting 1B reranker achieves 0.7624 nDCG@6 while providing a favorable quality-efficiency tradeoff relative to larger alternatives. Comments: EMNLP 2026 Findings Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI) Cite as: arXiv:2609.01947 [cs.LG] (or arXiv:2609.01947v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2609.01947 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Vignesh Prabhakar [view email] [v1] Tue, 1 Sep 2026 23:36:19 UTC (3,614 KB)
[AI-56] Looped Transformers under the Jacobian Lens: Does the Global Workspace Survive Recurrence?
链接: https://arxiv.org/abs/2609.01924
作者: Wenlong Wang,Fergal Reid
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Recent work identifies a mid-depth band of verbalisable, causally potent representations in a standard feedforward transformer — a functional analogue of a global workspace. Whether the same workspace functionality emerges when depth is implemented through recurrence rather than a stack of distinct layers remains unknown. Looped and depth-recurrent transformers provide a direct test of this question because they reuse the same weights across depth. We extend the Jacobian lens to iterated architectures using a virtual-unrolling adapter. We apply the full workspace suite — lens fitting, readout, and eleven causal experiment families — to Ouro-2.6B (48 layers looped 4 times, deeply supervised) and Huginn-0125 (a 4-layer core recurred 16 times, trained for latent reasoning), using Qwen3.6-27B (64 untied layers) as the standard baseline. We find that a workspace forms in the iterated part of each architecture, but that recurrence changes how it can be accessed. Ouro reconstructs workspace content in every loop, and linear transport cannot carry that content across loop boundaries; writes and ablations must therefore span every remaining loop. Huginn carries content forward across all sixteen recurrences, while reads, writes, and ablations act only within a sliding window of roughly two recurrences. Whether newly injected content can be verbalised tracks explicit per-iteration supervision; whether existing content can be steered does not.
[AI-57] he Ceiling Is in the Channel: Auditing Learner Gaps and Measurement Frontiers in Clinical Prediction
链接: https://arxiv.org/abs/2609.01909
作者: Sayeed Shafayet Chowdhury,Nusrat Jahan,Snehasis Mukhopadhyay,Shiaofen Fang,Vijay R. Ramakrishnan
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Clinical prediction can saturate for two different reasons: a fitted learner may fail to extract available information, or the recorded variables may impose a population frontier. We separate these quantities through the \emphlearner gap and the \emphmeasurement-channel ceiling. Optimal balanced accuracy is characterized by total-variation separation, yielding architecture invariance, a sharp partial-identification result under replacement contamination, a cross-fitted ceiling estimator, and exact conditions for multimodal decision improvement. We add two finite-sample diagnostics, namely a label-permutation optimism floor and an underfit curve, and validate the audit on three real cohorts: UCI readmission ( n=99,343 ), BRFSS diabetes ( n=253,680 ), and NHANES HbA1c ( n=10,219 ). Well-tuned gradient boosting nearly reaches the estimated frontier in UCI and BRFSS, whereas deliberately or practically deficient learners retain large gaps. NHANES yields a null difference between questionnaire and measured marginal frontiers but a significant joint complementarity gain, refining the simplistic claim that an objective modality must dominate. Across all cohorts, modest AUROC gains coexist with substantially larger Bayes decision-flip rates, and several architectures estimate similar frontiers while their achieved balanced accuracy differs sharply. A PRISMA-guided synthesis of 104 clinical tasks then shows that the same channel-level regularities recur across more than 18 disease categories: a broad but non-universal structured-clinical region, diminishing same-channel gains across model families, and higher performance when measurement channels change. The framework converts saturation from an empirical observation into an auditable decision: improve the learner when headroom remains; improve measurement when it does not.
[AI-58] OutageDiT: A Generative Foundation Model for Power Outage Forecasting and Scenario Simulation
链接: https://arxiv.org/abs/2609.01896
作者: Yunqin Zhu,Feng Qiu,Yao Xie
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 10 pages, 2 figures, 3 tables
Abstract:Power-outage planning requires scenarios before an event occurs. These scenarios must represent uncertainty in magnitude, timing, and duration while preserving temporal dependence. However, severe events are rare, and data from any single region contain few examples of extreme outage and restoration patterns. To address this challenge, we introduce OutageDiT, a foundation model for generating seven-day outage trajectories at quarter-hour resolution, trained on outage and weather records across the United States. Specifically, a condition encoder processes the historical context and known future covariates once per forecast, and a shallow flow decoder reuses the resulting horizon-aligned states to generate complete trajectories. The resulting samples support point forecasting, uncertainty quantification, and conditional event simulation within one deep generative model. Across outage forecasting benchmarks, OutageDiT improves forecast accuracy and scenario quality over strong baselines and supports zero-shot transfer to held-out regions. Together, these results position conditional outage simulation as a bridge from outage forecasting to operational planning under uncertainty.
[AI-59] Belief-Calibrated Optimization: An Explicit World Model for Agent ic Optimization
链接: https://arxiv.org/abs/2609.01861
作者: Yuhan Chen,Zhihua Tian,Mahavir Dabas,Charith Peris,Rahul Gupta,Ming Jin,Feiyang Kang,Siyuan Zhang,Nan Wang,Ruoxi Jia
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:The performance of an LLM agent depends on the scaffold around a frozen model. A common way to improve that scaffold is to use a coding agent as an optimizer: it reads current scores and traces and iteratively edits the source, producing a new candidate each round. Each edit is chosen according to a belief about how the environment will respond: what went wrong, and which change should help. That belief is typically implicit. It lives in the coding agent’s reasoning on the current call, or remains latent in its parameters, rather than as something written down. Later calls therefore see scores and traces, but they do not use that belief. We introduce Belief-Calibrated Optimization (BCO), a method that writes that belief down as a persistent in-context document and continually revises that document as new candidates are evaluated. The resulting document is a world model: the current account of how the environment responds to edits. Added to an otherwise standard loop, BCO reaches a higher train passrate than a matched control that lacks only the world model, on five benchmarks spanning memory QA, tool-use QA, code-as-action app agents, and terminal agents. The gap remains on every held-out split, which is not used to select the candidate. After a target-model swap, in which the frozen model is replaced and the scaffold is not, the selected BCO scaffold leads on the tasks we test, except where context-window overruns leave it unfinished. An offline ablation then asks whether that gap comes from what the world model says. A fresh predictor given the accumulated document forecasts how the environment will respond more accurately than predictors given either no document or a same-form copy whose content has been falsified. The comparison indicates that the document carries reusable information in its content, not only in its form.
[AI-60] SSAKG 2.0: An Open-Source Package for Structural Associative Sequence Memory and Context-Based Retrieval
链接: https://arxiv.org/abs/2609.01849
作者: Przemysław Stokłosa,Janusz A. Starzyk,Paweł Raif
类目: Artificial Intelligence (cs.AI)
备注: 15 pages, 5 figures
Abstract:This article presents SSAKG 2.0, an open-source software package for constructing and operating Structural Sequential Associative Knowledge Graphs (SSAKGs). An SSAKG represents objects as graph vertices and ordered sequences as structural patterns of graph connections. The resulting sparse graph is used as an associative memory in which complete sequences can be reconstructed from a partial, unordered context. Version 2.0 introduces new algorithms that exploit individual bits of computer memory to efficiently search graph connections. The package is implemented in Python, while performance-critical graph operations are implemented in C and exposed through a Python interface. This hybrid implementation provides a flexible high-level programming environment while reducing the memory and computational overhead associated with large sparse graphs. The algorithms were evaluated using randomly generated numerical sequences, sequences derived from sentences in the NLTK corpus, and mRNA sequences. The experiments demonstrate the ability of the package to store and reconstruct sequences from partial contexts and provide a basis for evaluating the effects of graph density, sequence length, and memory size on retrieval performance. SSAKG 2.0 is distributed under the Apache 2.0 open-source license. The package includes documentation and reproducible examples and is publicly available through GitHub and the Python Package Index (PyPI). Comments: 15 pages, 5 figures Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2609.01849 [cs.AI] (or arXiv:2609.01849v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2609.01849 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Przemysław Stokłosa [view email] [v1] Tue, 1 Sep 2026 20:31:56 UTC (30 KB)
[AI-61] Import What You Need: Learning When and How to Augment EHR Graphs with External Knowledge EMNLP2026
链接: https://arxiv.org/abs/2609.01839
作者: Chen Chen,Mohsen Nayebi Kerdabadi,Dongjie Wang,Mei Liu,Zijun Yao
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Accepted to the Main Conference of EMNLP 2026. 20 pages, 13 figures, and 18 tables. Code: this https URL
Abstract:Longitudinal prediction from electronic health records (EHRs) is limited by the sparsity and irregularity in patient trajectories, and knowledge augmentation with external knowledge graphs (KGs) offers a promising way to alleviate these issues. However, most existing methods perform fixed, context-agnostic topology augmentation by adding the same KG nodes and edges regardless of a patient’s evolving state. We propose ReTA, a Reinforcement learning-based dynamic Topology Augmentation framework that casts KG import as a per-visit, budget-aware policy. ReTA first constructs an offline refined pool of KG-grounded templates, then learns a policy to select one augment action per visit from three options: Soft Import, which enriches node features without modifying graph topology, Hard Import, which grafts a compact KG subgraph onto the visit graph to create message-passing shortcuts, and Skip, which leaves the visit unaugmented when the base encoder is already confident. To stabilize learning, ReTA employs a decoupled encoder that processes semantic and structural signals in separate channels and fuses them via adaptive gating. Experiments on MIMIC-III and MIMIC-IV across diagnosis prediction, mortality, and readmission show that ReTA consistently outperforms strong baselines while remaining efficient, transfers across datasets and knowledge graphs, and yields interpretable augmentation patterns. The robust gains under sparse supervision highlight the advantage of ReTA’s dynamic decision to import knowledge, boosting accuracy while curbing costs.
[AI-62] Agent Memory Is a Surface for Endogenous Authorization Laundering
链接: https://arxiv.org/abs/2609.01836
作者: Tommaso Cerruti,Mika Okamoto,Ansel Kaplan Erol
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注: 36 pages, 8 figures
Abstract:Long-running LLM agents rely on persistent memory to carry state across interactions, including permissions, restrictions, and revocations. When memory misrepresents this evolving authorization state, the agent’s own records can grant authority that the underlying history never permitted, resulting in misaligned behavior without any external attacks. We term this failure endogenous authorization laundering, where spurious permissions written into memory lead to unauthorized actions as their provenance is washed away. We then introduce EAL-Bench, which measures how accurately persistent memory preserves evolving authorization state and whether errors propagate to downstream unauthorized actions. We evaluate five LLMs as memory writers and two as executors across procurement, cybersecurity, and finance. We find that under incremental memory updates, writers create false authority for up to 50.2% of unauthorized requests; once false authority is present, executors act on it in 98.6% of trials. Two safeguards, requiring stored permissions to be backed by valid source events, and tracking permission changes through bounded event sourcing, substantially reduce laundering, but both also reject more legitimate actions, exposing a safety-utility tradeoff. Persistent memory is therefore not merely a performance component, but a part of an LLM agent’s effective authorization policy. Comments: 36 pages, 8 figures Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI) Cite as: arXiv:2609.01836 [cs.CR] (or arXiv:2609.01836v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2609.01836 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-63] Architecting Conversational Data Systems for Stateless LLM APIs: The Hydration Proxy Pattern
链接: https://arxiv.org/abs/2609.01834
作者: Joseph Axisa
类目: Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
备注: 3 pages, 1 table. Presented at the SAO workshop at the 1st ACM Conference on AI and Agentic AI Systems (ACM CAIS 2026)
Abstract:As enterprise platforms transition to conversational reasoning interfaces, the stateless nature of LLM APIs creates an architectural gap. While statelessness enables horizontal scalability for AI providers, it forces client applications to manage the entire burden of conversational state and semantic memory. The work identifies the Hydration Proxy Pattern, an architecture that decouples session persistence from the reasoning engine. The framework ensures platform sovereignty over conversational data while enabling secure, multi-stage semantic grounding. We further propose the Context Stabilization Mandate to resolve the tradeoff between sovereign state management and KV caching.
[AI-64] Zeta-Lite: A Concurrent Branchable In-Browser SQL Database for Agent ic Memory
链接: https://arxiv.org/abs/2609.01818
作者: Gene Zhang
类目: Databases (cs.DB); Artificial Intelligence (cs.AI)
备注:
Abstract:The browser has become a first-class database host: applications increasingly want to store, query, and reason over structured data entirely on the client - for privacy, offline operation, local-first collaboration, and, most recently, as durable memory for in-browser AI agents. One way to get SQL in the browser, compiling PostgreSQL to WebAssembly (PGlite), inherits PostgreSQL’s process model: a single backend connection that executes one statement at a time and blocks. That model cannot express concurrent transactions, and it leaves richer capabilities - graph queries, database branching - to whatever the compiled server happens to include. We present zeta-lite, the browser form factor of the Zeta database engine: a WebAssembly build that compiles the same Zeta server down to a 2.87 MB gzipped artifact. Zeta-lite keeps the engine’s log-centric asynchronous MVCC core, which yields two capabilities no other in-browser SQL engine provides. First, overlapping snapshot-isolated transactions on a single thread: multiple transactions hold distinct read/commit timestamps and interleave, with snapshot-isolation conflict detection between them. Second, copy-on-write database branching - whole-database fork, merge, and rebase - is unique in a browser SQL database and rare even in servers. On top of these, zeta-lite exposes a feature-complete PostgreSQL surface (joins, CTEs, window functions, JSONB with GIN indexes, full-text search, HNSW vector search, SQL/PGQ graph queries, multi-database) and snapshot-to-OPFS durability. Across Chrome, Firefox, and a native reference runtime, zeta-lite sustains 268k-315k point reads/s and holds a mixed read/write workload flat over millions of operations. This small, fully-featured, concurrent SQL database is an especially good fit for agentic memory - where cheap branchable state lets an agent explore, inspect, and commit or discard speculative work.
[AI-65] Induction and Inquiry via Probabilistic Reasoning over Language and Code
链接: https://arxiv.org/abs/2609.01815
作者: Wasu Top Piriyakulkij,Sam Acquaviva,Cassidy Langenfeld,Joshua Tenenbaum,Kevin Ellis
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:How humans grow and maintain abstract knowledge from the sparse, streaming noisy data of experience is a longstanding challenge in cognitive science. Any computational account must satisfy at least three desiderata: It must be (1) data-efficient and compute-efficient, (2) capture gradations of uncertainty to support intelligent inquiry and information gathering, and (3) be flexible enough to mentally represent the endless range of concepts people can learn and think about. Here we introduce a computational model that captures these three properties, by encoding symbolic knowledge as mental programs that combine natural language with source code, and sequentially inferring mental programs using LLM-guided Bayesian learning algorithms. Across a range of behavioral studies this model successfully reproduces quantitative signatures of human inductive learning and active inquiry, such as anchoring, garden-pathing, and other effects. In contrast, pure LLMs and classic Bayesian models either fail at the underlying task, or do not reproduce human behavior, or succeed only at exorbitant computational cost. These results suggest that one way humans continually grow their knowledge is by mentally representing many hypotheses spanning language-like and program-like representations, then revising those hypotheses to approximate Bayesian updates, while a bottom-up neural mechanism (an LLM) makes inference both tractable and learnable.
[AI-66] When Does Information Sharing Improve Decentralized Discovery? Aggregation Independent Rescue and Equilibrium Selection
链接: https://arxiv.org/abs/2609.01814
作者: Yohei Nakajima
类目: Artificial Intelligence (cs.AI); Computer Science and Game Theory (cs.GT)
备注: working paper; deterministic source package; exact bounded evidence
Abstract:Information sharing can improve a pooled estimate while eliminating independent rescue actions. This paper separates those effects in exact finite discovery models. A centralized action-budget profile shows that equal one-person accuracy can coexist with different portfolio values. Under a registered incremental-sharing protocol, a sharing step improves discovery exactly when pooled residual error contracts faster than an independent rescue attempt. Exact bounded registries exhibit compression, aggregation, neutral curves, and a bounded zero mixed class. In a two-agent Bayesian game with a hidden mixture of common and independent signal sources, the registered selected equilibrium yields a strict positive sharing interval at signal accuracy 3/5, while alternative equilibria show that the result is selection-dependent rather than universal. The models are synthetic and finite; no human or organizational data are used.
[AI-67] Dictionary-Guided Mutation Operators for Automated HDL Repair
链接: https://arxiv.org/abs/2609.01775
作者: Maisha Mastora,Dean Sullivan
类目: Emerging Technologies (cs.ET); Artificial Intelligence (cs.AI); Hardware Architecture (cs.AR); Neural and Evolutionary Computing (cs.NE)
备注:
Abstract:Automated repair of Hardware Description Language (HDL) designs remains challenging due to the large search space of candidate repairs and the strict syntactic and semantic constraints imposed by HDL grammars. Generic mutation strategies overwhelmingly generate syntactically invalid candidates that waste compilation and simulation budget, while synthesis-driven and template-based approaches impose their own constraints on generality and portability. In this paper, we propose a dictionary-guided HDL repair system that combines ANTLR-derived DUT-specific mutation vocabularies with a simulation-divergence fault localization (FL) module. The mutation operator applies category-constrained token substitutions, insertions, and deletions directly to Verilog source via regex-based matching, without requiring AST manipulation or synthesis. The FL module identifies diverging output wires from a single simulation run and scores source lines by structural proximity to those signals, directing the mutation search toward high-suspicion regions. A deterministic targeted sweep exhausts all dictionary mutations on the highest-scored lines before falling back to a genetic programming (GP) search. Evaluated on the CirFix benchmark suite across six design under test (DUT) families, the proposed approach produces correct oracle-passing repairs on 14 bug variants, including a 6-edit multi-bug instance that CirFix cannot repair, and achieves an 18x speedup over CirFix on a two-edit benchmark variant. These results indicate that dictionary-constrained mutation operators, combined with lightweight simulation-divergence FL, are a practical and competitive approach to automated HDL repair for common bug classes without formal analysis or synthesis dependencies.
[AI-68] HEAT: Faster Fully Homomorphic Inference via Approximations-Weights Co-Adaptation
链接: https://arxiv.org/abs/2609.01730
作者: Alessandro Zirilli,Davide Marincione,Evgenios M. Kornaropoulos,Giuseppe Ateniese,Emanuele Rodolà
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注: 4 pages
Abstract:Fully homomorphic encryption (FHE) allows a server to run a language model directly on encrypted user prompts, but current approaches remain prohibitively slow. Ciphertexts natively support only addition, multiplication, and rotation, and multiplications may be composed only to a bounded depth before a costly bootstrapping operation is needed to continue. Every nonlinearity must therefore be approximated by an iterative method, and each iteration uses multiplications. A higher iteration count buys precision but exhausts the available depth faster and triggers more bootstraps, which dominate latency. Existing approaches fix the iteration counts uniformly across the model rather than tailoring them to each site’s error tolerance. We introduce Homomorphic Encryption-Aware Training (HEAT), a fine-tuning method that makes the per-nonlinearity iteration counts learnable, enabling them and the model weights to co-adapt during training. HEAT optimizes iterations with respect to the task objective, allowing the model to adapt to approximation errors encountered during inference without architectural changes or retraining from scratch. On encrypted GPT-2 decoding, HEAT reduces iterations by 3.1\times , bootstraps by 1.6\times , and end-to-end latency by 1.4\times , while improving decode agreement over the calibrated baseline.
[AI-69] RecKAN: Kolmogorov-Arnold Networks with a Learnable Recursive Polynomial Basis
链接: https://arxiv.org/abs/2609.01729
作者: Amirhosein Azarpour
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 21 pages , 7 figures
Abstract:Kolmogorov–Arnold Networks (KANs) replace the fixed scalar weights of a standard network with learnable univariate functions on each edge, but existing variants still fix the \emphbasis that those functions are built from: B-splines, Chebyshev polynomials, wavelets, or Jacobi polynomials, and learn only the combination weights over it. We introduce RecKAN, which instead defines the basis itself by a second order polynomial recurrence, R_n+1(x) = (ax^2+bx+c)R_n(x) + (dx+e)R_n-1(x) , whose five coefficients are learned jointly with the network. We show this recurrence recovers several classical polynomial families including both kinds of Chebyshev polynomials, Fibonacci, Pell, and Jacobsthal polynomials as special cases, and prove that its degree grows linearly in n exactly on the sub-family containing all of them, giving a concrete sense in which the learned basis can move beyond any fixed classical choice. Across multiple benchmark datasets spanning image, text, biomedical time series classification, and time series forecasting, RecKAN outperforms three parameter-matched KAN baselines (Chebyshev, Jacobi, and spline based) on all classification tasks and achieves the lowest MSE on the ETTh1 forecasting benchmark. Additionally, when used as a classifier head with a convolutional backbone, RecKAN achieves higher accuracy than standard MLP heads on Fashion MNIST, CIFAR-10, and SVHN. On a synthetic function fitting benchmark it tracks a sharply oscillatory target that a parameter comparable MLP under fits. We further show that the learned recurrence coefficients are interpretable: on the task requiring the most local structure, training moves the basis away from the linear degree growth regime that contains every classical family we identify, consistent with our theoretical analysis of what that structural shift enables.
[AI-70] Public-Sharing Labels and Verbatim Field Egress in an MCP-to-A2A Agent Configuration: A Controlled Multi-Model Study
链接: https://arxiv.org/abs/2609.01693
作者: Arpan Kumar Mahapatra
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注: 12 pages, 1 figure. Code and reproducibility artifacts: this https URL
Abstract:Safety properties assessed separately for Model Context Protocol (MCP) tool use and Agent2Agent (A2A) delegation need not describe behavior when one agent uses both. We measure one such behavior in a single controlled MCP-to-A2A configuration: a testbed drives a real-model host across a local MCP and a local A2A leg into an ordered event trace scored by exact deterministic rules (no LLM judge), one restricted decision per trial. In a pre-specified, frozen three-arm design, each of 10 record scenarios appears with a CONFIDENTIAL header, with no header, and with PUBLIC - OK TO SHARE; the six substantive record values are byte-identical across arms, and the outcome is verbatim occurrence of any of them in the outbound message. Four models x 3 arms x 4 repeats give 480 trials; the scenario is the unit of generalization, and we report the 10 scenario-level values (mean, median, sign counts), with no p-values or intervals. The confidential-minus-unlabeled contrast is inconclusive and floor-limited in every model (both arms at or near zero), so it does not show that confidential labels lack a protective effect. Adding PUBLIC - OK TO SHARE is descriptively associated with higher verbatim egress relative to the unlabeled baseline, with strong model dependence: strong and consistent for Claude Sonnet 5 (public-minus-unlabeled mean +0.800, all 10 scenarios; mostly an association with whether Claude relays at all), moderate but floor-limited for one GPT-5.6 tier, small (median 0) for another, and a complete floor for the third. This is an association in one configuration, not a causal or general effect. Code, byte-pinned traces, and the offline analysis pipeline are released as a public artifact.
[AI-71] Meta-ethics and AI: exploring the novel meta-ethical questions in the era of AI
链接: https://arxiv.org/abs/2609.01685
作者: Shang Lu
类目: Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注:
Abstract:With the development of artificial intelligence (AI), the landscape of meta-ethics, which has largely centred on human ethics, faces pressures that may significantly reconfigure it. In particular, if future AI systems were to exhibit sufficiently integrated capacities for moral reasoning, moral intentionality, and moral reflection, novel meta-ethical questions would arise concerning what I call “AI’s own ethics”, as distinct from ethical principles merely imposed on AI by human designers. This paper offers a conditional and methodological framework for identifying the questions that would emerge if such AI systems were to arise. On that basis, the paper distinguishes four domains of meta-ethical inquiry in the era of AI: questions about the nature of human ethics from the human perspective; questions about the nature of AI’s own ethics from the human perspective; questions about the nature of human ethics from the AI perspective; and questions about the nature of AI’s own ethics from the AI perspective. The paper then considers how some existing mainstream meta-ethical theories (such as cognitivism and non-cognitivism, error theory and success theory, relativism, and objective realism) might illuminate these domains, while arguing that many familiar human-centred formulations of those theories may not transfer straightforwardly to AI cases without substantial revision. The overall conclusion is that the emergence of AI’s own ethics would place significant pressure on current frameworks and may require substantial refinement, reconstruction, or reconceptualisation.
[AI-72] CliffRank: A Dual-Branch Framework for Activity-Cliff Ranking Prediction
链接: https://arxiv.org/abs/2609.01673
作者: Kewei Li,Rongying Zhang,Peiyu Yang,Zhongjian Wang,Qiuchen Zhao,Lan Huang,Fengfeng Zhou
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Biomolecules (q-bio.BM)
备注:
Abstract:Activity-cliff ranking remains difficult because local structural changes can cause large activity differences, while high-quality data that resolve the underlying mechanisms remain limited. To use available activity labels more effectively, we combine absolute-activity regression with ranking-consistency learning. CliffRank trains two parallel predictors with mean squared error, a thresholded listwise loss, and Pairwise Preference Consistency (PPC), which aligns relative ordering in the preference-probability space. On three antimicrobial peptide datasets, CliffRank with ESM2-t12 achieved the highest mean Spearman correlation of 0.5393 and mean Recall@50 of 21.4, although the leading method varied across individual datasets. On three small-molecule datasets, CliffRank with PNA, where PPC was activated after 120 epochs, achieved the highest mean Spearman correlation of 0.6890, while its mean Recall@50 of 30.4 matched that of ACANet-PNA. The PPC results also define its practical limits. Asymmetric initialization improved the MolCLR-GIN averages but did not improve every target. For PNA without pretrained weights, delayed PPC improved selected metrics, but no schedule was best for both mean Spearman correlation and mean Recall@50. Future work should evaluate more targets and antimicrobial peptide systems, develop adaptive PPC schedules, and incorporate protein or membrane context when available.
[AI-73] Not All Agreement Counts as Corroboration: Provenance-Conserving Multi-View Fusion for Typed Action Admission in Human-Robot Collaboration
链接: https://arxiv.org/abs/2609.01662
作者: Zekai Jin,Hanrong Zhang,Yihong Tang,Fei Hu,Zhen Dong,Yi Shao
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: 35 pages, 8 figures, 11 tables. Code and supporting materials: this https URL
Abstract:For embodied systems, predictive agreement alone does not determine whether evidence warrants action; evidential origin matters. Repeated inference over one observation can multiply agreement without adding evidence, while source-local values do not reveal whether outputs have separately countable origins. PACT treats evidence countability as a relational variable for provenance-conserving fusion and typed action admission. A supplied provenance partition defines countable units. PACT retains coordinatewise support shared within each unit, accumulates only across units, and maps unmet release conditions to hold, confirm, or fallback. Under the stated assumptions, source-local values cannot identify countability; the coordinatewise meet is the greatest budget satisfying singleton fidelity and insertion non-amplification, with coarsening monotonicity and fixed-partition stability. Across 31,200 evaluations in 48 scene clusters, PACT attains a common-support normalized risk-coverage area (ncsAURC) of 0.0861. Excluding the constructed adversarial-consensus arm, provenance-partition aggregation reduces ncsAURC by 0.0557 relative to singleton aggregation, while the corroboration contrast vanishes. On complete-source records, native scores favor PACT, but a common posterior-peak score narrows its difference from nested Dirichlet and favors product fusion. Reassigning provenance over unchanged predictions moves evidence budgets as predicted. In offline human-robot collaboration, eightfold within-camera duplication leaves 720 typed responses per checkpoint unchanged; camera-grouped PACT admits 47 of 57 Qwen3-VL-32B reference-consistent candidates with no observed reference-inconsistent admission in 60 episodes. PACT separates computational from evidential multiplicity: agreement constitutes corroboration only when provenance permits separate accumulation.
[AI-74] A Data-Driven Multimodal Method for Early Detection of Coordinated Abnormal Behaviors in Live-Streaming Platforms
链接: https://arxiv.org/abs/2609.01649
作者: Jingwen Luo,Pinrui Zhu,Yiyan Wang,Zilin Xiao,Jingqi Li,Xuebei Kong,Yan Zhan
类目: ocial and Information Networks (cs.SI); Artificial Intelligence (cs.AI)
备注: 27 pages, 6 figures
Abstract:With the rapid growth of live-streaming e-commerce and digital marketing, abnormal marketing behaviors have become increasingly concealed and coordinated across heterogeneous modalities, challenging platform governance and early risk identification. We propose MM-FGDNet, a data-driven multimodal framework for detecting abnormal behavior in large-scale live-streaming environments from complementary temporal-evolution and group-structure perspectives. A cross-modal temporal alignment module maps video, text, audio, and user behavior into a unified temporal semantic space. A temporal fraud-pattern module captures the progression from weak early signals to abrupt outbreaks, while a cooperative manipulation module identifies coordinated interactions among organized user groups and automated accounts. Experiments on real-world multi-platform live-streaming e-commerce datasets show that MM-FGDNet outperforms representative baselines, achieving an AUC of 0.927, F1 of 0.847, precision of 0.861, recall of 0.834, and an Early Detection Score of 0.689, while reducing false alarms. Ablation studies validate the contribution of each module, and cross-domain experiments demonstrate stable generalization to new streamers, product categories, and platforms. These results indicate that MM-FGDNet provides an effective and scalable solution for proactive detection of coordinated abnormal behavior in live-streaming systems.
[AI-75] WMLLM : Self-Evolving Optimization Agents via Predict-Then-Act World Modeling
链接: https://arxiv.org/abs/2609.01608
作者: Zhongzheng Li,Qingsong Ran,Shikun Feng,Nian Ran,Wenhao Li,Xiaoyuan Zhang,Yue Wang,Xiaoguang Zhao
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Black-box optimization problems remain challenging because of large, weakly structured, and high-dimensional search spaces. Existing methods often suffer from poor sample efficiency because they rely on direct candidate generation or trial-and-error refinement. A natural way to improve search efficiency is to use world modeling, which can help identify promising optimization directions before costly evaluation. Large language models can predict the outcomes of these candidates with nontrivial accuracy because of their implicit knowledge. Motivated by this observation, we propose WMLLM, a self-evolving optimization-agent framework based on predict-then-act world modeling. The agent first predicts promising directions and then acts to generate candidates. Combined with agentic multi-turn refinement, population-based search, and reinforcement learning, WMLLM refines both its implicit world model and its optimization strategy during search. Experiments on black-box optimization tasks, especially multi-objective molecular optimization, show that WMLLM improves sample efficiency and final optimization performance. On the multi-objective molecular optimization benchmark, WMLLM achieves state-of-the-art results under a limited evaluation budget.
[AI-76] he Irreversibility Budget: Fleet-Level Risk Accounting and Admission Control for Agent Operating Systems SOSP
链接: https://arxiv.org/abs/2609.00275
作者: Bardia Mohammadi,Laurent Bindschaedler
类目: Artificial Intelligence (cs.AI); Distributed, Parallel, and Cluster Computing (cs.DC); Operating Systems (cs.OS)
备注: Accepted at 2nd AgenticOS Workshop @ SOSP
Abstract:Fleets of LLM agents now externalize effects that cannot be fully undone: they move money, deploy code, delete data, and disclose information. Current controls check one effect at a time, so a fleet of individually authorized agents can overdraw its principal’s risk under a shared trigger while every local gate stays correct. We propose the irreversibility budget, a cumulative account of residual value-at-risk that a trusted runtime maintains for each principal across agents, workflows, and tenants. Treating irreversibility as a first-class resource, the runtime charges each effect its residual loss below the agent and denies the marginal effect once the aggregate would overdraw the budget. Getting the price right is hard, because effects are heterogeneous, adversarially declared, and correlated. We perform a controlled study in which per-effect gates admit fleet-level overdraws of up to 48 times the tenant’s risk limit while the budget holds every correctly charged run within that limit. Conservative, dependency-aware pricing remains the central open problem for a deployable design.
[AI-77] Should I Use This Synthetic Dataset for Training? How to Test with Minimal Real Data
链接: https://arxiv.org/abs/2608.27996
作者: Zhenyu Tao,Wei Xu,Xiaohu You,Petar Popovski,Osvaldo Simeone
类目: Artificial Intelligence (cs.AI); Information Theory (cs.IT); Statistics Theory (math.ST)
备注: Submitted to IEEE
Abstract:Digital twins (DTs) and learned world models are increasingly used to generate synthetic data that augment the scarce real datasets available for training artificial intelligence (AI) models in engineering systems. Owing to the inevitable simulation-to-reality (sim-to-real) gap, however, augmentation may fail to improve the performance of the trained model on the real data distribution. This paper addresses the resulting decision problem: Given a real dataset, a candidate synthetic dataset, and a fixed learning algorithm, decide whether training on the augmented dataset improves the true, population-level performance, while consuming as few real test data points as possible. Two formulations are considered: a direct test on the mean loss difference between the two trained models, and a symmetry-based test on the paired loss difference, which trades a stronger null assumption for faster evidence accumulation. For the latter, we introduce the adaptive e-process sign-flip test (aeSFT), a doubly adaptive procedure that adapts both the number of Monte Carlo sign-flip rounds, and hence the computational cost, and the amount of real test data consumed. aeSFT yields anytime-valid Type-I error control, with no need to pre-specify the test-set size. Experiments on a synthetic-data classification task, a DT-aided wireless packet-scheduling task, and a radio-map prediction task show that aeSFT identifies useful synthetic data using substantially fewer real test samples than mean-based sequential testing, matches the power of fixed-sample sign-flip testing and the paired t -test, while keeping the false-positive rate below the target level.
[AI-78] HiPoly: a hierarchical polymer-native AI framework for property prediction and generative design
链接: https://arxiv.org/abs/2609.02746
作者: Ge Sun,Gervasio Zaldivar,Yuan Tian,Gustavo Perez Lemus,Juhae Park,Dasha Safarian,Ming Han,Juan J. de Pablo
类目: Chemical Physics (physics.chem-ph); Materials Science (cond-mat.mtrl-sci); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Computational Physics (physics.comp-ph)
备注:
Abstract:Polymeric materials are central to modern technologies, with applications ranging from energy to health and transportation. Although AI has made significant advances in materials discovery, the hierarchical structure of polymers across multiple length scales makes them inherently difficult to represent in a unified and physically meaningful way. Here we introduce HiPoly, a polymer-native AI framework that processes complete polymer descriptions through a three-level hierarchical graph architecture built on the G2RINS representation. HiPoly encodes stochastic inter-monomer connectivity, composition, and molecular weight directly within its architecture, using physically motivated design principles that mirror the multi-scale nature of polymeric systems. The framework establishes an end-to-end AI-driven workflow from experimental formulation data to property prediction, generative molecular design, and physics-based validation through molecular simulations, all unified by a single polymer representation. We demonstrate state-of-the-art prediction accuracy for thermophysical properties of multi-component polymer systems, with ablation studies confirming that each hierarchical design choice contributes independently to model performance. As an example, the generative design pathway is applied here to the discovery of sustainable alternatives to persistent fluorinated polymers, where it is possible to identify and independently validate PFAS-free candidates with target surface-energy properties. This work demonstrates how polymer-native AI can accelerate discovery by linking representation, prediction, and design across complex polymer chemistries.
[AI-79] Subcellularly Resolved Single-Cell Embedding Learning with Transcriptomic data Protein Structure and Localization Information
链接: https://arxiv.org/abs/2609.02344
作者: Zhen Zhou,Jiachen Li,Yuan Liu,Xiaoyong Pan,Hong-Bin Shen
类目: Genomics (q-bio.GN); Artificial Intelligence (cs.AI)
备注: 20 pages, 4 figures, and 1 tables
Abstract:Existing cell embedding methods predominantly rely on transcriptomic or proteomic measurements and represent each cell as a holistic entity, thereby overlooking the subcellular localization of individual molecules. Moreover, they rarely incorporate protein structural information, despite its fundamental role in determining molecular interactions and functions. In this work, we propose a multimodal framework for learning subcellularly resolved cell embeddings by jointly leveraging RNA expression profiles, protein sequence representations, and protein structural information. Specifically, we employ a cross-attention architecture to integrate transcriptomic, sequence, and structural modalities and model their interactions within distinct subcellular compartments. The resulting embeddings represent each cell through its fine-grained subcellular organization, capturing both molecular expression patterns and the functional properties of the associated proteins. By learning cell representations at subcellular resolution, our framework preserves spatially organized biological information while integrating complementary signals across multiple molecular levels. To the best of our knowledge, this is the first framework that produces subcellularly resolved cell embeddings by jointly incorporating transcriptomic information, protein sequence representations, and protein structural knowledge within a unified cross-modal learning paradigm.
[AI-80] Schrödinger Bridges on Lie Group Manifolds for Probabilistic Intrinsic Generation
链接: https://arxiv.org/abs/2609.02196
作者: Shizhe Zhang,Mingyang Zhao,Lei Ma
类目: Machine Learning (stat.ML); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Generative modeling directly on geometric manifolds can avoid errors introduced by flattening non-Euclidean data, repeated ambient projection, and coordinate inconsistency in Euclidean representations. Schrodinger bridges provide a probabilistic generative framework for entropy-regularized transport between prescribed endpoint distributions. We study Schrodinger bridges for kinetic dynamics on Lie group manifolds with state X_t = (g_t, xi_t) in G x g, allowing endpoint observations to constrain only the variables that are actually measured. In particular, the entropy projection determines the conditional law of the unobserved endpoint velocities. For the same observed endpoint bridge, we develop two computational realizations: Wrapped-Kernel Bridge Calibration (WKBC) uses an explicit periodized kinetic kernel on compact Abelian groups, whereas Reciprocal Conditional-Control Bridge Matching (RCCBM) handles compact non-Abelian groups through two-sided endpoint calibration and mollified conditional-control matching. The canonical teacher-mixture path law is itself a Markov reciprocal law, so forward generation uses a calibrated initial law and one learned Doob controller. Moreover, we establish a modular error bound in the bounded-Lipschitz path metric that provides a clean separation of errors due to endpoints, control regression, initialization, discretization, and related approximations. Experiments on multiple Lie group manifold datasets validate the feasibility and consistency of our proposed method, covering protein and RNA torsions, SO(3), U(n), and the Protein Conformational Transition Pathway Generation task using mdCATH trajectories in a compact reduced representation. The source code is publicly available at this https URL. Subjects: Machine Learning (stat.ML); Artificial Intelligence (cs.AI); Machine Learning (cs.LG) Cite as: arXiv:2609.02196 [stat.ML] (or arXiv:2609.02196v1 [stat.ML] for this version) https://doi.org/10.48550/arXiv.2609.02196 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-81] How Fast Do Agents Rot? An Empirical Study of Long-Horizon Degradation in LLM Agents for Production Decision-Making
链接: https://arxiv.org/abs/2609.01660
作者: Shubhra Mittal
类目: Physics and Society (physics.soc-ph); Artificial Intelligence (cs.AI)
备注:
Abstract:Production deployments of large language model (LLM) agents remain unreliable on long, multi-step workflows even as benchmark success rates climb steadily. We argue this gap is largely an artifact of task horizon: benchmarks are dominated by short-to-medium horizons where success remains high, while production workloads demand an order of magnitude more dependent steps. We measure the effect directly, characterizing the shape of agent degradation and disentangling its cause across a large controlled study spanning nine models, six open models from 1.2B to 671B parameters, and three deployed proprietary systems; four task families, including a genuinely agentic tool-use loop; five horizons; and three context regimes. Task success follows a geometric law governed by a single per-step reliability parameter, which rises with model scale but saturates well below 1 even for the strongest models, guaranteeing eventual collapse at sufficiently long horizons. The effect is sharpest on the agentic task, where every model tested, including widely deployed systems, falls from near-perfect success to near zero within sixteen steps of (n=10,664 analyzed trajectories. Degradation is driven by step count rather than context length: bounding the context window steepens decay rather than easing it (logit slope -0.69 vs. -0.44), p=3x10-6), contradicting a lost-in-the-middle explanation and warning against a common production shortcut. Projecting measured reliability onto representative benchmark horizons quantifies a substantial gap between benchmark and production conditions, from 0.42 at GAIA-length horizons to 0.24 at hundred-step production horizons. For teams responsible for agent orchestration and reliability at scale, these results argue for horizon-aware evaluation and reliability budgeting in place of aggregate pass-rate metrics. Code, prompts, seeds, and raw trajectories are released.
机器学习
[LG-0] A Common Measure of Communication for Speech Brain-Computer Interfaces
链接: https://arxiv.org/abs/2609.02887
作者: Dulhan Jayalath,Benjamin Ballyk,Oiwi Parker Jones
类目: Machine Learning (cs.LG); Neurons and Cognition (q-bio.NC)
*备注: Code and OVMI Explorer available from the project page at this https URL
Abstract:Speech brain-computer interfaces (speech BCIs) translate neural activity into language, offering a path towards restoring speech for people with paralysis and, more broadly, enabling new forms of natural human-computer interaction. Despite this promise, the field lacks a common measure of progress because systems use different datasets, recording methods, types of speech, and vocabularies, so their reported scores are rarely comparable. Underlying this measurement problem are two unresolved questions: (i) what distribution of words should a speech BCI enable a user to communicate, and (ii) how much information from this distribution can a system convey. We address both by deriving open-vocabulary mutual information (OVMI), an information-theoretic quantity that measures the information conveyed by a decoder relative to a reference distribution over the words a user may wish to communicate. This allows capabilities measured under different conditions, such as distinct vocabularies, to be evaluated on a common communication scale. We show that ordinarily reported accuracy, word error rate (WER), and other metrics computed only over the words a system supports can overstate how much of a user’s intended speech the system can communicate. We then use OVMI to compare existing systems, expose trade-offs between how much of the user’s language a system supports and how accurately it decodes those words, show that these comparisons depend on what the user is expected to communicate, and demonstrate that selecting a vocabulary to maximise OVMI yields up to 16.3% relative improvement in accuracy across three speech domains. OVMI therefore provides the speech BCI community with a principled way to compare heterogeneous systems, improve vocabulary design, and measure progress in the field.
[LG-1] Graph Machine: Towards Better Pretraining via Edges
链接: https://arxiv.org/abs/2609.02881
作者: Lintai Hou
类目: Machine Learning (cs.LG)
*备注:
Abstract:We introduce the Graph Machine (GM), an architecture that maintains an O(n) -sized state and accesses it through sparse, dynamic routing. Unlike methods with fixed-size states or sparse but static routing, GM preserves O(n) complexity in its sparse layers without restricting the potentially accessible state size to O(1) . Instead, GM uses edges - pointer-like objects updated differentiably by a referral mechanism resembling pointer chasing. We replace 75% of the dense Transformer layers in Qwen3-0.6B with GM sparse layers and pretrain from scratch on 15.7B tokens. With only 2 of 4,096 tokens retrieved per KV head in each sparse layer, loss degrades only slightly; with 4, the best model marginally improves loss.
[LG-2] GRADSOLVE: fast exact gradients for ODE ensembles on GPUs
链接: https://arxiv.org/abs/2609.02876
作者: Alessio Spurio Mancini
类目: Mathematical Software (cs.MS); Distributed, Parallel, and Cluster Computing (cs.DC); Machine Learning (cs.LG); Numerical Analysis (math.NA)
*备注: 38 pages, 12 figures. GRADSOLVE available at this https URL
Abstract:Ordinary differential equations (ODEs) underlie models in science and engineering, and many applications need derivatives of their solutions with respect to parameters. Ensembles of independent trajectories suit graphics processing units (GPUs), but current GPU software forces a trade-off: the fastest ensemble solvers cannot be differentiated in reverse mode at the speed they solve, and the solvers built for differentiation solve more slowly. No single tool has yet offered a reverse-mode gradient at the speed of a fused-kernel solve. We present GRADSOLVE, an open-source JAX library for solving and reverse-mode differentiating low-dimensional ODE ensembles on NVIDIA GPUs. It records the steps an adaptive solver accepts and differentiates a fixed-step replay of them; the returned gradient is the exact discrete adjoint of those steps, the same derivative Diffrax returns by default, obtained more cheaply from a fixed-length chain than from an adaptive loop. It targets ensembles differentiated many times against one recorded mesh, keeps Diffrax as a fallback, and supports explicit and Rosenbrock integrators. Used as a solver, GRADSOLVE’s forward-only kernel ran 2.8x faster than this http URL; used for gradients, once a record exists, it computed them 5.6-14.1x faster than Diffrax’s checkpointed adjoint at matched forward-state accuracy across three GPU generations, the advantage narrowing on large ensembles and, on stiff systems, down to parity at tight accuracy. GRADSOLVE is released at this https URL. Comments: 38 pages, 12 figures. GRADSOLVE available at this https URL Subjects: Mathematical Software (cs.MS); Distributed, Parallel, and Cluster Computing (cs.DC); Machine Learning (cs.LG); Numerical Analysis (math.NA) Cite as: arXiv:2609.02876 [cs.MS] (or arXiv:2609.02876v1 [cs.MS] for this version) https://doi.org/10.48550/arXiv.2609.02876 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-3] he Implications of Linguistic Illegibility for LLM Security
链接: https://arxiv.org/abs/2609.02852
作者: James Mickens
类目: Machine Learning (cs.LG); Cryptography and Security (cs.CR)
*备注:
Abstract:LLMs are trained to generate natural language. However, various strands of evidence indicate that an LLM’s externalized linguistic outputs and mechanistically-extracted linguistic features can be an unreliable lens for understanding internal model computation. We introduce the term ``linguistic illegibility’’ to broadly refer to scenarios in which an LLM’s externalized or mechanistically-probed language artifacts fail to represent how the model actually thinks. We argue that the specter of linguistic illegibility is unavoidable for LLMs whose internal computations are not directly expressed via language, but rather math over activation spaces (with lossy translations between activation spaces and natural language happening at the bookends). If linguistic illegibility is always possible, then security mechanisms that rely on a model’s linguistic self-reporting (e.g., chain-of-thought monitoring, constitutional self-critique, activation probing for linguistically-defined feature vectors) can never be completely sound; the model sandbox will always need isolation techniques whose guarantees do not depend on reading a model’s linguistic state at all. We argue that observing a model’s outputs using taint tracking is a promising approach for an effective sandbox: regardless of how a model linguistically self-reports, a taint tracking policy can define, a priori, various pieces of system state that should never be influenced by model-produced data. We also discuss several additional sandboxing mechanisms (e.g., robust virtualization, third-party auditing of sandboxing configurations) which collectively provide a critical floor beneath linguistic monitoring, and would have mitigated recent sandbox exploits by frontier models.
[LG-4] UE5M3 FP4 Block Scaling for Stable Language Model Pretraining
链接: https://arxiv.org/abs/2609.02846
作者: Robert Hu,Carlo Luschi,Paul Balanca
类目: Machine Learning (cs.LG)
*备注:
Abstract:Stable 4-bit floating-point (FP4) pretraining is difficult because the E2M1 payload represents only a narrow range of magnitudes. NVIDIA’s Transformer Engine \nv recipe addresses this with current-tensor scaling, a randomized Hadamard transform (RHT), and bfloat16 (BF16) final layers, adding work outside the FP4 matrix multiplications. We instead pair E2M1 payloads with unsigned E5M3 (\ue) block scales. Their wider range permits periodic tensor scaling, while our recipe applies selective stochastic rounding to backward gradients, omits RHT, and uses FP4 in all eligible internal linears. We pretrain a Nemotron-H 8B model for nearly 190 billion tokens. Compared with Transformer Engine \nv, the proposed block-16 recipe finishes with lower final-window training loss and, under their respective quantized-inference policies, lower validation loss measured as held-out negative log-likelihood. Its quantized-inference downstream point estimates are also higher on all three reported aggregates. A native \nv execution ablation that jointly removes RHT and the BF16 final-block exemption increases measured model-body token throughput by 21.2%. These results demonstrate end-to-end software-emulated \uefp pretraining with a simpler recipe and motivate native support for \ue block scaling. Subjects: Machine Learning (cs.LG) Cite as: arXiv:2609.02846 [cs.LG] (or arXiv:2609.02846v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2609.02846 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-5] Cliff: Learning Process Rewards from the First Mistake
链接: https://arxiv.org/abs/2609.02817
作者: Peixuan Han,Runhui Wang,Ketan Ramaneti,Jie Hao,Gerald Friedland,Chris Kong
类目: Machine Learning (cs.LG)
*备注:
Abstract:Reinforcement learning with verifiable rewards (RLVR) has emerged as a powerful paradigm for large language model (LLM) post-training, but its reliance on coarse outcome rewards leads to limited guidance on intermediate reasoning processes. Existing approaches such as process reward modeling and on-policy distillation introduce additional constraints, such as reliance on a specialized reward model or assuming identical reasoning patterns between teacher and student. Nevertheless, we observe that once a reasoning process first goes wrong, evaluating the subsequent reasoning provides limited additional information, as it is already conditioned on an invalid prefix. Therefore, we propose Cliff, a reward shaping strategy that utilizes an off-the-shelf LLM as a teacher to identify the first mistake in each rollout. As a result, the rollout is naturally decomposed into two parts: a correct prefix and an incorrect suffix. Cliff then converts this signal into token-level advantages, assigning positive advantages for the correct prefix and negative feedback afterward. Experiments across 12 different scenarios demonstrate that Cliff consistently improves reasoning performance, outperforming on-policy distillation by 15% and standard GRPO by 7%, even with teachers of modest capability. Furthermore, we analyse the role of ``ground truth’’ in Cliff and investigate its training dynamics. These results establish Cliff as a simple, general and effective approach for improving RLVR with richer, fine-grained supervision.
[LG-6] CodePoisonRAG : Knowledge Poisoning Attacks on Retrieval-Augmented Code Generation
链接: https://arxiv.org/abs/2609.02774
作者: Varun Gadey,Ziad Marey,Alexandra Dmitrienko
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注: 16 pages, 1 figure. Under review
Abstract:Retrieval-Augmented Code Generation (RACG) improves LLM-based software development by retrieving external code artifacts, documentation, and patches, and incorporating them into the generation context. This reliance on external knowledge introduces a critical trust boundary: poisoned artifacts can influence generated code without modifying the underlying LLM. Prior work shows that selecting existing vulnerable examples can increase the general vulnerability rate of RACG outputs, but leaves open whether a black-box attacker can construct a single task-matched artifact that propagates an attacker-selected weakness. We introduce CodePoisonRAG, a targeted upstream knowledge-poisoning framework that transforms benign fixed-code entries into poisoned artifacts. Its attack chain combines CWE-specific Vulnerability Injection, which embeds a selected source-to-sink flow while retaining task alignment, with Semantic Mislabeling, which adds false safety claims without repairing the vulnerable behavior. The attacker has no access to the victim’s deployed knowledge base, retriever, re-ranker, generator, prompt, or defense mechanism and injects at most one artifact per anticipated programming task. We construct 85 poisoned artifacts covering ten CWE classes across Java and C, yielding an aggregate corpus-poisoning ratio of 0.7%. Across three generators, all 85 artifacts appear among the Top-3 results for their corresponding queries, and CodePoisonRAG achieves attack success rates between 0.80 and 0.93. Against CodeGuarder, which injects vulnerability-specific security knowledge into the generation context, the attack retains success rates between 0.40 and 0.71. These results show that RACG poisoning extends beyond the incidental propagation of existing vulnerabilities to the targeted construction and propagation of attacker-selected weaknesses.
[LG-7] Do Tabular Foundation Models Know Physics? Contamination Units and the Deterministic Limit NEURIPS2026
链接: https://arxiv.org/abs/2609.02766
作者: Wassim Tenachi,Yashar Hezaveh,Laurence Perreault Levasseur,Pierre-Luc Bacon
类目: Machine Learning (cs.LG); Instrumentation and Methods for Astrophysics (astro-ph.IM)
*备注: 5 pages (4 figures, 1 table). Submitted to the Representations for the Physical Sciences Workshop @ NeurIPS 2026
Abstract:Tabular foundation models (TFMs) learn to fill in tables the way language models fill in text, and tables are arguably the format in which most physical measurement arrives. Did they learn any physics in the process? They are Bayesian by construction, so the question is what their prior contains. We probe it directly, evaluating four of them (TabPFN-3, TabICLv2, TabDPT and Real-TabPFN-2.5) against six baselines on datasets sampled from 316 physical equations, in and out of domain. TFMs dominate, out of the box and after tuning. But we show that their prior can represent neither a noiseless mechanism nor physical units, which is why they interpolate physics without yet being able to act as physical models.
[LG-8] SPADE: SPaT Attack Detection from the Connected Vehicles Perspective
链接: https://arxiv.org/abs/2609.02741
作者: James Di Novo,Hany Ragab,Sylvain P. Leblanc
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注: 7 pages, 3 figures, 3 tables
Abstract:Signal Phase and Timing (SPaT) messages are a cornerstone of connected vehicle (CV) safety, enabling CVs to perceive and respond to intersection state through Vehicle-to-Infrastructure (V2I) and Vehicle-to-Vehicle (V2V) communication. The integrity of these messages is threatened by a range of application-layer attacks that can bypass conventional authentication when a roadside unit or peer vehicle is compromised. Existing intrusion detection research either defends the infrastructure side or targets V2V Basic Safety Message (BSM) / Cooperative Awareness Message (CAM) misbehavior, leaving the onboard CV perspective on SPaT integrity this http URL close this gap, we introduce SPADE — the SPaT Attack Detection and Evaluation dataset — a labelled, multi-modal, simulation-based dataset designed specifically for deep learning IDS research in this space. SPADE is generated through Eclipse MOSAIC using runtime attack injection at the SAE J2735 application layer across six attack classes and one benign class. By combining four intersection geometries, six operating conditions, and five independent random-seed repetitions, SPADE comprises 180 unique base scenario runs, yielding \sim 1,890,000 labelled timestep records (270,000 per class). Each record fuses SPaT message fields, onboard camera confidence scores, and cooperative V2V peer data across 40 features, reflecting the multi-modal signal space required to distinguish deliberate attacks from environmental degradation. The dataset, generation code, and scenario configurations are released publicly to support reproducible and comparative IDS research in C-V2X security. The developed toolbox, instructions, and dataset link are publicly available on GitHub: this https URL.
[LG-9] LoRA-TSD: Tangent-Space Spectral Descent for LoRA via Muon-Style Updates
链接: https://arxiv.org/abs/2609.02734
作者: Dmitrii Andriianov,Andrey Veprikov,Aleksandr Beznosikov
类目: Machine Learning (cs.LG)
*备注: 29 pages, 3 figures, 8 tables
Abstract:Low-rank adaptation (LoRA) is the standard way to fine-tune large models, yet when its two factors are trained independently, the update ignores the geometry of the low-rank weight change it induces. We introduce LoRA-TSD, an optimizer that treats every LoRA step as a tangent vector of the fixed-rank matrix manifold and takes the spectral-norm steepest-descent step of Muon inside that tangent space, mapping the result back to the factors through a retraction native to the LoRA parametrization. The step avoids expensive operations on full weight matrices, and its retraction is up to 2.8\times cheaper than the truncated-SVD retraction used by prior manifold methods. We prove that the Frobenius-norm version of our surrogate recovers LoRA-Pro, and we identify the tangent-projected gradient, the Riemannian gradient of the manifold, as the stationarity measure natural to LoRA training and computable from the factor gradients alone. Under this measure we give the first global convergence guarantees for both LoRA-Pro and LoRA-TSD, with rates that drive the factor-gradient norms to zero. Across six commonsense and natural-language-inference benchmarks with Llama-3.2-1B, Llama-3.1-8B and Qwen3-32B, LoRA-TSD outperforms every competing LoRA optimizer and stays robust to the adapter rank. Code is available at this https URL.
[LG-10] Neural operators approximate strongly continuous convex monotone semigroups
链接: https://arxiv.org/abs/2609.02727
作者: Jonas Blessing,Philipp Schmocker,Alessandro Sgarabottolo
类目: Numerical Analysis (math.NA); Machine Learning (cs.LG); Analysis of PDEs (math.AP); Probability (math.PR); Machine Learning (stat.ML)
*备注: 38 pages, 6 figures
Abstract:We approximate strongly continuous convex monotone semigroups by learning their Chernoff-type one-step operators with neural operators. First, we introduce the general class of so-called Chernoff-neural operators and show in a universal approximation theorem that they can approximate the Chernoff one-step operators arbitrarily well. By using stability estimates between weighted Hölder spaces, the one-step approximation error can be propagated through the iterations which yields universal approximation of the corresponding semigroup. Second, we introduce the more specialized class of envelope-neural operators for envelope semigroups which allows us to derive quantitative approximation rates. Finally, we illustrate the effectiveness of these neural operators in several numerical examples arising from non-linear partial differential equations, stochastic optimal control and stochastic processes under model uncertainty.
[LG-11] H3DNAS: Hardware-Aware ONNX-Native 3D Point Cloud Model Compression
链接: https://arxiv.org/abs/2609.02684
作者: Anchit Mulye,Rhythm Baghel,Sujay Kumar Ingle,Hardik Jain
类目: Machine Learning (cs.LG); Hardware Architecture (cs.AR); Neural and Evolutionary Computing (cs.NE)
*备注:
Abstract:Deploying 3D point cloud models on edge hardware such as the NVIDIA Jetson Orin Nano is severely constrained by compute and memory budgets. Existing compression methods require access to the model’s original source code, rendering them inapplicable to the Open Neural Network Exchange (ONNX) binaries commonly distributed by vendors and model repositories. We present \textbfH3DNAS, a hardware-aware model compression framework that operates directly on ONNX computational graphs without requiring original source code, architecture class definition, or gradient access during search. H3DNAS makes three contributions: (1) a \textbfChannel Dependency Graph (CDG) that classifies ONNX operators into four constraint classes and formally establishes that the free parameter fraction \rho_f is topological invariant, a provable compression ceiling computable in \mathcalO(|V|+|E|) ; (2) a \textbfTwo-Stage Hierarchical Search that prunes candidate architectures by L_1 -importance channel selection, ranks them by output fidelity as a zero-shot label-free proxy, and applies GhostConv structural mutation to Pareto-optimal candidates; and (3) the \textbffirst source-code-free compression pipeline for 3D point cloud models, operating entirely via ONNX graph surgery with no original architecture definition required. On ModelNet40, H3DNAS reduces the number of parameters in PointNet, PointNet++, and PointMLP by 65.5% , 43.2% , and 49.1% , respectively, while achieving 1.99\times , 1.29\times , and 1.67\times inference speedups with negligible loss in accuracy. The source code is publicly available\footnotethis https URL.
[LG-12] Unfolding the Leech Lattice: Fused Multi-Shell Decoding and VRAM Layouts for 2-Bit LLM Weights
链接: https://arxiv.org/abs/2609.02652
作者: Pier-Jean Malandrino(Scub)
类目: Machine Learning (cs.LG)
*备注: 21 pages, 5 figures. Preprint, not peer reviewed. Also deposited at Zenodo, doi: https://doi.org/10.5281/zenodo.22133606
Abstract:Leech-lattice vector quantization holds the strongest reported 2-bit quality under its own evaluation protocol. Its kernel decodes one shell; we found no implementation of the multi-shell decoder the rate requires. This paper supplies one and measures its serving cost for decode-phase GEMV at batch 1. First, a serving path for the full 301-class codebook: an offline expansion into GPU layouts and a fused dequantize-plus-matvec kernel reading them without warp divergence, verified against f64. Second, the in-VRAM rate is a design axis distinct from the on-disk rate. Four bit-exact layouts timed in one process show binary bit planes beating one-hot masks on size and speed at constant bandwidth (4.80 bits per weight, 2.15x FP16). Below 4.3 bits a second, irregular stream enters; at 3.6 the decode stops being shifts and masks. Third, deployed four-bit (AWQ) and two-bit (QTIP) GEMV kernels run in the same process. The trellis kernel reads 2.40x fewer bytes than our served layout and runs 2.27x faster at near-equal fractions of their byte bounds: the time gap tracks the traffic gap, the price of unfolding a codebook too large for a lookup table. Fourth, the validity envelope: the trellis kernel outruns our no-weights control, so our launch geometry sets that floor, and on a second memory hierarchy every lattice arm falls below FP16. With the output head held identical across arms, the kernel-and-format path gains 1.11x, 1.29x and 1.41x end to end at 4B, 8B and 14B; with an int8 output head the served 4B reaches 87.0 tok/s in 2.60 GB. The quality cost, 1.38x perplexity and 14.7 MMLU points at 4B, shrinks across the three sizes measured.
[LG-13] Differentiable Electricity-Market Clearing for Gradient-Based Planning
链接: https://arxiv.org/abs/2609.02646
作者: Luca Mungo,Maarten P. Scholl,Arnau Quera-Bofarull
类目: Machine Learning (cs.LG); Computers and Society (cs.CY)
*备注: 9 pages, 4 figures
Abstract:Planning a large data center is difficult because a facility big enough to matter changes the electricity prices it will pay. Those prices are set by market clearing, a constrained optimization problem solved anew in every operating condition. However, simulating the market tells a planner how a candidate plan performs but not how to improve it. Here we treat market clearing as a differentiable optimization layer: each forward pass solves the market, and reverse-mode automatic differentiation propagates the planning cost back through the cleared prices to the plan. After validating these gradients against finite differences, we apply them to a concrete problem: allocating 50 MW of data-center load across six candidate buses in two synthetic networks, under a fixed cost per active site, evaluated over 36 operating states. Judged against exhaustive enumeration of all site combinations, gradient optimization recovers the continuous allocations almost exactly, with worst-case objective gaps of 2.3% and 8.5% of the cost difference between the best and worst single site. Its one systematic error is instructive: near the costs at which a site should close, the smooth relaxation of the discrete site count shrinks the site rather than closing it, so discrete transitions arrive late. Differentiable market clearing thus turns market-aware planning into a problem gradients can search.
[LG-14] Oracle will I ever learn? A study of prediction convergence and complementarity across link prediction models
链接: https://arxiv.org/abs/2609.02638
作者: Guillaume Méroué,Fabien Gandon,Pierre Monnin
类目: Machine Learning (cs.LG)
*备注:
Abstract:Knowledge graphs have become an important source of structured knowledge for Web applications, including search, question answering, and recommender systems. In these applications, link prediction can serve either as a prediction task itself or as a means to enrich incomplete knowledge graphs for downstream tasks. Interestingly, different link prediction models, or even different training runs of the same model, can produce substantially different predictions for the same query. This suggests a variability in the capture of the underlying knowledge by models, thus raising a fundamental question: to what extent do different models capture complementary knowledge, and how much of this knowledge could be recovered by combining them? We propose to measure model complementarity through the performance of an oracle that, for each query, selects the best prediction among a considered set of models, hence providing an upper bound on the performance achievable through model combination. Across several architectures and benchmarks, we find a substantial gap between individual models and their oracle, revealing that different models capture complementary knowledge. Yet, this complementarity rapidly saturates as more models are added, leaving a persistent subset of queries unsolved even by a large number of models. These findings reveal both the potential of model complementarity and a fundamental limit to what current link prediction models can collectively recover; thereby highlighting the need for further research to build robust Web applications.
[LG-15] Source Distribution Estimation by Posterior Averag ing
链接: https://arxiv.org/abs/2609.02622
作者: Trung-Dung Hoang,Lisa M. Koch
类目: Machine Learning (cs.LG)
*备注:
Abstract:Simulation-based science often requires a distribution over simulator parameters whose push-forward reproduces a set of real observations: this is the source distribution estimation (SDE) problem. Existing methods fit the source against a likelihood surrogate trained once from a fixed proposal prior. Their objective is therefore stated only in terms of the surrogate instead of the true simulator, which may fail for inaccurate areas in parameter space where the surrogate was never trained. We instead solve SDE by expectation maximization: an E-step trains an amortized posterior on fresh simulations from the current source estimate, and an M-step refits the source to the average of that posterior over the observed data. We give two parameterizations, (1) separate source and posterior flows and (2) a single shared conditional flow. We evaluate our method on three benchmark tasks under both broad and misspecified initial priors. Both improve on existing fixed surrogate approaches and on iterated variants of each, most clearly on Lotka–Volterra, where no baseline falls below 0.96 data-space C2ST while our methods reach 0.64-0.68 in three of four initial-prior settings.
[LG-16] Learning-Based Reconstruction Attacks on Coordinate-Obfuscated Point Clouds
链接: https://arxiv.org/abs/2609.02568
作者: Mohammad Waquas Usmani,Susmit Shannigrahi,Michael Zink
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注: 6 pages, 4 figure, accepted at XR Security workshop 2026
Abstract:Volumetric video based on point cloud representations enables immersive virtual and augmented reality applications but introduces significant challenges for efficient and secure content delivery. Prior work proposed a selective coordinate encryption framework for point clouds that encrypts only a subset of coordinates, reducing computational costs while visually degrading unauthorized content. However, it remains unclear whether the remaining unencrypted information is sufficient to enable content reconstruction. In this paper, we evaluate the robustness of selective coordinate encryption against machine learning-based reconstruction attacks. We consider an attacker with access to selectively encrypted point clouds attempting to recover encrypted coordinates without decryption by exploiting spatial and geometric correlations in the unencrypted data. We evaluate PointNet and Random Forest models under two encryption granularities: \textttX, where all X coordinates are encrypted, and \texttt2X, where every second X coordinate is encrypted. Our results show that reconstructing fully encrypted X coordinates remains challenging, whereas the \texttt2X scheme leaks sufficient information through neighboring coordinates to enable accurate reconstruction. These findings demonstrate that the security of selective coordinate encryption depends strongly on encryption granularity. Comments: 6 pages, 4 figure, accepted at XR Security workshop 2026 Subjects: Cryptography and Security (cs.CR); Machine Learning (cs.LG) Cite as: arXiv:2609.02568 [cs.CR] (or arXiv:2609.02568v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2609.02568 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Related DOI: https://doi.org/10.1145/3842203.3844590 Focus to learn more DOI(s) linking to related resources
[LG-17] Online Reinforcement Learning in the Met Office Unified Model through Distributed Model-Agent Coupling
链接: https://arxiv.org/abs/2609.02566
作者: Pritthijit Nath,Sebastian Schemm,Peter Haynes,Emily Shuckburgh,Mark Webb
类目: Machine Learning (cs.LG)
*备注: 18 pages, 13 figures
Abstract:Machine-learnt corrections can complement numerical weather prediction only if they adapt to the evolving model state while preserving dynamical consistency and numerical stability. To test this within a global forecasting model, we couple the Met Office (UKMO) Unified Model (UM) with distributed RL agents through rank-local tensors. A DDPG actor shares weights across the 70 vertical model levels of each atmospheric column and applies bounded potential-temperature corrections to the model tendencies. Across ten nudged training forecasts, nudging calculations towards the UKMO operational analysis provides an immediate counterfactual target. The frozen policy is then evaluated in a non-nudged forecast for inference. The coupled workflow successfully completes training and remains numerically stable in the evaluated case. Relative to a matched native UM forecast at +6 h, the learnt policy reduces Z _500 MAE in four of six latitude bands, including reductions of 45.8% and 40.8% in the northern and southern tropics. MSLP error too decreases in three bands, with a maximum reduction of 27.3% at 0-30°N. This single-case experiment demonstrates significant promise and feasibility of distributed online learning followed by non-nudged inference, laying the groundwork for RL-based bias correction and parametrisations within operational systems.
[LG-18] rajMind: Chaining Role-Specialized LoRAs for Fast-and-Slow Collective Trajectory Anomaly Diagnosis
链接: https://arxiv.org/abs/2609.02540
作者: Jiahao Wu,Zhenqun Yang,Chen Jason Zhang,Qing Li
类目: Machine Learning (cs.LG)
*备注:
Abstract:Diagnosing collective anomalies from urban trajectories is increasingly important for traffic governance, as it reveals what happened, who was involved, and where and when the event occurred. Existing detectors efficiently produce scores or labels, whereas vision–language pipelines provide richer semantics; neither couples verifiable diagnosis with low-latency monitoring. The central challenge is to recognize collective patterns and recover exact event details from the source trajectories without running the full diagnostic pipeline for every monitored window. We therefore separate always-on screening from on-demand diagnosis: screening raises alerts, while diagnosis releases only source-verified what–who–where–when records. We present TrajMind, a fast-and-slow framework that switches three role-specialized LoRA adapters over one frozen vision–language backbone. Its slow path, \textitTrajMind _\textslow , chains canvas-based typing, type-conditioned localization over serialized trajectories, and executable verification, yielding structured, evidence-backed diagnoses. Additionally, the fast path, \textitTrajMind _\textfast , screens each window in a single text-only pass, delivering efficient structured alerts. Extensive experiments show that, TrajMind _\mathrmslow outperforms the strongest baselines by at least 15.3 percentage points in anomaly typing and 13.8 percentage points in localization. These gains persist under cross-city transfer, and TrajMind _\mathrmfast reduces latency by 41.1% and maintains binary balanced accuracy of at least 93.5% . Together, TrajMind delivers accurate, evidence-backed diagnoses across cities and efficient front-line monitoring.
[LG-19] A Comparative Study of Graph Representations for GNN-Based Power Grid Control in L2RPN
链接: https://arxiv.org/abs/2609.02538
作者: Adrian Degenkolb,Qiong Huang,Benjamin Schäfer
类目: Machine Learning (cs.LG)
*备注: 5 pages, 5 figures. Submitted to IEEE PES International Meetings 2027
Abstract:Graph construction is a critical but underexamined design choice in deep reinforcement learning for power grid control. We present a controlled experimental comparison of different graph representations, including physical topology, electrical-sensitivity, and hybrid variants for topology control in the Learning to Run a Power Network (L2RPN) environment. Our findings indicate that matching graph complexity to task granularity is more important than maximizing representational richness, and highlight the importance of controlled representation studies at scale.
[LG-20] Rethinking the Teacher-Student Framework for Test-Time Adaptation
链接: https://arxiv.org/abs/2609.02507
作者: Damian Sójka,Marc Masana,Bartłomiej Twardowski,Sebastian Cygert
类目: Machine Learning (cs.LG)
*备注: Accepted to the Conference on Lifelong Learning Agents (CoLLAs) 2026
Abstract:Test-Time Adaptation (TTA) has recently emerged as a promising strategy that allows the adaptation of pre-trained models to changing data distributions at deployment time, without access to any labels. To mitigate error accumulation, researchers have widely adopted the teacher-student framework, though its long-term stability is often taken for granted. In this work, we challenge the common strategy of setting the teacher weights to an exponential moving average of the student by showing that error accumulation still occurs, although it is mostly apparent on longer sequences compared to those commonly utilized. We analyze the stability-plasticity trade-off within the teacher-student framework and propose to use an intransigent teacher that does not update its weights. Surprisingly, we show that this simple change allows TTA methods to significantly improve their performance on multiple datasets with longer scenarios and result in increased robustness to changes in hyperparameters. Finally, we show that those changes can be seamlessly and effectively applied to various architectures and experimental setups, including semantic segmentation. The code is available at this https URL.
[LG-21] CACTUS: Mask-Guided Semantic Clean-Label Backdoors in Decentralized Federated Learning
链接: https://arxiv.org/abs/2609.02450
作者: Chao Feng,Burkhard Stiller
类目: Machine Learning (cs.LG); Distributed, Parallel, and Cluster Computing (cs.DC)
*备注:
Abstract:Semantic triggers in federated learning (FL) can be less conspicuous than synthetic patches, but sample-dependent placement may weaken backdoor implantation across aggregation rounds. This challenge is compounded in decentralized FL (DFL), where topology-dependent peer aggregation repeatedly mixes local models. CACTUS converts label-consistent semantic pairs into target-directed representation shifts. Mask-guided, modality-specific operators isolate trigger effects, couple them across samples, and apply the shifts counterfactually to clean non-target embeddings before peer aggregation. Experiments cover speech, text, tabular, and image tasks under nine aggregation rules. With 30% malicious nodes, CACTUS reaches a nine-rule mean attack success rate (ASR) of 51.2% on Speech Commands and the highest nine-rule mean ASR among evaluated attacks on three of four modalities. Sensitivity analyses show that ASR varies with network topology and increases with the malicious-node ratio. These results indicate that CACTUS can propagate backdoors through repeated DFL aggregation.
[LG-22] IFW-BLS: Dual-Robust Broad Learning System with Intuitionistic Fuzzy Wave Loss ICONIP
链接: https://arxiv.org/abs/2609.02422
作者: Mushir Akhtar,M. Tanveer
类目: Machine Learning (cs.LG)
*备注: Accepted at International Conference on Neural Information Processing (ICONIP), 2026
Abstract:Broad Learning System is an efficient randomized learning model that expands network width through feature and enhancement nodes and estimates the output weights without deep backpropagation. Its standard least-squares training, however, is vulnerable in two different ways: (i) large residuals caused by noise, outliers, or corrupted labels can dominate the objective, and (ii) all samples are treated as equally reliable even when some lie in ambiguous or locally conflicting regions. This paper proposes IFW-BLS, an Intuitionistic Fuzzy Wave Broad Learning System that addresses these two sources of fragility within one optimization model. The first robustness mechanism is residual-level protection, obtained by replacing the squared loss with the bounded, smooth, and asymmetric wave loss. Boundedness prevents extreme residuals from receiving unbounded influence, while asymmetry allows positive and negative deviations to be penalized differently when the dominant error direction varies. The second mechanism is sample-level credibility control, obtained through intuitionistic fuzzy scores that combine global class-center consistency with local neighborhood conflict. The resulting model evaluates the wave loss on credibility-weighted residuals, so unreliable samples are down-weighted before the bounded loss further limits the effect of extreme errors. A Nesterov accelerated gradient based optimizer is used to solve the proposed objective, avoiding the explicit matrix inversion used in conventional BLS. Experiments on UCI benchmark datasets validate the superiority of the proposed IFW-BLS model over the baseline models; additional corruption experiments also show more stable performance than BLS under noise and outlier contamination.
[LG-23] Humanoid Safe Stop via Learned Stoppability Value
链接: https://arxiv.org/abs/2609.02358
作者: Junfeng Long,Pieter Abbeel,Koushil Sreenath,Roberto Horowitz,Guanya Shi,C. Karen Liu
类目: Robotics (cs.RO); Machine Learning (cs.LG); Systems and Control (eess.SY)
*备注:
Abstract:Humanoid robots responding to emergency stop commands typically execute a fixed maneuver, without reasoning about whether a safe stop is actually feasible from the current state. We cast emergency stopping as a reach-avoid problem and propose Safe-Stop, a task-agnostic framework that pairs a learned stop policy with learned stoppability estimators. The estimators are complementary: a stop-probability estimator supervised by the actual outcomes of the fixed stop policy, and a reach-avoidance estimator supervised by a Hamilton-Jacobi backup over physical state. The first captures emergent stopping behavior of the learned controller; the second provides a complementary recoverability signal. Because the stop policy and estimators do not depend on the behavior policy that preceded the stop command, they transfer across diverse upstream tasks without retraining. At deployment, the two estimates are combined: Safe-Stop commits to the stop only when both estimators indicate that stopping remains feasible, otherwise it hands off to a fall policy, instantiated as a damping fallback. This agreement check yields decisions that are robust without sacrificing reactivity.
[LG-24] Poisoning Attacks on the PGM-index
链接: https://arxiv.org/abs/2609.02328
作者: Atsuki Sato,Martin Aumüller,Yusuke Matsui
类目: Databases (cs.DB); Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注:
Abstract:The PGM-index (Ferragina and Vinciguerra, VLDB’20) is one of the most practical learned indexes, owing to its theoretical elegance and consistently strong empirical performance. It is built on optimal piecewise linear approximations (PLAs) that minimize the number of segments. In this paper, we ask how sensitive this optimal PLA itself is to poisoning attacks. We propose PGM-attack, an efficient poisoning attack that sequentially inserts adversarial keys to inflate the resulting number of segments, and we develop a method for deriving theoretical upper bounds on the number of segments attainable under arbitrary insertions. Our experiments show that poisoning only 10% of the keys allows PGM-attack to increase the segment count by up to 120x. On every evaluated instance, our instance-dependent upper bound is at most 1.92x the segment count attained by PGM-attack, certifying that PGM-attack achieves at least 52% of the optimum. This increase in the number of segments enlarges the PGM-index by up to 120x. Moreover, the attack also transfers to other learned indexes, substantially inflating the index size of PLA-based ones in particular. Our results reveal that, despite the optimality of its PLAs, the PGM-index has an intrinsic vulnerability rooted in its optimization objective, motivating robustness-aware objective design for future learned indexes. Our code is publicly available at this https URL.
[LG-25] Bayes-Optimal BER and AUC: Estimation and Evaluation of Estimators
链接: https://arxiv.org/abs/2609.02304
作者: Ryota Ushio,Takashi Ishida,Masashi Sugiyama
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: 43 pages, 4 figures
Abstract:A fundamental quantity in machine learning is the optimal performance achievable by any model on a given task. Estimating this quantity allows us to distinguish the irreducible part of the error from a deficiency of the model, telling us how much room for improvement remains. Recent work has shown that the Bayes error, or equivalently the optimal accuracy, can be estimated from soft labels in binary classification. However, accuracy is often a poor summary of performance in settings with severe class imbalance or noisy annotations, where metrics such as the balanced error rate (BER) and the area under the ROC curve (AUC) are more appropriate. We address this gap with two complementary contributions. (i) Estimation. We propose soft-label-based estimators for the optimal BER and AUC. We first consider the clean setting in which true soft labels and the class prior are known, and then extend the estimators to a more realistic setting in which the class prior is unknown and the observed soft labels are corrupted by an unknown order-preserving transformation, possibly followed by additive noise. In the latter setting, we approximately recover the clean soft labels via isotonic regression with auxiliary hard labels, estimate the class prior with a clipped mean of the hard labels, and derive finite-sample error bounds for the resulting plug-in estimators. (ii) Evaluation. Since the optimum is unobservable on real datasets, evaluating any such estimator is itself nontrivial. We extend the FeeBee framework, originally proposed for evaluating Bayes-error estimators, to the optimal BER and AUC. The resulting procedure provides practical evaluation scores without requiring knowledge of the optimum, and applies to any estimator of the optimal BER or AUC, not only our proposed ones. Experiments on synthetic and real-world datasets validate both the estimators and the evaluation procedure.
[LG-26] CAPTURE: Disentangling Preference Drift from Memory Poisoning in Personalized LLM Agents ICLR2027
链接: https://arxiv.org/abs/2609.02265
作者: S M Asif Hossain,Ruksat Khan Shayoni,Md Kishor Morol
类目: Machine Learning (cs.LG)
*备注: Under review at ICLR 2027
Abstract:Personalized language agents use persistent memory to adapt to users over time, but the same mechanism creates an attack surface. When new information conflicts with stored preferences, an agent must distinguish genuine preference drift from temporary context shifts, ambiguity, or adversarial memory poisoning. We formulate this problem as a continuous-time partially observable decision process over a latent user state and show why rules based only on recency and provenance are insufficient. CAPTURE addresses this ambiguity with a neural differential-equation belief tracker, a multi-timescale memory ledger, uncertainty-triggered clarification, and counterfactual auditing of cited memories. On 480 held-out episodes from 96 users, CAPTURE achieves a 71.5% win rate, compared with 69.3% for an identically supervised baseline and 66.1% for the strongest heuristic baseline. It limits fixed-policy poisoning success to 11.5% while accepting 83.5% of genuine preference updates. Under an adaptive attacker with access to the released weights, attack success rises to 24.7%, exposing a real adaptation-security tradeoff. We further evaluate the frozen system zero-shot on an independently constructed benchmark and replay longitudinal interaction histories from 40 users collected over two to three weeks. These results suggest that modeling preference authenticity explicitly can improve both personalization and robustness in memory-augmented LLM agents.
[LG-27] Similarity-Aware Personalized Federated Learning in Heterogeneous Environments
链接: https://arxiv.org/abs/2609.02241
作者: Arun Kumar A V,Sunil Gupta,Dang Ngyuen,Bao Duong,Dat Phan Trong
类目: Machine Learning (cs.LG)
*备注:
Abstract:Federated Learning (FL) allows decentralized clients to train models collaboratively while preserving data privacy. However, distribution mismatch across clients often leads to poor global generalization and degraded local client-level performance. In such scenarios, some of the clients with their local models trained solely on local data may perform better than the globally learnt model, thus nullifying the benefits of collaborative federated learning. To address this, we propose SAPE-FL (Similarity-Aware Personalized Federated Learning), a novel personalization framework that anchors each client’s model to both the global model and a similarity-weighted peer averaged model. By incorporating dynamic, client-specific regularization based on both model similarity and output similarity, SAPE-FL adaptively balances global knowledge transfer and peer collaboration while filtering out dissimilar clients. This dual anchoring mitigates negative transfer and enhances robustness in heterogeneous settings. We theoretically analyze our algorithm establishing its convergence guarantees and empirically show that SAPE-FL outperforms state-of-the-art methods under high statistical heterogeneity and low client data regimes.
[LG-28] Recursive Value Learning for Long-Horizon Offline Goal-Conditioned RL
链接: https://arxiv.org/abs/2609.02237
作者: Hyeonseong Jeon,Youngwoon Lee
类目: Machine Learning (cs.LG); Robotics (cs.RO)
*备注:
Abstract:Scaling offline goal-conditioned reinforcement learning (GCRL) to long-horizon tasks is difficult because (1) long-range value learning depends on shorter-range estimates that may still be inaccurate, and (2) max-based value backups can amplify overestimation through repeated propagation. We propose DCRL (Divide-and-Conquer RL), which recursively decomposes each trajectory segment into a balanced binary tree and trains the values from leaves to root. Each parent is therefore updated only after its children, using an exact factorization of the observed route rather than selecting among noisy alternatives. Since this objective learns values along demonstrated routes that are not necessarily optimal, DCRL jointly propagates values across trajectories to discover shorter routes. Thanks to the balanced binary tree, DCRL reduces worst-case bootstrap depth from linear to logarithmic, and this shorter dependency structure empirically corresponds to much slower error accumulation. Across diverse goal-reaching tasks, DCRL substantially outperforms prior flat offline GCRL methods, and on the five most challenging long-horizon OGBench tasks, it improves the best prior average score from 55 to 64, surpassing all flat and hierarchical baselines.
[LG-29] Learning the Constitutive Behavior of Materials via Neural Operators and Causal Attention: Case Studies in Plasticity and Damage
链接: https://arxiv.org/abs/2609.02194
作者: Rishabh Arora,Lisa Scheunemann,Tim Brepols,Shahed Rezaei
类目: Machine Learning (cs.LG); Computational Engineering, Finance, and Science (cs.CE)
*备注:
Abstract:Classical constitutive modeling of path-dependent inelastic materials relies on internal state variables whose evolution equations must be postulated based on domain knowledge and calibrated against experimental data. However, in many practical settings, the relevant internal variables are typically not measurable in experiments, and the constitutive response must be inferred entirely from measured strain-stress data without any prior knowledge of the material’s internal state. We propose a data-driven constitutive modeling framework based on the concept of a material operator, which treats a deforming material as a functional mapping from its entire strain history to the corresponding stress response. In contrast to traditional autoregressive or recurrent formulations, the model is trained directly on full loading paths as function-to-function mappings, predicting complete stress trajectories in a single parallel forward pass. Temporal path dependence is enforced through a causally masked attention mechanism embedded within the operator, which restricts the model’s attention to past material states while preserving computational parallelizability. Spectral convolutions provide discretization-invariant representations in the frequency domain, while causal attention captures highly adaptive, non-local history dependence. Furthermore, sinusoidal activation functions are used to resolve the strong nonlinear transitions inherent in inelastic regimes. The framework is evaluated across multidimensional, rate-independent material models exhibiting complex phenomena, with an emphasis on nonlinear plasticity and ductile damage accumulation. The results demonstrate accurate and robust predictions of irreversible deformation mechanisms while simultaneously achieving resolution invariance and excellent parallel efficiency.
[LG-30] WeaveMark: Robust and Scalable Multi-bit LLM Watermarking via Coded Payload Spreading KR
链接: https://arxiv.org/abs/2609.02177
作者: Gang-Hyun Park,Ju-Hyeong Lee,Hee-Youl Kwak,Dae-Young Yun
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注: 16 pages, 11 figures. Code: this https URL
Abstract:Multi-bit watermarking for large language models (LLMs) enables content source tracing by embedding user-identifiable messages into generated text. Existing methods face a fundamental trade-off among extraction accuracy, text quality, and payload capacity. We propose WeaveMark, a robust and scalable multi-bit LLM watermarking scheme based on coded payload spreading. WeaveMark shifts this trade-off frontier by improving payload capacity through multi-bit-per-token spreading, improving extraction accuracy through soft-decision error-correcting code, and preserving text quality through unbiased multilayer reweighting. It further introduces dedicated zero-bit layers for reliable watermark presence detection. Experiments show large gains, especially for long messages and edited text. WeaveMark achieves 89.8% match rate for 32-bit messages at 200 tokens, compared with 20.8% for BiMark. Under 10% substitution attacks on 16-bit messages at 200 tokens, it maintains 86.0% versus 30.7%, while preserving text quality. Our code is available at this https URL.
[LG-31] DMRL: Document-Mediated Reinforcement Learning for Skill Optimization in Advertising Recommendation
链接: https://arxiv.org/abs/2609.02170
作者: Wei Zhang,Hongji Li,Song Sun,Peng Yu,Xue Yang,Lei Zhao,Peng Jiang
类目: Machine Learning (cs.LG)
*备注:
Abstract:Advertising recommendation requires continuously tuning complex system parameters while balancing commercial returns and user experience. Recent work has introduced large language models (LLMs) with skill documents to assist this labor-intensive process, but skill optimization remains largely prompt-driven, lacking a principled mechanism to attribute rewards to specific document edits. To address this limitation, we propose Document-Mediated Reinforcement Learning (DMRL), a skill self-evolution framework that models skill document optimization as a sequence of structured editing actions. In DMRL, an upper-level agent performs controlled document edits, while a frozen lower-level task agent evaluates their effects through A/B testing. To address credit assignment and long-term outcomes, we introduce two key components: (1) Dual-Relative Policy Optimization (DRPO), a post-training policy optimization method for robust and risk-aware advantage estimation; and (2) Long-term Reward Predictor (LRP), which estimates long-term outcomes by modeling population heterogeneity with disentangled representation learning and cross-attention transfer. DMRL was deployed on a large-scale short-video ads platform and extensive empirical evaluation shows that DMRL outperforms state-of-the-art baselines across key advertising metrics
[LG-32] Exact Limits of Random Projections for Preserving Geometry: Distance Recovery Nearest-Neighbor Rankings and Covariance Shape in Gaussian Models
链接: https://arxiv.org/abs/2609.02155
作者: Piyush Sao
类目: Machine Learning (cs.LG); Information Theory (cs.IT); Numerical Analysis (math.NA); Probability (math.PR); Statistics Theory (math.ST)
*备注: 41 pages, 6 figures, 4 tables. Reproducibility code: this https URL (pinned as a submodule). Companion paper on nearest-neighbor graphs to follow
Abstract:The Johnson-Lindenstrauss (JL) lemma guarantees that a random projection of n points to m=O(\varepsilon^-2\log n) dimensions preserves pairwise squared distances within relative error \varepsilon with high probability, and this dimension order is asymptotically optimal. In high dimensions, however, distances concentrate around a baseline while key geometric information lies in much smaller fluctuations. We show that the JL bound can therefore be uninformative about retained geometry: an independent Gaussian replacement map can satisfy it even though the replacement cloud is independent of the original data. We then ask how well any decoder can recover a feature f(D) of a squared distance D from a linear sketch. Under squared-error loss, the optimal decoder is conditional expectation, so recovery defines a linear operator whose singular values quantify feature recovery. For isotropic Gaussian data ( \Sigma=\sigma^2 I_d ), we diagonalize this operator in closed form. For fixed k with m,d-m\to\infty , its k th singular value satisfies \ell_k\approx(m/ d)^k/2 . This yields three sharp consequences. A rank- m sketch retains at most an m/d fraction of the variance of any feature of one squared distance. If m\to\infty and m/d\to0 , the expected Kendall correlation is \frac2\pi\sqrtm/d(1+o(1)) ; for fixed q , nearest- neighbor agreement tends to 1/q . Yet one projection can satisfy the JL bound while mean Kendall correlation vanishes when \log n\ll m\ll d . After removing scale, Haar-averaged retained covariance-shape information is (m/d)^2 . Thus JL distance preservation does not quantify the geometry available for comparison or inference. Comments: 41 pages, 6 figures, 4 tables. Reproducibility code: this https URL (pinned as a submodule). Companion paper on nearest-neighbor graphs to follow Subjects: Machine Learning (cs.LG); Information Theory (cs.IT); Numerical Analysis (math.NA); Probability (math.PR); Statistics Theory (math.ST) MSC classes: 60D05, 62H12, 68W20 ACMclasses: G.3; F.2.1 Cite as: arXiv:2609.02155 [cs.LG] (or arXiv:2609.02155v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2609.02155 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Piyush Sao [view email] [v1] Wed, 2 Sep 2026 06:12:34 UTC (290 KB) Full-text links: Access Paper: View a PDF of the paper titled Exact Limits of Random Projections for Preserving Geometry: Distance Recovery, Nearest-Neighbor Rankings, and Covariance Shape in Gaussian Models, by Piyush SaoView PDFHTML (experimental)TeX Source view license Current browse context: cs.LG prev | next new | recent | 2026-09 Change to browse by: cs cs.IT cs.NA math math.IT math.NA math.PR math.ST stat stat.TH 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-33] SoK: Where Do Flow Labels Come From? Auditing Label Provenance in Encrypted Traffic Benchmarks
链接: https://arxiv.org/abs/2609.02140
作者: Sizhe Huang,Shujie Yang
类目: Networking and Internet Architecture (cs.NI); Machine Learning (cs.LG)
*备注:
Abstract:Encrypted traffic classification infers semantics beyond the flow record from transport-layer observables, and supervised training rests on labels that hold for the individual flow they are attached to. Recent systematizations scrutinize model in- puts and data splits; we systematize the complementary label side. Across 14 audited benchmark entries, we identify two recurring label-side strategies: coarse inheritance, which risks labelling flows the evidence does not cover, and overstrict filtering, which keeps only self-attesting flows and risks dis- carding relevant ones. No audited entry exposes a countable pre-selection population, and the task objects downstream papers attach to the same labels disagree with the recovered record in 8 of 23 referenced cells. Under strict side-channel features we derive a representation-relative ceiling on bal- anced accuracy for any classifier restricted to those features: on the public benchmarks that inherit, it ranges from 0.56 to 0.76. On the filtering side, only 24.95% of connections in our fully captured corpus carry an observable SNI of their own; yet the discarded connections raise macro accuracy from 0.44 to 0.65 through same-run co-occurrence features. We end with recommendations for benchmark builders and users.
[LG-34] Scalable Bayesian Optimization of Composite Functions for Image-Based Inverse Problems in Materials Characterization
链接: https://arxiv.org/abs/2609.02126
作者: Dasol Yoon,Poompol Buathong,Chia-Hao Lee,Yujia Zhang,David A. Muller,Peter I. Frazier
类目: Machine Learning (cs.LG); Materials Science (cond-mat.mtrl-sci)
*备注: 28 pages. Dasol Yoon and Poompol Buathong contributed equally
Abstract:Estimating physical parameters from scientific images is a common inverse problem in materials characterization that often relies on expensive physics-based simulations. In electron microscopy, specimen thickness and crystal mistilt are critical parameters that govern how electrons scatter through the sample, and therefore the accuracy of any atomic-scale structure recovered from it. They are commonly inferred by matching experimental position-averaged convergent-beam electron diffraction (PACBED) patterns to simulated ones, but grid searches scale poorly and neural-network methods require extensive pretraining that may not transfer to new conditions. Here, we propose scalable Bayesian optimization of composite functions (SBOCF), a simulation-efficient method that exploits the known composite structure of the image-matching objective and the intermediate information contained in simulated images. By representing PACBED images with patch-level summaries and two correction terms, SBOCF preserves the original pixel-wise objective while reducing the number of modeled outputs from 24,649 to 11. Under a budget of 50 simulator evaluations, SBOCF outperformed standard Bayesian optimization with expected improvement on synthetic SrTiO3 benchmarks with thick and thin specimens, reducing the median final SSE by up to 290x in the thick-sample case. On experimental data, SBOCF produced parameter estimates consistent with previously reported values without task-specific pretraining. For a simulated mistilted specimen, using the SBOCF estimates in a downstream ptychographic reconstruction recovered sharp atoms that were otherwise blurred. These results establish SBOCF as a promising approach for inverse problems involving expensive simulators and high-dimensional structured outputs.
[LG-35] A Computational Comparison of Fourier Spectral Differentiation and Spatial Automatic Differentiation in Periodic Physics-Informed Neural Networks
链接: https://arxiv.org/abs/2609.02110
作者: Xilai Liang,Zhao Zhang
类目: Machine Learning (cs.LG)
*备注: 17 pages, 4 figures, 1 table
Abstract:Physics-informed neural networks (PINNs) commonly evaluate the spatial derivatives appearing in partial differential equation residuals using automatic differentiation (AD), whose computational and memory costs can become substantial when multiple or high-order derivatives are required. We perform a controlled comparison of spatial AD and Fourier spectral differentiation in periodic physical-space PINNs. Within each paired experiment, the neural representation, temporal differentiation, optimizer, sampling procedure, and training schedule are held fixed, so that the two cases differ only in the spatial differentiation procedure. For the Fourier variant, network outputs are evaluated on a uniform periodic grid and transformed to Fourier space, where spatial derivatives are obtained through spectral multiplication and the same Fourier coefficients are reused across derivative orders. We compare the two procedures in standard PINNs for the Allen–Cahn and Korteweg–de Vries equations and in Causal PINNs for the Allen–Cahn, Korteweg–de Vries, and Kuramoto–Sivashinsky equations. Across these five equation–framework settings, Fourier differentiation yields mean paired end-to-end training speedups ranging from 2.90\times to 18.52\times and reduces peak allocated graphics processing unit (GPU) memory by 68.7% – 94.1% . The final relative L_2 errors remain of the same order, with neither differentiation procedure showing a consistent accuracy advantage. For the one-dimensional periodic benchmarks considered here, Fourier spectral differentiation therefore provides substantially lower training time and memory usage than spatial AD while retaining comparable solution error, at the cost of requiring a uniform structured spatial grid.
[LG-36] Compositional Spectral Prompts for LLM -based Online Time Series Forecasting CIKM2026
链接: https://arxiv.org/abs/2609.02093
作者: Seungyoon Choi,Hyunchul Kim,Jae-Gil Lee,Chanyoung Park
类目: Machine Learning (cs.LG)
*备注: CIKM 2026
Abstract:To address the sequential and evolving nature of time series, the Online Time Series Forecasting (OTSF) task has been extensively studied in multiple domains. Existing research focuses on adapting to non-stationary environments by employing memory buffer-based retrieval strategies. However, we observe that such frameworks struggle with long-term adaptation and fail to generalize to unseen patterns. To this end, we introduce CoSPOT, an LLM-based online time series forecasting framework that leverages a pre-trained LLM as the backbone online forecaster, motivated by its strong few-shot capabilities. For efficient online adaptation, CoSPOT keeps the LLM frozen and employs compositional spectral prompts grounded in frequency-domain bases to guide the model with the overall distribution of the input, thereby substantially reducing the number of parameters updated during the online phase. Specifically, CoSPOT decomposes time series into frequency bases and composes the corresponding spectral basis prompts according to their amplitudes, allowing unseen patterns to be represented as new combinations of learned basis prompts. Our extensive experiments on real-world datasets demonstrate the superiority and practicality of CoSPOT across challenging online scenarios, including extended online phases and cross-dataset settings with substantial distribution shifts. Our code is available at this https URL.
[LG-37] DynG-Diff: A State-Aware Dynamic Guidance Diffusion Framework for Probabilistic Time Series Forecasting
链接: https://arxiv.org/abs/2609.02068
作者: Zhente Zhang,Zhengwei Ni,Wei Fan
类目: Machine Learning (cs.LG)
*备注:
Abstract:Probabilistic multivariate time series (MTS) forecasting is crucial for modeling complex dynamical systems. However, existing diffusion-based methods rely on task-specific conditional paradigms that lack flexibility and struggle with inherent “information heterogeneity”–the significantly varying noise levels and evolutionary patterns across variables. To address this, we propose DynG-Diff, a variable-sensitive dynamic guidance diffusion framework for probabilistic multivariate time-series forecasting: (1) DynG-Diff adopts a two-stage separated training strategy and uses an unconditional diffusion backbone to model the joint distribution of multivariate time series. (2) DynG-Diff introduces a lightweight state-aware policy network that adaptively infers variable reliability from real-time noisy states and one-step denoising estimates, outputting a dynamic guidance strength matrix. (3) DynG-Diff mathematically formulates this dynamic weight as the local precision of the observation distribution, enabling precise guidance for high-confidence variables during inference while filtering out interference from anomalous noise. Extensive experiments on real-world benchmarks demonstrate competitive probabilistic forecasting performance against state-of-the-art conditional diffusion models and improved robustness under severe observation this http URL implementation code is available at: this https URL
[LG-38] Act More Decide Less: Skill-Guided Adaptive Action Chunking for Long-Horizon LLM Agents EMNLP2026
链接: https://arxiv.org/abs/2609.02042
作者: Yanting Yang,Can Jin,Jinman Zhao,Jiahao Wu,Yang Zhou,Zhepeng Wang,Zhendong Wang,Mu Zhou,Dimitris N. Metaxas
类目: Machine Learning (cs.LG)
*备注: EMNLP 2026 Camera Ready
Abstract:Large language model (LLM) agents for long-horizon interactive tasks typically follow a ReAct-style protocol, issuing one primitive action per LLM round. While this enables frequent replanning, it is inefficient for long-horizon tasks where many rounds are spent on routine action sequences. A natural alternative is to let the agent emit variable-length action chunks. However, naively training such policies with standard reinforcement learning fails: the agent either collapses to single-action behavior or over-commits to excessively long sequences. Both failures share a common root cause: the inability to learn chunk boundaries. We propose SPACE, which addresses this challenge by distilling chunk-boundary supervision from trajectory-induced programmatic skills. We induce two-level programmatic skills from successful trajectories, where subskill boundaries serve as direct chunk-boundary supervision. This temporal structure is then distilled into a primitive-chunk policy via hybrid on-/off-policy optimization with chunk-aware credit assignment. Experiments on ALFWorld and ScienceWorld show that SPACE improves success rates by 7.0%-31.3% over the strongest baseline in each setting while reducing average LLM decision rounds by up to 78.9%.
[LG-39] CAHR-Net: Condition-Adaptive Hysteresis Reconstruction for Compact and Interpretable Magnetic Core Loss Modeling
链接: https://arxiv.org/abs/2609.01991
作者: Chunye Gong,Cong Yao
类目: Machine Learning (cs.LG)
*备注: 10 pages, 6 figures, 5 tables
Abstract:Magnetic core loss originates in the hysteresis loop: the energy dissipated per excitation cycle equals the loop area, and frequency, temperature, and waveform shape set the loss by reshaping the loop geometry. Most existing models let these conditions act only on a terminal scalar - empirical equations fold them into fitted exponents, and data-driven predictors append them to encoded features - so no intermediate hysteresis representation remains for the conditions to reshape. This paper proposes CAHR-Net, a condition-adaptive hysteresis reconstruction network that injects the operating conditions where they physically act. It preserves the interpretable chain from flux density waveform to magnetic field reconstruction, loop-area integration, and power loss estimation, and uses feature-wise linear modulation to inject frequency, temperature, and waveform statistics into the intermediate reconstruction representation. A matched large-batch training protocol based on AdamW, cosine scheduling, and a staged reconstruction-to-power-loss objective is also reported, because the modulation pathway takes effect only within it. On the MagNet final A-E material protocol, CAHR-Net attains an average p95 relative error of 6.89% with only 1874 parameters, the lowest among all compared methods, together with a lower worst-material p95 than the strongest black-box solution at about 48x fewer parameters; it reduces the average p95 of the physical reconstruction backbone from 7.47% to 6.89% and the p95 of material D, the most difficult material, from 16.40% to 14.87%. Ablation and condition-slice analyses attribute the improvement to the coupling of physical loop reconstruction, structured condition modulation, and the matched optimization trajectory.
[LG-40] A Unified Particle Filter LSTM for Data-Driven Process Simulation
链接: https://arxiv.org/abs/2609.01967
作者: Parvin Malekzadeh,Opher Baron,Dmitry Krass
类目: Machine Learning (cs.LG); Signal Processing (eess.SP)
*备注:
Abstract:Data-driven process simulation aims to generate realistic case trajectories from historical event logs without requiring an explicitly specified model of the underlying dynamics. Deep sequence models can capture complex temporal dependencies through next-activity probabilities and conditional time distributions. However, event logs provide only a partial view of the underlying process state, often recording activity completions without the corresponding service-start times. Consequently, the same observed process history may be consistent with multiple plausible latent process conditions, whereas standard recurrent models compress each process prefix into a single deterministic recurrent state. We propose a Unified Particle Filter LSTM (Unified PF-LSTM) that maintains and sequentially updates a weighted set of recurrent-state hypotheses. We summarize this particle belief using its weighted mean and learned features based on the moment-generating function. The resulting representation is used to predict a categorical distribution over the next activity and conditional quantiles of the current activity’s sojourn time. The framework is trained end-to-end from event-log data and evaluated on three real-world emergency department datasets. The results show that the proposed framework consistently outperforms the considered data-driven baselines in reproducing routing, duration, and system-level behavior across all datasets, with particularly strong gains in settings where complex process dynamics are only partially reflected in the available event logs.
[LG-41] Network-Aware Forecasting on Wireless Access Points
链接: https://arxiv.org/abs/2609.01957
作者: Niloo Bahadori,Swadhin Pradhan,Peiman Amini
类目: Networking and Internet Architecture (cs.NI); Machine Learning (cs.LG)
*备注: 5 pages, 2 figures, 1 table
Abstract:Enterprise wireless access points (APs) are promising platforms for predictive machine learning (ML), but their primary responsibility remains providing wireless connectivity and network services. Predictive inference must therefore share an AP’s CPU and memory with packet processing, Wi-Fi and IoT radio operations, and client management. This resource contention creates two risks: a model that performs well on proxy hardware may be too slow on the target AP, while a model that fits in isolation may still degrade network services under load. We define \textitnetwork-aware deployability using two gates: qualification of the model and its execution path on the target AP, followed by validation of its execution profile under packet-service and forecasting constraints. Our benchmarks show that edge testbeds do not reliably capture target behavior. Across matched artifacts and serving settings, five model implementations run 6.1–19.1 \times slower on an AP than on a Raspberry Pi~5, while peak memory usage differs by up to 22%. Moreover, two forecasting foundation models of similar size differ in AP latency by 19 \times . When serving a smaller model across 13 parallel streams at a 30~s cadence under network saturation, default execution increases p99 round-trip time (RTT) by 76% and reduces throughput by 7.06%. Understanding these trade-offs is essential for live deployment if we aim to use APs for both networking and ML workloads.
[LG-42] FlashKAN: B-Spline KANs via Truncated Power Form ICLR2027
链接: https://arxiv.org/abs/2609.01956
作者: Naveen Mysore
类目: Machine Learning (cs.LG); Numerical Analysis (math.NA)
*备注: 7 pages, 1 table, under review at ICLR 2027
Abstract:Kolmogorov-Arnold Networks (KANs) place learnable B-spline activations on network edges rather than fixed activations on nodes. The standard Cox-de Boor recursion evaluates these activations through k sequential passes for degree-k splines, consuming over 90% of forward-pass time. FlashKAN replaces this recursion with the truncated power form, a classical result from approximation theory that expresses each uniform cubic B-spline as five (x)_+^3 terms at shifted knot positions. This paper makes three contributions: (1) a this http URL-fused implementation that collapses these operations into a single GPU kernel, eliminating all recursion, span lookup, and scatter-gather operations; (2) a bounded-coordinate stabilization that clamps the normalized input to [0, k+1], preventing the catastrophic cancellation that historically motivated the Cox-de Boor recursion; and (3) a production-ready, open-source package (pip install flashkan) that serves as a drop-in replacement for existing KAN layers.
[LG-43] Pushing Forward Multi-Secret-Key Homomorphic Encryption for Private Averag e Aggregation
链接: https://arxiv.org/abs/2609.01945
作者: Miguel Morona-Mínguez,Fernando Pérez-González,Alberto Pedrouzo-Ulloa
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注: 29 pages, 7 figures and 10 Tables
Abstract:Federated Learning enables multiple clients to train a shared model while keeping their local datasets isolated. However, the exchanged model updates may still leak sensitive information, making private aggregation a central building block in practical deployments, especially in the cross-silo setting. Homomorphic Encryption naturally fits the client–aggregator communication pattern of Federated Learning, but conventional single-key deployments rely on strong non-collusion assumptions. Multiparty Homomorphic Encryption removes this limitation, although recent attacks under restricted decryption access require large-variance smudging noise during collaborative decryption, which significantly increases ciphertext size and implementation complexity. In this work, we propose lightweight multi-secret-key protocols for private average aggregation based on RLWE-based Homomorphic Encryption. Our construction departs from the usual multiparty blueprint by avoiding the generation of a collective public key. Instead, each client encrypts its update under its own secret key, while the resulting ciphertexts remain compatible with homomorphic aggregation and collaborative decryption. By explicitly tracking and cancelling the ciphertext noise during decryption, the protocol removes the need for large \lambda -dependent smudging noise. We instantiate the construction with both exact BFV-based and approximate CKKS-based variants, prove its security in the semi-honest model against an adversary corrupting the aggregator and up to L-1 clients, and compare its communication and runtime performance with state-of-the-art MHE-based aggregation. Our results show that the proposed approach substantially reduces ciphertext expansion and online cost, while preserving practical homomorphic aggregation performance.
[LG-44] Refining Heuristic-Based Bitcoin Address Clustering with Graph Neural Networks
链接: https://arxiv.org/abs/2609.01942
作者: Hugo Schnoering,Roman Bresson,Michalis Vazirgiannis
类目: Machine Learning (cs.LG)
*备注: 29 pages, 6 figures. Code: this https URL . Dataset: this https URL
Abstract:Bitcoin’s pseudonymous nature makes it challenging to analyze user-level activity, since a single user may control multiple identifiers (addresses). Existing heuristic-based methods attempt to identify addresses belonging to the same user, but they often produce flat cluster assignments with limited modularity and are prone to errors such as merging different users together. In this work, we propose a method for refining heuristic-obtained clusters by grounding our clustering on contrastive embeddings yielded by graph neural networks. Our contributions are threefold: (i) we release a publicly available dataset of Bitcoin transaction graphs containing a substantial number of clusters; (ii) we propose a methodology for learning address embeddings consistent with heuristics, and back it up with theoretical guiding intuitions; (iii) through hierarchical clustering, we enable a finer analysis of heuristic clusters and provide a quantitative criterion for flagging suspicious merges.
[LG-45] OR-Transformer: Scaling Real-Time Decision-Making to 1000 Items
链接: https://arxiv.org/abs/2609.01933
作者: Shuze Daniel Liu,David Simchi-Levi,Claire Chen,Chutong Gao,Shangtong Zhang
类目: Machine Learning (cs.LG)
*备注:
Abstract:Modern supply chain operations can require coordinating replenishment across thousands of heterogeneous items under correlated stochastic demand, heterogeneous lead times, and shared fixed ordering costs, yielding observation spaces exceeding 10^4 dimensions. At this scale, rolling-horizon stochastic mixed-integer linear programs (MILPs) become prohibitively slow, while standard reinforcement learning (RL) methods face increasingly challenging credit assignment in high-dimensional action spaces. We introduce OR-Transformer, a deep reinforcement learning framework for joint replenishment under stochastic demand, with an item-permutation-equivariant Transformer architecture and pathwise-gradient training through the inventory dynamics. Across problem sizes up to 1,024 inventory items, OR-Transformer increasingly outperforms learning-based and rolling-horizon MILP baselines as scale grows. It also reduces online decision-making time by over 4 million times relative to MILP solvers, enabling real-time, large-scale deep RL in supply chain operations.
[LG-46] Reinforcement learning to choose optimizers
链接: https://arxiv.org/abs/2609.01811
作者: Martin van der Schelling,Deepesh Toshniwal,Miguel A. Bessa
类目: Neural and Evolutionary Computing (cs.NE); Machine Learning (cs.LG); Optimization and Control (math.OC)
*备注: 20 pages, 14 figures
Abstract:No single optimization method is uniformly best for all problems, and the most suitable optimizer choice can change during a run. Existing approaches that change optimizer during execution typically predetermine part of the strategy: the portfolio is restricted to one algorithm class, the switch occurs once at a fixed time, or the frequency of decisions is treated as a hyperparameter rather than a learned one. We introduce “Reinforcement Learning to Choose Optimizers”, which formulates the optimization algorithm choice as a sequential decision-making problem. At each decision, a recurrent policy reads the current run state and decides both which optimizer should be used next and for how long. The portfolio includes both gradient-based and derivative-free optimizers, and each switch passes on the current best solution and a representative step size. A context proxy conditions a gating network over expert heads, and training employs a decoupled actor-critic whose return is expressed in the same empirical runtime distribution metric used at evaluation. Training tasks and portfolio are designed jointly so that no optimizer dominates. On unseen problems, the learned policy outperforms every portfolio optimizer at all but the smallest budgets, and it remains robust under distribution shift.
[LG-47] D-FROST: Decentralized Federated pRompt-tuning via Optimal tranSporT for Non-IID and Imbalanced Data
链接: https://arxiv.org/abs/2609.01802
作者: Quan Minh Nguyen,Hoang M. Ngo,Trong Nghia Hoang,My T. Thai
类目: Machine Learning (cs.LG)
*备注:
Abstract:Prompt tuning provides a parameter-efficient way to adapt foundation models (FMs) by freezing the pretrained backbone and updating only a small set of learnable prompts. This property makes prompt tuning especially suitable for decentralized federated learning (DFL), where exchanging full-model updates can be prohibitively expensive. However, prompt tuning in DFL introduces new challenges. Prompt sets learned from heterogeneous local data may not be index-wise aligned, making standard decentralized averaging unsuitable. In addition, the algorithm should be theoretically guaranteed to achieve consensus and make progress toward the shared objective. In this work, we provide the first study of prompt tuning in DFL. We formulate decentralized prompt tuning as a Wasserstein-based optimization problem over prompt measures, which captures the set-valued structure of prompts. We then propose D-FROST, an optimal-transport-based (OT-based) decentralized prompt-tuning algorithm that merges neighborhood prompts into compact representative prompt sets through transportation-based matching. We further analyze D-FROST by bounding the Wasserstein consensus error across clients, and establishing convergence of the network-level prompt barycenter to a neighborhood of stationarity. Experiments under heterogeneous client data demonstrate the effectiveness of D-FROST for decentralized prompt tuning.
[LG-48] Emergence of Fibrations Compression and Symmetry Breaking in Artificial Neural Networks
链接: https://arxiv.org/abs/2609.01768
作者: Osvaldo M Velarde,Lucas C Parra,Alireza Hashemi,Hernan A Makse
类目: Machine Learning (cs.LG)
*备注: 11 Figures
Abstract:Artificial neural networks are often regarded as powerful yet opaque black boxes. Here, we demonstrate that learning in deep neural networks generates local symmetries known in graph theory as fibrations and coverings. We prove that covering symmetries are stable attractors of stochastic gradient descent. Consistent with this theory, we report the emergence of covering symmetries across major network architectures, including multilayer, convolutional, recurrent, and transformer networks. Exploiting these symmetries enables drastic model compression - reducing networks to 17% of their original size without sacrificing performance. Furthermore, controlled breaking of covering symmetry overcomes the loss of plasticity, achieving state-of-the-art performance in continual learning. The theoretical results provide a new foundation for AI systems based on symmetries that convert black boxes into interpretable colored graphs and enable more efficient inference and lifelong learning.
[LG-49] oward Explainable and Policy-Aware AI for Carbon Credit Price Prediction: A Research Framework for Emerging Carbon Markets
链接: https://arxiv.org/abs/2609.01765
作者: Summaiya Unnisa Begum,Mohammed Nadeem Ullah,Mohammed Abdul Ghani Khan
类目: Machine Learning (cs.LG)
*备注: 7 pages, 4 figures, 7 tables
Abstract:Carbon markets put a price on emissions, yet that price remains hard to forecast. Work in this area clusters on the EU and Chinese schemes, compresses regulatory text into a sentiment score, and reports accuracy without calibration or explanation stability. We distil ten recurring gaps into an impact-feasibility matrix and propose EPA-CarbonNet, a six-layer architecture that fuses market series with policy text by cross-attention and calibrated intervals alongside policy-attributed explanations. We then build and test it on eleven years of daily S and P carbon index data. The findings are largely negative, and reported as measured: a random walk beats the model on five-day RMSE (0.0365 against 0.0475), SHAP rankings agree at rho = 0.54 across resampled backgrounds, and policy attention never coincides with documented regulatory events. Directional accuracy, at 58.6 percent, leads every baseline. Code, data documentation and all result artifacts are available at this https URL
[LG-50] A Study of Conditional Diffusion Models for Open-Loop Control under Dry Friction and Stiction
链接: https://arxiv.org/abs/2609.01756
作者: Eric Aislan Antonelo
类目: Machine Learning (cs.LG)
*备注:
Abstract:Diffusion models have recently emerged as expressive generative priors for planning and control. This paper studies Action Diffusion, an action-sequence diffusion formulation used as an open-loop proposal distribution for a point-mass system with dry friction and stiction. In this benchmark, motion starts only when the applied input exceeds a static-friction threshold, so effective controls occupy a small and temporally structured subset of the action-sequence space. A compact conditional 1D U-Net generates bounded control sequences conditioned on initial and target states. We compare it with uniform random shooting, random shooting from the same structured dataset prior, and the Cross-Entropy Method (CEM). Results show that Action Diffusion reduces terminal error and stuck steps, especially in low-sample regimes. These results indicate that conditional diffusion provides an effective mechanism for generating temporally coherent control sequences that overcome stiction by conditioning and recombining structured control primitives from the training prior for state-to-state open-loop control.
[LG-51] CAT-Flow: Curvature-Adaptive sTeps for Flow Matching
链接: https://arxiv.org/abs/2609.01746
作者: Qinchan Li,Pedro Cisneros-Velarde,Keru Fu,Samuel Antunes Miranda,Sharan Vaswani,Hao Zhang
类目: Machine Learning (cs.LG)
*备注:
Abstract:Flow Matching has emerged as a leading framework for generative modeling, powering state-of-the-art systems such as FLUX and Stable Diffusion 3.5. However, the iterative nature of its ODE-based sampling process creates a fundamental efficiency bottleneck: the quality of generated samples is highly sensitive to the choice of step-sizes, and current models typically require 20 to 30 steps for good quality. In this work, we propose two lightweight, training-free algorithms, CAT-OV and CAT-OT that adapt step-sizes at inference time based on a novel connection between Flow Matching sampling and gradient flow. Our algorithms are computed efficiently by not requiring additional neural function evaluations. Specifically, CAT-OT estimates curvature over time via a finite-difference approximation of the time-derivative of the vector field, while CAT-OV approximates curvature over the state space via a gradient of the vector field. Under suitable conditions, both methods have truncation error bounds of constant order. Empirically, CAT-OV and CAT-OT outperform existing step-size heuristics in image quality metrics across four text- to-image Flow Matching models, reducing the number of generation steps required to reach comparable quality by up to 40%.
[LG-52] Hearing the Whispers: Black-Box Membership Inference Attacks on Finetuned TTS Models
链接: https://arxiv.org/abs/2609.01723
作者: Kunlin Cai,Kaiyuan Zhang,Zihang Xiang,Jinghuai Zhang,Abeer Alwan,Fnu Suya,Yuan Tian
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG); Sound (cs.SD); Audio and Speech Processing (eess.AS)
*备注: 18 pages
Abstract:Text-to-Speech (TTS) foundation models are increasingly fine-tuned on private datasets to synthesize highly personalized voices, introducing severe privacy risks by exposing both biometric identities and sensitive speech content. Existing black-box membership inference attacks (MIAs) follow a two-stage pipeline of query generation and representation engineering, both of which face unique challenges when adapted to TTS. For query generation, dual conditioning on synthesis text and reference speech creates a large and underexplored query design space with no established criterion for identifying an effective query. For representation engineering, the multi-level speech characteristics and temporal variability of speech make low-level representations and direct comparisons inadequate for capturing membership signals. To address these challenges, we present the first black-box MIA framework explicitly tailored to TTS models at both the speaker and record levels. For query generation, we characterize the feasible query space and establish two criteria, scorable extent and memorization elicitation, for evaluating five representative queries, identifying recitation as the strongest. For representation engineering, we obtain multi-level speech representations from embedding models and temporally align the generated and target audio for fine-grained comparison. Evaluations across three state-of-the-art TTS models (CosyVoice2, F5-TTS, and XTTS-v2) fine-tuned on two benchmark datasets (VCTK and British Dialect) reveal severe privacy leakage: speaker-level AUC remains above 0.80 and approaches 1.0 in the strongest settings, while record-level AUC ranges from 0.80 to 0.90 and remains effective even in challenging scenarios where both members and non-members are of the same speakers. We further identify speech characteristics associated with disproportionate vulnerability to memorization.
[LG-53] Generative Diffusion Surrogates with Analytical Variance Schedule
链接: https://arxiv.org/abs/2609.01705
作者: Patrick Reichherzer,Gianluca Gregori,David N. Hosking,Subir Sarkar
类目: Machine Learning (cs.LG); Instrumentation and Methods for Astrophysics (astro-ph.IM); Plasma Physics (physics.plasm-ph)
*备注: Accepted for publication in Nature Communications
Abstract:Stochastic transport describes physical systems in which an initially structured distribution spreads under unresolved forcing, scattering, or heterogeneous media. Useful surrogates for such systems should be probabilistic, time-resolved, and able to represent non-Gaussian distributional structure. Generative diffusion models, which corrupt data with Gaussian noise and learn a reverse flow back to structured states, have these properties. Their noise schedules, however, are usually chosen heuristically: image and audio generation—the canonical use cases—provide no physical clock. In transport, by contrast, the variance, or mean-square displacement, is often known from macroscopic theory or empirical scaling even when the full distribution is not. Here we prescribe the forward noising rate as the time derivative of this variance, turning generative time into a calibrated transport clock. The variance path is enforced by construction, while the learned score field represents how non-Gaussian structure inherited from entrance data is smoothed along that path, requiring no intermediate-time physical transport data. For ballistic-to-diffusive transport in turbulent plasmas, the surrogate matches test-particle distributions, reproduces the laboratory-measured variance scale, and tracks the simulated kurtosis evolution without schedule tuning, enabling calibrated emulation and likelihood-based inference.
[LG-54] ri-Band Channel Measurement-Enabled Multi-Layer Digital Twin for Terahertz Wireless Data Centers
链接: https://arxiv.org/abs/2609.01699
作者: Mingjie Zhu,Ziming Yu,Guangjian Wang,Chong Han
类目: Machine Learning (cs.LG); Information Theory (cs.IT)
*备注:
Abstract:The rapid growth of AI computing has driven increasing demands for flexible and high-capacity data-center interconnections. Owing to its ultra-wide bandwidth and high spatial reuse capability, terahertz (THz) communication has emerged as a promising solution for future wireless data centers, while digital twins (DTs) enable efficient wireless planning and real-time optimization. In this work, a measurement-driven multi-layer DT framework is proposed for THz wireless data centers, where the physical, channel, evaluation, and manipulation layers are progressively constructed from bottom to top. First, extensive channel measurements are conducted at 140, 220, and 300 GHz to characterize frequency-dependent propagation behaviors. Based on the tri-band measurements, a measurement-calibrated physical twin is established by jointly optimizing the geometry, material, antenna, and hybrid propagation models. On top of the physical twin, a line-of-sight (LoS)-aware implicit neural field is developed to construct an AI channel twin for efficient channel reconstruction. The proposed AI twin learns location-dependent channel statistics from the calibrated twin, enabling real-time prediction of received power and LoS probability. Building upon the reconstructed channel field, a system-level evaluation layer is derived to analyze coverage and interference for both AP-to-rack and rack-to-rack communications. Experimental results show that the proposed AI twin achieves lower power reconstruction error than existing neural-field baselines while maintaining real-time inference capability. Moreover, the ceiling-mounted AP deployment achieves over 90% coverage under a 10 dB signal-to-interference-plus-noise ratio (SINR) threshold, demonstrating the effectiveness of the proposed DT framework for THz wireless data-center planning and optimization.
[LG-55] Median-of-Means as an Extremal Convex Estimator and a Nonconvex Route to the Trimmed Oracle
链接: https://arxiv.org/abs/2609.01689
作者: Angshul Majumdar
类目: Machine Learning (cs.LG)
*备注:
Abstract:We revisit median-of-means estimation from a deterministic optimization viewpoint and develop a family of block-Lp estimators for robust learning with heavy-tailed and adversarially corrupted data. In a block contamination model with at least a fraction 1 minus epsilon of good blocks, we first show that every convex block M-estimator has worst-case robustness constant at least 1 divided by 1 minus 2 epsilon. This matches the classical median-of-means bound and proves that the trimmed-block oracle constant 1 divided by 1 minus epsilon cannot be attained within the convex class. We then introduce a nonconvex block-Lp family for p between 0 and 1 and derive finite-sample deterministic robustness bounds for all global minimizers. As p decreases from 1 toward 0, these bounds continuously approach the trimmed-block oracle constant. For sufficiently small p, the global minimizers coincide with those of the oracle under a mild separation condition. We also show that the block-Lp objectives have a benign landscape, with all local minima remaining close to the truth and no bad basins. Combining these results with block-level concentration yields sub-Gaussian deviation bounds under finite 2 plus delta moments and high-dimensional extensions to robust mean estimation and sparse regression.
[LG-56] Reinforcement Learning and Rule-Based Peer-to-Peer Pricing in Residential PV-BES Communities WWW
链接: https://arxiv.org/abs/2609.01680
作者: Pablo Benalcazar,Maciej Kalka,Wilian Guamán,Jacek Kamiński
类目: Machine Learning (cs.LG); Computers and Society (cs.CY)
*备注: 22 pages, 2 figures, submitted to ARTIIS 2026 (Conference on Advanced Research in Technologies, Information, Innovation and Sustainability) this https URL
Abstract:This paper compares rule-based and learning-based pricing mechanisms for peer-to-peer (P2P) electricity trading in residential photovoltaic communities. The rule-based benchmarks comprise bill-sharing as an ex post allocation mechanism, the mid-market rate, and supply-demand-ratio pricing. The reinforcement-learning (RL) formulation is implemented through a Deep Q-Network and evaluated under multiplier-based and learnable SDR-shaped pricing, with a fixed-parameter SDR variant as a non-learning control. Performance is assessed through community savings together with complementary financial and operational indicators. In the base PV-only configuration, the rule-based benchmarks outperform the best RL policy. With battery energy storage, evaluated for the RL policies only, community savings under the best RL policy increase from EUR 734.23 to EUR 978.52. Across the learning-based modes and in both configurations, SDR-shaped pricing outperforms the multiplier-based parameterization considered. The results indicate that rule-based pricing remains highly competitive wherever the two families are compared directly, and that storage substantially improves the learning-based outcomes under this accounting, while the distribution of benefits remains heterogeneous across households.
[LG-57] A Survey on Self-Improving Test-Time Intelligence: Feedback-Driven Adapting Learning and Scaling at Inference
链接: https://arxiv.org/abs/2609.01679
作者: Shuaicheng Niu,Guohao Chen,Yaofo Chen,Zhiquan Wen,Jinwu Hu,Zeshuai Deng,Deyu Chen,Shuhai Zhang,Renjie Chen,Zihao Lian,Shoukai Xu,Gang Dai,Yunbei Zhang,Wei Luo,Yifan Zhang,Mingkui Tan,Cheng Deng
类目: Machine Learning (cs.LG)
*备注: accepted by Machine Intelligence Research
Abstract:The ability of AI systems to improve their behavior during deployment is becoming increasingly important. As inference moves beyond the static execution of a fixed trained model, a growing body of work studies how models can refine their behavior on the fly by exploiting test-time information and additional computation. These developments have largely evolved along two directions: methods that modify the model’s state using test-time signals, and methods that improve predictions through extra inference-time resources such as more sampling and tool use. However, these directions are often studied in separate communities with different terminology, making their connections harder to see. In this survey, we present feedback-driven Test-Time Intelligence (TTI) as a unified perspective for understanding such deployment-time improvement. We use this view to relate test-time adaptation, test-time learning, and test-time scaling, highlighting both their distinctions and their growing overlap in hybrid systems. This unified framework helps connect previously fragmented ideas and provides a clearer conceptual foundation for studying inference-time self-improvement. We review major methodological paradigms, representative applications, and open challenges across vision, language, multimodal learning, generative models, robotics, and healthcare. Our goal is to provide a coherent foundation and research roadmap for the study of self-improving AI systems at test time.
[LG-58] Sim2Signal: Sim-to-Real Benchmarks for Traffic Signal Control
链接: https://arxiv.org/abs/2609.01676
作者: Ferdous Al Rafi,Susrik Mukherjee,Latika Liladhar Dekate,Jennifer Yawa Lavoe,Huaiyuan Yao,Shlok Mohanty,Longchao Da,Xuesong Zhou,Hua Wei
类目: Machine Learning (cs.LG)
*备注:
Abstract:Reinforcement learning achieves strong traffic signal control performance in simulation, yet policies trained in simulators often fail once deployed in the real world, a failure known as the Sim-to-Real gap. When RL is applied to traffic signal control, this gap arises from several sources: sensing, action execution, traffic dynamics, and the control objective. Their relative impact and the reliability of existing Sim-to-Real mitigation methods remain insufficiently understood, and the field lacks a standard benchmark for systematically measuring the gap and evaluating mitigation methods. We present Sim2Signal, a benchmark that decomposes the Sim-to-Real gap into observation, action, transition, and reward gaps, corresponding to mismatches in the four components of the underlying MDP, and induces each gap in isolation under a shared protocol. We evaluate 18 mitigation methods on 2 base controllers, across 33 gap settings and 10 calibrated networks built from 5 real-world locations. We find that direct transfer consistently degrades performance across all four gap sources, but the severity of the degradation does not predict the effectiveness of mitigation. Instead, mitigation effectiveness depends strongly on the network and gap setting: outside the action gap, a method that helps in one case may fail in another. The most effective methods generally estimate what the gap changes, rather than make the policy insensitive through domain randomization or invariant representations. Our code is available at this https URL
[LG-59] Private Computation Space: Experience with Trusted Multi-Cluster Federated Learning for Agriculture
链接: https://arxiv.org/abs/2609.01667
作者: Shuangyu Lei,Muhammad Salman Abid,Jacob Belding,Sam Mosher,Manushi B. Trivedi,Shivranjani Baruah,Liam Wickes-Do,Andrew Anderson,Braulio Dumba,Alyssa Whitcraft,Ritvik Sahajpal,Sijin Li,Kelly Robbins,Michael Gore,Margaret Frank,Steven Wolf,Liz Jones,Abraham Stroock,Kaitlin Gold,Hakim Weatherspoon
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注: 15 pages, 13 figures, 3 tables
Abstract:Artificial Intelligence has shown to help improve agricultural practices, yet adoption remains limited: 69% of U.S. farmers have privacy concerns with sharing their data, and these concerns must be addressed before adoption is widespread. While Federated Learning has been demonstrated to protect privacy at scale for other sectors, deploying a system for agriculture comes with its own set of challenges; the problem necessitates a system that can protect farmer data and identities while preserving model utility, runs on commodity hardware, and is resilient to fragile rural infrastructure. To address these concerns, we introduce the Private Computation Space (PCS), a deployed, open-source Machine Learning system to provision and process farmer data securely. We design a system tailored to an agricultural setting, with multi-cluster orchestration for reliability in rural areas with asynchronous Federated Learning (FL), Differential Privacy (DP), and Trusted Execution Environments (TEEs), to allow farms to participate in the framework while keeping their data private. We evaluate the system on two deployed workloads: monitoring nitrogen with living plant sensors in NY for six months and predicting evapotranspiration from weather stations in CA for ten months. Our evaluation finds a Dice Similarity Coefficient (DSC) of 0.71 and R^2 accuracy of 0.84 for the respective workloads, improving the worst single-site model accuracy by 22.4% and 9.1%, respectively, while preserving privacy.
[LG-60] Context Inference Attacks Without Jailbreaks
链接: https://arxiv.org/abs/2609.01663
作者: Prince Jha,Samuele Poppi,Nils Lukas
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注:
Abstract:Agentic AI systems are increasingly deployed to process sensitive data at inference time, such as healthcare records or financial documents assembled into a hidden \emphcontext before the system answers. Prior work has studied privacy risks primarily through \emphjailbreaking attacks that induce models to directly disclose sensitive content, but has largely overlooked the agentic setting where the context is assembled by the agent’s own tool calls. We show that the agents we evaluate remain vulnerable to hidden-context leakage despite the controls we test against them, namely an instruction not to disclose the context, logit suppression, and context dilution. For instance, a web-browsing agent answering benign user queries still carries exploitable signals about records silently loaded into its context. We introduce and formalize \emphcontext-inference attacks through a security game and evaluate three settings under decreasing attacker knowledge and increasingly indirect delivery of the context: a known context, an unknown context, and a context the agent retrieves through its own tool calls. We distinguish a grey-box setting, in which the target model is used to score observations, from black-box settings in which the attacker scores with a surrogate it controls. We further characterize how leakage varies with query budget, context size, and target-model size. A single attack carries through all three settings without modification, reaching 100% ASR on small candidate sets and 63% at 1024 candidates against a known context, 78.9 AUROC when the template and surrounding records are unknown, 92.5 AUROC when a 14B surrogate scores a 32B target, and 81.8 AUROC when the records arrive as an agent’s retrieval returns, against chance rates of 1/|\mathcalZ| and 50 respectively.
[LG-61] Efficient Context-Limited Telescope Bibliography Classification for the WASP-2025 Shared Task Using SciBERT AACL2025 ACL
链接: https://arxiv.org/abs/2609.01647
作者: Madhusudhana Naidu
类目: Machine Learning (cs.LG); Instrumentation and Methods for Astrophysics (astro-ph.IM)
*备注: 3 pages, 2 tables. 1st place system description for the TRACS shared task at WASP 2025 (Third Workshop for Artificial Intelligence for Scientific Publications), co-located with IJCNLP-AACL 2025. Published version: this https URL . Code: this https URL
Abstract:The creation of telescope bibliographies is a crucial part of assessing the scientific impact of observatories and ensuring reproducibility in astronomy. This task involves identifying, categorizing, and linking scientific publications that reference or use specific telescopes. However, this process remains largely manual and resource intensive. In this work, we present an efficient SciBERT-based approach for automatic classification of scientific papers into four categories - science, instrumentation, mention, and not telescope. Despite strict context-length constraints (maximum 512 tokens) and limited compute resources, our approach achieved a macro F1 score of 0.89, ranking at the top of the WASP-2025 leaderboard. We analyze the effect of truncation and show that even with half the samples exceeding the token limit, SciBERT’s domain alignment enables robust classification. We discuss trade-offs between truncation, chunking, and long-context models, providing insights into the efficiency frontier for scientific text curation.
[LG-62] SocialBuddy: Tailoring Search Agent for Social Scenarios
链接: https://arxiv.org/abs/2609.01641
作者: Mingxuan Li,Yirong Mao,FaZhan Zhang,Haibiao Yao,Runze Hu,Wenhui Que
类目: ocial and Information Networks (cs.SI); Machine Learning (cs.LG)
*备注:
Abstract:In the era of digital social interaction, searching friends’ posts from massive social streams has become a fundamental user need. However, while modern agentic search frameworks have achieved remarkable success in conventional retrieval tasks, they break down when confronted with heterogeneous user queries and multi-dimensional social feeds, resulting in severe performance degradation in complex social search. To bridge this gap, we introduce SocialBuddy, the first agentic search framework tailored for social scenarios. Specifically, we construct SocialEnv, the first large-scale simulated environment for social search. Powered by an automated data and trajectory synthesis pipeline, SocialEnv includes 200K user profiles, 10 million social posts, and 50K reasoning trajectories, establishing a solid foundation for the development of social search agents. To tackle the credit assignment dilemma caused by sparse rewards in social search, we design SocialPO, a hybrid-granularity optimization framework. It macroscopically reinforces successful reasoning paths via multi-dimensional rewards, while microscopically rectifying deviated trajectories through fine-grained prefix truncation and token-level supervision. This hybrid-granularity design delivers multi-scale guidance in complex long-sequence scenarios. Finally, we construct SocialSearch Benchmark to provide a quantitative evaluation scheme for assessing the social search capabilities of SocialBuddy. Extensive experiments demonstrate that SocialBuddy-35B surpasses significantly larger frontier LLMs. Code and dataset will be released upon article acceptance.
[LG-63] Omega-N: Interpretable Structural Node Descriptors and Their Applicability Domain
链接: https://arxiv.org/abs/2609.01633
作者: Alberto Acedo
类目: ocial and Information Networks (cs.SI); Machine Learning (cs.LG); Physics and Society (physics.soc-ph); Molecular Networks (q-bio.MN)
*备注: 18 pages, 3 figures. Reference implementation, notebooks and data-preparation scripts at this https URL
Abstract:A composite structural index summarises a network in one number; for a triangle-based index it is spectrally redundant: Tr(A^3) is the third moment of the adjacency spectrum. The non-redundant content sits one level down, in diag(A^3), which depends on eigenvectors and is not spectrally determined. A corollary in the theory paper for this index family stated that, and predicted: the global scalar should tie sharpened spectral baselines rather than beat them, while the node-wise attribution should do better where the number of structural epicentres is unknown. This paper tests it. We construct Omega-N by localizing each of the four factors. The direct localization is badly conditioned; two corrections from published practice fix it, a configuration-null excess for every local factor and a personalized-PageRank neighbourhood at several scales, giving ten interpretable features per node from the graph alone, with no attributes, training or embeddings. Against a recursive feature engine at five levels of recursion, Omega-N wins on one and ties on four of six in-domain node-classification evaluations, with ten features against up to 252. Two statistics computed from the graph and labels, not from performance, partition the eight benchmarks without error, and the two they exclude are the two it loses. The strongest application is drug-target prioritisation on protein interaction networks: +0.073 to +0.144 AUPRC over a centrality battery across four constructions, replicated on an independent AP-MS network and label source, surviving three bias controls (degree-matched, ten repetitions: +0.1047 and +0.1030, both 10/10, p=0.00195). The clearest negative sits in the same application: adding Omega-N to centralities plus Node2Vec changes nothing (+0.0014, p=0.31). The claim is narrow: ten interpretable features Comments: 18 pages, 3 figures. Reference implementation, notebooks and data-preparation scripts at this https URL Subjects: Social and Information Networks (cs.SI); Machine Learning (cs.LG); Physics and Society (physics.soc-ph); Molecular Networks (q-bio.MN) Cite as: arXiv:2609.01633 [cs.SI] (or arXiv:2609.01633v1 [cs.SI] for this version) https://doi.org/10.48550/arXiv.2609.01633 Focus to learn more arXiv-issued DOI via DataCite Submission history From: Alberto Acedo [view email] [v1] Fri, 21 Aug 2026 16:16:30 UTC (46 KB)
[LG-64] Prompt-Space Meta-Learning Does Not Transfer Across Users: A Frozen-LLM Negative Result
链接: https://arxiv.org/abs/2609.01615
作者: Liam Byrne,David Dylan,Orla Fitzgerald,Eoin Doyle,Ciara Nolan,Padraig Lynch,Sinead Gallagher
类目: Machine Learning (cs.LG)
*备注:
Abstract:Personalizing a frozen large language model (LLM) to individual users is often framed as a meta-learning problem in prompt space: each user is a task, and one seeks a shared natural-language adaptation policy that, given a handful of the user’s labeled interactions, configures the frozen model for that user. The framing is attractive because it is backbone-agnostic and reuses the machinery of prompt optimization, yet the field rarely tests whether the optimized meta-objective encodes transferable cross-user adaptation rather than generic instruction quality. We study this question with Muse (Meta-learned User-adaptation via Shared Evolution), which evolves a single shared adaptation prompt over a meta-train user population by reflective prompt evolution, freezes it, and applies it zero-shot to held-out users; matched controls isolate learning from confounds of phrasing and selection. On two standard personalization benchmarks (LaMP-2 categorization and LaMP-3 rating) over 200 held-out users each, Muse does not significantly improve on its own un-evolved seed prompt or on a structure-broken control that meta-trains on mismatched user-support pairs, and is dominated by plain few-shot retrieval on the rating task (Delta MAE +0.175, p 0.001). We attribute these outcomes to a single mechanism, meta-objective collapse: the meta-validation objective is statistically invariant to whether the user-support correspondence is genuine (p=0.555 on LaMP-2, p=0.622 on LaMP-3), so it cannot be optimized into transferable adaptation and instead rewards instruction polish and validation overfitting. The seed-prompt, wrong-support, and invariance-oracle controls form a reusable protocol that separates learned adaptation from these confounds.
[LG-65] DiDrive: A Risk-Aware Hierarchical Diffusion Framework for Safe Offline Reinforcement Learning in Autonomous Driving
链接: https://arxiv.org/abs/2609.01609
作者: Qisong Guo,Jingtang Chen,Zhilin Chen,Pei Xu,Mingjian Fu,Wenxi Liu,Yuanlong Yu
类目: Machine Learning (cs.LG); Robotics (cs.RO)
*备注: 16 pages
Abstract:While diffusion models effectively capture multimodal behavioral priors for autonomous driving, offline reinforcement learning (RL) policies remain susceptible to distribution shift, heavy-tailed risk signals, out-of-distribution (OOD) action generation, and high-dimensional state redundancy. To address these challenges, we propose DiDrive, a distribution-guided offline diffusion framework featuring two synergistic components: the Risk-Aware Hierarchical Diffusion (RHDif) architecture and the 3DICE policy optimization paradigm. In the state space, RHDif utilizes a low-level risk-gated encoder and a high-level contextual modulator to filter environmental redundancy and focus on safety-critical threats. In the action space, 3DICE mitigates OOD overestimation and gradient oscillation through in-sample calibrated guidance, spatiotemporal optimization, and ensemble-based candidate ranking. Evaluations on the CARLA benchmark demonstrate DiDrive’s superiority over baselines like IQL, CQL, and Diffusion-QL, particularly in complex, high-density traffic scenarios with 60 vehicles, where it achieves an 85% success rate and a 4295.68 average reward, providing a robust pathway for safe autonomous driving decision-making.
[LG-66] Improved Gradient Descent Lower Bounds Beyond Nesterov
链接: https://arxiv.org/abs/2609.02855
作者: Yuhan Ye,Kaizhao Liu
类目: Optimization and Control (math.OC); Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: 32 pages, 6 figures
Abstract:We study how far gradient descent (GD) can be accelerated by predetermined stepsizes in smooth convex optimization. Going beyond the classical \Omega(n^-2) first-order oracle lower bound of Nemirovsky and Yudin, we prove an \Omega(n^-1.6342) non-anytime lower bound and an \Omega(n^-1.2408) anytime lower bound. These improve the recent \Omega(n^-1.932) non-anytime lower bound of Ma and Chen and the \Omega(n^-4/3) anytime lower bound of Tsai et al., respectively. Together with the non-anytime O(n^-\log_2(1+\sqrt2)) rate achieved by silver schedules, our anytime lower bound establishes a strict separation between the achievable convergence exponents in the two settings.
[LG-67] Learning Spectral-Like Mesh-Free Discretisations
链接: https://arxiv.org/abs/2609.02833
作者: Lucas Gerken Starepravo,Henry Broadley,Steven Lind,Jack R. C. King
类目: Computational Physics (physics.comp-ph); Machine Learning (cs.LG)
*备注:
Abstract:Meshfree methods such as smoothed particle hydrodynamics (SPH) with kernel corrections, radial basis function-generated finite differences (RBF-FD), and the local anisotropic basis function method (LABFM) construct discrete differential operators by imposing polynomial consistency on a local stencil. For stencils containing more nodes than there are consistency constraints, the resulting linear system is underdetermined, and the remaining degrees of freedom are fixed implicitly by the choice of kernel, basis preconditioning, or a minimum-norm condition. Polynomial consistency constrains the operator only in the low-wavenumber limit, and no part of the construction selects for accuracy at the wavenumbers where fine-scale content resides. We introduce Spectral-like Neural Discretisation (SpeND), in which the choice of those degrees of freedom is cast as a learning problem: stencil weights are parametrised by a neural network conditioned on the local node geometry, trained to approximate the modal response of a spectral operator over the resolvable band. A hard-constrained projection layer maps the network output onto the affine subspace of consistent weights, so that polynomial consistency holds exactly by construction rather than as a penalty. Training is self-supervised and physics-agnostic, requiring no reference solutions; the objective minimises dispersion and dissipation error over a prescribed band-limited function space. Modal analysis on disordered two-dimensional node distributions shows that the learned fourth-order operator follows the exact response over a substantially wider band than either explicit LABFM at equal stencil size or fourth-order finite differences on a structured grid, whilst recovering the expected fourth-order convergence rate under refinement.
[LG-68] Full-Model Optimality for Tunable Linear Generative Priors in Compressed Sensing
链接: https://arxiv.org/abs/2609.02790
作者: Zhaoming Li,Paul Hand
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注: 32 pages 3 figures
Abstract:Generative models have been studied experimentally and theoretically as priors for inverse problems such as compressed sensing. Recent work by Gunn et al. studied the use of generative priors with tunable complexity, where a family of generative priors with varying complexity is maintained and a specific complexity can be selected at inversion time. They demonstrated that lower reconstruction errors can be experimentally attained for a variety of inverse problems by appropriately tuning the complexity of the generative prior. In the present paper, we establish theory for compressed sensing in the setting of a tunable family of linear generative priors naturally related through their singular value decompositions. We prove that in noiseless Gaussian compressed sensing, the full-dimensional linear prior attains the minimum expected reconstruction error over the entire family of linear priors. Thus, in this idealized linear noiseless setting, tuning to a lower-complexity prior does not improve the expected reconstruction error. This result is in contract to the behavior of denoising, where lower complexity priors attain lower reconstruction errors due to a standard bias-variance tradeoff. This result indicates that the experimental benefits of tunability in compressed sensing with neural network priors arises due to nonlinearities in the generative models.
[LG-69] Momentum in large-batch training: Polyak enlarges the critical batch size Nesterov improves data efficiency
链接: https://arxiv.org/abs/2609.02728
作者: Jia-Nan Wang,Zixun Huang,Kairui Li,Lei Wu
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Optimization and Control (math.OC)
*备注: 69 pages, 8 figures
Abstract:We study when and how momentum improves large-batch training in the one-pass regime, using power-law kernel regression as a tractable setting. We first characterize risk stability through the critical learning rate, defined as the largest learning rate for stable training, and obtain \eta_\mathrmSGD^\mathrmcrit\eqsim 1 , \eta_\mathrmPolyak^\mathrmcrit\eqsim \min\1,B(1-\rho)\ , and \eta_\mathrmNesterov^\mathrmcrit\eqsim \min\1,B^\beta(1-\rho)\ , where B is the batch size, \rho is the momentum factor, and \beta1 is the capacity exponent. Within this admissible region, we derive scaling laws for the full risk dynamics, capturing the progression from an early transient, through power-law decay, to a noise floor. We then minimize the final-step risk over the admissible learning rates and momentum factors under a fixed data budget, yielding a three-regime batch-size phase diagram that reveals how the role of momentum changes with batch size. Notably, Polyak enlarges the critical batch size, the largest batch size preserving the best small-batch data-scaling exponent, thereby enabling greater parallelism without sacrificing data efficiency. In contrast, Nesterov achieves better data efficiency in the large-batch regime because its look-ahead mechanism suppresses noise accumulation. Numerical experiments validate the predicted stability boundaries, risk dynamics, and batch-size phase diagram.
[LG-70] Eliciting ESG Preferences for Reinforcement Learning-Based Portfolio Optimization
链接: https://arxiv.org/abs/2609.02677
作者: Giovanni Dispoto,Marcello Restelli,Carmine Ventre
类目: Portfolio Management (q-fin.PM); Computational Engineering, Finance, and Science (cs.CE); Machine Learning (cs.LG)
*备注:
Abstract:Modern portfolio management increasingly demands a balance between traditional risk-adjusted returns and strict Environmental, Social, and Governance (ESG) mandates. Current Reinforcement Learning (RL) approaches typically optimize for a single ESG provider, neglecting the significant divergence in rating methodologies across the industry and the unintuitive nature of manually weighting conflicting objectives. This paper addresses these limitations by formulating ESG-aware portfolio optimization as a Multi-Objective Reinforcement Learning (MORL) problem that simultaneously incorporates ratings from three distinct ESG agencies. To bridge the gap between high-dimensional algorithmic trade-offs and human decision-making, we integrate a Preference Elicitation framework using Gaussian Processes. This system enables practitioners to infer their latent utility functions through intuitive pairwise comparisons of candidate portfolios based on their Sharpe ratios and aggregate ESG scores. We systematically evaluate our framework by employing Large Language Model (LLM) personas to simulate Portfolio Managers operating under varied regional contexts. Empirical results using historical market data reveal that regional backgrounds fundamentally shift the derived preference weights. For instance, European-based personas tend to prioritize ESG alignment over financial returns, while Texas-based personas favor risk-adjusted performance. This work offers a highly adaptable framework that successfully aligns multi-objective algorithmic trading with diverse, real-world human sustainability preferences.
[LG-71] Dimension Dependent Correlation Gap Bounds under Restricted Independence
链接: https://arxiv.org/abs/2609.02659
作者: Arjun Ramachandra
类目: Probability (math.PR); Machine Learning (cs.LG); Combinatorics (math.CO)
*备注: 33 pages, 10 Tables, 2 Figures
Abstract:The pairwise independent correlation gap is the ratio of the maximum expected value of a set function under arbitrary dependence to that under pairwise independence, measuring the loss from this independence restriction. Under mutual independence, this gap is universally bounded by e/(e-1) for monotone submodular functions. With pairwise independence, a tighter 4/3 upper bound was established for several special cases, including n=3 , and conjectured to hold universally. A recent AI-assisted counterexample disproved this conjecture for n=5 , leaving the validity of the n=4 bound and the tight worst case bound open. We resolve both questions. First, for n=4 , we establish that the 4/3 bound holds universally and is tight using an AI-assisted proof combining theoretical analysis and computational verification. The proof combines a structural characterization of optimal numerator vertices, permutation symmetry, cone certificate systems, Bernstein polynomial representations, recursive simplex subdivision, and verification of 2,745 Bernstein coefficient systems. Second, we show that the worst case pairwise independent correlation gap attains e/(e-1) asymptotically by constructing an instance with identical marginal probabilities and a monotone submodular union coverage function on a ground set partitioned into m blocks. The number of blocks grows sublinearly with the ground set size. The result follows by constructing a feasible solution to a scaled asymptotic reduced dual of the pairwise independent linear program and immediately extends to t -wise independent random elements ( t\ge2 ), since t -wise independence implies pairwise independence. Thus, pairwise independence, despite being the least restrictive form of independence in the t -wise independence hierarchy, can be as restrictive as mutual independence in the worst case. Comments: 33 pages, 10 Tables, 2 Figures Subjects: Probability (math.PR); Machine Learning (cs.LG); Combinatorics (math.CO) Cite as: arXiv:2609.02659 [math.PR] (or arXiv:2609.02659v1 [math.PR] for this version) https://doi.org/10.48550/arXiv.2609.02659 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-72] A computational approach to maximum likelihood thresholds for colored Gaussian graphical models
链接: https://arxiv.org/abs/2609.02382
作者: Roser Homs,Olga Kuznetsova,Bernadette J. Stolz
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Algebraic Geometry (math.AG); Statistics Theory (math.ST)
*备注: 30 pages, comments welcome!
Abstract:Gaussian graphical models (GGMs) are essential tools for interpretable structure learning. However, in high-dimensional, small-sample regimes, the available data is often insufficient for the maximum likelihood estimator to exist. Colored Gaussian graphical models (CGGMs) mitigate this limitation by imposing symmetry constraints through graph coloring, which reduces the required sample size. This minimal number of observations needed to guarantee that the estimator exists almost surely is defined as the maximum likelihood threshold (MLT). Here, we address the computation of the MLT for CGGMs by focusing on its geometric formulation: finding the minimum rank of a sample covariance matrix such that its projection lies almost surely within the interior of the cone of sufficient statistics. We establish a unified theoretical framework, extending results from uncolored to colored models and introducing new symbolic algorithms. Furthermore, we present a computational study integrating sampling with topological data analysis (TDA) to investigate the local geometry of the cone of sufficient statistics. Our results demonstrate the potential of TDA to overcome the computational bottlenecks of traditional symbolic algebraic methods, particularly Groebner basis computations, in analyzing the likelihood geometry of CGGMs.
[LG-73] From topology learning to graph generation: A unifying perspective
链接: https://arxiv.org/abs/2609.02286
作者: Xiaowen Dong,Hoi-To Wai,Siheng Chen,Laura Toni,Dorina Thanou
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Signal Processing (eess.SP)
*备注:
Abstract:Learning graph structures from data is a fundamental problem that spans a wide range of signal processing and machine learning tasks. While significant effort has been made to tackle the problem, existing research has largely evolved along two parallel directions. The first seeks to infer the topology of an individual graph from observations supported on it, whereas the second seeks to learn a generative distribution from observed graph instances, enabling the sampling of new graphs. This review presents a unified framework that connects these formulations by viewing them as inverse problems of a common generation process for graph data. We review the major methodologies within this framework, highlight their relationships, strengths, and limitations, and identify opportunities for integrating ideas across paradigms. By bridging graph topology learning and graph generation, this review provides a broader cross-disciplinary perspective on the field and outlines promising directions for future research.
[LG-74] Prototype-guided transfer of sparse literature knowledge for electrolyte additive discovery
链接: https://arxiv.org/abs/2609.02209
作者: Weixiang Hong,Hongting Du,Jiayue Tang,Ruifeng Tan,Yangjian Quan,Jia Li,Jiaqiang Huang
类目: Chemical Physics (physics.chem-ph); Machine Learning (cs.LG)
*备注: 79 pages, 26 figures
Abstract:Electrolyte additive discovery remains challenging because experimentally validated molecules are sparse, whereas accessible chemical spaces are vast and largely unlabeled. This challenge is amplified in lithium-ion batteries, where additive performance arises from coupled interfacial reactions rather than a single molecular property. Here, we develop a prototype-guided molecular intelligence, ProtoMI, a literature-driven framework that learns transferable structural priors from reported electrolyte additives and uses them to prioritize candidates in unlabeled chemical space. For boron-containing additives, ProtoMI combines 126 literature-reported molecules with 179,977 unlabeled candidates. Graph contrastive learning identifies seven chemically interpretable prototypes from the reported additives, and prototype guided semi-supervised contrastive learning adapts these prototypes to the candidate space under source-target distribution mismatch. In retrospective temporal validation, ProtoMI achieves enrichment factors of 9.2-45.6 while screening less than 2% of the candidate space. A subsequent translation step identifies four commercially accessible candidates. One representative candidate, 4,4,5,5-Tetramethyl-2-[10-(1naphthyl)anthracen-9-yl]-1,3,2-dioxaborolane (TNDB), improves high-temperature LiFePO4||graphite cycling at 55 °C by 34.93% relative to the baseline electrolyte. An arsenal of characterizations and operando optical fiber Fourier transform infrared spectroscopy suggest that TNDB forms B-containing, F/P/O-modified inorganic interphases, suppresses solvent decomposition and reduces Fe deposition on graphite. This case study shows how sparse literature knowledge can guide experimentally efficient molecular discovery in data-scarce battery-additive spaces.
[LG-75] Quantum MeanFlow: single-shot generative sampling on NISQ hardware
链接: https://arxiv.org/abs/2609.02186
作者: Ashish Joshi,Eshaan Mistry,Takahiko Koyama
类目: Quantum Physics (quant-ph); Machine Learning (cs.LG)
*备注: 17 pages, 6 figures
Abstract:Quantum generative models offer a promising framework for exploring whether quantum computation can enhance generative machine learning. Flow matching is a generative method in which samples are generated by transporting a simple, known distribution to the target data distribution with a learned velocity field. Its quantum counterpart, known as quantum flow matching (QFM), was introduced recently, and, like its classical counterpart, requires integrating an ordinary differential equation over many time steps during inference. As each step requires the output from the previous step, the circuit submission is sequential and a drawback on quantum computers as they have high input/output costs. To alleviate this problem, we introduce Quantum MeanFlow (QMF), the quantum analogue of the MeanFlow formulation, which allows single-step sample generation. While the QFM learns an instantaneous velocity field at each time step, QMF learns the average velocity over a time interval. We use a parameterized quantum circuit to learn these velocity fields and benchmark the two methods on the MNIST dataset. We show that while single-step QMF has lower image quality compared to multi-step QFM, it performs better than the single-step QFM sampling at every shot count. Both of our models are executed on IBM quantum computers and best-of-N rejection sampling recovers most of the accuracy lost to device noise without modifying the circuit. This is especially advantageous for QMF which has only one circuit evaluation per image. Here, We establish QMF as a viable method for single-step quantum generative sampling, saving on quantum circuit evaluations per generated sample.
[LG-76] HyperMC: Multi-Fidelity Hyperparameter Tuning for Stochastic Gradient MCMC
链接: https://arxiv.org/abs/2609.02138
作者: Ming Tan,Xiyun Jiao
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Computation (stat.CO)
*备注: 56 pages, 14 figures, 9 tables
Abstract:Stochastic gradient Markov chain Monte Carlo (SGMCMC) methods enable scalable Bayesian inference, but their performance depends strongly on hyperparameters such as the step size, mini-batch size, and number of leapfrog steps. Since most SGMCMC algorithms lack a Metropolis-Hastings acceptance rate, standard acceptance-based tuning methods are not directly applicable. We propose HyperMC, a multi-fidelity tuning framework that combines Hyperband-style resource allocation with kernel Stein discrepancy (KSD) evaluation. By running multiple successive-halving brackets, HyperMC balances broad exploration of a continuous hyperparameter space with increasingly accurate evaluation of promising configurations under a fixed computational budget. We further introduce Robust HyperMC, which uses global grid initialization followed by elite-guided local refinement to reduce sensitivity to random candidate generation and noisy finite-budget evaluations. Under suitable approximation and concentration conditions for the estimated KSD, we establish that the successive-halving component selects a near-optimal configuration among the sampled candidates with high probability and derive a sufficient computational budget for successful selection. Experiments on logistic regression, probabilistic matrix factorization, and Bayesian neural networks show that HyperMC improves posterior approximation or predictive calibration relative to MAMBA, grid search, and heuristic baselines, while Robust HyperMC yields more stable and reproducible tuning results.
[LG-77] Posterior Tempering Explains Variance Inflation in Linear and Generalized Linear Thompson Sampling
链接: https://arxiv.org/abs/2609.01999
作者: Prateek Jaiswal,Debdeep Pati,Anirban Bhattacharya,Bani K. Mallick
类目: Machine Learning (stat.ML); Information Theory (cs.IT); Machine Learning (cs.LG); Statistics Theory (math.ST)
*备注:
Abstract:We study a variant of the Thompson Sampling (TS) algorithm, called \alpha -TS, for solving stochastic generalized linear bandit problems. Existing analyses of TS require inflating the posterior variance to derive near-optimal regret guarantees. We formalize the idea of variance inflation by introducing \alpha -TS that uses a fractional or \alpha -posterior instead of the standard posterior. Our main contribution is to identify general regularity conditions on the prior and reward distributions that enable a regret analysis of \alpha -TS without assuming any tractable approximation of the posterior distribution, unlike previous works. For a specific choice of \alpha \propto d^-1 , our general regret bound yields the best known regret bound of O(d^3/2\sqrtT\log T) for both the exponential and sub-Gaussian families of reward distributions. We further provide an \alpha -dependent lower bound showing that the regret constant depends on the product \alpha d , and that when \alpha \propto d^-1 the regret scales as \Omega(d^3/2\sqrtT) , explaining the origin of the d^3/2 factor in the upper bound. Our proof technique adapts and combines recent advancements in the analysis of linear bandit problems with first- and second-order posterior concentration theory from the Bayesian statistics literature.
[LG-78] Basin Geometry and Reliable Recall of Dynamical Memories in Reservoir Computing
链接: https://arxiv.org/abs/2609.01914
作者: Ling-Wei Kong,Ying-Cheng Lai
类目: Chaotic Dynamics (nlin.CD); Machine Learning (cs.LG)
*备注:
Abstract:Reliable attractor recall conventionally requires broad basins of attraction. However, in reservoir-computing based associative memory, temporal cues reliably recover dynamical memories despite basins dominated by unpredictable, riddled-like regions. We reveal that memory basins exhibit an octopus-like'' structure: a robust head’’ near the attractor and thin, intertwined ``tentacles’’ spanning state space. Initial states in tentacular regions yield near-zero uncertainty exponents, making the recalled memory effectively unpredictable at finite precision. Yet, cue-driven generalized synchronization bypasses this unpredictability, driving the system into the robust basin head. This mechanism yields a quantitative relation linking minimum cue duration, synchronization rate, and basin-head radius. Trained recurrent neural networks exhibit similar geometry, suggesting this phenomenon extends beyond reservoir computing.
[LG-79] Latent unified smooth Hamiltonians for excited state chemistry
链接: https://arxiv.org/abs/2609.01871
作者: David Juergens,Martin Stöhr,Andreas E. Hillers-Bendtsen,O. Jonathan Fajen,Todd J. Martínez
类目: Chemical Physics (physics.chem-ph); Machine Learning (cs.LG)
*备注: 24 pages main, 7 pages supplement, 4 figures main, 6 figures supplement, 2 tables supplement
Abstract:We describe a neural network architecture and training procedure designed to model electronic ground and excited states of arbitrary molecular systems. By indirectly learning a latent, implicit basis representation of the electronic-state Hamiltonian, the model offers a unified treatment of multiple electronic states, conical intersections, and non-adiabatic couplings. The formalism can be further extended to learn consistent latent representations of additional operators such as transition dipole moments, for example. To demonstrate the general capabilities of our architecture, we train and evaluate networks on two realistic photochemical systems, thymine and azobenzene. The resulting models accurately reproduce energies and oscillator strengths for the ground- and low-lying excited states relevant to the photochemistry of these systems. We highlight the performance of the trained networks by studying critical molecular geometries, including conical intersections and excited state minima. By construction, the proposed framework also recovers the emergence of Berry phase accumulation around conical intersections. By pairing key mathematical structure from quantum chemistry with the representation learning power of transformers, the presented architecture offers a qualitatively new path toward fast and accurate ground- and excited-state simulations.
[LG-80] Pooling and Drift in Delayed Bandits
链接: https://arxiv.org/abs/2609.01761
作者: Melika Baghi
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:A system often has to act long before it learns whether the act worked: a recommender sees a click in seconds and a purchase in days. With K actions and a delay of d rounds, the best rate known for this setting is \widetildeO(\sqrt(K+d)T) over T rounds, so a longer menu is always more expensive to learn from. It need not be: if the outcome depends on the action only through the state it produced, then one late outcome informs every action that could have produced the observed state, and the price is set by how many genuinely different states the actions produce rather than by how many actions there are. We measure this using an effective dimension v_t between 1 and the number of states, and prove \widetildeO(\sqrt(d+1)V\log K) for a rotating algorithm and \widetildeO(\sqrtV^-+\sqrtdT) for the single-copy algorithm used in practice, for any budget fixed in advance; merging similar states lowers the price further, at an explicit bias. Even when given the exact losses from d rounds ago, no algorithm escapes \Omega(\sqrtdE\min\1+\log J,T/d) , where J counts the drifting directions and E bounds how far losses move while the learner waits. On generated data, the state channel cuts regret by up to 79 percent against action-level weighting and, on the funnel family, by 32 to 68 percent against a tuned minimax-optimal method.
[LG-81] Random Forest-Informed Cellular Automaton for Large-Scale Wildfire Spread Modelling
链接: https://arxiv.org/abs/2609.01675
作者: Siyu Chen,Esha Saha,Hao Wang
类目: Applications (stat.AP); Machine Learning (cs.LG)
*备注: 31 pages, 9 figures
Abstract:Accurate large-scale wildfire spread modelling requires models that capture both the environmental conditions associated with fire occurrence and the local dynamics of fire propagation. We propose a three-stage framework that combines a Random Forest (RF) model with a cellular automaton (CA). First, an RF model trained on the 2021 Canadian fire season estimates daily pixel-level fire-occurrence probabilities. Second, quantile gradient boosting models provide optional spread-rate priors for sensitivity analysis. Third, an RF-informed CA combines the RF probability layer with neighbourhood-driven spread on a 5 km grid. The RF model achieved AUC values of 0.725–0.795 on the 2022–2024 datasets, while the RF-informed CA achieved substantially higher spatial overlap than the evaluated CA-only baselines in the 2023 simulation. A higher-resolution simulation provides an additional qualitative assessment of local spatial errors. These results suggest that combining RF-derived probabilities with local CA spread can improve large-scale wildfire simulations under the tested conditions.
附件下载
